DNA Breaks in Hypermutating Immunoglobulin Genes: Evidence for a Break-and-Repair Pathway of Somatic Hypermutation

Corresponding author: Nancy Maizels, Departments of Immunology and Biochemistry, University of Washington Medical School, Rm. H474A HSB, 1959 NE Pacific St., Box 357650, Seattle, WA 98195-7650., maizels{at}u.washington.edu (E-mail)

Communicating editor: N. ARNHEIM

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

To test the hypothesis that immunoglobulin gene hypermutation in vivo employs a pathway in which DNA breaks are introduced and subsequently repaired to produce mutations, we have used a PCR-based assay to detect and identify single-strand DNA breaks in 1 genes of actively hypermutating primary murine germinal center B cells. We find that there is a two- to threefold excess of breaks in 1 genes of hypermutating B cells, relative to nonhypermutating B cells, and that 1.3% of germinal center B cells contain breaks in the 1 gene that are associated with hypermutation. Breaks were found in both top and bottom DNA strands and were localized to the region of 1 that actively hypermutates, but duplex breaks accounted for only a subset of breaks identified. Almost half of the breaks in hypermutating B cells occurred at hotspots, sites at which two or more independent breaks were identified. Breaksite hotspots were associated with characteristic sequence motifs: a pyrimidine-rich motif, either RCTYT or CCYC; and RGYW, a sequence motif associated with hypermutation hotspots. The sequence motifs identified at breaksite hotspots should inform the design of substrates for characterization of activities that participate in the hypermutation pathway.

SOMATIC hypermutation is a highly regulated process that specifically targets immunoglobulin genes for very high levels of mutation. The great majority of the mutations are single base substitutions, with more transitions than transversions. The mutation rate is about one mutation per kilobase pair per generation, which is 10⁶-fold higher than the typical mutation rate in mammalian somatic cells. In mice and human, hypermutation occurs after challenge with antigen and is coupled with selection for high-affinity B cell clones. Following mutation and selection, a B cell can produce antibody with greatly increased affinity for antigen. Hypermutation thus increases the efficiency of the humoral immune response.

Extensive study of hypermutation has produced some understanding of the biological requirements for this process (reviewed by KLEIN et al. 1998 ; KONG et al. 1998A ; NEUBERGER et al. 1998 ; STORB et al. 1998 ; WINTER and GEARHART 1998 ). Transcriptional activation is necessary but not sufficient to target an immunoglobulin gene for hypermutation, and hypermutated sequences are restricted to the region 1–2 kb downstream of the promoter. Hypermutation occurs in specialized physiological microenvironments called germinal centers (KELSOE 1996 ). A huge database of hypermutated sequences has been compiled, and comparison of these sequences has revealed common hotspots for hypermutation. The sequence motif RGYW (R = A or G; Y = C or T; W = A or T) is frequently found at hypermutated sites (ROGOZIN and KOLCHANOV 1992 ; BETZ et al. 1993 ; JOLLY et al. 1996 ; ROGOZIN et al. 1996 ). Reporter genes (ß-globin, gpt, and neo) can be targeted for hypermutation when they are substituted for immunoglobulin variable regions in transgenic constructs, and the RGYW motif is also a hotspot for mutation in these reporter genes (YELAMOS et al. 1995 ).

Hypermutation alters immunoglobulin loci in vertebrates from nurse sharks to mammals, but the regulation and products of hypermutation differ from species to species. In chicken (REYNAUD et al. 1987 ; THOMPSON and NEIMAN 1987 ), rabbit (BECKER and KNIGHT 1990 ; WEINSTEIN et al. 1994 ; MAGE 1998 ), pig (BUTLER et al. 1996 ), and cattle (PARNG et al. 1995 , PARNG et al. 1996 ), hypermutation occurs prior to encounter with antigen to diversify the preimmune repertoire, and most mutations are templated. In sheep, hypermutation also occurs prior to encounter with antigen, but mutations are untemplated (REYNAUD et al. 1991 ). In mouse (BETZ et al. 1993 ; WAGNER et al. 1995 ; JOLLY et al. 1996 ; KONG et al. 1998B ), human (DORNER et al. 1998 ; DUNN-WALTERS et al. 1998 ; VARADE et al. 1998 ; FOSTER et al. 1999 ), and the distantly related nurse shark, a cold-blooded vertebrate (DIAZ et al. 1998 , DIAZ et al. 1999 ), hypermutation occurs after antigen stimulation and most mutations are untemplated. In an attempt to reconcile the apparent paradox of closely related organisms using distinct mechanisms to achieve hypermutation, we have proposed that hypermutation in all organisms occurs by a break-and-repair pathway (MAIZELS 1995 ; KONG et al. 1998A ; HARRIS et al. 1999 ). First, a nick or double-strand break (DSB) disrupts the DNA duplex, and then repair by either gene conversion or unfaithful copying would produce either templated or untemplated mutations. This model generates a testable hypothesis: that DNA lesions related to hypermutation can be found in immunoglobulin genes of primary hypermutating B cells.

One line of evidence that breaks occur as part of the immunoglobulin gene hypermutation mechanism in humans and mice comes from the identification of deletion and insertion mutations produced as a by-product of hypermutation in vivo (GOOSSENS et al. 1998 ; WILSON et al. 1998A , WILSON et al. 1998B ). These sorts of mutations are likely to result from inaccurate repair of DNA breaks. Another line of evidence comes from analysis of immunoglobulin loci in lymphoid malignancies. The involvement of DNA breaks in many B cell lymphomas is well documented, and while some of these breaks appear to result from aberrant V(D)J recombination or switch recombination, others appear to result from breaks induced during hypermutation and implicate aberrant hypermutation events as contributing to some B cell malignancies (GOOSSENS et al. 1998 ; KUPPERS et al. 1999 ).

Direct evidence for the involvement of DNA breaks in hypermutation first came from analysis of a human B cell lymphoma line, Ramos, which carries out constitutive immunoglobulin gene hypermutation in cell culture, although at a level approximately fivefold below that of primary germinal center B cells (SALE and NEUBERGER 1998 ). SALE and NEUBERGER 1998 found an increase in the level of insertion and deletion mutations in the light chain gene V region in Ramos cells, as well as untemplated nucleotide insertion in transfectants expressing terminal deoxynucleotidyl transferase. Recently, two groups used ligation-mediated PCR to identify DSBs in immunoglobulin genes in B cells carrying out hypermutation. PAPAVISILIOU and SCHATZ 2000 identified DSBs in the immunoglobulin V regions of the Ramos cell line, and by analysis of a reporter transgene confirmed that breaks could also be identified in hypermutating B cells in vivo. BROSS et al. 2000 also identified DSBs in a "knock-in" mouse model, generated by targeting a modified immunoglobulin heavy chain gene to the heavy chain locus. Both groups showed that the appearance of breaks is dependent upon transcriptional activation, as would be predicted by the dependence of hypermutation on proximal enhancer elements.

We have assayed for the presence of breaks in murine immunoglobulin 1 light chain genes using a PCR-based assay designed to identify single-strand breaks. We find that the level of single-strand DNA breaks is two- to threefold higher in hypermutating B cells than in nonhypermutating B cells isolated from the same mouse. We measured breaks in both the endogenous 1 gene of wild-type C57BL/6 mice and in a 1 transgene that hypermutates actively (KONG et al. 1998B ) and find similar levels and patterns of breaks in the endogenous gene and the transgene. Sequence analysis shows that a significant fraction of the breaks in hypermutating B cells are clustered at distinct hotspots, which contain one of two pyrimidine-rich consensus motifs, RCTYT or CCYC. The hypermutation hotspot sequence motif, RGYW, was also found at or near most of the breaksite hotspots. Our observations support the hypothesis that immunoglobulin gene hypermutation depends upon a break-and-repair pathway. They also provide direct evidence for DNA breaks at an endogenous immunoglobulin locus in vivo. The consensus sequence motifs at breaksite hotspots may define the specificity of one or more activities essential to the mechanism of immunoglobulin gene hypermutation.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of PNA^hi and PNA^lo B cells and DNA preparation:
The endogenous 1 gene was analyzed in wild-type C57BL/6 mice. A rearranged 1-E_2-4 transgene, which undergoes active hypermutation, was analyzed in the LZ15-90 transgenic line (KONG et al. 1998B ). Mice were immunized intraperitoneally with 100 µg alum-precipitated NP-CGG and 2 x 10⁹ Baccilus pertussis. Secondary immunization was with 10 µg NP-CGG in saline (PEAKMAN and MAIZELS 1998 ). Single cell suspensions were prepared from spleen, lymph nodes, and Peyer's patches. Cells were stained with PNA-FITC and B220-RPE as previously described (KONG et al. 1998B ) and sorted using a FACSVantage cell sorter (Beckton Dickinson, San Jose, CA).

Identification of breaksites:
We developed a method for rapid identification of breaksites in small numbers of mammalian cells, referred to as breaksite batch mapping (KONG and MAIZELS 2001 ). This method considerably increases the speed, throughput, and convenience of breaksite identification by direct sequence analysis of ligation-mediated PCR (LM-PCR) products in batch, with no gel purification or cloning steps. This method also includes two minor modifications that enhance the sensitivity and specificity of standard LM-PCR protocols: use of a biotinylated primer in the extension step and use of nested PCR primers in the amplification steps. This method, which is described in detail elsewhere (KONG and MAIZELS 2001 ), was used for identification of breaksites in murine 1 genes. Briefly, genomic DNA was prepared from sorted B cells, carefully quantitated, and a small amount of DNA (2000 cell-equivalents) was denatured and annealed to a biotinylated primer specific for the region of interest. The primer was extended with Sequenase (United States Biochemical, Cleveland) to produce duplex DNA extension products that terminated at the site of the break with a blunt end. (A minority of Sequenase extension products may not be blunt but instead contain a nontemplated 3' A, resulting in a modest underestimation of total breaks.) After ligation to a duplex linker containing one blunt and one staggered end (generated by annealing oligonucleotides LL3 and LP2), biotinylated extension products were isolated on streptavidin-magnetic beads (Dynal, Great Neck, NY). The recovered extension products were diluted and aliquoted so that subsequent amplification reactions contained 250 cell-equivalents for analysis of the endogenous 1 gene and 50 cell-equivalents for analysis of the 1 transgene, which is present at six to eight copies/cell (KONG et al. 1998B ). DNAs were amplified by nested PCR using primer pairs specific to the region of interest or the linker. PCR products were analyzed by Southern blotting or sequence analysis. In some cases, LM-PCR products were separated by gel electrophoresis and sequenced after subcloning. In most cases, the final PCR product mixtures were sequenced directly on an ABI 3700 sequencer, and the breakpoints were identified according to the position of the linker sequence on the sequencing profiles.

The duplex linker was generated by annealing oligonucleotides LL3 (CGAGTTCAGTCCGTAGACCATGGAGATCTGAA TTC) and LP2 (GAATTCAGATCTCC). Linker-specific primers for nested PCR were LL2 (GTAGACCATGGAGATCT GAATTC) and LL4 (CGAGTTCAGTCCGTAGAC). For identification of breaks in the endogenous 1 gene top strand, extension was carried out using the biotinylated primer LRB (BBBBATGCTCTTGCTGTCAGG, B = biotin), and 1-specific primers for nested PCR were LR2 (GTAGAAATCAGTGAT CGTAC) and LR3 (ACAGGGTGACTGATGGCGAAG). Extension and amplification primers for the endogenous 1 gene bottom strand were LFB2 (BBBBGATAGTGGGTGTTTATG), wLF2 (ACTCTGGATAAGCCTGAAC), and wLF3 (GATGAT TAATGCCCCTGAGCTC). For the 1 transgene, top strand-specific primers were LRB3 (BBBBGAATGTTCTGTGCTCTC), LR4d (CTGTGTCTCTCTCTATGAC), and LR4e (GTTTTCCT TCTCCATGAGATAGC); and bottom strand-specific primers were LFB (BBBBCATGAGATCACTGTTCTC), LF2 (CAGTT ACGGAGCACACAG), and LF3c2 (GCAACAATGCGCATC TTGTCTC). LR3a (CATCTTTCAGTAGCAATCCTGG), LR5a (TATTCCTCACAAAGTTCACCAGC), LF3c3 (GATTTGCTA CTGATGACTGG), and wLF3a (CAGTGTAGTAGATTTCACA TGAC) were end-labeled and used as probes for hybridization following extension and amplification with the LRB, LRB3, LFB, and LFB2 primer sets, respectively. Primers used for sequencing were LR3a, LR3b (CCTTGCCATTGACCTCCAA TAC), LR4a (TATGCCTTCTGGGTACAAG), LR4c (CTAGGAG CCCAGCTTCCAAAC), LR4d, LR4e, LR5a, LR5b (GAGATTA GACATGAAAGGCTACAG), LR6 (TGGTTGCTGTACCATA GAG), and LR7 (TGAGTCACAACAGCCTG) for top strand breaks; and LF3c3, wLF3a, LF4 (CAATGCGCATCTTGTCTC), LF5a (ACCGAGCTCCAGGTGTTCC), LF5b (GAACCAAACT GACTGTCC), LF7 (TCTCATGGAGAAGGAAAACC), LF8 (GAGAAAAAGTTCAAGCGAGTG), and LF9 (CCAGGATTG CTACTGAAAG) for bottom strand breaks.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

DNA breaks in 1 light chain genes:
Somatic hypermutation occurs in sequestered microenvironments, called germinal centers (reviewed by PRZYLEPA et al. 1998 ). Germinal centers are dynamic structures that form in secondary lymphoid tissues like the spleen, lymph nodes, and Peyer's patches. In the germinal center, B cells make intimate contact with antigen, proliferate actively, and undergo selection for production of high-affinity antibodies. To ask if breaks are present in 1 genes of hypermutating primary B cells, DNA was prepared from hypermutating germinal center B cells (PNA^hi B220⁺) and nongerminal center B cells (PNA^lo B220⁺), which had been sorted by FACS from secondary lymphoid tissues of individual immunized mice. Break sites were detected using 1-specific primers, as described in MATERIALS AND METHODS. Break sites were analyzed not only in the endogenous 1 gene of wild-type C57BL/6 mice but also in a 1 transgene that had been shown to undergo active transcription and hypermutation (KONG et al. 1998B ).

Fig 1 shows examples of one set of amplification reactions, in which breaks in DNA preparations from spleen, lymph nodes, or Peyer's patches of a single C57BL/6 mouse were amplified, resolved by gel electrophoresis, blotted, and hybridized with a 1-specific oligonucleotide probe. In each of the tissues analyzed, essentially every band visible by ethidium bromide staining hybridized with the probe (Fig 1A), verifying the specificity of the PCR-based assay. Specificity of other primer sets was comparable (data not shown). As illustrated by the examples in Fig 1, there were more distinct product bands produced upon amplification of samples from germinal center B cells than nongerminal center B cells, indicative of more DNA breaks in this cell type. This was the case for samples from spleen (Fig 1A), lymph node (Fig 1B), and Peyer's patches (Fig 1C).

View larger version (99K): In this window In a new window Download PPT slide   Figure 1. Identification of DNA breaks in germinal center (PNAhi) and nongerminal center (PNAlo) primary B cells. DNA was analyzed in B cells that originated from three different lymphoid tissues: (A) spleen; (B) lymph node; and (C) Peyer's patches. Breaks were identified as described in MATERIALS AND METHODS, using the primer set LRB, LR2, and LR3, which amplifies breaks in the top (nontranscribed) DNA strand, and template DNA from germinal center (PNAhi) and nongerminal center (PNAlo) B cells from a C57BL/6 mouse. The DNA products were separated by agarose gel electrophoresis and hybridized with 32P-labeled oligonucleotide LR3a, complementary to the J1-C1 intron. The ethidium bromide-stained gel is shown at left (M1, 1-kb ladder; M2, 100-bp ladder), and the corresponding Southern blot at right. A diagram of the rearranged 1 gene at the left of A–C provides an approximate map of where breaks occurred. L, leader; V, variable region; C, constant region.

Table 1 summarizes experiments that compared the number of single-strand DNA breaks in the top (nontranscribed) strand of the 1 gene in germinal center and nongerminal center B cells from five individual mice, three wild type (mice 1–3), and two transgenics (mice 4 and 5). The ratio of single-strand breaks in germinal center and nongerminal center B cells ranged from 2.3 to 3.3 and was similar in wild-type and transgenic mice. The absolute number of breaks was higher in the transgenic mice because the 1 transgene is present at six to eight copies per cell (KONG et al. 1998B ). The average level of top strand breaks per 100 cells in the single copy endogenous 1 genes of the wild-type mice was 1.0 in the germinal center B cells, and 0.35 in the nongerminal center B cells. Thus, in the 1 genes of primary germinal center B cells there was an excess of 0.65 single-strand breaks on the top strand per 100 cells. Similar results were obtained when DNA break frequencies on the bottom (transcribed) strand were analyzed (data not shown). We therefore estimate that 1.3% of primary germinal center B cells contain breaks associated with hypermutation in either strand of the 1 genes.

View this table: In this window In a new window   Table 1. Top strand DNA beaks/100 cells in 1 genes of germinal center (PNAhi) and nongerminal center (PNAlo) primary B cells

Breaksite hotspots in 1 genes from germinal center B cells:
The precise sites of DNA breaks were determined by sequencing LM-PCR products. Sequences of 183 breaks were determined, 84 from the endogenous 1 genes and 99 from the 1 transgenes. Breaksites were identified in DNA from both germinal center (PNA^hi) and nongerminal center (PNA^lo) B cells. The breaksites are mapped in Fig 2. Strikingly, in the samples from the germinal center B cells, there were numerous examples of multiple independent breaks at single sites. In contrast, there was only one example of a site with more than one break in the samples from nongerminal center B cells, and there were a total of only 2 breaks at this site.

View larger version (22K): In this window In a new window Download PPT slide   Figure 2. Sites of DNA breaks in 1 genes in germinal center (PNAhi) and nongerminal center (PNAlo) primary B cells. The positions of break-sites were determined by sequencing LM-PCR products. Separate maps are shown for breaks identified in PNAhi and PNAlo cells from (A) the endogenous 1 gene in wild-type C57BL/6 mice and (B) the 1 transgene in the LZ15-90 line. Breaks on the top (nontranscribed) strand are shown above the line, and breaks on the bottom (transcribed) strand below the line. L, leader region; VJ, variable region; C, constant region.

Table 2 compiles these data, showing the number of breaks and breaksites in DNA samples from germinal center and nongerminal center B cells. Of the total of 133 breaks identified in germinal center B cells, 55 (or 41%) were at sites of 2 or more breaks. This contrasts with the results from nongerminal center B cells, where only 2 of 50 breaks (4%) mapped to the same site. If breaks occurred randomly, for example, during DNA preparation, some small fraction of repeated hits would be expected in the dataset, and this fraction would increase with the number of breaks analyzed. Nonetheless, the difference in the number of breaksites identified in the PNA^hi and PNA^lo cell samples is not sufficient to explain the large difference in multiple hits at individual breaksites. Instead, these results suggest that germinal center B cells carry breaks at specific sites in the 1 gene.

Identical breaksite hotspots in the endogenous and transgenic 1 genes:
Multiple independent breaks were identified at nine sites in the endogenous 1 gene, and nine sites in the 1 transgene. The sequences of these sites and the number of breaks identified at each site are shown in Fig 3. For simplicity, we refer to sites at which multiple independent breaks were identified as breaksite hotspots. Two top strand and two bottom strand breaksite hotspots were identical in the endogenous 1 gene and the transgene and are marked with asterisks in Fig 3, at positions 236 (bottom strand) and 377 (top strand). The breaksite hotspot at position 236 is in the V1 region, 33 bp upstream of CDR3. Nine breaks at this site were independently identified, 7 in samples from wild-type mice and 2 in samples from the transgenic line. The breaksite hotspot at position 377 is in the J1-C1 intron. Ten breaks at this site were independently identified in the top strand, 5 in samples from wild-type mice, and 5 in samples from the transgenic line. Three breaks were also identified in the bottom strand at the same position. As discussed below, identification of breaks at the same site in top and bottom strands is consistent with the possibility that at least some of the breaks are DSBs. In addition, 3 top strand breaks were identified one nucleotide downstream (position 378), for a total of 16 breaks clustered at positions 377–378. A cluster of breaks was also observed in the bottom strand (positions 256–257) 12 bp upstream of CDR3. These breaksite hotspot clusters are indicated by braces in Fig 3.

View larger version (31K): In this window In a new window Download PPT slide   Figure 3. Sequences of breaksite hotspots. Sequences flanking all sites at which multiple breaks were identified are shown. Breaks are indicated by gaps. The number of times each breaksite was independently identified is shown in the second column, and the third column shows the position of the break, using the numbering system of GONZÁLEZ-FERNÁNDEZ et al. (1994), in which the first nucleotide of the V region is number 1 and the last nucleotide in the J region is position 328. Breaksites identified on both top and bottom strands are indicated by double daggers (). Breaksites identified in samples from both wild-type and transgenic mice are indicated by asterisks (*), and numbers identified in wild-type and transgenics, respectively, are shown in parentheses. In two cases, breaks were separated by only a single base. These clustered breaksites are indicated by brackets, and the breaksite sequences are aligned. Pyrimidine-rich consensus sequence motifs RCTYT and CCYC are underlined. Matches to the hypermutation hotspot RGYW and its reverse complement, WRCY, are in lightface type.

The breaks in the endogenous 1 gene and the 1 transgene were amplified with different primers. Identification of identical breaksite hotspots in these samples therefore supports the premise that these breaks reflect specific cleavage in vivo and are not artifacts of PCR amplification. The fact that identical breaksite hotspots were found in samples from both wild-type C57BL/6 mice and 1 transgenics also eliminates chromosomal location or details of gene organization as possible contributors in determining breaksites.

A duplex breaksite hotspot:
Breaksite hotspots are mapped in Fig 4A. As the map shows, the most frequently targeted breaksite hotspots identified in this dataset map at the 3' end of the V1 region or 5' end of the J1-C1 intron. This region of the 1 gene is highly mutated in vivo (GONZALEZ-FERNANDEZ et al. 1994 ; AZUMA 1998 ). One breaksite hotspot (position 236) maps just upstream of CDR3. Another (position 269) is within CDR3 and just 1 bp upstream of a hypermutation hotspot at Ala89, characterized by frequent mutation of the G in the GCT triplet (GONZALEZ-FERNANDEZ et al. 1994 ; JOLLY et al. 1996 ). The most heavily targeted hotspot in this collection maps in the intron (position 377–378), just downstream of the 3' end of the J1 region. This breaksite in the J1-C1 intron is 30 bp upstream of a region of the intron that hypermutates very actively (GONZALEZ-FERNANDEZ et al. 1994 ). Breaks at this site occurred in both top and bottom strands (see Fig 3 and Fig 4A) and in samples from both wild-type mice and transgenics.

View larger version (18K): In this window In a new window Download PPT slide   Figure 4. Breaksites correlate with somatic hypermutation. (A) Breaksite hotspots mapped in the 1 gene. The murine 1 locus is shown. Arrows denote positions of breaksite hotspots listed in Fig 3. Numbers above arrows denote the number of breaks at each position among the 133 breaksites identified from PNAhi cells. (B) Breaks localize to the region of the J1-C1 intron that hypermutates. The numbers and ratios of breaks in two intervals within the intron are compared in germinal center (PNAhi) and nongerminal center (PNAlo) B cells. The upstream interval is 400 bp in length, undergoes active hypermutation, and includes a hypermutation hotspot (GONZALEZ-FERNANDEZ et al. 1994 ) that is denoted by a star. The downstream interval is 753 bp long in wild-type mice and 639 bp in the transgenics and mutates at much lower levels.

Recurrent sequence motifs at breaksite hotspots:
RGYW has been identified as a consensus site at which hypermutation frequently occurs (ROGOZIN and KOLCHANOV 1992 ; BETZ et al. 1993 ; JOLLY et al. 1996 ). Almost all breaksites contain at least one match to either the RGYW motif or to its reverse complement, WRCY (Fig 3). Matches to the WRCY motif are in most cases at the breaksite.

All of the breaksite hotspots also contain one of two pyrimidine-rich motifs (Fig 3). An RCTYT consensus motif occurs at or just downstream of all breaksite hotspots on the top strand and at hotspots upstream of the leader on the bottom strand. This motif is in many cases preceded by a T, which fuses the WRCY hotspot motif (as TRCT) and the RCTYT motif to generate an expanded consensus motif, TRCTYT. A CCYC consensus motif occurs at or near hotspots in the variable region and the J1-C1 intron on the bottom strand. The same motifs are found in samples from both transgenic and wild-type germinal center B cells.

The potential for the DNA sequences spanning the breaksite hotspots to form secondary structures was analyzed by using the Mfold program (http://mfold.wustl. edufolder/dna/form1.cgi). While many breaksite hotspots had structural potential, this analysis did not identify structural motifs consistently associated with cleavage sites. Nonetheless, because predictions of secondary structure are sensitive to parameters such as sequence length and temperature, this does not rule out the possibility that DNA must become transiently structured for cleavage.

Breaks localize to the region of the 1 gene that hypermutates:
To ask if the breaks in primary PNA^hi B cells reflected generalized degradation of genomic DNA rather than a feature of the hypermutation mechanism, we wanted to compare breaks in regions of DNA that mutate at very different rates. We chose for comparison two regions of the J1-C1 intron. The upstream region of this intron hypermutates actively and includes a hotspot that hypermutates at a level comparable to the complementarity-determining regions (CDRs) within the V segments; and the downstream region mutates at a much lower level (GONZALEZ-FERNANDEZ et al. 1994 ; AZUMA 1998 ). The J1-C1 region has a further advantage for this analysis, namely, that it does not encode protein and is not subject to clonal selection. Moreover, by comparing hypermutation of different zones within the same gene, rather than two different genes, the comparison could be made independently of differences in levels of transcription. We separately analyzed breaks in two regions within the J1-C1 intron: a 400-bp upstream interval, with its 5' border at the J1 boundary, and a downstream interval containing the remainder of the intron, with the 3' boundary at C1. The downstream interval is 753 bp in the wild-type mice and 639 bp in the transgenics. This difference is due to a 114-bp deletion introduced into the intron during transgene construction (HENGSTSCHLAGER et al. 1994 ).

Fig 4B shows that in DNA from germinal center B cells there were 60 breaks in the upstream interval and 14 breaks in the downstream interval and that in nongerminal center B cells there were 22 breaks in the upstream interval and 14 breaks in the downstream interval. The ratio of breaks in germinal center (PNA^hi) to nongerminal center (PNA^lo) cells is therefore 2.7 in the upstream interval and 1.0 in the downstream interval. Because the same PCR primers and protocols were used to amplify DNA from both cell populations, the data are very unlikely to reflect particulars of the PCR analysis. We conclude that the excess of breaks in germinal center B cells is confined to regions of DNA that actively hypermutate. This provides further evidence that the breaks are associated with the hypermutation mechanism.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have identified single-strand breaks in 1 genes of actively hypermutating primary germinal center B cells. These breaks appear to be intermediates in the hypermutation pathway, for the following reasons: (1) There are more breaks in 1 genes from germinal center B cells than from nongerminal center B cells; (2) a significant fraction (41%) of the breaks in germinal center B cells occur at breaksite hotspots, sites at which two or more independent breaks were identified; (3) characteristic motifs could be identified at breaksite hotspots, including the RGYW motif (or its reverse complement, WRCY), which is associated with hotspots for somatic hypermutation; and (4) breaks are concentrated in DNA regions that hypermutate.

Fig 5 presents a simple model for how repair of a DNA break could produce mutations. In the first step, DNA undergoes cleavage, shown here as producing a single-strand break. In the second step, DNA is resected; next, error-prone repair introduces mutations; and in the last step the duplex is religated. The figure shows resection occurring by a 3' 5' exonuclease to dramatize the fact that the 5' end of a DNA break may persist unaltered between the initial cleavage event and the final religation step. In contrast, the 3' end of the DNA break is effectively a moving target, which may be attacked by an exonuclease during resection, and which will necessarily travel in the 5' 3' direction during repair by DNA polymerase. The model in Fig 5 is similar in outline to the model originally proposed by BRENNER and MILSTEIN 1966 to explain the origin of antibody variability. The pathway outlined in Fig 5 could readily accommodate initial cleavage to produce a double-strand break, resection by a nuclease with opposite polarity, or strand displacement that accompanies repair synthesis. Cleavage followed by strand transfer rather than error-prone repair could result in templated mutation, as previously proposed by our laboratory (MAIZELS 1995 ; KONG et al. 1998A ; HARRIS et al. 1999 ).

View larger version (12K): In this window In a new window Download PPT slide   Figure 5. DNA break-and-repair model for somatic hypermutation. A single-strand break introduced into the DNA is shown (top line). The broken strand is resected by a 3'–5' exonuclease (second line); mutations (stars) are introduced by an error-prone polymerase (third line); and the DNA is religated (last line). Errors introduced into one DNA strand can be fixed by subsequent DNA replication or by mismatch repair. A similar model can also be drawn for initiation of hypermutation by a double-strand break. This polarity for resection is shown to emphasize how a 5' end may persist during both resection and repair, but our data do not preclude resection by a nuclease of opposite polarity or strand displacement by a repair polymerase.

The LM-PCR assay we used ligates an identifying linker to a free 5' end. The assay thus reports both single-strand breaks and DSBs. If 5' 3' exonucleolytic degradation accompanies either repair or strand displacement, the dynamic changes in the position of the 5' end will also be identified as nicks at distinct sites; but the assay will not register 3' 5' exonucleolytic digestion or extension of the 3' end by a repair polymerase. Thus, some of the breaksites we have identified in V1 regions of primary germinal center B cells may represent initial sites of DNA cleavage, and others may represent sites at which resection or repair is ongoing. Several of the breaksite hotspots were clustered (Fig 3). Clustering is consistent with a mechanism in which initial cleavage is imprecise or in which limited processing of the breaksite occurs prior to linker ligation.

Other laboratories identified DSBs in immunoglobulin genes in hypermutating B cells (BROSS et al. 2000 ; PAPAVISILIOU and SCHATZ 2000 ). Because the assay we used detects breaks on both strands, DSBs would be expected to comprise a subset of the breaks we identified. At least one breaksite, at positions 377–378, almost certainly reflects cleavage of the DNA duplex, because breaks were identified on both top and bottom strands at this site (see Fig 3 and Fig 4A). This is the most predominant breaksite identified in the 1 gene: breaks at this site account for 12% of the total breaks in germinal center B cells and 29% of the breaks at hotspots. Breaks were found at this site in both the endogenous 1 gene and the transgene. This site is only 30 bp upstream from the region of the J1-C1 intron that hypermutates very actively (positions 406–431; GONZALEZ-FERNANDEZ et al. 1994 ), and it is possible that initial cleavage at 377–378 is followed by repair that preferentially alters the sequence at this downstream region. Other breaksite hotspots we identified may also correspond to sites of duplex cleavage, but the fact that we did not detect breaks on both strands at some of the most frequently cleaved sites makes it unlikely that all breaks are DSBs. We estimate that 1.3% of primary germinal center B cells contain breaks in the 1 regions associated with hypermutation (Table 1). DSBs were estimated to be present in the V_H region of 4% of Ramos cells (PAPAVISILIOU and SCHATZ 2000 ). The higher fraction of Ramos cells with breaks may reflect longer persistence of breaks in cultured cells, either because breaks are repaired less rapidly in this cell line or because cultured cells cycle less rapidly than primary germinal center B cells.

Several lines of evidence suggest that there may be two or more components to the hypermutation apparatus (MILSTEIN et al. 1998 ; RADA et al. 1998 ; SALE and NEUBERGER 1998 ; EHRENSTEIN and NEUBERGER 1999 ; SPENCER et al. 1999 ; COWELL and KEPLER 2000 ). It is possible that one distinguishing feature of these two components is creation or persistence of single-strand breaks and DSBs. The possibility that one component of this process involves single-strand breaks is consistent with considerable evidence pointing to the involvement of mismatch repair in somatic hypermutation (CASCALHO et al. 1998 ; FREY et al. 1998 ; KIM and STORB 1998 ; PHUNG et al. 1998 , PHUNG et al. 1999 ; RADA et al. 1998 ; WINTER et al. 1998 ; KONG and MAIZELS 1999 ; VORA et al. 1999 ; WIESENDANGER et al. 2000 ) and with the absence of a requirement for DNA protein kinase (BEMARK et al. 2000 ) in hypermutation. The presence of single-strand breaks is also consistent with observations of BROSS et al. 2000 , who identified DSBs in heavy chain genes in murine germinal center B cells. This group found that the DNA end at the promoter-proximal side of breaks was more readily ligated to blunt duplex linkers than the promoter-distal end, leading them to suggest that a single-strand lesion may be processed to produce a blunt end.

Specific sequence motifs could be identified at or near most breaksite hotspots. Matches to the hypermutation consensus, RGYW, and either of two pyrimidine-rich consensus motifs were also evident at hotspots. The RCTYT consensus motif was found at top strand breaksite hotspots and at bottom strand hotspots upstream of the beginning of the leader intron. The CCYC motif was found at bottom strand breaksite hotspots within and downstream of the V region. The sites at which multiple independent breaks were identified in the V2 region of the human B cell line Ramos (PAPAVISILIOU and SCHATZ 2000 ) also contain matches to either the RCTYT or the CCYC motif. Although our data do not bear directly on the question of whether transcription plays a role in production of DNA breaks, the observation that different sequence motifs predominated at breaksite hotspots on the transcribed and nontranscribed strands is intriguing in light of the possibility that the hypermutation apparatus differentiates between the transcribed and nontranscribed DNA strands.

The consensus motifs at hotspots are especially compelling evidence in support of the notion that the breaks reflect some aspect of the hypermutation mechanism. The breaksite hotspot sequence motifs identified in our experiments may represent sites of endonuclease cleavage or sites at which the exonucleases or polymerases pause during DNA repair. If the breaksite hotspot motifs do define the specificity of one or more activities essential to the hypermutation mechanism, then it should be possible to identify these enzymes and establish their functions.

	FOOTNOTES

¹ Present address: Division of Neuropathology, Department of Pathology, Case Western Reserve University, 2085 Adelbert Rd., Cleveland, OH 44106.

	ACKNOWLEDGMENTS

We are grateful to many members of the Maizels laboratory for helpful discussions. This research was supported by National Institutes of Health grant R01GM41712.

Manuscript received January 9, 2001; Accepted for publication February 12, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AZUMA, T., 1998  Somatic hypermutation in mouse chains. Immunol. Rev. 162:97-105[Medline].

BECKER, R. S. and K. L. KNIGHT, 1990  Somatic diversification of immunoglobulin heavy chain VDJ genes: evidence for somatic gene conversion in rabbits. Cell 63:987-997[Medline].

BEMARK, M., J. E. SALE, H. J. KIM, C. BEREK, and R. A. COSGROVE et al., 2000  Somatic hypermutation in the absence of DNA-dependent protein kinase catalytic subunit (DNA-PK(cs)) or recombination-activating gene (RAG)1 activity. J. Exp. Med. 192:1509-1514[Abstract/Free Full Text].

BETZ, A. G., M. S. NEUBERGER, and C. MILSTEIN, 1993  Discriminating intrinsic and antigen-selected mutational hotspots in immunoglobulin V genes. Immunol. Today 14:405-411[Medline].

BRENNER, S. and C. MILSTEIN, 1966  Origin of antibody variation. Nature 211:242-243[Medline].

BROSS, L., Y. FUKITA, F. MCBLANE, C. DEMOLLIERE, and K. RAJEWSKY et al., 2000  DNA double-strand breaks in immunoglobulin genes undergoing somatic hypermutation. Immunity 13:589-597[Medline].

BUTLER, J. E., J. SUN, I. KACSKOVICS, W. R. BROWN, and P. NAVARRO, 1996  The V_H and C_H immunoglobulin genes of swine: implications for repertoire development. Vet. Immunol. Immunopathol. 54:7-17[Medline].

CASCALHO, M., J. WONG, C. STEINBERG, and M. WABL, 1998  Mismatch repair co-opted by hypermutation. Science 279:1207-1210[Abstract/Free Full Text].

COWELL, L. G. and T. B. KEPLER, 2000  The nucleotide-replacement spectrum under somatic hypermutation exhibits microsequence dependence that is strand-symmetric and distinct from that under germline mutation. J. Immunol. 164:1971-1976[Abstract/Free Full Text].

DIAZ, M., A. S. GREENBERG, and M. F. FLAJNIK, 1998  Somatic hypermutation of the new antigen receptor gene (NAR) in the nurse shark does not generate the repertoire: possible role in antigen-driven reactions in the absence of germinal centers. Proc. Natl. Acad. Sci. USA 95:14343-14348[Abstract/Free Full Text].

DIAZ, M., J. VELEZ, M. SINGH, J. CERNY, and M. F. FLAJNIK, 1999  Mutational pattern of the nurse shark antigen receptor gene (NAR) is similar to that of mammalian Ig genes and to spontaneous mutations in evolution: the translesion synthesis model of somatic hypermutation. Int. Immunol. 11:825-833[Abstract/Free Full Text].

DORNER, T., H. P. BREZINSCHEK, S. J. FOSTER, R. I. BREZINSCHEK, and N. L. FARNER et al., 1998  Comparable impact of mutational and selective influences in shaping the expressed repertoire of peripheral IgM+/CD5- and IgM+/CD5+ B cells. Eur. J. Immunol. 28:657-668[Medline].

DUNN-WALTERS, D. K., A. DOGAN, L. BOURSIER, C. M. MACDONALD, and J. SPENCER, 1998  Base-specific sequences that bias somatic hypermutation deduced by analysis of out-of-frame human IgV_H genes. J. Immunol. 160:2360-2364[Abstract/Free Full Text].

EHRENSTEIN, M. R. and M. S. NEUBERGER, 1999  Deficiency in Msh2 affects the efficiency and local sequence specificity of immunoglobulin class-switch recombination: parallels with somatic hypermutation. EMBO J. 18:3484-3490[Medline].

FOSTER, S. J., T. DORNER, and P. E. LIPSKY, 1999  Targeting and subsequent selection of somatic hypermutations in the human V kappa repertoire. Eur. J. Immunol. 29:3122-3132[Medline].

FREY, S., B. BERTOCCI, F. DELBOS, L. QUINT, and J. C. WEILL et al., 1998  Mismatch repair deficiency interferes with the accumulation of mutations in chronically stimulated B cells and not with the hypermutation process. Immunity 9:127-134[Medline].

GONZÁLEZ-FERNÁNDEZ, A., S. K. GUPTA, R. PANNELL, M. S. NEUBERGER, and C. MILSTEIN, 1994  Somatic mutation of immunoglobulin lambda chains: a segment of the major intron hypermutates as much as the complementarity-determining regions. Proc. Natl. Acad. Sci. USA 91:12614-12618[Abstract/Free Full Text].

GOOSSENS, T., U. KLEIN, and R. KÜPPERS, 1998  Frequent occurrence of deletions and duplications during somatic hypermutation: implications for oncogene translocations and heavy chain disease. Proc. Natl. Acad. Sci. USA 95:2463-2468[Abstract/Free Full Text].

HARRIS, R. S., Q. KONG, and N. MAIZELS, 1999  Somatic hypermutation and the three R's: repair, replication and recombination. Mutat. Res. 436:157-178[Medline].

HENGSTSCHLÄGER, M., M. WILLIAMS, and N. MAIZELS, 1994  A 1 transgene under the control of a heavy chain promoter and enhancer does not undergo somatic hypermutation. Eur. J. Immunol. 24:1649-1656[Medline].

JOLLY, C. J., S. D. WAGNER, C. RADA, N. KLIX, and C. MILSTEIN et al., 1996  The targeting of somatic hypermutation. Semin. Immunol. 8:159-168[Medline].

KELSOE, G., 1996  The germinal center: a crucible for lymphocyte selection. Semin. Immunol. 8:179-184[Medline].

KIM, N. and U. STORB, 1998  The role of DNA repair in somatic hypermutation of immunoglobulin genes. J. Exp. Med. 187:1729-1733[Free Full Text].

KLEIN, U., T. GOOSSENS, M. FISCHER, H. KANZLER, and A. BRAEUNINGER et al., 1998  Somatic hypermutation in normal and transformed human B cells. Immunol. Rev. 162:261-280[Medline].

KONG, Q. and N. MAIZELS, 1999  PMS2-deficiency diminishes hypermutation of a 1 transgene in young but not older mice. Mol. Immunol. 36:83-91[Medline].

KONG, Q. and N. MAIZELS, 2001  Breaksite batch mapping, a rapid method for assay and identification of DNA breaksites in mammalian cells. Nucleic Acids Res. 29:E33.

KONG, Q., R. S. HARRIS, and N. MAIZELS, 1998a  Recombination-based mechanisms for somatic hypermutation. Immunol. Rev. 162:67-76[Medline].

KONG, Q., L. ZHAO, S. SUBBAIAH, and N. MAIZELS, 1998b  A 3' enhancer drives active and untemplated somatic hypermutation of a 1 transgene. J. Immunol. 161:294-301[Abstract/Free Full Text].

KUPPERS, R., T. GOOSSENS, and U. KLEIN, 1999  The role of somatic hypermutation in the generation of deletions and duplications in human Ig V region genes and chromosomal translocations. Curr. Top. Microbiol. Immunol. 246:193-198[Medline].

MAGE, R. G., 1998  Diversification of rabbit VH genes by gene-conversion-like and hypermutation mechanisms. Immunol. Rev. 162:49-54[Medline].

MAIZELS, N., 1995  Somatic hypermutation: how many mechanisms diversify V region sequences? Cell 83:9-12[Medline].

MILSTEIN, C., M. S. NEUBERGER, and R. STADEN, 1998  Both DNA strands of antibody genes are hypermutation targets. Proc. Natl. Acad. Sci. USA 95:8791-8794[Abstract/Free Full Text].

NEUBERGER, M. S., M. R. EHRENSTEIN, N. KLIX, C. J. JOLLY, and J. YÉLAMOS et al., 1998  Monitoring and interpreting the intrinsic features of somatic hypermutation. Immunol. Rev. 162:107-116[Medline].

PAPAVISILIOU, F. N. and D. G. SCHATZ, 2000  Cell-cycle-regulated DNA double-strand breaks in somatic hypermutation of immunoglobulin genes. Nature 408:216-221[Medline].

PARNG, C. L., S. HANSAL, R. A. GOLDSBY, and B. A. OSBORNE, 1995  Diversification of bovine lambda-light chain genes. Ann. NY Acad. Sci. 764:155-157[Medline].

PARNG, C. L., S. HANSAL, R. A. GOLDSBY, and B. A. OSBORNE, 1996  Gene conversion contributes to Ig light chain diversity in cattle. J. Immunol. 157:5478-5486[Abstract].

PEAKMAN, M.-C. and N. MAIZELS, 1998  Localization of splenic B cells activated for switch recombination by in situ hybridization with I1 switch transcript and Rad51 probes. J. Immunol. 161:4008-4015[Abstract/Free Full Text].

PHUNG, Q. H., D. B. WINTER, A. CRANSTON, R. E. TARONE, and V. A. BOHR et al., 1998  Increased hypermutation at G and C nucleotides in immunoglobulin variable genes from mice deficient in the MSH2 mismatch repair protein. J. Exp. Med. 187:1745-1751[Abstract/Free Full Text].

PHUNG, Q. H., D. B. WINTER, R. ALREFAI, and P. J. GEARHART, 1999  Hypermutation in Ig V genes from mice deficient in the MLH1 mismatch repair protein. J. Immunol. 162:3121-3124[Abstract/Free Full Text].

PRZYLEPA, J., C. HIMES, and G. KELSOE, 1998  Lymphocyte development and selection in germinal centers. Curr. Top. Microbiol. Immunol. 229:85-104[Medline].

RADA, C., M. R. EHRENSTEIN, M. S. NEUBERGER, and C. MILSTEIN, 1998  Hot spot focusing of somatic hypermutation in MSH2-deficient mice suggests two stages of mutational targeting. Immunity 9:135-141[Medline].

REYNAUD, C. A., V. ANQUEZ, H. GRIMAL, and J. C. WEILL, 1987  A hyperconversion mechanism generates the chicken light chain preimmune repertoire. Cell 48:379-388[Medline].

REYNAUD, C. A., C. R. MACKAY, R. G. MULLER, and J. C. WEILL, 1991  Somatic generation of diversity in a mammalian primary lymphoid organ: the sheep ileal Peyer's patches. Cell 64:995-1005[Medline].

ROGOZIN, I. B. and N. A. KOLCHANOV, 1992  Somatic hypermutagenesis in immunoglobulin genes. II. Influence of neighbouring base sequences on mutagenesis. Biochim. Biophys. Acta 1171:11-18[Medline].

ROGOZIN, I. B., N. E. SREDNEVA, and N. A. KOLCHANOV, 1996  Somatic hypermutagenesis in immunoglobulin genes. III. Somatic mutations in the chicken light chain locus. Biochim. Biophys. Acta 1306:171-178[Medline].

SALE, J. E. and M. S. NEUBERGER, 1998  TdT-accessible breaks are scattered over the immunoglobulin V domain in a constitutively hypermutating B cell line. Immunity 9:859-869[Medline].

SPENCER, J., M. DUNN, and D. K. DUNN-WALTERS, 1999  Characteristics of sequences around individual nucleotide substitutions in IgVH genes suggest different GC and AT mutators. J. Immunol. 162:6596-6601[Abstract/Free Full Text].

STORB, U., A. PETERS, E. KLOTZ, N. KIM, and H. M. SHEN et al., 1998  Cis-acting sequences that affect somatic hypermutation of Ig genes. Immunol. Rev. 162:153-160[Medline].

THOMPSON, C. B. and P. E. NEIMAN, 1987  Somatic diversification of the chicken immunoglobulin light chain gene is limited to the rearranged variable gene segment. Cell 48:369-378[Medline].

VARADE, W. S., J. A. CARNAHAN, P. D. KINGSLEY, and R. A. INSEL, 1998  Inherent properties of somatic hypermutation as revealed by human non-productive VH6 immunoglobulin rearrangements. Immunology 93:171-176[Medline].

VORA, K. A., K. M. TUMAS-BRUNDAGE, V. M. LENTZ, A. CRANSTON, and R. FISHEL et al., 1999  Severe attenuation of the B cell immune response in Msh2-deficient mice. J. Exp. Med. 189:471-482[Abstract/Free Full Text].

WAGNER, S. D., C. MILSTEIN, and M. S. NEUBERGER, 1995  Codon bias targets mutation. Nature 376:732[Medline].

WEINSTEIN, P. D., A. O. ANDERSON, and R. G. MAGE, 1994  Rabbit IgH sequences in appendix germinal centers: VH diversification by gene conversion-like and hypermutation mechanisms. Immunity 1:647-659[Medline].

WIESENDANGER, M., B. KNEITZ, W. EDELMANN, and M. D. SCHARFF, 2000  Somatic hypermutation in MutS homologue (MSH)3-, MSH6-, and MSH3/MSH6-deficient mice reveals a role for the MSH2–MSH6 heterodimer in modulating the base substitution pattern. J. Exp. Med. 191:579-584[Abstract/Free Full Text].

WILSON, P., Y. J. LIU, J. BANCHEREAU, J. D. CAPRA, and V. PASCUAL, 1998a  Amino acid insertions and deletions contribute to diversify the human Ig repertoire. Immunol. Rev. 162:143-151[Medline].

WILSON, P. C., O. DE BOUTEILLER, Y. J. LIU, K. POTTER, and J. BANCHEREAU et al., 1998b  Somatic hypermutation introduces insertions and deletions into immunoglobulin V genes. J. Exp. Med. 187:59-70[Abstract/Free Full Text].

WINTER, D. B. and P. J. GEARHART, 1998  Dual enigma of somatic hypermutation of immunoglobulin variable genes: targeting and mechanism. Immunol. Rev. 162:89-96[Medline].

WINTER, D. B., Q. H. PHUNG, A. UMAR, S. M. BAKER, and R. E. TARONE et al., 1998  Altered spectra of hypermutation in antibodies from mice deficient for the DNA mismatch repair protein PMS2. Proc. Natl. Acad.Sci. USA 95:6953-6958[Abstract/Free Full Text].

YÉLAMOS, J., N. KLIX, B. GOYENECHEA, F. LOZANO, and Y. L. CHUI et al., 1995  Targeting of non-Ig sequences in place of the V segment by somatic hypermutation. Nature 376:225-229[Medline].

This article has been cited by other articles:

				M. Mierau, G. A. Drexler, A. Kutzera, K. Braunschmidt, J. Ellwart, F. Eckardt-Schupp, E. Fritz, J. Bachl, and B. Jungnickel Non-conservative homologous recombination in human B lymphocytes is promoted by activation-induced cytidine deaminase and transcription Nucleic Acids Res., October 1, 2008; 36(17): 5591 - 5601. [Abstract] [Full Text] [PDF]

				D. Ronai, M. D. Iglesias-Ussel, M. Fan, Z. Li, A. Martin, and M. D. Scharff Detection of chromatin-associated single-stranded DNA in regions targeted for somatic hypermutation J. Exp. Med., January 22, 2007; 204(1): 181 - 190. [Abstract] [Full Text] [PDF]

				E. S. Tang and A. Martin NHEJ-deficient DT40 cells have increased levels of immunoglobulin gene conversion: evidence for a double strand break intermediate Nucleic Acids Res., December 4, 2006; 34(21): 6345 - 6351. [Abstract] [Full Text] [PDF]

				K. Machida, K. T.-H. Cheng, N. Pavio, V. M.-H. Sung, and M. M. C. Lai Hepatitis C Virus E2-CD81 Interaction Induces Hypermutation of the Immunoglobulin Gene in B Cells J. Virol., July 1, 2005; 79(13): 8079 - 8089. [Abstract] [Full Text] [PDF]

				S. A. Martomo, D. Fu, W. W. Yang, N. S. Joshi, and P. J. Gearhart Deoxyuridine Is Generated Preferentially in the Nontranscribed Strand of DNA from Cells Expressing Activation-Induced Cytidine Deaminase J. Immunol., June 15, 2005; 174(12): 7787 - 7791. [Abstract] [Full Text] [PDF]

				H. Nagaoka, S. Ito, M. Muramatsu, M. Nakata, and T. Honjo DNA cleavage in immunoglobulin somatic hypermutation depends on de novo protein synthesis but not on uracil DNA glycosylase PNAS, February 8, 2005; 102(6): 2022 - 2027. [Abstract] [Full Text] [PDF]

				J. F. Ruiz, D. Lucas, E. Garcia-Palomero, A. I. Saez, M. A. Gonzalez, M. A. Piris, A. Bernad, and L. Blanco Overexpression of human DNA polymerase {micro} (Pol {micro}) in a Burkitt's lymphoma cell line affects the somatic hypermutation rate Nucleic Acids Res., November 1, 2004; 32(19): 5861 - 5873. [Abstract] [Full Text] [PDF]

				S. S. Brar, M. Watson, and M. Diaz Activation-induced Cytosine Deaminase (AID) Is Actively Exported out of the Nucleus but Retained by the Induction of DNA Breaks J. Biol. Chem., June 18, 2004; 279(25): 26395 - 26401. [Abstract] [Full Text] [PDF]

				J. S. Rush, S. D. Fugmann, and D. G. Schatz Staggered AID-dependent DNA double strand breaks are the predominant DNA lesions targeted to S{micro} in Ig class switch recombination Int. Immunol., April 1, 2004; 16(4): 549 - 557. [Abstract] [Full Text] [PDF]

				K. Machida, K. T.-N. Cheng, V. M.-H. Sung, S. Shimodaira, K. L. Lindsay, A. M. Levine, M.-Y. Lai, and M. M. C. Lai Hepatitis C virus induces a mutator phenotype: Enhanced mutations of immunoglobulin and protooncogenes PNAS, March 23, 2004; 101(12): 4262 - 4267. [Abstract] [Full Text] [PDF]

				N. Catalan, F. Selz, K. Imai, P. Revy, A. Fischer, and A. Durandy The Block in Immunoglobulin Class Switch Recombination Caused by Activation-Induced Cytidine Deaminase Deficiency Occurs Prior to the Generation of DNA Double Strand Breaks in Switch {micro} Region J. Immunol., September 1, 2003; 171(5): 2504 - 2509. [Abstract] [Full Text] [PDF]

				D. B. Winter, Q. H. Phung, X. Zeng, E. Seeberg, D. E. Barnes, T. Lindahl, and P. J. Gearhart Normal Somatic Hypermutation of Ig Genes in the Absence of 8-Hydroxyguanine-DNA Glycosylase J. Immunol., June 1, 2003; 170(11): 5558 - 5562. [Abstract] [Full Text] [PDF]

				R. Bransteitter, P. Pham, M. D. Scharff, and M. F. Goodman Activation-induced cytidine deaminase deaminates deoxycytidine on single-stranded DNA but requires the action of RNase PNAS, April 1, 2003; 100(7): 4102 - 4107. [Abstract] [Full Text] [PDF]

				A. J. Rattray, B. K. Shafer, C. B. McGill, and J. N. Strathern The Roles of REV3 and RAD57 in Double-Strand-Break-Repair-Induced Mutagenesis of Saccharomyces cerevisiae Genetics, November 1, 2002; 162(3): 1063 - 1077. [Abstract] [Full Text] [PDF]

				N. D'Avirro, D. Truong, M. Luong, R. Kanaar, and E. Selsing Gene Conversion-Like Sequence Transfers Between Transgenic Antibody V Genes Are Independent of RAD54 J. Immunol., September 15, 2002; 169(6): 3069 - 3075. [Abstract] [Full Text] [PDF]

				Y. I. Pavlov, I. B. Rogozin, A. P. Galkin, A. Y. Aksenova, F. Hanaoka, C. Rada, and T. A. Kunkel Correlation of somatic hypermutation specificity and A-T base pair substitution errors by DNA polymerase eta during copying of a mouse immunoglobulin kappa light chain transgene PNAS, July 23, 2002; 99(15): 9954 - 9959. [Abstract] [Full Text] [PDF]

				H. Nagaoka, M. Muramatsu, N. Yamamura, K. Kinoshita, and T. Honjo Activation-induced Deaminase (AID)-directed Hypermutation in the Immunoglobulin S{micro} Region: Implication of AID Involvement in a Common Step of Class Switch Recombination and Somatic Hypermutation J. Exp. Med., February 19, 2002; 195(4): 529 - 534. [Abstract] [Full Text] [PDF]

				H.-F. Tsai, N. D'Avirro, and E. Selsing Gene conversion-like sequence transfers in a mouse antibody transgene: antigen selection allows sensitive detection of V region interactions based on homology Int. Immunol., January 1, 2002; 14(1): 55 - 64. [Abstract] [Full Text] [PDF]

				X. Chen, K. Kinoshita, and T. Honjo Variable deletion and duplication at recombination junction ends: Implication for staggered double-strand cleavage in class-switch recombination PNAS, November 20, 2001; 98(24): 13860 - 13865. [Abstract] [Full Text] [PDF]

				V. Poltoratsky, C. J. Woo, B. Tippin, A. Martin, M. F. Goodman, and M. D. Scharff Expression of error-prone polymerases in BL2 cells activated for Ig somatic hypermutation PNAS, June 20, 2001; (2001) 141222198. [Abstract] [Full Text] [PDF]

				V. Poltoratsky, C. J. Woo, B. Tippin, A. Martin, M. F. Goodman, and M. D. Scharff Expression of error-prone polymerases in BL2 cells activated for Ig somatic hypermutation PNAS, July 3, 2001; 98(14): 7976 - 7981. [Abstract] [Full Text] [PDF]

				L. Bross, M. Muramatsu, K. Kinoshita, T. Honjo, and H. Jacobs DNA Double-Strand Breaks: Prior to but not Sufficient in Targeting Hypermutation J. Exp. Med., May 6, 2002; 195(9): 1187 - 1192. [Abstract] [Full Text] [PDF]

				F. N. Papavasiliou and D. G. Schatz The Activation-induced Deaminase Functions in a Postcleavage Step of the Somatic Hypermutation Process J. Exp. Med., May 6, 2002; 195(9): 1193 - 1198. [Abstract] [Full Text] [PDF]

				K. F. Chua, F. W. Alt, and J. P. Manis The Function of AID in Somatic Mutation and Class Switch Recombination: Upstream or Downstream of DNA Breaks J. Exp. Med., May 6, 2002; 195(9): F37 - F41. [Full Text] [PDF]