Based on its substrate specificity, activation-induced cytidine deaminase can directly induce C:G mutations in Ig genes. However the origin of A:T mutations, which occur in a similar proportion in germinal center (GC) B cells, is unclear. Genetic evidence suggests that the induction of A:T mutations requires the components of the mismatch repair system and DNA polymerase η (POLH). We found that fibroblasts and GC B cells expressed similar levels of the mismatch repair components, but nonetheless the fibroblasts failed to generate a significant proportion of A:T mutations in a GFP reporter gene even after POLH overexpression. To investigate whether the ability to generate A:T mutations is dependent on the cellular environment (i.e., GC B cell or fibroblast) or the target gene (i.e., Ig or GFP), we developed a mutation detection system in a human GC-like cell line. We introduced a GFP gene with a premature stop codon into Ramos cells and compared the activation-induced cytidine deaminase-induced mutations in the endogenous VH and the transgenic GFP genes. Remarkably, a high proportion of A:T mutations was induced in both genes. Ectopic expression of POLH did not further increase the proportion of A:T mutations but diminished the strand bias of these mutations that is normally observed in VH genes. Intriguingly, the total mutation frequency in the GFP gene was consistently one-fifth of that in the VH gene. These results demonstrate that the ability to generate A:T mutations is dependent on the GC B cell environment but independent of the mutation frequency and target gene location.

The Ig genes in germinal center (GC)2 B cells undergo a high frequency of point mutations at both C:G and A:T base pairs. This somatic hypermutation (SHM) process is essential for the diversification of Ig genes, resulting in the production of high affinity Abs (1). SHM is initiated by the activation-induced cytidine deaminase (AID) (2), which catalyzes the deamination of cytosine to uracil and generates a U:G DNA lesion (3). Mutations are thought to be introduced during the replication and error-prone processing of the AID-triggered U:G lesions (4, 5, 6). According to the model proposed by Di Noia and Neuberger (1), direct replication of the U:G mismatch could result in C to T and G to A transitional mutations because the structure of U is similar to that of thymine, which normally pairs with adenine. In addition, uracil can be excised by the uracil DNA glycosylase, which initiates base excision repair, and the replication of the resulting abasic site would generate both transitions and transversions at C:G pairs.

Although C:G mutations can be generated by replication of the AID-triggered U:G lesion or its repair intermediate, the abasic site, it remains largely unknown how the same U:G lesion leads to mutations at nondamaged A:T pairs in GC B cells. Genetic and biochemical evidence suggests that components of the mismatch repair (MMR) system, including MutS homolog (MSH) 2, MSH6, exonuclease (EXO) 1, and proliferating cell nuclear Ag (PCNA), as well as DNA polymerase η (POLH), are required for the induction of A:T mutations (7, 8, 9, 10, 11, 12, 13, 14, 15, 16). It has been proposed that the AID-triggered U:G lesion is recognized by the MSH2/6 heterodimer, followed by EXO-1-mediated strand degradation that generates a gap in the damaged DNA strand. Nonspecific mutations are then thought to be introduced during the gap-filling reaction catalyzed by POLH, a low fidelity DNA polymerase that frequently incorporates the incorrect nucleotide opposite a template A or T.

In contrast to GC B cells, in which AID-triggered U:G lesions in Ig genes lead to mutations at C:G and A:T pairs at roughly the same frequency, in fibroblasts the mutations in a GFP reporter gene induced by ectopic AID expression are predominantly C:G mutations (17). To explore why GC B cells but not fibroblasts are able to generate a high proportion of A:T mutations, we have compared the expression levels of the MMR components and POLH between GC B cells and fibroblasts and analyzed whether ectopic POLH expression might lead to increased A:T mutations in fibroblasts. In addition, we have established an efficient mutagenesis system in the human GC-like cell line Ramos and compared the mutation frequency and patterns in the endogenous, rearranged Ig VH gene and a GFP reporter gene. Our results suggest that the MMR components and POLH are not sufficient for the induction of A:T mutations in fibroblasts and that additional GC-specific mechanisms are required. Furthermore, we demonstrate that the ability to generate A:T mutations is dependent on the GC B cell environment but independent of mutation frequency and target gene location.

GC B cells were sorted from immunized mice as described (18). RNA was extracted using TRIzol reagent (Invitrogen) and first-strand cDNA was synthesized with SuperScript III reverse transcriptase and random primers. The following primers were used in RT-PCR analyses: Msh2/s520, 5′-GTTGGAGTTGGGTATGTGGA-3′; Msh2/as1756, 5′-TTAACAATGGCATCCTGGGC-3′; Msh6/s350, 5′-GGTTTGGGCTAAGATGGAAG-3′; Msh6/as831, 5′-TCAGAGCCACCAATGTCACT-3′; Exo-1/s256, 5′-GCTATTGCTTGTGCTGAAAA-3′; Exo-1/as757, 5′-TCCGTGAATACATCCCCAAG-3′; Pcna/s454, 5′-GAAGCACCAAATCAAGAGAA-3′; Pcna/as844, 5′-TTATACTCTACAACAAGGGG-3′; Polh/s1137, 5′-GAATGGGTCTTGGCTGTATG-3′; Polh/as1514, 5′-GCAGAGGAAGAGCATTGTGA-3′; Aicda/s119, 5′-TTCAAAAATGTCCGCTGGGC-3′; Aicda/as548, 5′-CCTTCCCAGGCTTTGAAAGT-3′ (the Aicda primers annealed with both mouse Aicda and human AICDA genes); β-actin/s80, 5′-ATGGATGACGATATCGCT-3′; β-actin/as630, 5′-ATGAGGTAGTCTGTCAGGT-3′; Gapdh/s316, 5′-GTGCTGAGTATGTCGTGGAG-3′; Gapdh/as746, 5′-ACACATTGGGGGTAGGAACA-3′; POLH/s1170, 5′-ACCCAGGCAACTACCCAAAAC-3′; POLH/as1807, 5′-GGGCTCAGTTCCTGTACTTTG-3′; GAPDH/s628, 5′-ACCACAGTCCATGCCATCAC-3′; and GAPDH/as1060, 5′-TCCACCACCCTGTTGCTGTA-3′. RT-PCR was performed using Taq polymerase (Toyobo) at 95°C for 2 min followed by 25 or 30 cycles of amplification under the following conditions: Msh2, 95°C for 5 s, 60°C for 10 s, and 72°C for 2 min; Msh6, Aicda and Gapdh, 95°C for 5 s, 60°C for 10 s, and 72 °C for 1 min; Exo-1 and Pcna, 95°C for 5 s, 55°C for 10 s, and 72°C for 1 min; Polh, 95°C for 5 s, 58°C for 10 s, and 72°C for 1 min; β-actin, 95°C for 5 s, 54°C for 10 s, and 72°C for 1 min; AICDA, same conditions as Aicda; POLH, 95°C for 5 s, 61°C for 10 s, and 72 °C for 1 min; GAPDH, 95°C for 5 s, 60°C for 10 s, and 72°C for 1 min.

Rabbit polyclonal Abs against POLH were provided by Prof. F. Hanaoka, Gakushuin University (Tokyo, Japan). Goat polyclonal Abs against AID (C-20) were purchased from Santa Cruz Biotechnology. Rabbit polyclonal Abs against actin were obtained from Sigma-Aldrich. Immunoblot analysis was performed as described (19).

The AID-mediated mutagenesis system in NIH3T3 has been described previously (17). We used NIH3T3-pI clone 19, which contained one copy of the GFP reporter gene with a premature stop codon under the control of the tetracycline (Tet)-responsive promoter. Upon transduction with a mouse AID-expressing retrovirus, the premature stop codon can be reverted to allow translation of GFP. Mutations can be analyzed in sorted GFP+ revertants. To investigate the role of POLH in the induction of A:T mutations, we constructed a pMX-AID-IRES-POLH retroviral vector (where IRES is internal ribosome entry site), which allowed simultaneous expression of mouse AID and mouse POLH. NIH3T3-pI clone 19 was transduced with retrovirus expressing AID or AID-IRES-POLH and the GFP+ revertants were sorted 10 days later. The GFP gene was amplified using KOD-plus polymerase (Toyobo) under conditions described previously (17). Mutations were detected by DNA sequencing. The mutations that had reverted the premature stop codon (TAG) were not included in the analysis because they represented selected mutations that were, by definition, enriched in the GFP+ revertants.

To mimic the NIH3T3 cell mutagenesis system, we established a Ramos subline containing a Tet-controlled GFP gene with the same premature stop codon as that used in the NIH3T3-pI clone 19 cells by using a Tet-off gene expression system (Takara Bio). Briefly, Ramos cells were first transfected with a pTet-Off vector by electroporation and G418-resistant clones were isolated. Sixty-eight stable clones were analyzed by transient transfection of pTRE2hyg-Luc along with an internal control pTK-Luc vector followed by luciferase assays, and one clone (no. 52) that exhibited low background and high Tet-dependent induction was used in these studies. This clone was further transfected by electroporation or Nucleofector (Amaxa Biosystems) with a pTRE-Tight vector expressing a GFP gene with a premature stop codon along with a linearized hygromycin selection vector. After selection, we eventually established one stable line from each transfection (Ramos/TAG and Ramos/TAG/Amaxa) in which GFP gene transcription could be rapidly induced by removing Tet (supplemental figure 1).3 The Ramos/TAG line had a lower background GFP expression than did Ramos/TAG/Amaxa and was primarily used in the present study. Quantitative genomic PCR revealed that the Ramos/TAG cells harbored two integrated copies of the GFP gene (data not shown).

Mutation analysis in Ramos/TAG cells was performed in the same way as in NIH3T3-pI clone 19 cells. Briefly, cells were transduced with retroviruses expressing AID or AID-IRES-POLH in the absence of Tet and the GFP+ revertants were sorted 10 days later. The GFP gene was amplified using CMVPF (5′-TAGCCTCCATAGAAGACACCG-3′) forward and HygGFP (5′-GCGAAAAAGAAAGAACAATC-3′) reverse primers at 94 °C for 2 min followed by 30 cycles of amplification at 94°C for 30 s, 57°C for 30 s, and 68°C for 1 min using KOD-plus polymerase. Mutations that reverted the premature stop codon (TAG) were not included in the analysis. The endogenous Ig VH gene was amplified using RVHFOR forward and JOL48 reverse primers as described (20) at 95°C for 2 min followed by 30 cycles of amplification at 95°C for 15 s, 58°C for 15 s, and 68°C for 1 min using KOD-plus polymerase. In addition, we also transduced Ramos/TAG cells with retrovirus expressing AID-IRES-GFP and sorted the GFP+ (i.e., virus-transduced) cells 2 days later. These GFP+ cells, which expressed GFP protein derived from the retrovirus, were cultured for an additional 8 days (a total of 10 days of culture) and analyzed for mutations in the endogenous VH and the GFP reporter genes. Note that the CMVPF and HygGFP primers were designed to amplify only the GFP reporter gene but not the GFP gene in the retrovirus, because the CMVPF primer could only anneal with the CMV promoter located upstream of the GFP reporter gene.

To understand why AID-triggered U:G lesions result in a high proportion of A:T mutations in GC B cells but not fibroblasts, we first compared the expression of the known mediators of A:T mutations in Ig genes, including Msh2, Msh6, Exo-1, Pcna, and Polh. Semiquantitative RT-PCR analysis revealed that the Msh2, Msh6, and Exo-1 transcripts were similar between the GC B and NIH3T3-pI clone 19 cells (Fig. 1). The Pcna transcripts were increased ∼3-fold in GC B as compared with NIH3T3 cells, consistent with the fact that GC B cells undergo rapid cell division and PCNA is up-regulated in the S phase of the cell cycle. The Polh transcript level was increased ∼3-fold in GC B as compared with NIH3T3 cells (Fig. 1), in agreement with a previous observation that Polh expression was elevated in GC B cells relative to resting B cells (21). These results suggest that the inability of NIH3T3 to generate A:T mutations is unlikely due to the lack of ubiquitously expressed MMR components but could be due to limited POLH expression.

FIGURE 1.

Expression of the mismatch repair components (Msh2, Msh6, Exo-1, Pcna) and Polh in GC B and NIH3T3-pI clone 19 cells. RT-PCR for the indicated genes was performed as described in Materials and Methods. The β-actin gene was used as a cDNA loading control. The amplification was conducted for 30 cycles except for β-actin, which was amplified for 25 cycles.

FIGURE 1.

Expression of the mismatch repair components (Msh2, Msh6, Exo-1, Pcna) and Polh in GC B and NIH3T3-pI clone 19 cells. RT-PCR for the indicated genes was performed as described in Materials and Methods. The β-actin gene was used as a cDNA loading control. The amplification was conducted for 30 cycles except for β-actin, which was amplified for 25 cycles.

Close modal

To determine whether A:T mutations could be increased in fibroblasts by elevating Polh expression, we used an efficient AID-mediated mutagenesis system in NIH3T3 cells (17). Transduction of the AID retrovirus resulted in AID protein expression as detected by immunoblot analysis (Fig. 2,A), and 0.47% of the cells reverted the stop codon to express GFP protein (Fig. 2,C) on day 10 after virus transduction. The efficiency of virus transduction was ∼80% as monitored by using a retrovirus expressing AID-IRES-GFP (supplemental figure 2A). Sequence analyses of the GFP genes in the sorted GFP+ revertants revealed a predominance of C:G mutations (Fig. 2,D), a result consistent with previous observations (17). Transduction of retroviruses expressing POLH (Fig. 2,C) or a mutant AID lacking the cytidine deamination motif (supplemental figure 2B) did not induce GFP+ cells, indicating that POLH alone cannot induce mutations and that mutation requires catalytically active AID. To examine the role of POLH in the generation of A:T mutations, we transduced fibroblasts with a retrovirus expressing AID-IRES-POLH, which allowed simultaneous expression of AID and POLH from the same mRNA. Immunoblot (Fig. 2,A) and semiquantitative RT-PCR (Fig. 2,B) analyses confirmed the expression of both AID and POLH, and 0.28% of the cells reverted to express GFP (Fig. 2,C). Sequence analyses of the sorted GFP+ cells revealed that, again, mutations occurred predominantly at C:G with no significant increase in A:T mutations (Fig. 2,E). Notably, the transcript levels of Aicda and Polh in NIH3T3 transduced with AID-IRES-POLH were ∼10-fold greater than in the GC B cells (Fig. 2,B), indicating that both genes were expressed at sufficiently high levels. The total mutation frequency was 0.466 and 0.365% in NIH3T3 transduced with AID and AID-IRES-POLH, respectively (Table I). These results demonstrate that elevated expression of POLH is not sufficient to induce A:T mutations in a GFP reporter gene in fibroblasts.

FIGURE 2.

Elevated POLH expression is not sufficient to induce A:T mutations in fibroblasts. A, Expression of AID and POLH proteins in NIH3T3-pI clone 19 cells transduced with retrovirus expressing AID, POLH, or AID-IRES-POLH. Expression of endogenous actin was used as a loading control. Immunoblot analysis was performed on day 2 after virus transduction. B, Comparison of the transcript levels of Aicda and Polh in GC B cells, mock-transduced NIH3T3, and NIH3T3 transduced with retrovirus expressing AID-IRES-POLH. Total RNA was extracted on day 2 after virus transduction, and the first-strand cDNA was subjected to RT-PCR analysis. The Gapdh gene was used as a cDNA loading control. The amplification was performed for 25 cycles (for Aicda and Gapdh) or 30 cycles (Polh). C, Induction of GFP+ revertants in NIH3T3-pI clone 19 cells by transduction of retrovirus expressing AID or AID-IRES-POLH but not POLH alone. FACS analysis was performed on day 10 after virus transduction. D and E, Mutation patterns of the GFP reporter gene after transduction of AID (D) or AID-IRES-POLH (E). The GFP+ cells were sorted on day 10 after virus transduction. Data are corrected for base composition (A = 23.21%, T = 14.76%, C = 33.56%, G = 28.44%). See Table I for mutation frequencies.

FIGURE 2.

Elevated POLH expression is not sufficient to induce A:T mutations in fibroblasts. A, Expression of AID and POLH proteins in NIH3T3-pI clone 19 cells transduced with retrovirus expressing AID, POLH, or AID-IRES-POLH. Expression of endogenous actin was used as a loading control. Immunoblot analysis was performed on day 2 after virus transduction. B, Comparison of the transcript levels of Aicda and Polh in GC B cells, mock-transduced NIH3T3, and NIH3T3 transduced with retrovirus expressing AID-IRES-POLH. Total RNA was extracted on day 2 after virus transduction, and the first-strand cDNA was subjected to RT-PCR analysis. The Gapdh gene was used as a cDNA loading control. The amplification was performed for 25 cycles (for Aicda and Gapdh) or 30 cycles (Polh). C, Induction of GFP+ revertants in NIH3T3-pI clone 19 cells by transduction of retrovirus expressing AID or AID-IRES-POLH but not POLH alone. FACS analysis was performed on day 10 after virus transduction. D and E, Mutation patterns of the GFP reporter gene after transduction of AID (D) or AID-IRES-POLH (E). The GFP+ cells were sorted on day 10 after virus transduction. Data are corrected for base composition (A = 23.21%, T = 14.76%, C = 33.56%, G = 28.44%). See Table I for mutation frequencies.

Close modal
Table I.

Mutation frequency in the GFP reporter gene in the GFP+ revertants of NIH3T3 cells transduced with retrovirus expressing AID or AID-IRES-POLH

AIDAID-IRES-POLH
No. of sequences 57 64 
Mutated sequences (%) 49 (86.0%) 52 (81.3%) 
Sequences with additional mutations (%)a 48 (84.2%) 47 (73.4%) 
Sequences with unique mutationsb 48 41 
Total length of sequences with unique mutations 40,320 34,440 
No. of unique mutationsc 223 171 
Mutation frequency (%)d 0.553 0.497 
Total mutation frequencye 0.466 0.365 
AIDAID-IRES-POLH
No. of sequences 57 64 
Mutated sequences (%) 49 (86.0%) 52 (81.3%) 
Sequences with additional mutations (%)a 48 (84.2%) 47 (73.4%) 
Sequences with unique mutationsb 48 41 
Total length of sequences with unique mutations 40,320 34,440 
No. of unique mutationsc 223 171 
Mutation frequency (%)d 0.553 0.497 
Total mutation frequencye 0.466 0.365 
a

Of the mutated sequences, sequences that had acquired additional mutations other than those reverting the premature TAG stop codon.

b

Of the sequences with additional mutations, identical sequences were excluded.

c

Mutations that reverted the premature stop codon were not included in the analysis since they represented selected mutations that were enriched in the GFP+ revertants.

d

Number of unique mutations divided by the total length of sequences with unique mutations.

e

Mutation frequency × percentage of sequences with additional mutations.

One possible explanation for the inability of fibroblasts to generate A:T mutations in a GFP reporter gene is that the VH but not the GFP gene may have a particular chromatin structure that allows for the induction of A:T mutations. To determine whether the generation of A:T mutations is dependent on the target gene (i.e., Ig or VH) or on the GC B cell environment, we established a similar AID-mediated mutagenesis system in the GC-like human B cell line, Ramos. We introduced a Tet-controlled GFP gene with the same premature stop codon as that used in the NIH3T3 system into Ramos (Ramos/TAG). Because Ramos has been shown to undergo a low level of spontaneous mutation in the Ig genes, we first analyzed 24 sequences from Ramos/TAG cells that were mock transduced and then cultured for 10 days. We found that 8 of 24 sequences contained a C to T transition at position 66 when compared with the published Ramos VH sequence (20). However, sequence analysis of Ramos/TAG before virus transduction revealed the same C to T mutation in 11 of 24 sequences, indicating that this C to T mutation already existed in the Ramos/TAG line, probably having arisen during the establishment of the stable line. These results indicate that the Ramos/TAG line established in the present study had a relatively low spontaneous mutation frequency. This mutagenesis system allowed us to compare AID-induced mutations in the GFP genes between Ramos and NIH3T3 cells. In addition, this assay system allowed us to directly compare the mutation frequency and patterns in the endogenous Ig VH gene and the GFP reporter gene in Ramos.

Ramos/TAG cells were mock transduced or transduced with retroviruses expressing AID or AID-IRES-POLH. Semiquantitative RT-PCR analyses using a primer set that annealed with both human AICDA and mouse Aicda genes revealed a low level expression of the endogenous AICDA (Fig. 3,A). Transduction of the AID and AID-IRES-POLH retroviruses resulted in ∼9- and 4-fold increases in the Aicda transcript levels, respectively (Fig. 3,A). Taking into consideration that the transduction efficiency was ∼17% (supplemental figure 2C), the Aicda transcript levels among the virus transduced cells were actually much higher, ∼50- and ∼20-fold that of the endogenous AICDA levels, respectively. Repeated attempts to detect AID protein expression in Ramos cells transduced with retroviruses expressing AID or AID-IRES-POLH have been unsuccessful. Using Gapdh and GAPDH as controls and taking into consideration the efficiency of virus transduction, the transcript levels of retrovirus-derived Aicda in NIH3T3 were estimated to be ∼2-fold greater than those in Ramos (Figs. 2,B and 3,A). It is notable that Ramos transduced with AID consistently expressed higher levels of Aicda transcripts than those transduced with AID-IRES-POLH (Fig. 3,A and data not shown). It appeared that the presence of the 2.1-kb mouse Polh cDNA in the retroviral vector decreased the efficiency of retrovirus-mediated gene expression. Because Ramos already expressed high levels of the endogenous POLH, the level of the virus-derived mouse Polh transcripts was only ∼50% that of POLH after correcting for the virus transduction efficiency (Fig. 3 A).

FIGURE 3.

A high proportion of A:T mutations is found in both the VH and the GFP genes in the GFP+ revertants of Ramos/TAG transduced with retroviruses expressing AID or AID-IRES-POLH. A, Transcript levels of Aicda, Polh, and endogenous POLH in mock-transduced Ramos, Ramos transduced with AID, or AID-IRES-POLH. The GAPDH gene was used as a cDNA loading control. The PCR primers for Aicda also annealed with the human AICDA gene. Total RNA was extracted on day 2 after virus transduction, and the first-strand cDNA was subjected to RT-PCR analysis. The amplification was performed for 25 cycles. B, Induction of GFP+ revertants in Ramos cells after transduction of retrovirus expressing AID or AID-IRES-POLH. FACS analysis was performed on day 10 after virus transduction. C, Mutation patterns in Ramos cells transduced with a retrovirus expressing AID. Upper panel, the VH gene; lower panel, the GFP reporter gene. D, Mutation patterns in Ramos transduced with AID-IRES-POLH. Upper panel, the VH gene; lower panel, the GFP gene. GFP+ revertants were sorted on day 10 after virus transduction and subjected to sequence analysis as described in Materials and Methods. Data are corrected for base composition (For the VH gene, A = 22.28%, T = 20.82%, C = 25.20%, G = 31.69%; for the GFP gene, A = 23.04%, T = 16.08%, C = 32.98%, G = 27.90). Mutations at A:T in both the VH and the GFP genes were significantly increased in Ramos in comparison to fibroblasts (p < 0.001, χ2 test). See Table III for mutation frequencies.

FIGURE 3.

A high proportion of A:T mutations is found in both the VH and the GFP genes in the GFP+ revertants of Ramos/TAG transduced with retroviruses expressing AID or AID-IRES-POLH. A, Transcript levels of Aicda, Polh, and endogenous POLH in mock-transduced Ramos, Ramos transduced with AID, or AID-IRES-POLH. The GAPDH gene was used as a cDNA loading control. The PCR primers for Aicda also annealed with the human AICDA gene. Total RNA was extracted on day 2 after virus transduction, and the first-strand cDNA was subjected to RT-PCR analysis. The amplification was performed for 25 cycles. B, Induction of GFP+ revertants in Ramos cells after transduction of retrovirus expressing AID or AID-IRES-POLH. FACS analysis was performed on day 10 after virus transduction. C, Mutation patterns in Ramos cells transduced with a retrovirus expressing AID. Upper panel, the VH gene; lower panel, the GFP reporter gene. D, Mutation patterns in Ramos transduced with AID-IRES-POLH. Upper panel, the VH gene; lower panel, the GFP gene. GFP+ revertants were sorted on day 10 after virus transduction and subjected to sequence analysis as described in Materials and Methods. Data are corrected for base composition (For the VH gene, A = 22.28%, T = 20.82%, C = 25.20%, G = 31.69%; for the GFP gene, A = 23.04%, T = 16.08%, C = 32.98%, G = 27.90). Mutations at A:T in both the VH and the GFP genes were significantly increased in Ramos in comparison to fibroblasts (p < 0.001, χ2 test). See Table III for mutation frequencies.

Close modal

Transduction of Ramos with the AID retrovirus induced 0.24% GFP+ revertants on day 10 (Fig. 3,B, middle panel). These GFP+ cells were sorted and analyzed for mutations in the endogenous VH and the GFP genes. A high proportion of A:T mutations was observed both in the VH (36.9%, Fig. 3,C, upper panel) and the GFP (28.4%, Fig. 3,C, lower panel) genes. Although the proportion of A:T mutations in the GFP gene was lower than that observed in the Ig genes of GC B cells in vivo (∼50%) (18), it was significantly higher than that seen in NIH3T3 cells (p < 0.001; χ2 test). As was the case for Ig gene SHM, the RGYW/WRCY motifs were preferentially targeted both in the endogenous VH gene (supplemental figure 3) and the GFP gene (supplemental figure 4). Moreover, the base substitution patterns in the GFP and VH genes were quite similar (Fig. 3 C), suggesting that both genes were mutated by a similar mechanism.

Transduction of the AID-IRES-POLH retrovirus induced 0.17% GFP+ cells (Fig. 3,B, right panel). Similarly, a high proportion of A:T mutations was observed in both the VH and the GFP genes in the GFP+ revertants. The proportion of A:T mutations in the VH gene was 36.7% (Fig. 3,D, upper panel), which was similar to that observed in Ramos transduced with AID alone (Fig. 3,C, upper panel). The proportion of A:T mutations in the GFP reporter gene was 38.2% (Fig. 3,D, lower panel), which appeared higher than that observed in Ramos transduced with AID alone (Fig. 3 C, lower panel), but the difference did not reach statistical significance (p > 0.1; χ2 test). These results demonstrate that GC-like Ramos B cells are able to generate A:T mutations in the GFP reporter gene as efficiently as in the VH gene and that ectopic POLH expression did not further increase the proportion of A:T mutations. Collectively, these findings suggest that the ability to generate A:T mutations is an intrinsic property of the GC-like Ramos B cells and is independent on the target gene.

A ∼2-fold increase of mutations at A relative to T has been observed in Ig genes when recorded from the upper (nontranscribed) strand (22), suggesting that there is a strand bias in generating A:T mutations. Analysis of AID-mediated mutations confirmed the ∼2-fold increase of mutations at A vs T in both the VH and the GFP genes (Fig. 3,C). Intriguingly, whereas the overall proportion of A:T mutations in the VH gene in Ramos was not further increased by ectopic POLH expression, the strand bias at A:T disappeared (Fig. 3,D, upper panel). A:T mutations are known to occur frequently at the WA hotspot on the upper strand (W is A or T and the mutated position is underlined). The same WA motif on the lower strand, corresponding to the TW motif on the upper strand, is less frequently mutated (23). To further confirm the effect of ectopic POLH expression on eliminating the strand bias of A:T mutations, we compared the number of mutations that occurred at WA vs TW motifs. There were 20 WA and nine TW motifs mutated in the VH gene in AID-transduced Ramos (Table II and supplemental figure 3), confirming that the WA motif on the upper strand was more frequently mutated than that on the lower strand. In contrast, there were 16 WA and 20 TW motifs mutated in the VH gene in Ramos transduced with AID-IRES-POLH. The WA/TW ratio was 2.2 and 0.8 in Ramos transduced with AID and AID-IRES-POLH, respectively, a statistically significantly difference (p < 0.05; χ2 test). Therefore, the WA motifs on both the upper and lower strands were similarly mutated in Ramos expressing ectopic POLH. Curiously, ectopic expression of POLH did not affect the strand bias of A:T mutations in the GFP gene (Fig. 3 D, lower panel, and supplemental figure 4).

Table II.

Mutations at WA and TW motifs in the VH gene of Ramos

MotifsNumber of Mutationsa
AIDAID-IRES-POLH
WA 20 16 
T20 
Ratio (WA/TW) 2.2 0.8b 
MotifsNumber of Mutationsa
AIDAID-IRES-POLH
WA 20 16 
T20 
Ratio (WA/TW) 2.2 0.8b 
a

The number of mutations occurring at WA or TW motifs. The mutated positions are underlined.

b

Significant difference compared to Ramos transduced with AID only (p < 0.05; χ2 test).

Analysis of the VH gene in the GFP+ revertants of AID-transduced Ramos cells revealed that 86 of 159 (54.1%) sequences were mutated and on average there were 2.3 mutations per each 341-bp VH sequence (Table III). The mutation frequency reached 0.673%, which is comparable to that observed in the JH4 intronic region of Ig genes in GC B cells on day 14 after immunization with a T-dependent Ag (18). The total mutation frequency, taking into consideration the percentage of the mutated clones (54.1%), was 0.364%. Analysis of the GFP gene in the GFP+ revertants of AID-transduced Ramos revealed that 207 of 224 (92.4%) sequences were mutated. Of the mutated sequences, 87 (38.8%) had acquired additional mutations other than those reverting the premature stop codon (Table III). The mutation frequency in the GFP gene was 0.192%, which was lower than that in the VH gene. When taking into consideration the percentage of sequences with additional mutations (38.8%), the total mutation frequency in the GFP gene was 0.075%, approximately one-fifth of that observed in the VH gene.

Table III.

Mutation frequency in the VH and GFP reporter genes in the GFP+ revertants of Ramos B cells transduced with retroviruses expressing AID or AID-IRES-POLH

AIDAID-IRES-POLH
VHGFPVHGFP
No. of sequences 159 224 301 314 
Mutated sequences (%) 86 (54.1%) 207 (92.4%) 110 (36.5%) 275 (87.0%) 
Sequences with additional mutations (%)a NAb 87 (38.8%) NAb 77 (24.4%) 
Sequences with unique mutationsc 75 83 73 61 
Total length of sequences with unique mutations 25,575 71,712 24,893 52,704 
No. of unique mutationsd 172 138 143 90 
Mutation frequency (%)e 0.673 0.192 0.574 0.171 
Total mutation frequencyf 0.364 0.075 0.210 0.042 
AIDAID-IRES-POLH
VHGFPVHGFP
No. of sequences 159 224 301 314 
Mutated sequences (%) 86 (54.1%) 207 (92.4%) 110 (36.5%) 275 (87.0%) 
Sequences with additional mutations (%)a NAb 87 (38.8%) NAb 77 (24.4%) 
Sequences with unique mutationsc 75 83 73 61 
Total length of sequences with unique mutations 25,575 71,712 24,893 52,704 
No. of unique mutationsd 172 138 143 90 
Mutation frequency (%)e 0.673 0.192 0.574 0.171 
Total mutation frequencyf 0.364 0.075 0.210 0.042 
a

Of the mutated GFP sequences, sequences that had acquired additional mutations other than those reverting the premature TAG stop codon.

b

Not applicable.

c

Identical sequences were excluded.

d

Mutations that reverted the premature stop codon were not included in the analysis since they represented selected mutations that were enriched in the GFP+ revertants.

e

No. of unique mutations divided by the total length of sequences with unique mutations.

f

Mutation frequency × percentage of mutated sequences (for VH) or percentage of sequences with additional mutations (for GFP).

Ramos transduced with AID-IRES-POLH expressed lower levels of Aicda transcripts as compared with Ramos transduced with AID (Fig. 3,A). Consistently, a lower percentage of the VH sequences (110 of 301; 36.5%) were mutated, and the total mutation frequency was reduced to 0.21% (Table III). Therefore, the mutation frequency appeared to depend on the AID expression level, consistent with earlier observations (24). Intriguingly, mutation in the GFP reporter gene also decreased proportionally. As a result, the total mutation frequency in the GFP gene (0.042%) was again one-fifth of that in the VH gene (0.21%). These findings demonstrate that the Ig VH gene is more efficiently mutated than the GFP reporter gene and that the relative mutation frequency in the VH and GFP genes remains constant regardless of the AID expression levels.

The above results obtained with the GFP+ revertants (Fig. 3 and Table III) revealed a high proportion of A:T mutations both in the endogenous VH and in the GFP genes. These GFP+ revertants, however, represented only ∼0.2% of the total cells (Fig. 3,B) or ∼1% of the virus-transduced cells, taking into consideration the efficiency of virus transduction (∼17%, supplemental figure 2C). To ensure that these GFP+ revertants did not represent a biased population of the virus-transduced Ramos cells, we next transduced Ramos/TAG with retrovirus expressing AID-IRES-GFP (Fig. 4,A) and simply sorted the GFP+, i.e., virus-transduced cells 2 days later (Fig. 4,B). These GFP+ cells (which expressed wild-type GFP protein derived from the retrovirus) were cultured for an additional 8 days (a total of 10 days of culture) and analyzed for mutations in the endogenous VH and the GFP reporter genes. The results confirmed a high proportion of A:T mutations with a similar bias to the RGYW/WRCY motifs both in the VH and the GFP genes (Fig. 4,C and supplemental figures 5 and 6). Once again, the total mutation frequency in the GFP gene was one-fifth that of the VH gene (Table IV).

FIGURE 4.

Induction of a high proportion of A:T mutations both in the VH and the GFP genes in AID-transduced Ramos/TAG cells. A and B, Ramos cells were transduced with retrovirus expressing AID-IRES-GFP and analyzed 2 days later for Aicda transcript levels by RT-PCR (A) and GFP protein expression by FACS (B). The PCR was conducted for 25 cycles. The GFP+ (i.e., virus transduced) cells were sorted and cultured for an additional 8 days (a total of 10-days of culture) before mutation analysis. C, Mutation patterns in the VH (upper panel) and the GFP reporter (lower panel) genes. See Table IV for mutation frequencies.

FIGURE 4.

Induction of a high proportion of A:T mutations both in the VH and the GFP genes in AID-transduced Ramos/TAG cells. A and B, Ramos cells were transduced with retrovirus expressing AID-IRES-GFP and analyzed 2 days later for Aicda transcript levels by RT-PCR (A) and GFP protein expression by FACS (B). The PCR was conducted for 25 cycles. The GFP+ (i.e., virus transduced) cells were sorted and cultured for an additional 8 days (a total of 10-days of culture) before mutation analysis. C, Mutation patterns in the VH (upper panel) and the GFP reporter (lower panel) genes. See Table IV for mutation frequencies.

Close modal
Table IV.

Mutation frequency in the VH and GFP reporter genes in Ramos cells transduced with retrovirus expressing AID-IRES-GFP

VHGFP
No. of sequences 87 220a 
Mutated sequences (%) 68 (78.2%) 89 (40.5%) 
Mutated sequences with unique mutationsb 64 82 
Total length of sequences with unique mutations 21,824 70,848 
Number of unique mutations 150 198 
Mutation frequency (%)c 0.687 0.279 
Total mutation frequencyd 0.537 0.113 
VHGFP
No. of sequences 87 220a 
Mutated sequences (%) 68 (78.2%) 89 (40.5%) 
Mutated sequences with unique mutationsb 64 82 
Total length of sequences with unique mutations 21,824 70,848 
Number of unique mutations 150 198 
Mutation frequency (%)c 0.687 0.279 
Total mutation frequencyd 0.537 0.113 
a

Of 220 sequences, only five had mutations that reverted the premature stop codon and all the mutations were included in the analysis because they were unselected mutations.

b

Of the mutated sequences, identical sequences were excluded.

c

Number of unique mutations divided by the total length of sequences with unique mutations.

d

Mutation frequency × percentage of mutated sequences.

A major unsolved issue in SHM is how AID-triggered U:G lesions lead to a high proportion of mutations at undamaged A:T pairs in Ig genes of GC B cells. It has been proposed that the AID-triggered U:G lesion is initially recognized by the MMR components and that mutations are ultimately introduced during a gap-filling reaction catalyzed by POLH (1). We found in the present study that fibroblasts and GC B cells expressed equivalent levels of the MMR components, but the fibroblasts failed to generate A:T mutations in a GFP reporter gene even after ectopic POLH expression. In contrast, GC-like Ramos cells were able to generate a high proportion of A:T mutations in a similar Tet-controlled GFP reporter gene. Notably, mutations were induced by a similar retrovirus-driven AID expression strategy in both the fibroblast and the Ramos cell systems. Although our results indicate that Ramos but not NIH3T3 cells are able to generate A:T mutations, one might argue that the differences could be due to the different integration sites of the GFP gene in these cells. However, Yoshikawa et al. had analyzed three independent clones of NIH3T3 and all of them showed low levels of A:T mutations (17). In the case of Ramos, we had obtained another stable line, Ramos/TAG/Amaxa, by an independent transfection (supplemental figure 1). Although not extensively analyzed, this line also showed a high proportion of A:T mutations (supplemental figure 7). Therefore, the difference in the ability to generate A:T mutations between Ramos and NIH3T3 is unlikely due to the integration sites, although additional clones and cell types need to be analyzed to completely rule out this possibility. Because NIH3T3 is a mouse cell line whereas Ramos is derived from a human Burkitt’s lymphoma, it is also conceivable that species differences might affect the proportion of A:T mutations.

We consider it a more likely possibility that the ability to generate A:T mutations is determined by the GC B environment rather than by the target gene or the integration site. This conclusion is in agreement with our recent findings obtained in a lacZ-transgenic system where a high proportion of A:T mutations was observed in the lacZ genes in GC B cells but not in other B cells or nonlymphoid cells (25). Our results also conform with a very recent study showing that tonsil and Ramos cell lysates generated significantly more A:T mutations than did HeLa cell lysates in a cell-free MMR system (26). What, then, makes GC B cells different from other cell types? GC B cells may express additional factors that are required for efficient induction of A:T mutations. For instance, GC B cells might express a factor or factors that facilitate the use of POLH instead of other polymerases to catalyze the gap-filling reaction during the processing of AID-triggered U:G lesions. Alternatively, the components of MMR, POLH, or AID itself might undergo specific modifications in GC B cells and allow for the induction of A:T mutations. In fact, it has been shown that B cells may use a mechanism to limit the activity of AID (27, 28). Additionally, GC B cells are known to undergo a very rapid rate of cell division, and the presence of many AID-induced U:G lesions may disturb the function of normal repair pathways and cause aberrant usage of polymerases during DNA replication and repair, leading to mutations at nondamaged A:T base pairs. Indeed, Ramos B cells have a shorter cell cycle time compared with NIH3T3 cells, which might be the reason for the increased A:T mutations in Ramos. Additional experimental approaches are required to explore these nonexclusive potential mechanisms.

Both Ramos and NIH3T3 exhibited a fairly high frequency of transversion mutations at C:G regardless of exogenous POLH expression. These transversion mutations are thought to be introduced by error-prone polymerases that are able to replicate over the abasic site generated during base excision repair of the AID-triggered U:G lesion. Therefore, other low fidelity polymerases, but not POLH, are likely involved in the generation of transversion mutations at C:G in both Ramos and NIH3T3 cells. This interpretation is consistent with our recent finding that POLH is not involved in C:G mutations (29). In addition, the results suggest that even in nonlymphoid cells excessive DNA lesions induced by AID overexpression may exhaust the normal repair capacity of the cell, leading to aberrant replication of the repair intermediates. Although transversion mutations at C:G were induced in both Ramos and NIH3T3, a high proportion of A:T mutations was observed only in Ramos. These observations suggest that dysregulated DNA repair due to excess DNA lesions is not sufficient for the induction of A:T mutations and that additional GC B-specific mechanisms are required.

It is intriguing that ectopic expression of POLH in Ramos did not further increase the proportion of A:T mutations but diminished the strand bias of these mutations normally observed in the VH gene. POLH expression has been shown to be elevated during the S phase of the cell cycle, coinciding with the kinetics of thymidine uptake (30). The expression of the retrovirus-derived POLH is under the control of the virus long-terminal repeat and is thus regulated differently than the endogenous POLH. The dysregulated expression of the exogenous POLH might lead to its aberrant usage and the induction of A:T mutations on the normally less mutated strand. We have recently generated a catalytically inactive POLH, and it will be interesting to investigate how the mutant enzyme might affect the proportion and strand bias of A:T mutations.

The total mutation frequency in the GFP reporter gene was consistently one-fifth that of the endogenous VH gene. This was true in Ramos cells transduced with either AID alone or AID-IRES-POLH and then analyzed for mutations in the sorted GFP+ revertants, and in Ramos cells transduced with AID-IRES-GFP and then analyzed for mutations in the total virus-transduced cells. Previous studies have suggested that beyond a certain threshold of expression, there is a correlation between a decrease in gene expression and an increase in SHM (31). We found that the transcript levels of the VH gene were ∼3-fold greater than that of the GFP gene (supplemental figure 8). Therefore, at least in Ramos, we did not observe any inverse correlation between mRNA levels and SHM frequency. The results of the present study suggest that the relative efficiency of AID targeting to the VH and GFP genes remains constant in the Ramos/TAG cells. These results do not support the idea that AID is first targeted to Ig gene locus and that other genes are targeted only when there is excess AID. Rather, the relative targeting efficiency of AID to different genes may be determined by their epigenetic status, for example, chromatin structure, rate of transcription, and possibly other yet-to-be identified features. A variety of genes undergo AID-dependent somatic mutation at various frequencies in GC B cells (32). Our results suggest that these genes are not randomly targeted by AID. Rather, each of them may be targeted at a given frequency that correlates with its status. In fact, Bcl-6 has been shown in several independent studies to be the most frequently mutated gene, with a mutation frequency as high as one-tenth of that in Ig genes (32, 33, 34). The mutation frequency in the GFP reporter gene observed in our studies of Ramos/TAG is one-fifth of that in the endogenous VH gene, and therefore as high as the most frequently mutated GC gene in vivo. It is unclear at this point what determines the susceptibility of a gene to AID targeting. The high mutation frequencies observed in the Tet-controlled GFP reporter genes both in NIH3T3 and Ramos suggest that Tet-induced transcription may predispose a gene to AID targeting.

In our AID-mediated in vitro mutagenesis system in Ramos, the mutation frequency in the VH gene reached >0.5% in just 10 days, comparable to that observed in GC B cells after immunization with T-dependent Ags. In addition, mutations were biased to the RGYW/WRCY motif with a relatively high proportion of A:T mutations. Ramos cells have been shown to undergo constitutive Ig gene SHM at a low frequency (20, 24, 35, 36). Analysis of subclones of Ramos revealed that the mutation frequency correlated with the level of the endogenous AID (24). Ectopic expression of AID in Ramos by using plasmid-based expression vectors only slightly increased the mutation frequency (37, 38). This is likely due to a low level of AID expression with these vectors and/or inactivation of the exogenous AID during the establishment of the stable transfectants, similar to what has been observed in the AID-transgenic mice that constitutively express AID in B cells (27, 28). In these earlier studies, Ramos cells were cultured for a prolonged period (1–2 mo) and usually only a limited number of mutations were analyzed, with most of them occurring at C:G. It is possible that a sufficiently high mutation frequency might be required for the induction of a high proportion of A:T mutations. The results of the present study demonstrate that the retrovirus-mediated AID expression in Ramos is an efficient mutagenesis system and recapitulates two fundamental features of the hypermutation process of Ig genes: a high mutation frequency in a short time period and a high proportion of A:T mutations.

A:T mutations are reduced by >80% in POLH-deficient mice (13, 14, 15, 16, 18). We recently found that these mice produced decreased serum titers of high-affinity Abs against a T-dependent Ag and that this was associated with a reduced frequency and altered patterns of amino acid substitutions in the complementary determining region of the Ig VH genes in the responding B cells (29). Therefore, mutations at A:T are important for efficient diversification of Ig genes and affinity maturation of Abs. Intriguingly, previous studies suggested that mutations at A:T were increased in tissues of aged mice (39). More recently, analyses of the mutation patterns in 210 diverse human cancers revealed that A:T mutations were increased in certain malignancies (40). These observations suggest that the induction of A:T mutations in non-GC B cells may reflect dysregulated DNA repair associated with the processes of aging and tumorigenesis.

We thank Professor Tasuku Honjo and Dr. Reiko Shinkura for providing the in vitro mutagenesis system in NIH3T3 fibroblasts, the FACS Laboratory for cell sorting, the Immunogenomics Group for sequencing, and Kanae Fukui for secretarial assistance.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

2

Abbreviations used in this paper: GC, germinal center; AID, activation-induced cytidine deaminase; EXO, exonuclease; IRES, internal ribosome entry site; MMR, mismatch repair; MSH, MutS homolog; PCNA, proliferating cell nuclear Ag; POLH, DNA polymerase η; SHM, somatic hypermutation; Tet, tetracycline.

3

The online version of this article contains supplemental material.

1
Di Noia, J. M., M. S. Neuberger.
2007
. Molecular mechanisms of antibody somatic hypermutation.
Annu. Rev. Biochem.
76
:
1
-22.
2
Muramatsu, M., K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, T. Honjo.
2000
. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme.
Cell
102
:
553
-563.
3
Chaudhuri, J., M. Tian, C. Khuong, K. Chua, E. Pinaud, F. W. Alt.
2003
. Transcription-targeted DNA deamination by the AID antibody diversification enzyme.
Nature
422
:
726
-730.
4
Martomo, S. A., P. J. Gearhart.
2006
. Somatic hypermutation: subverted DNA repair.
Curr. Opin. Immunol.
18
:
243
-248.
5
Bransteitter, R., J. L. Sneeden, S. Allen, P. Pham, M. F. Goodman.
2006
. First AID (activation-induced cytidine deaminase) is needed to produce high affinity isotype-switched antibodies.
J. Biol. Chem.
281
:
16833
-16836.
6
Longerich, S., U. Basu, F. Alt, U. Storb.
2006
. AID in somatic hypermutation and class switch recombination.
Curr. Opin. Immunol.
18
:
164
-174.
7
Rada, C., M. R. Ehrenstein, M. S. Neuberger, C. Milstein.
1998
. Hot spot focusing of somatic hypermutation in MSH2-deficient mice suggests two stages of mutational targeting.
Immunity
9
:
135
-141.
8
Frey, S., B. Bertocci, F. Delbos, L. Quint, J. C. Weill, C. A. Reynaud.
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.
9
Martomo, S. A., W. W. Yang, P. J. Gearhart.
2004
. A role for Msh6 but not Msh3 in somatic hypermutation and class switch recombination.
J. Exp. Med.
200
:
61
-68.
10
Bardwell, P. D., C. J. Woo, K. Wei, Z. Li, A. Martin, S. Z. Sack, T. Parris, W. Edelmann, M. D. Scharff.
2004
. Altered somatic hypermutation and reduced class-switch recombination in exonuclease 1-mutant mice.
Nat. Immunol.
5
:
224
-229.
11
Langerak, P., A. O. Nygren, P. H. Krijger, P. C. van den Berk, H. Jacobs.
2007
. A/T mutagenesis in hypermutated immunoglobulin genes strongly depends on PCNAK164 modification.
J. Exp. Med.
204
:
1989
-1998.
12
Rada, C., J. M. Di Noia, M. S. Neuberger.
2004
. Mismatch recognition and uracil excision provide complementary paths to both Ig switching and the A/T-focused phase of somatic mutation.
Mol. Cell
16
:
163
-171.
13
Faili, A., S. Aoufouchi, S. Weller, F. Vuillier, A. Stary, A. Sarasin, C. A. Reynaud, J. C. Weill.
2004
. DNA polymerase η is involved in hypermutation occurring during immunoglobulin class switch recombination.
J. Exp. Med.
199
:
265
-270.
14
Martomo, S. A., W. W. Yang, R. P. Wersto, T. Ohkumo, Y. Kondo, M. Yokoi, C. Masutani, F. Hanaoka, P. J. Gearhart.
2005
. Different mutation signatures in DNA polymerase η- and MSH6-deficient mice suggest separate roles in antibody diversification.
Proc. Natl. Acad. Sci. USA
102
:
8656
-8661.
15
Delbos, F., A. De Smet, A. Faili, S. Aoufouchi, J. C. Weill, C. A. Reynaud.
2005
. Contribution of DNA polymerase η to immunoglobulin gene hypermutation in the mouse.
J. Exp. Med.
201
:
1191
-1196.
16
Delbos, F., S. Aoufouchi, A. Faili, J. C. Weill, C. A. Reynaud.
2007
. DNA polymerase η is the sole contributor of A/T modifications during immunoglobulin gene hypermutation in the mouse.
J. Exp. Med.
204
:
17
-23.
17
Yoshikawa, K., I. M. Okazaki, T. Eto, K. Kinoshita, M. Muramatsu, H. Nagaoka, T. Honjo.
2002
. AID enzyme-induced hypermutation in an actively transcribed gene in fibroblasts.
Science
296
:
2033
-2036.
18
Masuda, K., R. Ouchida, M. Hikida, T. Kurosaki, M. Yokoi, C. Masutani, M. Seki, R. D. Wood, F. Hanaoka, J. O. -Wang.
2007
. DNA polymerases η and θ function in the same genetic pathway to generate mutations at A/T during somatic hypermutation of Ig genes.
J. Biol. Chem.
282
:
17387
-17394.
19
Ouchida, R., S. Yamasaki, M. Hikida, K. Masuda, K. Kawamura, A. Wada, S. Mochizuki, M. Tagawa, A. Sakamoto, M. Hatano, et al
2008
. A lysosomal protein negatively regulates surface T cell antigen receptor expression by promoting CD3ζ-chain degradation.
Immunity
29
:
33
-43.
20
Sale, J. E., 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.
21
Zeng, X., D. B. Winter, C. Kasmer, K. H. Kraemer, A. R. Lehmann, P. J. Gearhart.
2001
. DNA polymerase η is an A-T mutator in somatic hypermutation of immunoglobulin variable genes.
Nat. Immunol.
2
:
537
-541.
22
Smith, D. S., G. Creadon, P. K. Jena, J. P. Portanova, B. L. Kotzin, L. J. Wysocki.
1996
. Di- and trinucleotide target preferences of somatic mutagenesis in normal and autoreactive B cells.
J. Immunol.
156
:
2642
-2652.
23
Rogozin, I. B., Y. I. Pavlov, K. Bebenek, T. Matsuda, T. A. Kunkel.
2001
. Somatic mutation hotspots correlate with DNA polymerase η error spectrum.
Nat. Immunol.
2
:
530
-536.
24
Zhang, W., P. D. Bardwell, C. J. Woo, V. Poltoratsky, M. D. Scharff, A. Martin.
2001
. Clonal instability of V region hypermutation in the Ramos Burkitt’s lymphoma cell line.
Int. Immunol.
13
:
1175
-1184.
25
Ouchida, R., A. Ukai, H. Mori, K. Kawamura, M. E. Dolle, M. Tagawa, A. Sakamoto, T. Tokuhisa, T. Yokosuka, T. Saito, et al
2008
. Genetic analysis reveals an intrinsic property of the germinal center B cells to generate A:T mutations.
DNA Repair
7
:
1392
-1398.
26
Pham, P., K. Zhang, and M. F. Goodman. 2008. Hypermutation at A/T sites during G·U mismatch repair in vitro by human B-cell lysates. J. Biol. Chem. In press.
27
Muto, T., I. M. Okazaki, S. Yamada, Y. Tanaka, K. Kinoshita, M. Muramatsu, H. Nagaoka, T. Honjo.
2006
. Negative regulation of activation-induced cytidine deaminase in B cells.
Proc. Natl. Acad. Sci. USA
103
:
2752
-2757.
28
Shen, H. M., G. Bozek, C. A. Pinkert, K. McBride, L. Wang, A. Kenter, U. Storb.
2008
. Expression of AID transgene is regulated in activated B cells but not in resting B cells and kidney.
Mol. Immunol.
45
:
1883
-1892.
29
Masuda, K., R. Ouchida, Y. Masayuki, F. Hanaoka, T. Azuma, and J.-Y. Wang. 2008. DNA polymerase η is a limiting factor for A:T mutations in Ig genes and contributes to the antibody affinity maturation. Eur. J. Immunol. In press.
30
Yamada, A., C. Masutani, S. Iwai, F. Hanaoka.
2000
. Complementation of defective translesion synthesis and UV light sensitivity in xeroderma pigmentosum variant cells by human and mouse DNA polymerase η.
Nucleic Acids Res.
28
:
2473
-2480.
31
Michael, N., H. M. Shen, S. Longerich, N. Kim, A. Longacre, U. Storb.
2003
. The E box motif CAGGTG enhances somatic hypermutation without enhancing transcription.
Immunity
19
:
235
-242.
32
Liu, M., J. L. Duke, D. J. Richter, C. G. Vinuesa, C. C. Goodnow, S. H. Kleinstein, D. G. Schatz.
2008
. Two levels of protection for the B cell genome during somatic hypermutation.
Nature
451
:
841
-845.
33
Shen, H. M., A. Peters, B. Baron, X. Zhu, U. Storb.
1998
. Mutation of BCL-6 gene in normal B cells by the process of somatic hypermutation of Ig genes.
Science
280
:
1750
-1752.
34
Pasqualucci, L., A. Migliazza, N. Fracchiolla, C. William, A. Neri, L. Baldini, R. S. Chaganti, U. Klein, R. Kuppers, K. Rajewsky, R. Dalla-Favera.
1998
. BCL-6 mutations in normal germinal center B cells: evidence of somatic hypermutation acting outside Ig loci.
Proc. Natl. Acad. Sci. USA
95
:
11816
-11821.
35
Harris, R. S., D. S. Croom-Carter, A. B. Rickinson, M. S. Neuberger.
2001
. Epstein-Barr virus and the somatic hypermutation of immunoglobulin genes in Burkitt’s lymphoma cells.
J. Virol.
75
:
10488
-10492.
36
Xiao, Z., M. Ray, C. Jiang, A. B. Clark, I. B. Rogozin, M. Diaz.
2007
. Known components of the immunoglobulin A:T mutational machinery are intact in Burkitt lymphoma cell lines with G:C bias.
Mol. Immunol.
44
:
2659
-2666.
37
Martin, A., M. D. Scharff.
2002
. Somatic hypermutation of the AID transgene in B and non-B cells.
Proc. Natl. Acad. Sci. USA
99
:
12304
-12308.
38
Parsa, J. Y., W. Basit, C. L. Wang, J. L. Gommerman, J. R. Carlyle, A. Martin.
2007
. AID mutates a non-immunoglobulin transgene independent of chromosomal position.
Mol. Immunol.
44
:
567
-575.
39
Dolle, M. E., W. K. Snyder, D. B. Dunson, J. Vijg.
2002
. Mutational fingerprints of aging.
Nucleic Acids Res.
30
:
545
-549.
40
Greenman, C., P. Stephens, R. Smith, G. L. Dalgliesh, C. Hunter, G. Bignell, H. Davies, J. Teague, A. Butler, C. Stevens, et al
2007
. Patterns of somatic mutation in human cancer genomes.
Nature
446
:
153
-158.