Aid

Immunity, Vol. 20, 659–668, June, 2004, Copyright 2004 by Cell Press

AID: How Does It Aid Antibody Diversity?

Tasuku Honjo,1,* Masamichi Muramatsu,1 and Sidonia Fagarasan2 1 Department of Medical Chemistry and Molecular Biology Graduate School of Medicine Kyoto University Yoshida Sakyo-Ku, Kyoto 606-8501 2 RIKEN Research Center for Allergy and Immunology Tsurumi-ku, Yokohama Kanagawa 230-0045 Japan

Activation-induced cytidine deaminase (AID) is an essential enzyme to regulate class switch recombination (CSR), somatic hypermutation (SHM), and gene conversion (GC). AID is known to be required for DNA cleavage of S regions in CSR. However, its molecular mechanism is a focus of extensive debate. RNA editing hypothesis postulates that AID edits yet unknown mRNA to generate specific endonucleases for CSR and SHM. By contrast, DNA deamination hypothesis assumes that AID deaminates cytosine in DNA, followed by DNA cleavage by base excision repair enzymes. We discuss available evidence for the two proposed models. Recent findings, namely requirement of protein synthesis for DNA breakage and dispensability of U removal activity of uracil DNA glycosylase, force us to reconsider DNA deamination hypothesis.

Introduction Generation of immunoglobulin (Ig) diversity depends on four types of genetic alteration mechanism in Ig loci, namely VDJ recombination, somatic hypermutation (SHM), gene conversion (GC), and class switch recombination (CSR). The basic repertoire of Ig variable region (V) diversity is generated by VDJ recombination, which is tightly coupled with the differentiation steps of B lymphocytes (Sekiguchi et al., 2004). In bone marrow, B lymphocyte progenitors differentiate by stepwise assembly of subexon DNA segments of the heavy chain (H) V genes, followed by those of the light chain V genes. VDJ recombination is a site-specific recombination regulated by two recombinases, RAG-1 and RAG-2, and precisely programmed in parallel with B lymphocyte differentiation (Oettinger et al., 1990; Schatz et al., 1989). VDJ recombination is responsible for generation of a huge repertoire of immunoglobulin V, including those that are not particularly useful for the body defense. All B lymphocytes that successfully completed VDJ recombination express IgM on surface and migrate to the secondary lymphoid organs, where they are activated by encountering antigens. Antigen stimulation induces the other three types of genetic alterations, namely SHM, GC, and CSR, in activated B cells (Honjo et al., 2002; Stavnezer et al., 2004). *Correspondence: [email protected]

Review

SHM introduces point mutations in V genes without template, whereas GC does so with pseudo V genes as template (Arakawa and Buerstedde, 2004). When SHM is coupled with selection by limited amounts of antigen, affinity maturation takes place by enriching high-affinity Ig-producing cells. CSR is a region-specific recombination that switches isotypes from IgM to the other isotypes, such as IgG and IgE, adding diverse effector functions to Ig with a given antigen specificity. CSR takes place in S regions of the CH locus and proceeds by cleaving, juxtaposing, and joining S␮ and one of the other S regions that are located 5⬘ to each CH gene. CSR results in looping-out deletion of the intervening DNA segment from the chromosome and bringing the rearranged V gene from upstream of the C␮ gene to the proximity of the different CH gene. Antigen-induced genetic alterations are thus essential to generate antigen-specific antibodies. Enzymes responsible for the three genetic alteration reactions had not been revealed for about a decade after the discovery of RAG-1 and RAG-2, until activation-induced cytidine deaminase (AID) was found to be essential to both CSR and SHM (Muramatsu et al., 2000; Revy et al., 2000). Subsequently, AID was shown to be responsible for GC as well (Arakawa et al., 2002). In this review, we focus on the molecular mechanism of how AID regulates three genetic alterations required for antigen-specific Ig synthesis. Molecular and Physiological Properties of AID Isolation, Expression, and Structure of AID AID cDNA was isolated from a murine B lymphoma line CH12F3-2 by subtractive cDNA hybridization using mRNAs from class switch-stimulated and nonstimulated CH12F3-2 cells (Muramatsu et al., 1999). AID mRNA is found only in activated B cells, most prominently in the germinal center B cells. AID mRNA encodes a protein of 198 amino acid residues containing a unique cytidine deaminase motif at residues 55–94, which includes three essential residues (H56, C87, and C90) for zinc binding and catalytic activity. These residues are highly conserved by all members of the cytidine deaminase family, including metabolic cytidine deaminase in E. coli. The authologue of AID is found in vertebrates but not in nonvertebrates. The amino acid sequence of AID has the strongest homology with that of apolipoprotein (apo) B mRNA editing catalytic polypeptide 1 (APOBEC-1), a well-characterized RNA editing enzyme of apoB100 mRNA that encodes the cholesterol carrier protein in low-density lipoprotein (LDL) (Teng et al., 1993). APOBEC-1 recognizes the structure of apoB100 mRNA through a cofactor called APOBEC-1 complementation factor (ACF) (Mehta et al., 2000), which guides the APOBEC-1 catalytic center to the specific cytosine (C) at position 6666. APOBEC-1 converts this C to uracil (U) and generates apoB48 mRNA that encodes the chylomicron protein component, a carrier of triglyceride. Both APOBEC-1 and AID form dimer (Lau et al., 1994; Ta et al., 2003).

Immunity 660

AID Is Essential and Sufficient for CSR and SHM The physiological function of AID is clearly demonstrated by studies on AID-deficient animals and patients (Muramatsu et al., 2000; Revy et al., 2000). AID-deficient mice show the complete loss of class switching and accumulation of IgM in sera and feces. Patients with an autosomal recessive hereditary disease called hyperIgM syndrome type 2 (HIGM2) have severe defects in class switching. Genetic linkage analysis of the disease locus using polymorphic markers has revealed that the mutation is mapped on chromosome 12p13, which coincided with the human AID gene locus determined by the FISH analysis (Muto et al., 2000). Subsequent molecular studies demonstrated that all HIGM2 patients have mutations in the AID gene, and all these mutated AID cDNAs are recently shown to be defective in CSR by in vitro assays (see below) (Ta et al., 2003). The B lymphocytes from HIGM2 patients and AIDdeficient mice are unable to switch isotypes by in vitro stimulation. Surprisingly, AID-deficient memory B cells of HIGM2 patients do not have SHM, and repeated antigen stimulation of AID-deficient mice do not show accumulation of mutations in the antigen-specific V region (Muramatsu et al., 2000; Revy et al., 2000). Ectopic expression of AID induces class switching and hypermutation in non-B cells, such as fibroblasts (Okazaki et al., 2002; Yoshikawa et al., 2002), hybridomas (Martin et al., 2002; Martin and Scharff, 2002b), and T cells (Okazaki et al., 2002), which carry artificial constructs for measuring CSR and SHM. DT40 chicken B cells that mutated the AID gene cannot undergo GC, which can be recovered by transfection of the AID cDNA (Arakawa et al., 2002). These results clearly demonstrate that AID is essential and sufficient to all three different genetic alterations induced by antigen stimulation of B cells. The other enzymes and cofactors are probably expressed ubiquitously. Phenotypes in AID-Deficient Mice As compared with RAG-2⫺/⫺ mice, AID⫺/⫺ mice are relatively healthy under the SPF condition and resistant to infection with a virulent strain of influenza virus (Harada et al., 2003). However, AID-deficient mice are more susceptible to secondary infection with higher doses, indicating that nonmutated IgM has significant protection capacity against low-dose virus but tailored Igs with SHM and CSR are important for protection from infection with higher-dose virus. HIGM2 patients suffer from recurrent infections, which cause hyperthrophy of lymph nodes and enlarged germinal centers (Revy et al., 2000). AID-deficient mice also have enlarged germinal centers (Muramatsu et al., 2000) and accumulation of activated IgM⫹ B cells and IgM plasma cells in all lymphoid tissues, but especially in the gut lamina propria (Fagarasan et al., 2002). Accumulation of IgM⫹ B cells and plasma cells in the intestine of AID⫺/⫺ mice is explained by (1) blockade of in situ class switching in the gut lymphoid tissues of AID⫺/⫺ mice, where local IgM⫹ B cells ordinarily switch preferentially to IgA and differentiate to IgA plasma cells (Fagarasan et al., 2001), and (2) sustained activation of the immune system due to the absence of intestinal IgA, causing continuous recruitment of immune cells to gut, which leads to hyperthrophy of Peyer’s patches and

protrusion of isolated lymphoid follicles (Fagarasan et al., 2002). The absence of normal hypermutated intestinal IgA causes a profound disregulation of the gut microflora, especially an excessive proliferation of anaerobic bacteria. The anaerobes detected in all segments of AID⫺/⫺ small intestine are nonpathogenic, commensal strains usually found in flora of the large intestine (Fagarasan et al., 2002). Among them, the major population is represented by segmented filamentous bacteria, strict anaerobes that cannot be cultured, at least with available microbiogical techniques, and strongly attach to the mucosal epithelium (Suzuki et al., 2004). An antibiotic treatment inhibiting anaerobe expansion or reconstitution of IgA production in AID⫺/⫺ small intestine recovers the normal composition of gut flora and abolishes both local and systemic activation of the immune system (Fagarasan et al., 2002; Suzuki et al., 2004). Thus, IgA secreted into the gut lumen appears to function not only for protection against pathogenic bacterial or viral antigens but also for the homeostasis of the nonpathogenic gut flora, which is essential to prevent overstimulation of the nonmucosal immune system (Fagarasan and Honjo, 2003). Unmutated IgMs, although secreted into the gut lumen, cannot prevent the excessive and aberrant expansion of anaerobes. Tumorgenesis by Constitutive Expression of AID Transgenic mice with AID cDNA under the control of the chicken ␤-actin promoter develop T cell tumors and die by 85 weeks without an exception (Okazaki et al., 2003). The onset of tumors varies from 4 to 40 weeks, depending on the copy numbers of the transgene. By surface phenotype, T cell tumors appear to originate from either thymic or peripheral T cells. In rare cases, tumors develop in lung, liver, and muscle tissues (our unpublished data). To our surprise, no B lymphomas have been detected in the AID-transgenic mice. In both types of T lymphomas, a large number of mutations accumulate in the V gene (10⫺3 ) but very infrequently in the C gene (10⫺4 ) of the T cell receptor, with a distribution profile of mutations reminiscent of SHM accumulation in Ig V genes. The mutation accumulation is also identified in the c-myc gene. Mutation target genes are selective because there are many transcriptionally active genes, which do not have mutations. No common chromosomal translocation other than sporadic one is found in these tumors (Okazaki et al., 2003). Therefore, mutation accumulation leading to tumor development is not solely related with deficiency in DNA repair genes but also to an uncontrolled AID activity, thus identifying AID as the first active mutator in vertebrates. Molecular Mechanism for Regulation of CSR and SHM by AID AID Involvement in DNA Cleavage The molecular mechanism how AID regulates CSR, SHM, and GC is a puzzling and fascinating question to scientists in broad disciplines (Durandy and Honjo, 2001; Maizels, 2000; Martin and Scharff, 2002a; Reynaud et al., 2003). Although these DNA alterations are distinct in the overall mechanism and properties of the products, the initiation step of the three reactions is similar;

Review 661

Figure 1. Duplication and Deletion in an Inversion-Type CSR Substrate (A) Staggered nicking on both strands of DNA; (B) 3⬘ or 5⬘ overhang is processed to make it blunt end by exonuclease or DNA synthesis, respectively; (C) inversion-type recombination results in duplication or deletion at junctions. The inversion-type substrate is designed to select out deletion products (Chen et al., 2001).

namely, all depend on the DNA cleavage. There is no question about requirement of double-strand breakages (DSBs) in CSR. Study on an artificial inversion-type substrate of CSR strongly indicates that two single-strand breakages (SSBs) are triggering events in CSR (Chen et al., 2001). This is because inversion products of CSR often contain duplication or deletion at the recombination junctions (Figure 1). Two SSBs each on a separate strand (staggered nicks) generate 3⬘ or 5⬘ overhang at the cleaved ends, which have to be repaired by exonucleolytic digestion or gap-filling DNA synthesis. CSR is a region-specific recombination that requires NHEJ to repair cleaved ends (Casellas et al., 1998; Manis et al., 1998, 2002; Rolink et al., 1996). Since one of the recombination junctions is looped out during normal CSR, identification of deletion or duplication is only possible by the inversion-type substrate. SHM can be easily explained by SSB and error-prone DNA repair (Diaz et al., 2001; Faili et al., 2002; Yavuz et al., 2002; Zan et al., 2001; Zeng et al., 2001). Although DSB coupled with homologous recombination was recently proposed as the mechanism of SHM (Papavasiliou and Schatz, 2000; Zan et al., 2003), DT40 cells with a loss-of-function mutation of XRCC2/3 that is essential to homologous recombination show reduced GC and strongly enhanced SHM (Sale et al., 2001), suggesting that homologous recombination is involved in GC but not in SHM. GC is probably mediated by homologous recombination with pseudogenes, but SHM may not depend on any type of recombination. Thus, the three reactions that depend on AID have different repair mechanisms after DNA cleavage.

The direct evidence to support involvement of AID in the DNA cleavage rather than the repair was obtained by experiments using focus formation of a phosphorylated form of histone H2AX (␥-H2AX) in the IgH locus as a marker of DSB (Petersen et al., 2001). Cleaved ends of DNA may be monitored by ATM/ATR or DNA-PKcs, which phosphorylates histone H2AX, necessary for protection of DNA ends and for recruitment of many repair enzymes (Celeste et al., 2002). Since H2AX-deficient mice have severely reduced CSR, ␥-H2AX focus formation is probably an important intermediate event of CSR. ␥-H2AX focus formation is induced at the IgH locus of spleen B cells upon class switch stimulation (Petersen et al., 2001). However, such focus formation as well as CSR is defective in AID-deficient B cells stimulated for class switching, which indicates that DSB in the IgH locus is dependent on the AID protein. Hypotheses for the DNA Cleavage Mechanism Two models are proposed for DNA cleavage by AID, as schematically depicted in Figure 2. RNA editing hypothesis is originally proposed by evolutionary conservation between AID and RNA editing enzyme APOBEC-1 (Honjo et al., 2002; Muramatsu et al., 2000). RNA editing hypothesis predicts that a complex of AID and its cofactors recognizes an unidentified mRNA, and AID deaminates C to U at a specific position of the mRNA. The edited mRNA encodes a DNA endonuclease that cleaves the V or S region DNA. Since AID expression in non-B cells (fibroblasts, T cells, hybridoma, and CHO cells) can initiate the DNA cleavage for CSR and SHM (Martin et al., 2002; Okazaki et al., 2002; Yoshikawa et al., 2002), the RNA editing hypothesis postulates that

Immunity 662

Figure 2. Two Models for AID Function in CSR and SHM Detailed mechanisms are explained in the text.

target mRNA as well as cofactor proteins are expressed more or less ubiquitously. The cofactors may recognize the structure of the target mRNA, just like ACF recognizes the structure of apoB100 mRNA. On the other hand, Neuberger and his colleagues have proposed that AID deaminates C in target DNA and generates U:G mismatch base pair (DNA deamination hypothesis) (Petersen-Mahrt et al., 2002; Di Noia and Neuberger, 2002). Such U:G pairs are assumed to be eliminated and repaired by the base excision repair pathway, which are evolutionarily conserved from E. coli to mammals. The enzymes involved in the base excision repair pathway are uracil DNA glycosylase or uracilN-glycosylase (UNG), apurinic/apyrimidinic (AP) endonuclease, and DNA polymerase ␤. UNG hydrolyzes U base in DNA, generating an AP site. AP endonuclease then introduces SSB at the AP site, which triggers repair DNA synthesis by DNA polymerase ␤. DNA deamination model also postulates that SHM can be induced without DNA cleavage, by replication-coupled mutation after generation of U (Petersen-Mahrt et al., 2002; Di Noia and Neuberger, 2002) (Figure 2). Supporting Evidence for DNA Deamination Hypothesis DNA Deamination In Vitro and in E. coli. DNA deamination hypothesis is supported by the fact that expression

of AID induces widespread mutations in the E. coli genome (Diaz et al., 2003; Petersen-Mahrt et al., 2002; Ramiro et al., 2003; Sohail et al., 2003; Ta et al., 2003). Since it is unlikely that AID cofactors and target mRNA are conserved between E. coli and mammals, the observation strongly suggests that AID directly deaminates C in E. coli DNA. DNA deamination hypothesis gained further support by in vitro experiments showing that single-strand DNA can be deaminated by AID (Bransteitter et al., 2003; Chaudhuri et al., 2003; Dickerson et al., 2003; Pham et al., 2003; Sohail et al., 2003; Yu et al., 2004). However, it should be noted that in vitro DNA deamination activity of AID is very weak. The amount of AID required to deaminate single-strand DNA in vitro is in far excess as compared with in vivo. Furthermore, not only AID but also a bona fide RNA editing enzyme APOBEC-1 introduces mutations in E. coli. However, overexpression of APOBEC-1 in B cells does not induce SHM or CSR (Eto et al., 2003; Fugmann et al., 2004). Induction of mutations in E. coli is also shown by other APOBEC-1 family members, whose functions in vivo are not yet clear (Harris et al., 2002). Obviously, DNA deamination by AID in E. coli and in vitro has no specificity in primary sequences of DNA, which is required for SHM and CSR. Since AID requires cofactors for recognition of substrates (see below), it may be the cofactor

Review 663

but not AID that determines the substrate specificity for DNA or RNA. These findings suggest that the DNA deamination reaction in E. coli and in vitro may not necessarily represent the real function of AID. Defects of CSR and SHM in UNG Deficiency. CSR is severely defective in UNG-deficient mouse (Rada et al., 2002) and human B cells (Imai et al., 2003). UNG⫺/⫺ B cells stimulated in vitro show about one order magnitude less efficient switching, but serum IgG and IgA levels are not strikingly reduced, indicating that UNG plays a role in CSR but apparently not an essential one. In vitrostimulated B cells of UNG-deficient patients show more severe defects in CSR. The frequency of SHM is not reduced but the base specificity is perturbed (higher transition at G/C) in UNG⫺/⫺ mouse B cells and DT40 chicken B cells expressing a specific UNG inhibitor, Ugi (Di Noia and Neuberger, 2002). Similar base specificity change is also confirmed in patients with UNG mutations (Imai et al., 2003). Intriguingly, UNG-deficient mice do not show other phenotypes, except for slow development of B cell lymphomas (50% survival at 101 weeks for UNG⫺/⫺ versus 116 weeks for UNG⫹/⫹) (Nilsen et al., 2003). In addition, the overall mutation rate in UNG⫺/⫺ mice is only marginally increased as compared to UNG⫹/⫹ mice (Nilsen et al., 2000). These results indicate that UNG activity to remove U can be compensated by other enzymes that recognize U:G mismatch pairs, such as SMUG1 and MSH2/6 (Krokan et al., 2002; Wang et al., 1999). Then why is the phenotype of UNG null mutation on CSR so severe? One speculation is that UNG may have another function than U removal that is important to CSR. Supporting Evidence for RNA Editing Hypothesis Requirements of De Novo Protein Synthesis for Cleavage. One of the distinctions between DNA deamination and RNA editing hypotheses resides in requirement of translation of mRNA edited by AID (Figure 2). It is therefore important to test whether new protein synthesis is required for CSR and SHM in addition to the AID protein synthesis. The study was done using a fusion protein of AID and the hormone binding domain of the estrogen receptor (AIDER) to avoid the effect of a protein synthesis inhibitor on synthesis of AID protein (Doi et al., 2003). AIDER associates with heat shock protein 90 and remains inactive unless a hormone analog, tamoxiphen, dissociates heat shock protein 90 to activate AID activity. AID-deficient spleen B cells were infected by retrovirus containing AIDER and then a protein synthesis inhibitor (cycloheximide) was added before and after activation of AIDER by tamoxiphen. Cycloheximide, when added before activation of AIDER, almost completely blocked CSR. However, inhibition by cycloheximide when added after activation of AIDER was much less significant. The results strongly indicate de novo protein synthesis is required in the early phase of AIDdependent CSR. The next obvious question is whether protein synthesis is required before or after the DNA cleavage. Very recently, we found that de novo protein synthesis is required before the cleavage step, using histone ␥-H2AX focus formation as a marker of DNA cleavage (N. Begum and T.H., submitted). AID-dependent ␥-H2AX focus formation in the IgH locus of AIDER expressing CH12F3-2 cells was blocked when protein synthesis was inhibited

by cycloheximide, although it did not block ␥-H2AX focus formation induced by ␥ ray irradiation. The fact that AID-dependent DNA cleavage in CSR requires new protein synthesis is consistent with RNA editing hypothesis but cannot be explained by direct deamination of DNA by AID because base excision repair enzymes are ubiquitously expressed. Functional Homology of AID and APOBEC-1. AID and APOBEC-1 are homologous not only in structure but also in their functional properties. As described below, both appear to require cofactors for recognition of substrates and both form dimer. Another interesting feature of AID that strongly indicates functional similarity to APOBEC-1 was also recently reported (Ito et al., 2004). APOBEC-1 is known to be a nuclear shuttling protein (Chester et al., 2003). The N-terminal and C-terminal regions of APOBEC-1 interact with a nuclear transporter (importin) and nuclear exclusion machinery, respectively. The APOBEC-1/ACF complex is transported by importin into the nucleus, where it recognizes and edits target mRNA. Then, the complex is exported to the cytoplasm where edited mRNA is translated. Ito et al. (2004) showed that AID also contains an N-terminal nuclear localization signal and a C-terminal nuclear exclusion signal. The C terminus truncation protein of AID accumulates in the nucleus. An antibiotic, leptomycin B, blocks nuclear exclusion machinery and thus causes accumulation of AID in the nucleus. These observations show that AID has evolutionarily conserved structure and function of a nuclear-cytoplasm shuttling protein like APOBEC-1. Dispensability of U Removal Activity of UNG in CSR. Severe reduction in CSR efficiency in UNG-deficient B cells strongly indicates that UNG plays a role in CSR. Since U removal is the only known enzymatic activity of UNG, this finding was thought to support involvement of U generation by AID-mediated DNA deamination in CSR. If UNG is required for U removal activity to trigger DNA cleavage by base excision repair mechanism, UNG inhibition should block DNA cleavage. This possibility was tested by combining a specific inhibitor protein of UNG, Ugi, and ␥-H2AX focus formation to monitor DSB in activated B cells (N. Begum and T.H., submitted). When DNA binding activity of UNG was blocked by expression of Ugi, CSR in CH12F3-2 B cells was almost completely blocked. However, DNA cleavage assessed by the ␥-H2AX focus formation at the IgH locus was not inhibited by Ugi expression. The results strongly indicate that UNG is not involved in the DNA cleavage step of CSR but rather repair stages that include exonucleolytic cleavage, DNA synthesis, and DNA ligation. Consistently, studies on several UNG mutants, which are devoid of U removal activity, rescued CSR in UNG⫺/⫺ B cells, indicating that U removal activity of UNG is not required for CSR. These results indicate that some unknown function but not U removal activity of UNG is required for repair stages in CSR. This conclusion explains the drastic difference in the effects on CSR and the general mutation frequency in UNG null mice described above. Human B cells with UNG null mutations that were previously reported to be defective in activation induced DSB by the linker-mediated PCR assay (Imai et al., 2003). The difference from the above observation in mouse could be ascribed to the detection

Immunity 664

method of DSB or stimulation conditions. DSB detected by linker mediated PCR in activated B cells was reported to be not always AID dependent (Bross et al., 2002; Catalan et al., 2003; Faili et al., 2002; Papavasiliou and Schatz, 2002; Rush et al., 2004). Role of UNG at the Repair Stage in CSR The tertiary structures of human UNG and its complex with Ugi have been solved (Mol et al., 1995a, 1995b). Human and mouse UNG are ⬎95% homologous. UNG has the N-terminal PCNA/replication protein A (RPA) binding domain and the C-terminal catalytic domain that interacts with U and DNA. Ugi binds to UNG tightly and blocks its interaction with DNA. The primary function of UNG is removal of misincorporated U opposite to adenine during replication (Parikh et al., 2000). UNG in the PCNA complex at the replication folk scans newly synthesized DNA to flip out U and remove it by hydrolysis. Removal of U by UNG is followed by strand cleavage by AP endonuclease. The AP site is removed and repaired by gap-filling activity of DNA polymerase ␤. However, the initial sensor of U in the interphase DNA is not well understood (Krokan et al., 2002). It is clear that UNG is important but not essential to CSR because of the residual switching activity in UNG⫺/⫺ mice (Rada et al., 2002). DNA binding activity of UNG appears to be critical, because Ugi almost completely blocks UNG as well as CSR. Although the role of UNG in CSR remains elusive, the enzymatic activity of UNG is not required for CSR. We speculate that UNG may be involved in recruiting either DNA synthesis apparatus, including error-prone polymerases or other members of the repair machinery. It is also possible that UNG plays a role in holding the cleaved ends. Indeed, UNG without U removal activity is required for replication of vaccinia virus DNA (De Silva and Moss, 2003). A similar puzzle was reported for the DNA-PKcs null mutation and the SCID mutation of DNA-PKcs. The SCID mutant mouse has no catalytic activity of DNA-PKcs but shows about 50% CSR activity for all isotypes (Bosma et al., 2002). By contrast, B cells of mice lacking DNA-PKcs cannot switch to any isotypes except for IgG1 with 30%–50% efficiency of wild-type (Manis et al., 2002), suggesting that DNA-PKcs protein but not its enzymatic activity is important. Here, again, DNA-PKcs may play an essential role as recruiter of NHEJ repair proteins. Furthermore, B cells of ATPase-defective MSH2 knockin mice show milder reduction in CSR as compared with those of MSH2⫺/⫺ mice (Martin et al., 2003). Once again, dissociation of CSR phenotypes is evident between loss of the enzymatic activity and of protein per se. Distinction of CSR and SHM by Separate Cofactors The loss-of-function mutations of AID for CSR in HIGM2 syndrome patients scatter from the N terminus to the C terminus of the 198 amino acid residue AID protein (Ta et al., 2003). Such observation raises two possibilities: (1) this protein is very labile and a point mutation of every part of the molecule destroys the tertiary structure of the AID protein and (2) AID interacts with other proteins at different parts of the molecule. To explore this possibility, we assayed in vitro CSR and SHM activities of the AID mutant cDNAs of HIGM2. Most of them are loss-of-function mutations of both activities. However, three mutations (truncations or replacement) of the C terminus of AID abolish CSR activity but retain SHM activity. Such mutants also retain DNA deamination ac-

tivity in E. coli. The results can be most easily explained by the presence of cofactor that plays a specific role in CSR but not in SHM. Subsequently, it was found that other point mutations located in the N-terminal part of AID almost completely destroyed SHM activity but retained CSR activity (Shinkura et al., 2004). Thus, AID appears to interact with at least two separate cofactors, each specifically required to either SHM or CSR. The function of these cofactors remains unknown, but it is most likely that they are involved in recognition of AID target, RNA, or DNA. Although AID per se can deaminate C in mRNA or DNA, actual enzymatic activity of AID may require cofactors that target specific bases in nucleic acids. In the absence of such cofactors, AID may attack DNA or RNA in an uncoupled fashion as shown in E. coli and in vitro. Although AID is likely involved in cleavage of both V and S regions, CSR and SHM take place independently in activated B cells, in which both V and CH genes are transcribed. The results suggest that V and S regions must be recognized distinctively and AID itself is unlikely to distinguish V and S region DNA. According to DNA deamination hypothesis, AID requires CSR- and SHMspecific cofactors, which can recognize S and V region DNA, respectively. RNA editing hypothesis predicts that separate DNA endonucleases generated by RNA editing distinguish V and S regions and that AID cofactors specific to SHM and CSR recognize different mRNAs to be edited by AID. The evidence for DNA deamination hypothesis in SHM is increase in the transition rate in C/G mutations, namely C to T or G to A in UNG-deficient B cells. However, UNG deficiency does not reduce SHM frequency at all (Rada et al., 2002). On the other hand, MSH2-deficient mice show considerable reduction of SHM frequency in addition to a strong bias of target bases to GC (Martin et al., 2003; Phung et al., 1998; Rada et al., 1998). It is not clear whether mutated base specificity shift reflects a direct effect of UNG or MMR, because mutated base specificity can be modulated by many factors, including choices of error-prone polymerases. Both UNG and MSH3/6 have domains to interact with PCNA, implying that they may be involved in recruitment of DNA synthesis machinery. Change of mutation base specificity in UNG- and MSH2-deficient mice can be explained by the hypothesis that they recruit different error-prone DNA polymerases. It remains to be determined whether UNG plays a role in SHM as a U removal enzyme or as a recruiter of other DNA repair machinery. Summary and Perspective Recent data clearly indicate that U removal activity of UNG is not required for CSR and that another function of UNG is involved in the repair stage of CSR. Although these results do not exclude DNA deamination itself, the mechanism of DNA cleavage proposed by DNA deamination hypothesis has to be reconsidered. Taken together with other recent findings, including de novo protein synthesis requirement for cleavage, it is easier to explain AID-induced DNA cleavage according to RNA editing hypothesis. The most likely scenario for DNA cleavage based on the current data is as follows (Figure 3). AID together with class switch-specific cofactor or SHM-specific cofactor

Review 665

Figure 3. The Updated RNA Editing Model for CSR and SHM AID deaminates two mRNAs through recognition by SHM- and CSR-specific cofactors, and mRNAs thus edited are translated to a mutator and recombinase, respectively. SSB on V gene recruits error-prone repair system in association with MMR and UNG. Replication fixes newly introduced mutations. DSB on two S regions are recognized damage sensors, giving rise to ␥H2AX focus formation. DNA ends with ␥H2AX foci recruit NHEJ repair (Ku 70, 80, DNA-PKcs, XRCC4, lygase IV), UNG, MMR, and error-prone repair system. Exonuclease and polymerase activities make DNA blunt ends ready to NHEJ repair. X and Y are unknown enzymes involved in repair of V and S regions, respectively. M, mutation.

edits yet unidentified mRNAs, generating mRNAs encoding putative endonucleases, specific to either V or S region DNA. Both V- and S-specific endonucleases introduce nick on the target DNA. S-specific endonuclease introduces nicks rather frequently, generating staggered nick DSB. By contrast, V-specific endonuclease introduces nick rarely and generates SSB. After such cleavage, both V and S regions require repair systems to complete SHM and CSR, respectively. Cleaved S region is monitored by DSB sensor protein ATM/ATR and DNA-PKcs, which phosphorylate H2AX to initiate NHEJ mechanism. During this process, many other proteins, including MMR, exonuclease I, UNG, and DNA polymerases, are most likely recruited. UNG is required not as its U removal activity but probably as a recruiting protein of other enzymes. For SHM, the repair process is probably simpler and does not depend on NHEJ (Bemark et al., 2000). SHM is severely affected by MMR and UNG deficiency. Specificity of mutated bases may depend on many factors, including polymerases used during the repair process. Other enzymes shown to affect SHM include exonuclease 1 (Bardwell et al., 2004) and error-

prone polymerases (Diaz et al., 2001; Yavuz et al., 2002; Zan et al., 2001; Zeng et al., 2001). Obviously, there are many questions to be solved. The key question is how two endonucleases can distinguish V and S regions. Since there is no primary sequence specificity, it is likely that these two endonucleases recognize secondary structures, which may be formed during active transcription of these loci. Such recognition specificity will be clarified by isolation of target mRNA of AID. If this RNA editing scenario is verified by isolation of target mRNAs, genetic modifications in the immune system consist of tremendously complex layers of informational alteration systems, i.e., RNA editing, DNA recombination, and point mutations. It is really fascinating how the immune system has involved such complex layers of system to amplify genetic information encoded in the genome. Acknowledgments This study was supported by a Center of Excellence Grant from the Ministry of Education, Science, Sports, and Culture of Japan.

Immunity 666

References

(2003). AID mediates hypermutation by deaminating single stranded DNA. J. Exp. Med. 197, 1291–1296.

Arakawa, H., and Buerstedde, J.M. (2004). Immunoglobulin gene conversion: insights from bursal B cells and the DT40 cell line. Dev. Dyn. 229, 458–464.

Doi, T., Kinoshita, K., Ikegawa, M., Muramatsu, M., and Honjo, T. (2003). De novo protein synthesis is required for the activationinduced cytidine deaminase function in class-switch recombination. Proc. Natl. Acad. Sci. USA 100, 2634–2638.

Arakawa, H., Hauschild, J., and Buerstedde, J.M. (2002). Requirement of the activation-induced deaminase (AID) gene for immunoglobulin gene conversion. Science 295, 1301–1306. Bardwell, P.D., Woo, C.J., Wei, K., Li, Z., Martin, A., Sack, S.Z., Parris, T., Edelmann, W., and Scharff, M.D. (2004). Altered somatic hypermutation and reduced class-switch recombination in exonuclease 1-mutant mice. Nat. Immunol. 5, 224–229. Bemark, M., Sale, J.E., Kim, H.J., Berek, C., Cosgrove, R.A., and Neuberger, M.S. (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. Bosma, G.C., Kim, J., Urich, T., Fath, D.M., Cotticelli, M.G., Ruetsch, N.R., Radic, M.Z., and Bosma, M.J. (2002). DNA-dependent protein kinase activity is not required for immunoglobulin class switching. J. Exp. Med. 196, 1483–1495. Bransteitter, R., Pham, P., Scharff, M.D., and Goodman, M.F. (2003). Activation-induced cytidine deaminase deaminates deoxycytidine on single-stranded DNA but requires the action of RNase. Proc. Natl. Acad. Sci. USA 100, 4102–4107. Bross, L., Muramatsu, M., Kinoshita, K., Honjo, T., and Jacobs, H. (2002). DNA double-strand breaks: prior to but not sufficient in targeting hypermutation. J. Exp. Med. 195, 1187–1192. Casellas, R., Nussenzweig, A., Wuerffel, R., Pelanda, R., Reichlin, A., Suh, H., Qin, X.F., Besmer, E., Kenter, A., Rajewsky, K., et al. (1998). Ku80 is required for immunoglobulin isotype switching. EMBO J. 17, 2404–2411. Catalan, N., Selz, F., Imai, K., Revy, P., Fischer, A., and Durandy, A. (2003). 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 ␮ region. J. Immunol. 171, 2504–2509. Celeste, A., Petersen, S., Romanienko, P.J., Fernandez-Capetillo, O., Chen, H.T., Sedelnikova, O.A., Reina-San-Martin, B., Coppola, V., Meffre, E., Difilippantonio, M.J., et al. (2002). Genomic instability in mice lacking histone H2AX. Science 296, 922–927. Chaudhuri, J., Tian, M., Khuong, C., Chua, K., Pinaud, E., and Alt, F.W. (2003). Transcription-targeted DNA deamination by the AID antibody diversification enzyme. Nature 422, 726–730. Chen, X., Kinoshita, K., and Honjo, T. (2001). Variable deletion and duplication at recombination junction ends: implication for staggered double-strand cleavage in class switch recombination. Proc. Natl. Acad. Sci. USA 98, 13860–13865. Chester, A., Somasekaram, A., Tzimina, M., Jarmuz, A., Gisbourne, J., O’Keefe, R., Scott, J., and Navaratnam, N. (2003). The apolipoprotein B mRNA editing complex performs a multifunctional cycle and suppresses nonsense-mediated decay. EMBO J. 22, 3971– 3982. De Silva, F.S., and Moss, B. (2003). Vaccinia virus uracil DNA glycosylase has an essential role in DNA synthesis that is independent of its glycosylase activity: catalytic site mutations reduce virulence but not virus replication in cultured cells. J. Virol. 77, 159–166. Di Noia, J., and Neuberger, M.S. (2002). Altering the pathway of immunoglobulin hypermutation by inhibiting uracil-DNA glycosylase. Nature 419, 43–48. Diaz, M., Verkoczy, L.K., Flajnik, M.F., and Klinman, N.R. (2001). Decreased frequency of somatic hypermutation and impaired affinity maturation but intact germinal center formation in mice expressing antisense RNA to DNA polymerase ␨. J. Immunol. 167, 327–335. Diaz, M., Ray, M., Wheeler, L.J., Verkoczy, L.K., and Mathews, C.K. (2003). Mutagenesis by AID, a molecule critical to immunoglobulin hypermutation, is not caused by an alteration of the precursor nucleotide pool. Mol. Immunol. 40, 261–268. Dickerson, S.K., Market, E., Besmer, E., and Papavasiliou, F.N.

Durandy, A., and Honjo, T. (2001). Human genetic defects in classswitch recombination (hyper-IgM syndromes). Curr. Opin. Immunol. 13, 543–548. Eto, T., Kinoshita, K., Yoshikawa, K., Muramatsu, M., and Honjo, T. (2003). RNA-editing cytidine deaminase Apobec-1 is unable to induce somatic hypermutation in mammalian cells. Proc. Natl. Acad. Sci. USA 100, 12895–12898. Fagarasan, S., and Honjo, T. (2003). Intestinal IgA synthesis: regulation of front-line body defences. Nat. Rev. Immunol. 3, 63–72. Fagarasan, S., Kinoshita, K., Muramatsu, M., Ikuta, K., and Honjo, T. (2001). In situ class switching and differentiation to IgA-producing cells in the gut lamina propria. Nature 413, 639–643. Fagarasan, S., Muramatsu, M., Suzuki, K., Nagaoka, H., Hiai, H., and Honjo, T. (2002). Critical roles of activation-induced cytidine deaminase (AID) in the homeostasis of gut flora. Science 298, 1424– 1427. Faili, A., Aoufouchi, S., Gueranger, Q., Zober, C., Leon, A., Bertocci, B., Weill, J.C., and Reynaud, C.A. (2002). AID-dependent somatic hypermutation occurs as a DNA single-strand event in the BL2 cell line. Nat. Immunol. 3, 815–821. Fugmann, S.D., Rush, J.S., and Schatz, D.G. (2004). Non-redundancy of cytidine deaminases in class switch recombination. Eur. J. Immunol. 34, 844–849. Harada, Y., Muramatsu, M., Shibata, T., Honjo, T., and Kuroda, K. (2003). Unmutated immunoglobulin M can protect mice from death by influenza virus infection. J. Exp. Med. 197, 1779–1785. Harris, R.S., Petersen-Mahrt, S.K., and Neuberger, M.S. (2002). RNA editing enzyme APOBEC1 and some of its homologs can act as DNA mutators. Mol. Cell 10, 1247–1253. Honjo, T., Kinoshita, K., and Muramatsu, M. (2002). Molecular mechanism of class switch recombination: linkage with somatic hypermutation. Annu. Rev. Immunol. 20, 165–196. Imai, K., Slupphaug, G., Lee, W.I., Revy, P., Nonoyama, S., Catalan, N., Yel, L., Forveille, M., Kavli, B., Krokan, H.E., et al. (2003). Human uracil-DNA glycosylase deficiency associated with profoundly impaired immunoglobulin class-switch recombination. Nat. Immunol. 4, 1023–1028. Ito, S., Nagaoka, H., Shinkura, R., Begum, N., Muramatsu, M., Nakata, M., and Honjo, T. (2004). Activation-induced cytidine deaminase shuttles between nucleus and cytoplasm like apolipoprotein B mRNA editing catalytic polypeptide 1. Proc. Natl. Acad. Sci. USA 101, 1975–1980. Krokan, H.E., Drablos, F., and Slupphaug, G. (2002). Uracil in DNA— occurrence, consequences and repair. Oncogene 21, 8935–8948. Lau, P.P., Zhu, H.J., Baldini, A., Charnsangavej, C., and Chan, L. (1994). Dimeric structure of a human apolipoprotein B mRNA editing protein and cloning and chromosomal localization of its gene. Proc. Natl. Acad. Sci. USA 91, 8522–8526. Maizels, N. (2000). Secret sharers in the immune system: a novel RNA editing activity links switch recombination and somatic hypermutation. Genome Biol. 1, 1025. Manis, J.P., Gu, Y., Lansford, R., Sonoda, E., Ferrini, R., Davidson, L., Rajewsky, K., and Alt, F.W. (1998). Ku70 is required for late B cell development and immunoglobulin heavy chain switching. J. Exp. Med. 187, 2081–2089. Manis, J.P., Dudley, D., Kaylor, L., and Alt, F.W. (2002). IgH class switch recombination to IgG1 in DNA-PKcs-deficient B cells. Immunity 16, 607–617. Martin, A., and Scharff, M.D. (2002a). AID and mismatch repair in antibody diversification. Nat. Rev. Immunol. 2, 605–614. Martin, A., and Scharff, M.D. (2002b). Somatic hypermutation of the AID transgene in B and non-B cells. Proc. Natl. Acad. Sci. USA 99, 12304–12308.

Review 667

Martin, A., Bardwell, P.D., Woo, C.J., Fan, M., Shulman, M.J., and Scharff, M.D. (2002). Activation-induced cytidine deaminase turns on somatic hypermutation in hybridomas. Nature 415, 802–806. Martin, A., Li, Z., Lin, D.P., Bardwell, P.D., Iglesias-Ussel, M.D., Edelmann, W., and Scharff, M.D. (2003). Msh2 ATPase activity is essential for somatic hypermutation at a-T basepairs and for efficient class switch recombination. J. Exp. Med. 198, 1171–1178. Mehta, A., Kinter, M.T., Sherman, N.E., and Driscoll, D.M. (2000). Molecular cloning of apobec-1 complementation factor, a novel RNA-binding protein involved in the editing of apolipoprotein B mRNA. Mol. Cell. Biol. 20, 1846–1854. Mol, C.D., Arvai, A.S., Sanderson, R.J., Slupphaug, G., Kavli, B., Krokan, H.E., Mosbaugh, D.W., and Tainer, J.A. (1995a). Crystal structure of human uracil-DNA glycosylase in complex with a protein inhibitor: protein mimicry of DNA. Cell 82, 701–708. Mol, C.D., Arvai, A.S., Slupphaug, G., Kavli, B., Alseth, I., Krokan, H.E., and Tainer, J.A. (1995b). Crystal structure and mutational analysis of human uracil-DNA glycosylase: structural basis for specificity and catalysis. Cell 80, 869–878. Muramatsu, M., Sankaranand, V.S., Anant, S., Sugai, M., Kinoshita, K., Davidson, N.O., and Honjo, T. (1999). Specific expression of activation-induced cytidine deaminase (AID), a novel member of the RNA-editing deaminase family in germinal center B cells. J. Biol. Chem. 274, 18470–18476. Muramatsu, M., Kinoshita, K., Fagarasan, S., Yamada, S., Shinkai, Y., and Honjo, T. (2000). Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102, 553–563. Muto, T., Muramatsu, M., Taniwaki, M., Kinoshita, K., and Honjo, T. (2000). Isolation, tissue distribution and chromosomal localization of the human activation-induced cytidine deaminase (hAID) gene. Genomics 68, 85–88. Nilsen, H., Rosewell, I., Robins, P., Skjelbred, C.F., Andersen, S., Slupphaug, G., Daly, G., Krokan, H.E., Lindahl, T., and Barnes, D.E. (2000). Uracil-DNA glycosylase (UNG)-deficient mice reveal a primary role of the enzyme during DNA replication. Mol. Cell 5, 1059– 1065. Nilsen, H., Stamp, G., Andersen, S., Hrivnak, G., Krokan, H.E., Lindahl, T., and Barnes, D.E. (2003). Gene-targeted mice lacking the Ung uracil-DNA glycosylase develop B-cell lymphomas. Oncogene 22, 5381–5386. Oettinger, M.A., Schatz, D.G., Gorka, C., and Baltimore, D. (1990). RAG-1 and RAG-2, adjacent genes that synergistically activate V(D)J recombination. Science 248, 1517–1523. Okazaki, I.M., Kinoshita, K., Muramatsu, M., Yoshikawa, K., and Honjo, T. (2002). The AID enzyme induces class switch recombination in fibroblasts. Nature 416, 340–345. Okazaki, I.M., Hiai, H., Kakazu, N., Yamada, S., Muramatsu, M., Kinoshita, K., and Honjo, T. (2003). Constitutive expression of AID leads to tumorigenesis. J. Exp. Med. 197, 1173–1181. Papavasiliou, F.N., and Schatz, D.G. (2000). Cell-cycle-regulated DNA double-stranded breaks in somatic hypermutation of immunoglobulin genes. Nature 408, 216–221. Papavasiliou, F.N., and Schatz, D.G. (2002). The activation-induced deaminase functions in a postcleavage step of the somatic hypermutation process. J. Exp. Med. 195, 1193–1198. Parikh, S.S., Putnam, C.D., and Tainer, J.A. (2000). Lessons learned from structural results on uracil-DNA glycosylase. Mutat. Res. 460, 183–199. Petersen, S., Casellas, R., Reina-San-Martin, B., Chen, H.T., Difilippantonio, M.J., Wilson, P.C., Hanitsch, L., Celeste, A., Muramatsu, M., Pilch, D.R., et al. (2001). AID is required to initiate Nbs1/␥-H2AX focus formation and mutations at sites of class switching. Nature 414, 660–665. Petersen-Mahrt, S.K., Harris, R.S., and Neuberger, M.S. (2002). AID mutates E. coli suggesting a DNA deamination mechanism for antibody diversification. Nature 418, 99–104. Pham, P., Bransteitter, R., Petruska, J., and Goodman, M.F. (2003).

Processive AID-catalysed cytosine deamination on single-stranded DNA simulates somatic hypermutation. Nature 424, 103–107. Phung, Q.H., Winter, D.B., Cranston, A., Tarone, R.E., Bohr, V.A., Fishel, R., and Gearhart, P.J. (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. Rada, C., Ehrenstein, M.R., Neuberger, M.S., and Milstein, C. (1998). Hot spot focusing of somatic hypermutation in MSH2-deficient mice suggests two stages of mutational targeting. Immunity 9, 135–141. Rada, C., Williams, G.T., Nilsen, H., Barnes, D.E., Lindahl, T., and Neuberger, M.S. (2002). Immunoglobulin isotype switching is inhibited and somatic hypermutation perturbed in UNG-deficient mice. Curr. Biol. 12, 1748–1755. Ramiro, A.R., Stavropoulos, P., Jankovic, M., and Nussenzweig, M.C. (2003). Transcription enhances AID-mediated cytidine deamination by exposing single-stranded DNA on the nontemplate strand. Nat. Immunol. 4, 452–456. Revy, P., Muto, T., Levy, Y., Geissmann, F., Plebani, A., Sanal, O., Catalan, N., Forveille, M., Dufourcq-Labelouse, R., Gennery, A., et al. (2000). Activation-induced cytidine deaminase (AID) deficiency causes the autosomal recessive form of the Hyper-IgM syndrome (HIGM2). Cell 102, 565–575. Reynaud, C.A., Aoufouchi, S., Faili, A., and Weill, J.C. (2003). What role for AID: mutator, or assembler of the immunoglobulin mutasome? Nat. Immunol. 4, 631–638. Rolink, A., Melchers, F., and Andersson, J. (1996). The SCID but not the RAG-2 gene product is required for S␮-S⑀ heavy chain class switching. Immunity 5, 319–330. Rush, J.S., Fugmann, S.D., and Schatz, D.G. (2004). Staggered AIDdependent DNA double strand breaks are the predominant DNA lesions targeted to S ␮ in Ig class switch recombination. Int. Immunol. 16, 549–557. Sale, J.E., Calandrini, D.M., Takata, M., Takeda, S., and Neuberger, M.S. (2001). Ablation of XRCC2/3 transforms immunoglobulin V gene conversion into somatic hypermutation. Nature 412, 921–926. Schatz, D.G., Oettinger, M.A., and Baltimore, D. (1989). The V(D)J recombination activating gene, RAG-1. Cell 59, 1035–1048. Sekiguchi, J., Alt, F., and Oettinger, M.A. (2004). The mechanisms of V(D)J recombination. In Molecular Biology of B Cells, T. Honjo, F. Alt, and M. Neuberger, eds. (San Diego: Elsevier Academic Press), pp. 62–71. Shinkura, R., Ito, S., Begum, N., Nagaoka, H., Muramatsu, M., Kinoshita, K., Sakakibara, Y., Hijikata, H., and Honjo, T. (2004). Separate domains of AID are required for somatic hypermutation and class switch recombination. Nat. Immunol., in press. Sohail, A., Klapacz, J., Samaranayake, M., Ullah, A., and Bhagwat, A.S. (2003). Human activation-induced cytidine deaminase causes transcription-dependent, strand-biased C to U deaminations. Nucleic Acids Res. 31, 2990–2994. Stavnezer, J., Kinoshita, K., Muramatsu, M., and Honjo, T. (2004). Molecular mechanisms of class switch recombination. In Molecular Biology of B Cells, T. Honjo, F. Alt, and M. Neuberger, eds. (San Diego: Elsevier Academic Press), pp. 307–319. Suzuki, K., Meek, B., Doi, Y., Muramatsu, M., Chiba, T., Honjo, T., and Fagarasan, S. (2004). Aberrant expansion of segmented filamentous bacteria in IgA-deficient gut. Proc. Natl. Acad. Sci. USA 101, 1981– 1986. Ta, V.T., Nagaoka, H., Catalan, N., Durandy, A., Fischer, A., Imai, K., Nonoyama, S., Tashiro, J., Ikegawa, M., Ito, S., et al. (2003). AID mutant analyses indicate requirement for class-switch-specific cofactors. Nat. Immunol. 4, 843–848. Teng, B., Burant, C.F., and Davidson, N.O. (1993). Molecular cloning of an apolipoprotein B messenger RNA editing protein. Science 260, 1816–1819. Wang, H., Lawrence, C.W., Li, G.M., and Hays, J.B. (1999). Specific binding of human MSH2.MSH6 mismatch-repair protein heterodimers to DNA incorporating thymine- or uracil-containing UV light pho-

Immunity 668

toproducts opposite mismatched bases. J. Biol. Chem. 274, 16894– 16900. Yavuz, S., Yavuz, A.S., Kraemer, K.H., and Lipsky, P.E. (2002). The role of polymerase ␩ in somatic hypermutation determined by analysis of mutations in a patient with xeroderma pigmentosum variant. J. Immunol. 169, 3825–3830. Yoshikawa, K., Okazaki, I.M., Eto, T., Kinoshita, K., Muramatsu, M., Nagaoka, H., and Honjo, T. (2002). AID enzyme-induced hypermutation in an actively transcribed gene in fibroblasts. Science 296, 2033– 2036. Yu, K., Huang, F.T., and Lieber, M.R. (2004). DNA substrate length and surrounding sequence affect the activation-induced deaminase activity at cytidine. J. Biol. Chem. 279, 6496–6500. Zan, H., Komori, A., Li, Z., Cerutti, A., Schaffer, A., Flajnik, M.F., Diaz, M., and Casali, P. (2001). The translesion DNA polymerase ␨ plays a major role in Ig and bcl-6 somatic hypermutation. Immunity 14, 643–653. Zan, H., Wu, X., Komori, A., Holloman, W.K., and Casali, P. (2003). AID-dependent generation of resected double-strand DNA breaks and recruitment of Rad52/Rad51 in somatic hypermutation. Immunity 18, 727–738. Zeng, X., Winter, D.B., Kasmer, C., Kraemer, K.H., Lehmann, A.R., and Gearhart, P.J. (2001). DNA polymerase ␩ is an A-T mutator in somatic hypermutation of immunoglobulin variable genes. Nat. Immunol. 2, 537–541.