Evolution of Phosphorylation-Dependent Regulation of Activation-Induced Cytidine Deaminase

Molecular Cell

Short Article Evolution of Phosphorylation-Dependent Regulation of Activation-Induced Cytidine Deaminase Uttiya Basu,1 Yabin Wang,1 and Frederick W. Alt1,* 1Howard Hughes Medical Institute, The Children’s Hospital, Harvard Medical School, and the Immune Disease Institute, 300 Longwood Avenue, Boston, MA 02115, USA *Correspondence: [email protected] DOI 10.1016/j.molcel.2008.08.019

SUMMARY

Interaction of activation-induced cytidine deaminase (AID) with replication protein A (RPA) has been proposed to promote AID access to transcribed double-stranded (ds) DNA during immunoglobulin light chain and heavy chain class switch recombination (CSR). Mouse AID (mAID) interaction with RPA and transcription-dependent dsDNA deamination in vitro requires protein kinase A (PKA) phosphorylation at serine 38 (S38), and normal mAID CSR activity depends on S38. However, zebrafish AID (zAID) catalyzes robust CSR in mouse cells despite lacking an S38-equivalent PKA site. Here, we show that aspartate 44 (D44) in zAID provides similar in vitro and in vivo functionality as mAID S38 phosphorylation. Moreover, introduction of a PKA site into a zAID D44 mutant made it PKA dependent for in vitro activities and restored normal CSR activity. Based on these findings, we generated mAID mutants that similarly function independently of S38 phosphorylation. Comparison of bony fish versus amphibian and mammalian AIDs suggests evolutionary divergence from constitutive to PKAregulated RPA/AID interaction. INTRODUCTION B cells undergo two different forms of immunoglobulin (Ig) gene alteration following antigen activation. Somatic hypermutation (SHM) introduces point mutations into Ig light-chain and heavychain (IgH) variable-region exons, allowing generation of antibodies with increased affinity (Di Noia and Neuberger, 2007; Odegard and Schatz, 2006). IgH class switch recombination (CSR) replaces the IgH Cm exons, which encode the first IgH constant region (CH) expressed, with one of several downstream sets of CH exons, which encode different antibody classes (Chaudhuri et al., 2007; Honjo et al., 2002). Long repetitive switch (S) regions precede each set of CH exons. CSR involves introduction of DNA double strand breaks (DSBs) into the donor Sm and into a downstream acceptor S region, followed by joining of the two S regions (Chaudhuri et al., 2007). Both CSR and SHM require activation-induced cytidine deaminase (AID)

(Muramatsu et al., 2000; Revy et al., 2000). AID initiates CSR and SHM by catalyzing deamination of cytidines in S regions and variable-region exons, respectively, in a process that requires transcription of targeted sequences (Chaudhuri et al., 2007; Goodman et al., 2007). Subsequently, deaminated cytidines are processed by subversion of normal repair pathways to generate variable region point mutations or S region mutations and DSBs (Di Noia and Neuberger, 2007). Purified AID has cytidine deaminase activity on singlestranded (ss), but not double-stranded (ds), DNA (Bransteitter et al., 2003; Chaudhuri et al., 2003; Dickerson et al., 2003; Ramiro et al., 2003; Sohail et al., 2003). Several mechanisms have been implicated in AID access of duplex DNA in vivo (Chaudhuri et al., 2003). When transcribed in physiologic orientation, but not inverted orientation, mammalian S regions generate ssDNA within R loops (Tian and Alt, 2000; Yu et al., 2003). In this regard, biochemical and genetic studies have indicated that R loop formation in transcribed mammalian S regions can enhance AID access (Shinkura et al., 2003; Chaudhuri et al., 2003). However, R loop formation cannot fully explain AID access to S regions, since they still support reduced CSR when transcribed in nonphysiologic orientation. Also, R loops cannot explain AID access to transcribed variable-region exons, which do not form R loops. In the latter context, biochemical assays revealed that RPA, a trimeric ssDNA binding protein involved in replication and repair (Wold, 1997), promotes efficient AID deamination of in vitro transcribed sequences rich in SHM motifs (‘‘SHM substrates’’) that do not form R loops (Chaudhuri et al., 2004). While AID and SHM are present in tetrapods (bony fish), CSR first occurs evolutionarily in amphibians, (e.g., Xenopus), suggesting that CSR evolved after SHM (Stavnezer and Amemiya, 2004). Transcribed Xenopus Sm (XSm) does not form R loops, but can replace mouse Sg1 to affect CSR (Zarrin et al., 2004). XSm CSR junctions and AID deamination sites have been observed to occur within a region dense in AGCT sequences—a canonical SHM motif (Zarrin et al., 2004). Thus, AID CSR activities may have evolved from AID SHM functions (Barreto et al., 2005; Wakae et al., 2006). As mammalian S regions are rich in SHM motifs, AID/RPA targeting may function in mammalian CSR (Zarrin et al., 2004). A portion of mouse B cell AID (mAID) is phosphorylated at serine 38 (S38) (Basu et al., 2005; McBride et al., 2006), and mAID interaction with RPA is dependent upon S38 phosphorylation (Basu et al., 2005). The mAID S38 site is part of a RRX(S/T) cAMP-dependent protein kinase A (PKA) consensus motif (Basu et al., 2005). Correspondingly, mAID can be phosphorylated by

Molecular Cell 32, 285–291, October 24, 2008 ª2008 Elsevier Inc. 285

Molecular Cell AID Regulation by Phosphorylation

PKA in vitro at S38, and coexpression of PKA with mAID in fibroblasts enhances S38 phosphorylation (Basu et al., 2005; Basu et al., 2007; McBride et al., 2006; Pasqualucci et al., 2006). In vitro phosphorylation of nonphosphorylated mAID by PKA conferred ability to bind RPA and to mediate transcriptiondependent dsDNA deamination (Basu et al., 2005). In addition, mutation of mAID S38 to alanine (mAIDS38A) had no effect on ssDNA catalytic activity, but abolished AID phosphorylation by PKA and, correspondingly, impaired ability of AID to interact with RPA and deaminate transcribed SHM substrates (Basu et al., 2005). Moreover, mAIDS38A had significantly reduced ability to catalyze CSR when expressed in AID-deficient B cells (Basu et al., 2005, 2007; McBride et al., 2006; Pasqualucci et al., 2006), consistent with AID phosphorylation at S38 and RPA interaction augmenting AID function in CSR. Finally, mAIDS38A showed greatly reduced ability to carry out gene conversion (GCV) and SHM in chicken DT40 cells (Chatterji et al., 2007). AID has a PKA-consensus S38 phosphorylation site in all analyzed tetrapods (mouse, human, Xenopus), but not in any analyzed teleost (zebrafish, fugu, catfish), the latter of which undergo SHM, but not CSR (Figure 1A). However, zebrafish AID (zAID), despite lacking a clear counterpart of the mammalian AID S38 PKA site, catalyzes robust CSR in mouse B cells (Barreto et al., 2005; Wakae et al., 2006). These findings were viewed as a challenge to the model that S38 phosphorylation and RPA interaction play roles in physiological AID function (Shinkura et al., 2007; Pham et al., 2008). The mechanism by which zAID ‘‘bypasses’’ PKA phosphorylation has remained speculative (Pan-Hammarstrom et al., 2007), with one possibility being that zAID might interact with RPA via phosphorylation at a different site or via a phosphorylation-independent mechanism. In the latter context, AID from all analyzed teleost species has a conserved aspartate residue (D44 in zebrafish) in the region that harbors tetrapod S38 (Figure 1A) (Barreto et al., 2005; Wakae et al., 2006), suggesting that D44 might be a phosphorylation mimic in teleost AID (Basu et al., 2005). In this context, only catfish AID has a serine residue corresponding to mAID S38 (at position S41), but this residue is not found in a consensus PKA site (Figure 1A). In this study, we demonstrate that zAID D44 functionally mimics S38 phosphorylation in mAID by allowing constitutive association with RPA. In addition, we have engineered a zAID that interacts with RPA and deaminates transcribed dsDNA in a PKA phosphorylation-dependent fashion and an mAID that similarly functions in a phosphorylation-independent fashion. RESULTS Binding of zAID to RPA without Phosphorylation To test potential functions of the conserved D44 residue in zAID, we expressed N-terminal, flag-tagged, wild-type (WT) zAID (zAIDWT), and zAID in which D44 had been mutated to alanine (zAIDD44A) (Figure 1B) in HEK293 cells. Subsequently, we partially purified zAIDWT and zAIDD44A from HEK293T nuclear extracts and immunoprecipitated the partially purified proteins via anti-Flag antibodies in the presence of recombinant RPA plus or minus recombinant PKA (Figure 1C). First, we confirmed that immunoprecipitates had similar amounts of active AID, as

Figure 1. zAID Interacts with RPA In Vitro (A) Phylogenetic comparison of AID from Homo sapiens (human), Pan troglodytes (chimpanzee), Mus musculus (mouse), Xenopus laevis (amphibian), Gallus gallus (chicken), Danio rerio (zebrafish), Takifugu rubripes (pufferfish), and Ictalurus punctatus (catfish). Arrows indicate S38 in mammals, birds, and amphibian AID and aspartate 44 in fish AID. The conserved PKA site in tetrapods is indicated in bold and the conserved aspartic acid in telosts in red. (B) The zAIDWT, zAIDD44A, and zAIDG42S,D44A are represented with the mutated residues indicated. (C) Mutant-tagged zAID proteins were expressed in 293T cells, and nuclear extracts obtained were treated (+) or not treated ( ) with PKA and immunoprecipitated. The immunoprecipitates were washed and assayed for zAID-RPA interactions by western blotting with anti-RPA antibodies. Presence of AID in the immunoprecipitation reaction was estimated by ssDNA deamination as previously described (Chaudhuri et al., 2003) and presented at the bottom of (C). (D) Flag-HA-tagged mAID and Flag-HA-tagged zAIDWT, zAIDD44A, and zAIDG42SD44A were expressed in 293 cells and partially purified by immunoprecipitation, incubated with PKA in the presence of g-32P-ATP, and analyzed by SDS-PAGE followed by autoradiography (top panel). Presence of Flag-HA epitope-tagged mAID or zAID proteins in immunoprecipitates was determined via western blotting with anti-HA antibodies (bottom panel).

evidenced by ssDNA cytidine deamination activity (Figure 1C, bottom). Interaction of RPA with zAID then was assessed via SDS-PAGE followed by western blotting with anti-RPA antibodies. Notably, we observed constitutive zAID interaction with RPA whether or not the zAID was preincubated with recombinant PKA (Figure 1C; Figures S1 and S2A). In contrast, we did not find readily detectable interaction of zAIDD44A with RPA, either in the presence or absence of PKA and ATP (Figure 1C; Figures S1 and S2A). As a control, mutation of adjacent residues, including glycine 42 to serine (zAIDG42S) or proline

286 Molecular Cell 32, 285–291, October 24, 2008 ª2008 Elsevier Inc.

Molecular Cell AID Regulation by Phosphorylation

43 to alanine (zAIDP43A), did not appear to alter zAID binding to RPA, indicating the specificity of D44 for this activity (Figure S2A). Finally, following ectopic expression in B cells, mAIDWT and zAIDWT interact with RPA, whereas mAIDS38A and zAIDD44A do not (Figure S2D). We next generated and expressed in HEK293 cells a secondsite mutant of zAIDD44A in which glycine 42 was mutated to serine to form the zAIDG42S,D44A double mutant (Figure 1B). The G42S mutation reconstitutes a PKA phosphorylation site at a position corresponding to mAID S38 (Figure 1B). Immunoprecipitated zAIDG42S,D44A retained ssDNA deamination activity similar to that of zAIDWT or zAIDD44A (Figure 1C, lower panel). When partially purified zAIDWT, zAIDD44A, or zAIDG42SD44A was incubated with g-32P ATP in the presence of recombinant PKA and analyzed via SDS-PAGE, we observed that zAIDG42S,D44A clearly was phosphorylated by PKA, whereas zAIDWT and zAIDD44A were not detectably phosphorylated (Figure 1D, mAIDWT is shown as a control; Figures S3A and S3B). In these experiments, zAIDG42S,D44A appeared less phosphorylated than the mAIDWT control protein, potentially reflecting roles for other mAID sequences in enhancing PKA phosphorylation. Strikingly, while untreated zAIDG42S,D44A failed to show RPA interaction, PKA phosphorylation of this double mutant led to clear RPA interaction (Figure 1C; Figures S1 and S2A). RPA-Dependent zAID dsDNA Deamination Activity Because zAID constitutively associates with RPA and zAIDD44A lacks ability to associate with RPA (Figure 1C), we asked whether zAID, in the presence of RPA, also showed constitutive ability to deaminate transcribed dsDNA SHM substrates in vitro and whether such ability was dependent on integrity of D44. As expected, partially purified (from HEK293 cells) mAIDWT, mAIDS38A, zAIDWT, and zAIDD44A proteins demonstrated similar ability to deaminate ssDNA, again confirming that the respective mutations do not affect catalytic activity (Figure 2A). We then assayed each protein, following preincubation with recombinant PKA and ATP, for ability to deaminate a T7 RNA polymerase-transcribed dsDNA SHM substrate in the presence of RPA (Chaudhuri et al., 2004; see Figure S4 for schematic of the assay). Both the mAIDWT and zAIDWT were active in these assays, although zAIDWT appeared to have somewhat less activity (Figure 2B). However, similar to mAIDS38A versus mAIDWT, zAIDD44A, despite having normal catalytic activity on ssDNA, showed greatly impaired transcription-dependent dsDNA deamination activity relative to zAIDWT (Figure 2B). We next compared activities of partially purified zAIDWT, zAIDD44A, and zAIDG42S,D44A proteins in the transcription-dependent dsDNA cytidine deamination assay, before and after PKA treatment. The partially purified AID proteins used for these assays again had similar ssDNA deamination activities (Figure S5A) and were used in roughly similar amounts in each assay (Figure S5B). The presence or absence of PKA had no affect on zAIDWT dsDNA deamination activity (Figure 2C) or the lack of this activity observed for zAIDD44A (Figure 2C). In contrast, preincubation with PKA led to a striking increase in dsDNA deamination activity of zAIDG42S,D44A to levels substantially above the background observed in the absence of treatment (Figure 2C). In these assays, the activity of PKA-treated

Figure 2. zAID Deaminates Transcribed RGYW-Rich dsDNA in an RPA-Dependent Manner (A and B) mAIDWT, mAIDS38A, zAIDWT, and zAIDD44A were expressed in 293T cells and partially purified (through DEAE cellulose and glycerol gradient). The proteins were assayed for ssDNA deamination activity without PKA addition (A) or RPA-dependent dsDNA deamination activity after PKA addition (B). For (A) and (B), values represent the mean of three experiments; error bars represent SD. (C) Indicated WT and mutant AID proteins were expressed in 293T cells, partially purified as above, and standardized for ssDNA deamination activity. Subsequently, 50 mg of each partially purified AID protein, as present in the peak fraction of the glycerol gradient, was incubated with or without 50 units of PKA for 30 min at 37 C, followed by assay for RPA-dependent dsDNA deamination activity. Cytidine deamination activity for each protein was adjusted by subtracting the background value obtained with similarly processed cell extracts not transfected with AID. Deamination acivity is plotted as a % of total counts added. Background in individual assays varied from 0.3% to 1.5%. For (A), (B), and (C), values represent the mean from three experiments; error bars represent SD.

zAIDG42S,D44A was not fully restored to that of zAIDWT (Figures 1C and 1D). However, these findings make the clear and important qualitative point that the second-site G42S mutation simultaneously allowed the AIDD44A to serve as a substrate for PKA phosphorylation, to acquire PKA-dependent ability to bind RPA, and to acquire PKA-dependent dsDNA deamination activity.

Molecular Cell 32, 285–291, October 24, 2008 ª2008 Elsevier Inc. 287

Molecular Cell AID Regulation by Phosphorylation

Rescue of Impaired CSR of the zAIDD44A Protein by a Second-Site Mutation To assess effects of different AID mutations on IgH CSR, we retrovirally introduced HA-tagged zAIDWT, HA-tagged zAIDD44A, or HA-tagged zAIDG42,D44A into anti-CD40 plus IL-4-stimulated, AID-deficient mouse B cells and assayed ability of introduced AID proteins to rescue class switching to IgG1(Fagarasan et al., 2001). HA-tagged zAID proteins were employed because there are no available antibodies that react well with zAID in total cell extracts. Retrovirally expressed WT and mutant zAID proteins were expressed at roughly similar levels in B cells (Figure 3A; Figure S2C) and in HEK293 cells (data not shown). As expected, introduction of zAIDWT resulted in a substantial rescue of class switching to IgG1 to levels similar to those obtained with mAIDWT (Figures 3B and 3C; data not shown). In contrast, CSR levels obtained via introduction of zAIDD44A were, on average, about 40%–45% those of zAIDWT (based on six independent experiments), indicating that D44 integrity is required for optimal zAID CSR activity (Figures 3B and 3C; Figure S2B). This high residual level of CSR supported by zAIDD44A, along with its apparently robust activity in GCV and SHM in DT40 chicken cells (Chatterji et al., 2007), supports the notion that zAID also can access transcribed DNA via RPA-independent mechanisms (see Supplemental Discussion). Strikingly, the zAIDD44A,G42S double mutant showed significantly increased CSR activity compared with zAIDD44A, with levels that were, on average, similar to zAIDWT (Figures 3B and 3C; Figure S2B). A zAIDG42S mutant had CSR activity similar to that of zAIDWT, consistent with the AIDG42S mutation functioning to rescue the AIDD44A mutation, as opposed to simply activating overall CSR activity of AIDWT (Figure S2B). We conclude that introduction of the second-site G42S mutation into the zAIDD44A mutant restored essentially normal ability of the zAIDD44A mutant to catalyze CSR in vivo under assayed conditions. Phosphomimetic Mutation of mAID S38 Phosphorylation We confirmed prior studies that showed that an mAID mutant in which S38 was replaced with aspartic acid, to potentially provide a mimetic of constitutive phosphorylation, failed to generate normal CSR activity (McBride et al., 2006; Shinkura et al., 2007; data not shown). However, only positive results are clearly interpretable in such a study, as any amino acid replacement may negatively alter protein structure and function by various effects. Based on our finding that zAID employs D44 as a mimetic of mAID S38 phosphorylation, we predicted that replacement of mAID T40 (which corresponds in position to zAID D44; Figure 4A) might generate an mAID protein that functions independently of S38 phosphorylation and rescue activities lost in the context of the mAID S38A mutation. A purified mAIDS38A,T40D double mutant protein had similar ssDNA deamination activity as AIDWT, confirming retention of basic catalytic activity (Figure 4B). Strikingly, the mAIDS38AT40D protein bound RPA (Figures S7A and S7B) and catalyzed deamination of T7 polymerase-transcribed dsDNA SHM substrates even more robustly than PKA phosphorylated mAIDWT (Figure 4C), with both activities occurring independent of PKA phosphorylation (Figure 4C; Figure S7B). In agreement with restored biochemical activities, retrovirally introduced mAIDS38AT40D supported CSR in mouse B cells at

Figure 3. CSR Activity of zAID Mutants (A and B) AID-deficient B cells stimulated with anti-CD40 and IL-4 were infected with retrovirus expressing tagged zAIDWT, zAIDD44A, or zAIDG42S, D44A. Expression of HA-tagged zAID proteins were analyzed by loading 50 mg of total B cell extracts on an SDS-polyacrylamide gel followed by western blotting using anti-HA antibodies (A), and the level of CSR to IgG1 was evaluated by flow cytometry (B) as previously described (Basu et al., 2005). (C) The percentage of GFP-positive cells that had undergone CSR to IgG1 from multiple experiments is indicated as plotted. Individual sets of experiments are represented by different symbols. In (C), the mean of all the experiments of a particular zAID protein is represented by a line.

levels similar to those of mAIDWT (Figure 4D; Figures S7C and S7D). Finally, an mAIDS38G/T40D mutant also had CSR similar to that of mAIDWT (data not shown). Taken together, these findings indicate that the mAID T40D mutation serves as a mimetic of mAID phosphorylation on S38. Thus, using an evolutionary approach, we have identified a dominantly active mutant of mAID in the context of S38 phosphorylation. DISCUSSION The model that mAID CSR activity is augmented via S38 phosphorylation was challenged based on observations that zAID lacks a PKA site equivalent to S38, but still supports CSR in AID-deficient mouse B cells (Shinkura et al., 2007; Pham et al., 2008). However, we now show that the D44 residue in zAID mimics S38 phosphorylation in mAID. Thus, constitutive zAID association with RPA, constitutive zAID activity in the in vitro transcription-dependent dsDNA deamination assay, and normal zAID CSR activity in B cells depend fully on integrity of the zAID

288 Molecular Cell 32, 285–291, October 24, 2008 ª2008 Elsevier Inc.

Molecular Cell AID Regulation by Phosphorylation

modest increases in AID levels have disproportionately greater effects on AID mutator activities, such as promoting translocations, than on CSR per se (Dorsett et al., 2008; Teng et al., 2008). Therefore, evolution of CSR might have selected for evolutionary fixation of posttranslational control mechanisms to more tightly regulate AID activity. In this regard, AID regulation via S38 phosphorylation in tetrapods may have evolved from constitutively active AID in bony fish, or, alternatively, the two AIDs may have evolved from a common AID ancestor, as potentially suggested by both a classical PKA site and an adjacent aspartate in a dogfish (elasmobranch) AID EST (Conticello et al., 2005). The importance of fine tuning AID activity via S38 phosphorylation is supported by the growing list of posttranscriptional AID regulatory mechanisms, which also include control of AID mRNA stability by microRNA155 (Dorsett et al., 2008; Teng et al., 2008) and regulation of AID levels via nuclear transport and proteosomal degradation (Aoufouchi et al., 2008; McBride et al., 2004). Figure 4. mAID ‘‘PhosphoMimetic’’ Mutants (A) The mAIDWT, mAIDS38A, and mAIDS38AT40D are represented with the mutated residues indicated. (B) Partially purified AID protein indicated were assayed for ssDNA deamination activities. (C) The RPA- and transcription-dependent dsDNA deamination activities of AIDWT and AIDS38AT40D proteins were analyzed in the presence (+) or absence ( ) of PKA as previously described (Basu et al., 2005). (D) CSR activities of mAIDWT, mAIDS38A, and mAIDS38At40D mutants in ex vivo complementation assay. Assays are presented in the same format as corresponding assays in Figures 2 and 3 and Figure S6. In (B), (C), and (D), values represent the mean from three experiments; error bars represent SD.

D44 residue, as all are impaired in zAIDD44A even though the mutant protein retains normal catalytic activity on ssDNA. Remarkably, the zAIDD44A,G42S double mutant, in which we have reconstituted a PKA site within zAIDD44A, can be phosphorylated by PKA and, moreover, has PKA-dependent RPA association, PKA-dependent dsDNA deamination activity, and in vivo CSR activity equivalent to that of zAIDWT. The ability of this second-site zAIDD44A mutation to simultaneously restore all these biochemical and in vivo activities provides compelling evidence for their linkage. Furthermore, based on our zAID findings, we generated an mAIDS38A,T40D mutant that functions as a constitutively S38-phosphorylated mimetic in both biochemical and CSR assays. Together, our current mAID and zAID studies strongly support a significant role for S38 phosphorylation and RPA association with respect to AID function. Moreover, our ability to generate mAID mutants that mimic ‘‘constitutive’’ AID activation via S38 phosphorylation should allow ‘‘knockin’’ approaches to firmly elucidate roles of this form of AID regulation in vivo. Bony fish undergo SHM, but not CSR, whereas amphibians and higher tetrapods undergo both processes. Notably, Xenopus S regions do not form R loops like S regions of higher tetrapods, and appear to rely on abundant SHM motifs to target AID-induced DSBs. Thus, CSR may have evolved from SHM through a process that, in large part, required just the evolution of S regions as specialized AID targets for generating DSBs (Chaudhuri et al., 2007). DSBs are a dangerous form of cellular DNA damage, leading to cell death or oncogenic deletions and translocations if not properly repaired (Povirk, 2006). Moreover,

EXPERIMENTAL PROCEDURES In Vitro AID Phosphorylation Approximately 500 mg of 293T cell extract containing N-terminal Flag/HA epitope-tagged zAID was incubated with 50 units of recombinant PKA (Sigma) for 1 hr at 30 C in a buffer containing 20 mM Tris HCl (pH. 7.5), 80 mM NaCl, 1 mM DTT, 10 mM MgCl2, and 1 mM ATP. Extracts were immunoprecipitated with aFLAG and aAID antibodies and assayed by western blotting, as previously described (Basu et al., 2005). Antibodies and Reagents Sources of antibodies were: anti-RPA32 from Santa Cruz Biotechnology; anti-GFP from Clontech; anti-HA and anti-Flag from Sigma; and anti-AID as previously described (Chaudhuri et al., 2003). Anti-AID antibodies were used as previously described (Chaudhuri et al., 2003, 2004). Protein Preparation and Deamination Assays Ectopically expressed mAID or N-terminal Flag/HA epitope-tagged zAID were purified from 293 cells (Chaudhuri et al., 2003, 2004). Deamination assays on ssDNA or dsDNA containing RGYW motifs were as previously described (Chaudhuri et al., 2003, 2004). Recombinant Retroviral Production and Infection of B Cells Mutant AIDs were generated by mutation of AID open reading frame (ORF) cloned in pcDNA3.0 (Invitrogen) using site-directed mutagenesis (Stratagene). Mutant AID ORFs were subcloned in pMX-IRES-GFP with an N-terminal Flag/ HA epitope tag (Fagarasan et al., 2001). The plasmid was transfected into virus packaging cell line (phoenix cells; Orbigen) using the Trans-It reagent (Mirus). The virus was mixed with preactivated B cells (with anti-CD40 and IL4), which were assayed for CSR after 72 hr as previously described (Basu et al., 2005). Flow Cytometric Analysis The antibodies PE-Cy5 conjugated anti-mouse CD45R (B220) (eBioscience) and PE anti-mouse IgG1 (A85-1) (BD PharMingen) were used for staining B cells for flow cytometric analyses performed with a BD FACScaibur. AID Immunoprecipitation Assay for In Vitro AID Phosphorylation and RPA Interaction Nuclear extracts from mouse or zebrafish Flag-HA epitope-tagged AID cDNA-transfected 293 cells or B cells were prepared as previously described (Chaudhuri et al., 2003, 2004). A 500 mg aliquot of extract was incubated with 50 units of purified PKA (Sigma) and [g-32P] ATP for 1 hr at 30 C, incubated with protein A-sepharose for 6 hr at 4 C in buffer A (20 mM Tris, pH 7.5, 1 mM DTT, 10 mM ZnCl2, 0.5 mM EDTA, 500 mM NaCl, and 10% glycerol), and immunoprecipitated with a mix of anti-Flag (Sigma) and anti-AID antibodies (Chaudhuri et al., 2003, 2004). Immunoprecipitates were washed with buffer

Molecular Cell 32, 285–291, October 24, 2008 ª2008 Elsevier Inc. 289

Molecular Cell AID Regulation by Phosphorylation

A and analyzed by SDS gel electrophoresis followed by autoradiography. To determine RPA interaction, immunoprecipitates were incubated in buffer A plus 150 mM NaCl, with whole 293 cell extracts. An additional 50 units of PKA was also added to each reaction in the presence of 10 mM MgCl2. The immunoprecipitate was washed in the same buffer and analyzed by western blotting using a-RPA32 antibodies. The presence of AID in the immunoprecipitation reactions was analyzed in some experiments by ssDNA deamination activity and in others by western blotting using aHA antibodies or aAID antibodies, as indicated. SUPPLEMENTAL DATA The Supplemental Data include Supplemental Discussion and seven figures and can be found with this article online at http://www.molecule.org/ supplemental/S1097-2765(08)00607-2.

Conticello, S.G., Thomas, C.J., Petersen-Mahrt, S.K., and Neuberger, M.S. (2005). Evolution of the AID/APOBEC family of polynucleotide (deoxy)cytidine deaminases. Mol. Biol. Evol. 22, 367–377. Di Noia, J.M., and Neuberger, M.S. (2007). Molecular mechanisms of antibody somatic hypermutation. Annu. Rev. Biochem. 76, 1–22. Dickerson, S.K., Market, E., Besmer, E., and Papavasiliou, F.N. (2003). AID mediates hypermutation by deaminating single stranded DNA. J. Exp. Med. 197, 1291–1296. Dorsett, Y., McBride, K.M., Jankovic, M., Gazumyan, A., Thai, T.H., Robbiani, D.F., Di Virgilio, M., San-Martin, B.R., Heidkamp, G., Schwickert, T.A., et al. (2008). MicroRNA-155 suppresses activation-induced cytidine deaminasemediated Myc-Igh translocation. Immunity 28, 630–638. 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.

ACKNOWLEDGMENTS

Goodman, M.F., Scharff, M.D., and Romesberg, F.E. (2007). AID-initiated purposeful mutations in immunoglobulin genes. Adv. Immunol. 94, 127–155.

We thank David Schatz, Jayanta Chaudhuri, Andrew Franklin, Jing Wang, Vivian Choi, Abhishek Datta, and Ting Ting Zhang for evaluation of the manuscript. We thank Tasuku Honjo for AID-deficient mice, and Vasco Barretto and Michel Nussenzweig for a retroviral construct expressing WT zebrafish AID. Tiffany Borjeson and Lisa Aquaviva provided technical assistance in maintenance of AID-deficient mice. U.B. was a fellow of the Irvington Institute of Immunology program of the Cancer Research Institute, and is currently a Special Fellow of the Leukemia and Lymphoma Society of America. This work was supported by a National Institutes of Health grant to F.W.A., who is an investigator of the Howard Hughes Medical Institute.

Honjo, T., Kinoshita, K., and Muramatsu, M. (2002). Molecular mechanism of class switch recombination: linkage with somatic hypermutation. Annu. Rev. Immunol. 20, 165–196.

Received: March 18, 2008 Revised: July 15, 2008 Accepted: August 20, 2008 Published: October 23, 2008

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.

REFERENCES

Odegard, V.H., and Schatz, D.G. (2006). Targeting of somatic hypermutation. Nat. Rev. Immunol. 6, 573–583.

Aoufouchi, S., Faili, A., Zober, C., D’Orlando, O., Weller, S., Weill, J.C., and Reynaud, C.A. (2008). Proteasomal degradation restricts the nuclear lifespan of AID. J. Exp. Med. 205, 1357–1368. Barreto, V.M., Pan-Hammarstrom, Q., Zhao, Y., Hammarstrom, L., Misulovin, Z., and Nussenzweig, M.C. (2005). AID from bony fish catalyzes class switch recombination. J. Exp. Med. 202, 733–738.

McBride, K.M., Barreto, V., Ramiro, A.R., Stavropoulos, P., and Nussenzweig, M.C. (2004). Somatic hypermutation is limited by CRM1-dependent nuclear export of activation-induced deaminase. J. Exp. Med. 199, 1235–1244. McBride, K.M., Gazumyan, A., Woo, E.M., Barreto, V.M., Robbiani, D.F., Chait, B.T., and Nussenzweig, M.C. (2006). Regulation of hypermutation by activation-induced cytidine deaminase phosphorylation. Proc. Natl. Acad. Sci. USA 103, 8798–8803.

Pan-Hammarstrom, Q., Zhao, Y., and Hammarstrom, L. (2007). Class switch recombination: a comparison between mouse and human. Adv. Immunol. 93, 1–61. Pasqualucci, L., Kitaura, Y., Gu, H., and Dalla-Favera, R. (2006). PKAmediated phosphorylation regulates the function of activation-induced deaminase (AID) in B cells. Proc. Natl. Acad. Sci. USA 103, 395–400.

Basu, U., Chaudhuri, J., Alpert, C., Dutt, S., Ranganath, S., Li, G., Schrum, J.P., Manis, J.P., and Alt, F.W. (2005). The AID antibody diversification enzyme is regulated by protein kinase A phosphorylation. Nature 438, 508–511.

Pham, P., Smolka, M.B., Calabrese, P., Landolph, A., Zhang, K., Zhou, H., and Goodman, M.F. (2008). Impact of phosphorylation and phosphorylation-null mutants on the activity and deamination specificity of activation-induced cytidine deaminase. J. Biol. Chem. 283, 17428–17439.

Basu, U., Chaudhuri, J., Phan, R.T., Datta, A., and Alt, F.W. (2007). Regulation of activation induced deaminase via phosphorylation. Adv. Exp. Med. Biol. 596, 129–137.

Povirk, L.F. (2006). Biochemical mechanisms of chromosomal translocations resulting from DNA double-strand breaks. DNA Repair (Amst.) 5, 1199–1212.

Bransteitter, R., Pham, P., Scharff, M.D., and Goodman, M.F. (2003). Activation-induced cytidine deaminase deaminates deoxycytidine on singlestranded DNA but requires the action of RNase. Proc. Natl. Acad. Sci. USA 100, 4102–4107. Chatterji, M., Unniraman, S., McBride, K.M., and Schatz, D.G. (2007). Role of activation-induced deaminase protein kinase a phosphorylation sites in Ig gene conversion and somatic hypermutation. J. Immunol. 179, 5274–5280. 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. Chaudhuri, J., Khuong, C., and Alt, F.W. (2004). Replication protein A interacts with AID to promote deamination of somatic hypermutation targets. Nature 430, 992–998. Chaudhuri, J., Basu, U., Zarrin, A., Yan, C., Franco, S., Perlot, T., Vuong, B., Wang, J., Phan, R.T., Datta, A., et al. (2007). Evolution of the immunoglobulin heavy chain class switch recombination mechanism. Adv. Immunol. 94, 157–214.

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). Activationinduced cytidine deaminase (AID) deficiency causes the autosomal recessive form of the hyper-IgM syndrome (HIGM2). Cell 102, 565–575. Shinkura, R., Okazaki, I.M., Muto, T., Begum, N.A., and Honjo, T. (2007). Regulation of AID function in vivo. Adv. Exp. Med. Biol. 596, 71–81. Shinkura, R., Tian, M., Smith, M., Chua, K., Fujiwara, Y., and Alt, F.W. (2003). The influence of transcriptional orientation on endogenous switch region function. Nat. Immunol. 4, 435–441. Stavnezer, J., and Amemiya, C.T. (2004). Evolution of isotype switching. Semin. Immunol. 16, 257–275. 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.

290 Molecular Cell 32, 285–291, October 24, 2008 ª2008 Elsevier Inc.

Molecular Cell AID Regulation by Phosphorylation

Teng, G., Hakimpour, P., Landgraf, P., Rice, A., Tuschl, T., Casellas, R., and Papavasiliou, F.N. (2008). MicroRNA-155 is a negative regulator of activation-induced cytidine deaminase. Immunity 28, 621–629. Tian, M., and Alt, F.W. (2000). Transcription-induced cleavage of immunoglobulin switch regions by nucleotide excision repair nucleases in vitro. J. Biol. Chem. 275, 24163–24172. Wakae, K., Magor, B.G., Saunders, H., Nagaoka, H., Kawamura, A., Kinoshita, K., Honjo, T., and Muramatsu, M. (2006). Evolution of class switch recombination function in fish activation-induced cytidine deaminase, AID. Int. Immunol. 18, 41–47.

Wold, M.S. (1997). Replication protein A: a heterotrimeric, single-stranded DNA-binding protein required for eukaryotic DNA metabolism. Annu. Rev. Biochem. 66, 61–92. Yu, K., Chedin, F., Hsieh, C.L., Wilson, T.E., and Lieber, M.R. (2003). R-loops at immunoglobulin class switch regions in the chromosomes of stimulated B cells. Nat. Immunol. 4, 442–451. Zarrin, A.A., Alt, F.W., Chaudhuri, J., Stokes, N., Kaushal, D., Du Pasquier, L., and Tian, M. (2004). An evolutionarily conserved target motif for immunoglobulin class-switch recombination. Nat. Immunol. 5, 1275–1281.

Molecular Cell 32, 285–291, October 24, 2008 ª2008 Elsevier Inc. 291