The Journal of Immunology

Activation-Induced Splice Variants Are Defective Because of the Lack of Structural Support for the Catalytic Site

Febe van Maldegem, R. Aarti Jibodh, Remco van Dijk, Richard J. Bende, and Carel J. M. van Noesel

Recently, conflicting results were reported on the hypermutation activity of activation-induced cytidine deaminase (AID) splice variants. With the generation of single point , we studied the structure-function relationship of the amino acids that are commonly absent from all described splice variants. The results from this analysis pointed to several amino acids that are required for class switch recombination (CSR), without perturbing cellular localization or nucleocytoplasmic shuttling. A defect in deaminase activity was found to underlie this CSR deficiency. Interestingly, the most debilitating mutations concentrated on hydrophobic amino acids, suggesting a structural role for this part of the . Indeed, by generating homologous replacements, CSR activity could be restored. These results are in agreement with recent reports on the protein structure of the AID homolog APOBEC3G, suggesting a similar protein composition. In addition, the findings underscore that AID splice variants are unlikely to have preservation of catalytic activity. The Journal of Immunology, 2010, 184: 2487–2491.

y deaminating cytidines to in the Ig variable (IgV) site of exon 4. Two of these splice variants, designated AID-DE4 and switch regions, activation-induced cytidine de- and AID-ivs3, lack the entire C terminus encoded by exons 4 and 5, B aminase (AID) generates the key lesions required for somatic whereas a third isoform (named AID-DE4a) has a relatively small hypermutation (SHM) and class switch recombination (CSR), two deletion of the first 10 aa encoded by exon 4. These 10 aa, absent processes pivotal in the generation of high-affinity Abs (1). Although from all three splice variants, encode the start of the previously instrumental in adaptive immunity, these genome-modifying pro- designated “pseudocatalytic site”—in analogy to cesses entail the risk of unwanted DNA damage, which is reflected cytidine deaminase—and constitute a part of the APOBEC-like C- in the various translocations involving the IgV genes and switch re- terminal domain (CTD) of AID (9, 10). Although the C-terminal gions, characteristic for B cell non-Hodgkin’s lymphoma. In addition, amino acids of AID, encoded by exon 5, were found to contain the aberrant targeting and deregulated expression of AID was described nuclear export signal (NES) and are absolutely required for CSR but in human B cell lymphomas (2, 3). Indeed, several mouse models dispensable for SHM, very little is known about the function of the confirmed the oncogenic potential of AID (2). These findings have remaining part of the CTD of AID (11, 12). Studies on hyper-IgM raised fundamental questions with respect to regulation of expression, patients have described three point mutations leading to single targeting, posttranslational modifications, and the structure-function amino acid changes in this region (i.e., M139V,F151S, and R174S); relationship of AID. all three lacking both CSR and SHM activity (13). More recently, Studies related to AID expression have reported the existence of T140 was found to be subjected to phosphorylation, and of a number of AID splice variants (3–6). Considering a potential role this amino acid also led to a reduced SHM and CSR in vivo (14). By of alternative splicing in the regulation of AID, functional studies of functionally analyzing the effect of mutations introduced in the first these isoforms were desired. Recently, we and others reported that 10 aa of exon 4, we have aimed to gain more insight in the structure- three of these isoforms completely lack CSR activity (7, 8). Con- function relationship of this part of AID. Interestingly, among these cordantly, we found that recombinant isoform were in- are three putative phosphorylation substrates; two tyrosines (Y144 capable of deaminating a synthetic oligonucleotide substrate invitro and Y146) and one threonine (T150). and that this deficiency was accompanied by a disturbed cellular localization. Alternative splicing takes place at the intersection of Materials and Methods exons 3 and 4, reflecting the intrinsic weakness of the splice acceptor AID mutants and vectors AID wild-type (WT) and mutants were cloned into the following vectors: for the generation of into LZRS-linker-IRES-YFP [a gift from Department of Pathology, Academic Medical Center, University of Amsterdam, Am- H. Spits, Department of Cell Biology and Histology, Academic Medical sterdam, The Netherlands Center, Amsterdam, The Netherlands (15)]; for the nucleocytoplasmic Received for publication September 22, 2009. Accepted for publication December shuttling experiments into pcDNA3.1/CT-GFP-topo (Invitrogen, Carlsbad, 23, 2009. CA); and for generation of GST-AID recombinant protein into pGEX5X-1 Address correspondence and reprint requests to Dr. Carel J. M. van Noesel, Depart- (GE Healthcare, Little Chalfont, U.K.). ment of Cell Biology and Histology, Academic Medical Center, University of Am- Human AID-WT was obtained from Ramos cells by RT-PCR. The gen- sterdam, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands. E-mail address: eration of AID mutants was established using site-directed mutagenesis: [email protected] first, two fragments were created by PCR (with Phusion polymerase; Abbreviations used in this paper: AID, activation-induced cytidine deaminase; CSR, Finnzymes, Espoo, Finland), combining the AID-fw primer with a mutant class switch recombination; CTD, C-terminal domain; LMB, leptomycin B; NES, reverse primer or a mutant forward primer with an AID-rev primer (Table I). nuclear export signal; SHM, somatic hypermutation; WT, wild-type. These fragments were then used as input for a second PCR with the AID-fw and -rev primers, creating full-length mutant AID. AID-fw/rev1 primers Copyright Ó 2010 by The American Association of Immunologists, Inc. 0022-1767/10/$16.00 were used for direct cloning into pcDNA3.1/CT-GFP-topo, resulting in www.jimmunol.org/cgi/doi/10.4049/jimmunol.0903102 2488 AID SPLICE VARIANTS LACK SUPPORT FOR THE CATALYTIC SITE a fusion protein of AID and GFP connected through the linker sequence 60 U/ml DNase I (Roche Diagnostics). GST-fusion proteins were purified QGQFCRYPAQWRPLESR. The AID-fw and -rev2 primers were used for from the lysate with GSTrap HP 1-ml columns (GE Healthcare) and di- cloning into LZRS and pGEX, of which the latter primer was equipped with alyzed in 20 mM Tris-HCl (pH 7.5), 10 mM NaCl, 0.1 mM DTT, and 50% a NotI site to enable easy transfer to the target vectors. The PCR products glycerol. Protein concentrations were measured on a NanoDrop ND-1000 were cloned in Topo 2.1, verified using M13 primer-mediated sequencing, Spectrophotometer (Isogen lifescience, IJsselstein, The Netherlands) and and used in subsequent digestion and ligation into the target vectors: confirmed by Coomassie staining of DNA-PAGE gels loaded with adjusted BamHI/NotI for ligation into the LZRS and EcoRI/NotI for pGEX5X-1. In protein amounts. the resulting GST-AID fusion proteins, GST and AID were linked by Catalytic activity was assessed using an invitro deaminase assay, in which IEGRGIPEFALLDTT. 100 fmol of a FAM-labeled 60-nt substrate, containing one cytidine in a RGYW-hotspot motif (FAM, 59-TAAAGGTGAAGAGAGGAGAGAGA- production and titration AGTAAGCTGAAGAGAGAGAAGGAAGAGAGTGAAGGAG-39), was m LZRS vectors containing the AID-encoding sequences were transfected with incubated overnight at 37˚C with 1–2 g purified protein in 50 mM Tris (pH m m Fugene6 transfection reagent(Roche Diagnostics,Almere,The Netherlands) 7.5), 100 mM NaCl, and 2 M MgCl in a volume of 10 l. Uridine excision into Phoenix Ecotropic packaging cells (American Type Culture Collection, by recombinant uracyl DNA-glycosylase (New England Biolabs, Ipswich, Manassas, VA) to produce supernatant containing the retrovirus. The virus MA) and breakage of the abasic site by NaOH heat treatment were performed titer wasmeasuredby applyingthe supernatants to3T3cellsin serial dilutions as previously described by Larijani et al. (16, 17), with the modification that 1 m and in the presence of 10 mg/ml polybrene (Sigma-Aldrich, Zwijndrecht, g of RNase A was added to the reaction (18). Deamination The Netherlands), with subsequent analysis of the percentage of infected products were separated by DNA-PAGE and visualized using the Typhoon cells by flow cytometry. Trio imager (GE Healthcare). CSR assay AID-knockout mouse splenocytes were depleted from erythrocytes and Results sorted for B cells with anti-B220 magnetic beads (Miltenyi Biotec, AID mutants Utrecht, The Netherlands). Cells were cultured in IMDM with 10% FCS, 1% penicillin/streptomycin, 1% glutamine, BSA (100 mg/ml), 0.04‰ We generated AID point mutations (Table I) within the area of the 2-ME, 25 mg/ml LPS (Sigma-Aldrich), and 10 ng/ml murine IL-4 (Pe- deletion found in splice-variant AID-DE4a, ranging from D143 to proTech, Rocky Hill, NJ). The next day, cells were transduced with equal V152 (i.e., the first 10 aa encoded by exon 4), as shown in Fig. 1A. titers of the AID-WT, AID mutants, or empty vector as negative control The F151S mutation previously found in a hyper-IgM patient and in RetroNectin-coated plates (TaKara Bio, Shiga, Japan). On day 3 after transduction, cells were analyzed by flow cytometry, based on YFP ex- mapping to this part of AID was included as well. pression and immunofluorescence of monoclonal anti-mouse CD45R/ B220-APC (RA3-6B2; BD Biosciences, San Jose, CA) and polyclonal Functional analysis of AID point mutants goat anti-mouse IgG1-PE (Southern Biotechnology Associates, Bir- mingham, AL). To measure CSR activity, retroviral reconstitution of AID-knockout mouse B cells and stimulation with LPS and IL-4 was used to Cytoplasmic localization specifically induce CSR to IgG1. As with AID-WT, some of the CT-GFP-topo vectors were transfected into HEK293 cells with Fugene 6 mutants were able to restore CSR, either completely (F145A) or transfection reagent according to manufacturer’s instruction. After 2 d, partially (D143A, C147A, N149A, and T150A) (Fig. 1B,1C). Five half of the cells were incubated with 10ng/ml Leptomycin B (Sigma- mutants appeared completely defective for CSR activity (i.e., Aldrich, St. Louis, MO), for 3–5 h. AID-GFP was visualized in live cells on a Leica DM5000B microscope (Leica Microsystems, Deerfield, IL) Y144A, Y146A, W148A, F151A, and V152A). Several of the at 320 magnification, and recorded with a Leica DFC500 camera (Leica mutants were tested for inhibitory effects on AID-WT, using double Microsystems). transduction experiments, but dominant negativity was not found In vitro deaminase assay (data not shown). Catalytic competence was assessed for several of the mutants using pGEX5X-1 vectors were transformed into BL21 E. coli cells (GE Health- care) and grown until OD 0.6–0.8, at which the bacteria were induced with 1 an invitro deaminase assay, in which the presence of a 30-nt digestion mM IPTG, for 16 h at 16˚C. Concentrated bacteria suspensions were lysed product is a reflection of AID-mediated DNA deamination. Similar to by sonication in the presence of 100 mg/ml lysozyme (Sigma-Aldrich) and AID-WT, the mutants C147A, N149A, and T150A showed

Table I. Primers used for the generation of AID mutants

Mutant Forward Primer 59-39 Reverse Primer 59-39 AID-fw CTGGACACCACTATGGACAG AID-rev1 GAAGTCCCAAAGTACGAAATGC AID-rev2 TCGAGCGGCCGCCCTGGAAGTTGCTATCAAAGTC T140A CATCATGGCCTTCAAAGATTATTTTTAC GTAAAAATAATCTTTGAAGGCCATGATG D143A CATGACCTTCAAAGCTTATTTTTAC GCAGTAAAAATAAGCTTTGAAGGTC D143E CATGACCTTCAAAGAGTATTTTTAC GCAGTAAAAATACTCTTTGAAGGTC Y144A CAAAGATGCTTTTTACTGCTGG GCAGTAAAAAGCATCTTTGAAGG Y144F CCTTCAAAGATTTTTTTTACTGCTGG CCAGCAGTAAAAAAAATCTTTGAAGG F145A CAAAGATTATGCTTACTGCTGG GCAGTAAGCATAATCTTTGAAGG F145Y CCTTCAAAGATTATTATTACTGCTGG CCAGCAGTAATAATAATCTTTGAAGG Y146A GATTATTTTGCCTGCTGGAATAC GTATTCCAGCAGGCAAAATAATC Y146F GATTATTTTTTCTGCTGGAATAC GTATTCCAGCAGAAAAAATAATC C147A TTCAAAGATTATTTTTACGCCTGG TTCCAGGCGTAAAAATAATCTTTG W148A TTTACTGCGCGAATACTTTTGTAG AAAGTATTCGCGCAGTAAAAATAATC W148Y CTGCTATAATACTTTTGTAG CTACAAAAGTATTATAGCAG N149A TTTACTGCTGGGCTACTTTTG AGTAGCCCAGCAGTAAAAATAATC N149D TTTACTGCTGGGATACTTTTGTAG TTCTACAAAAGTATCCCAGCAG T150A CTGCTGGAATGCTTTTGTAG CTACAAAAGCATTCCAGCAG F151A CTGGAATACTGCTGTAGAAAACC GGTTTTCTACAGCAGTATTCCAG F151S TTTACTGCTGGAATACTTCTGTAG TTCTACAGAAGTATTCCAGCAG V152A GGAATACTTTTGCAGAAAACC GGTTTTCTGCAAAAGTATTCC The Journal of Immunology 2489

FIGURE 1. Functional analysis of AID mutants. A, The first 10 aa of exon 4 were each mutated to alanine for functional analysis. The hyper-IgM–mutant F151S was included as well (indicated with an asterisk). B, A representative experiment of CSR by reconstitution of AID-knockout B cells with AID-WT, + empty vector, and AID mutants. Depicted is the percentage of IgG1 cells among transduced cells. C, Relative CSR activity compared with AID-WT (set to 100%). Depicted is the mean of five measurements, with the SE of means illustrated in error bars. CSR activity was completely absent for the mutants of hydrophobic amino acids Y144A, Y146A, W148A, F151A, and V152A. D, In vitro deaminase activity measured for GST-coupled AID-WT and AID- mutants and GST-only as a control. Deamination at the cytidine in the RGYW hotspot is visible as the 30-nt product after treatment with recombinant UDG and subsequent alkaline cleavage. AID-WT, C147A, N149A, and T150A have preserved deaminase activity, whereas for GST, W148A, and F151S, no deaminase-product is evident. E, Cellular localization of GFP and AID-GFP fusion products, with and without treatment the CRM1 inhibitor of nuclear export LMB. All AID variants are by default localized in the cytoplasm but accumulate into the nucleus upon LMB treatment, demonstrating functional nucleocytoplasmic shuttling.

conversionofthesubstrate,whereasnoactivitywasfoundforW148A, by its C-terminal NES. This was demonstrated by incubation F151S, and V152A (Fig 1D). In the case of these three mutants, the with the CRM1-exportin inhibitor leptomycin B (LMB), lack of deaminating activity was in agreement with their null phe- resulting in entrapment of AID in the cell nucleus. To see notype for CSR. whether the catalytic defect of the mutants coincides with an Similar to APOBEC1, one of its homologs, AID has been aberrant cytoplasmic localization, as was previously found shown to shuttle between nucleus and cytoplasm, mediated for the splice variants, GFP-tagged AID mutants were 2490 AID SPLICE VARIANTS LACK SUPPORT FOR THE CATALYTIC SITE

sequence is confined to the most hydrophobic amino acids, which, when mutated to the neutral alanine, are defective for CSR and deaminase activity. By introducing point mutations, we have generated relatively subtle defects. This has revealed that the CSR deficiency of these mutants is not due to impairment in cellular localization or nucleocytoplasmic shuttling. When con- verting the hydrophobic amino acids to homologous hydropho- bic residues, the functionality was restored, indicating that these amino acids are important in maintaining structural integrity. FIGURE 2. Rescue of CSR activity by homologous mutation. CSR These results rule out a role for phosphorylation at Y144 and activity can be rescued when replacements are made with amino acids of Y146. In contrast, T150A showed only a partial defect in CSR, similar biochemical properties. which was not directly reflected in the cytidine deaminase assay. Hence, phosphorylation on this residue cannot be excluded, moreover, as similar results were recently reported for the expressed in HEK293 cells. AID-WT as well as all mutants phosphorylation site at T140 (14). showed normal cytoplasmic localization in untreated cells The 10-aa stretch is highly conserved in AID: 100% in other and nuclear accumulation upon LMB treatment, demon- vertebrates, such as mouse, rat, and the African clawed frog and to strating functional nucleocytoplasmic shuttling (Fig. 1E). a great extent in other cytidine deaminases, as is shown in the Interestingly, the mutants with the most severely defective phe- alignment with APOBEC2 and APOBEC3G in Fig. 3. On the notype corresponded to the most hydrophobic amino acids, sug- basis of the crystal structures of APOBEC2 and APOBEC3G, it gesting a function in maintaining structural integrity. To test this can be deduced that they constitute an a-helix (a5), positioned hypothesis, several additional mutants were generated in which parallel to the b-sheet, antiparallel to the a1-helix and opposite substitutions were made into amino acids with similar chemical to the side containing the DNA-binding and Zn-coordinating characteristics (i.e., D143E, Y144F, F145Y, Y146F, W148Y, and motifs (21–25). Chen, et al. (22) confirmed their APOBEC3G N149D). Indeed, these homologous mutants were able to rescue structure model by extensive functional mutant analysis. Among partial or complete functionality, as demonstrated by restoration of the numerous APOBEC3G mutants assayed for mutation fre- CSR (Fig. 2). quency in bacteria, five were analogous to the AID mutants presented in this study: F343A, C346A, W347A, F350A, and Discussion V351A in APOBEC3G, corresponded to Y144A, C147A, Recently, we showed that three of the reported AID splice variants W148A, F151A, and V152A in AID, respectively. Only one of encode truncated proteins that lack catalytic activity and have these five mutants displayed mutation activity detectable above abnormal cellular localization. Alternative splicing takes place background (i.e., C346A), whereas the other four lacked muta- downstream of exon 3, a sequence normally encoding a part of to tion activity, which matches with the outcome of our experi- the APOBEC-like C terminus. Although conserved throughout ments. The APOBEC3G a5-helix makes extensive stabilizing many of the APOBEC family members, the function of this hydrophobic contacts with the b-strand platform and the a1- domain is still largely unknown. The most C-terminal 10 aa helix, which together form a strong hydrophobic core, support- encode a functional NES, and mutation of this sequence resulted ing the catalytic site (21–23). Similarly, we can conclude that the in a defect in CSR, whereas SHM remained intact (11, 12). In hydrophobic amino acids of the a5-helixintheAPOBEC-likeC contrast, deletion of 10 aa from the upstream side of this domain, terminus of AID is of essential structural importance in enabling as with splice variant AID-DE4a, was sufficient to give a com- deaminase activity. The pivotal significance of the hydrophobic plete defect in catalytic activity in our experiments. By gener- core, as indicated by the single amino acid mutants, infers that ating artificial point mutants, we studied the function of this area the complete lack of this domain in the splice variants must have in more detail. Our results showed that the significance of this severe consequences for AID functionality (7).

1 1 22 hAID ------MDSLLMNRRKFLYQFKNVRWA K hAPOBEC3G-CTD ------MDPPTFTFNFNNEPWV R hAPOBEC2 MAQKEEAAVATEAASQNGEDLENLDDPEKLKELIELPPFEIVTGERLPANFFKFQFRNVEYS S : * ::*.* :

1 2 2' 2 23 75 FIGURE 3. Clustal-W sequence alignment of human hAID GRRETYLCYVVKRRDSATSFSLD--FGYLRNKN------GCHVELLFLRYISDWDLDPGRC hAPOBEC3G-CTD GRHETYLCYEVERMHNDTWVLLNQRRGFLCNQAPHKHGFLEGRHAELCFLDVIPFWKLDLDQD AID, the CTD of hAPOBEC3G and hAPOBEC2 (19, hAPOBEC2 GRNKTFLCYVVEAQGKGGQVQAS--RGYLEDEH------AAAHAEEAFFNTILP-AFDPALR 20). Asterisks indicate identical residues; colons and **.:*:*** *: . . . *:* :: . *.* *: * :* periods indicate conservation of strong and weak groups, respectively. The 10 aa subjected to mutation in 3 3 4 4 5 this study, are shaded gray. The predicted secondary 76 138 hAID YRVTWFTSWSPCYDCARHVADFLRGNPNLSLRIFTARLYFCEDRKAEPEGLRRLHRAGVQIAI structure of APOBEC3G-CTD is illustrated above the hAPOBEC3G-CTD YRVTCFTSWSPCFSCAQEMAKFISKNKHVSLCIFTARIY--DDQGRCQEGLRTLAEAGAKISI sequence (23). hAPOBEC2 YNVTWYVSSSPCAACADRIIKTLSKTKNLRLLILVGRLFMWEEPE-IQAALKKLKEA GCKLRI *.** :.* *** ** .: . : . :: * *:..*:: :: .*: * .** :: *

5 6 139 198 hAID MTFKDYFYCWNTFVEN---HERTFKAWEGLHENSVRLSRQLRRILLPLYEVDDLRDAFRTLG L hAPOBEC3G-CTD MTYSEFKHCWDTFVDH---QGCPFQPWDGLDEHSQDLSGRLRAILQ---NQEN------hAPOBEC2 MKPQDFEYVWQNFVEQEEGESKAFQPWEDIQENFLYYEEKLADILK------*. .:: : *:.**:: . .*:.*:.:.*: . :* ** The Journal of Immunology 2491

Acknowledgments and cytoplasm like mRNA editing catalytic polypeptide 1. Proc. Natl. Acad. Sci. USA 101: 1975–1980. We thank N. Blom for his expertise on protein mutagenesis and protein 12. McBride, K. M., V. Barreto, A. R. Ramiro, P. Stavropoulos, and modeling, D. Speijer for help with Typhoon imaging, A. Martin and M. M. C. Nussenzweig. 2004. Somatic hypermutation is limited by CRM1- Larijani for their help on the deaminase assay, H. Spits for the LZRS vector, dependent nuclear export of activation-induced deaminase. J. Exp. Med. 199: and T. Honjo for the AID-knockout mice. 1235–1244. 13. Durandy, A., N. Taubenheim, S. Peron, and A. Fischer. 2007. Pathophysiology of B-cell intrinsic immunoglobulin class switch recombination deficiencies. Adv. Disclosures Immunol. 94: 275–306. 14. McBride, K. M., A. Gazumyan, E. M. Woo, T. A. Schwickert, B. T. Chait, and The authors have no financial conflicts of interest. M. C. Nussenzweig. 2008. Regulation of class switch recombination and somatic mutation by AID phosphorylation. J. Exp. Med. 205: 2585–2594. 15. Heemskerk, M. H., B. Blom, G. Nolan, A. P. Stegmann, A. Q. Bakker, K. Weijer, References P. C. Res, and H. Spits. 1997. Inhibition of and promotion of natural killer 1. Di Noia, J. M., and M. S. Neuberger. 2007. Molecular mechanisms of antibody cell development by the dominant negative helix loop helix factor Id3. J. Exp. somatic hypermutation. Annu. Rev. Biochem. 76: 1–22. Med. 186: 1597–1602. 2. Okazaki, I. M., A. Kotani, and T. Honjo. 2007. Role of AID in tumorigenesis. 16. Larijani, M., and A. Martin. 2007. Single-stranded DNA structure and positional Adv. Immunol. 94: 245–273. context of the target cytidine determine the enzymatic efficiency of AID. Mol. 3. Albesiano, E., B. T. Messmer, R. N. Damle, S. L. Allen, K. R. Rai, and Cell. Biol. 27: 8038–8048. N. Chiorazzi. 2003. Activation-induced cytidine deaminase in chronic lympho- 17. Larijani, M., A. P. Petrov, O. Kolenchenko, M. Berru, S. N. Krylov, and cytic leukemia B cells: expression as multiple forms in a dynamic, variably sized A. Martin. 2007. AID associates with single-stranded DNA with high affinity and fraction of the clone. Blood 102: 3333–3339. a long complex half-life in a sequence-independent manner. Mol. Cell. Biol. 27: 4. McCarthy, H., W. G. Wierda, L. L. Barron, C. C. Cromwell, J. Wang, 20–30. K. R. Coombes, R. Rangel, K. S. Elenitoba-Johnson, M. J. Keating, and 18.Bransteitter,R.,P.Pham,M.D.Scharff,andM.F.Goodman.2003.Activa- L. V. Abruzzo. 2003. High expression of activation-induced cytidine deaminase tion-induced cytidine deaminase deaminates deoxycytidine on single-stranded (AID) and splice variants is a distinctive feature of poor-prognosis chronic DNA but requires the action of RNase. Proc.Natl.Acad.Sci.USA100: 4102– lymphocytic leukemia. Blood 101: 4903–4908. 4107. 5. Noguchi, E., M. Shibasaki, M. Inudou, M. Kamioka, Y. Yokouchi, 19. Subramaniam, S. 1998. The Biology Workbench—a seamless database and K. Yamakawa-Kobayashi, H. Hamaguchi, A. Matsui, and T. Arinami. 2001. analysis environment for the biologist. Proteins 32: 1–2. Association between a new polymorphism in the activation-induced cytidine 20. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: im- deaminase and atopic asthma and the regulation of total serum IgE levels. proving the sensitivity of progressive multiple sequence alignment through se- J. Allergy Clin. Immunol. 108: 382–386. quence weighting, position-specific gap penalties and weight matrix choice. 6. Oppezzo, P., F. Vuillier, Y. Vasconcelos, G. Dumas, C. Magnac, B. Payelle- Nucleic Acids Res. 22: 4673–4680. Brogard, O. Pritsch, and G. Dighiero. 2003. Chronic lymphocytic leukemia B 21. Chen, K. M., E. Harjes, P. J. Gross, A. Fahmy, Y. Lu, K. Shindo, R. S. Harris, and cells expressing AID display dissociation between class switch recombination H. Matsuo. 2008. Structure of the DNA deaminase domain of the HIV-1 re- and somatic hypermutation. Blood 101: 4029–4032. striction factor APOBEC3G. Nature 452: 116–119. 7. van Maldegem F., F. A. Scheeren, R. Aarti Jibodh, R. J. Bende, H. Jacobs, C. J. 22. Chen, K. M., N. Martemyanova, Y. Lu, K. Shindo, H. Matsuo, and R. S. Harris. van Noesel. 2009. AID splice variants lack deaminase activity. Blood 113: 1862– 1864; author reply 1864. 2007. Extensive mutagenesis experiments corroborate a structural model for the 8. Wu, X., J. R. Darce, S. K. Chang, G. S. Nowakowski, and D. F. Jelinek. 2008. DNA deaminase domain of APOBEC3G. FEBS Lett. 581: 4761–4766. Alternative splicing regulates activation-induced cytidine deaminase (AID): 23. Harjes, E., P. J. Gross, K. M. Chen, Y. Lu, K. Shindo, R. Nowarski, J. D. Gross, implications for suppression of AID mutagenic activity in normal and malignant M. Kotler, R. S. Harris, and H. Matsuo. 2009. An extended structure of the B cells. Blood 112: 4675–4682. APOBEC3G catalytic domain suggests a unique holoenzyme model. J. Mol. 9. Wedekind, J. E., G. S. Dance, M. P. Sowden, and H. C. Smith. 2003. Messenger Biol. 389: 819–832. RNA editing in mammals: new members of the APOBEC family seeking roles in 24. Holden, L. G., C. Prochnow, Y. P. Chang, R. Bransteitter, L. Chelico, U. Sen, the family business. Trends Genet. 19: 207–216. R. C. Stevens, M. F. Goodman, and X. S. Chen. 2008. Crystal structure of the 10. Jarmuz, A., A. Chester, J. Bayliss, J. Gisbourne, I. Dunham, J. Scott, and anti-viral APOBEC3G catalytic domain and functional implications. Nature 456: N. Navaratnam. 2002. An anthropoid-specific of orphan C to U RNA- 121–124. editing on 22. Genomics 79: 285–296. 25. Prochnow, C., R. Bransteitter, M. G. Klein, M. F. Goodman, and X. S. Chen. 11. Ito, S., H. Nagaoka, R. Shinkura, N. Begum, M. Muramatsu, M. Nakata, and 2007. The APOBEC-2 crystal structure and functional implications for the de- T. Honjo. 2004. Activation-induced cytidine deaminase shuttles between nucleus aminase AID. Nature 445: 447–451.