of August 4, 2016.This information is current as

HypermutationClass Switching and Reduced SomaticMutation Leading to Complete Loss of Ig Activation-Induced Cytidine DeaminaseSyndrome: A Spontaneous A Novel Mouse Model for the Hyper-IgM

Westerberg, Eva Severinson and Lena StrömLiz Torres, Elena M. Cortizas, Ramiro E. Verdun, Lisa S. Carin I. M. Dahlberg, Minghui He, Torkild Visnes, Magda

http://www.jimmunol.org/content/193/9/4732doi: 10.4049/jimmunol.1401242September 2014;

2014; 193:4732-4738; Prepublished online 24J Immunol 

MaterialSupplementary

2.DCSupplemental.htmlhttp://www.jimmunol.org/content/suppl/2014/09/24/jimmunol.140124

Referenceshttp://www.jimmunol.org/content/193/9/4732.full#ref-list-1

, 5 of which you can access for free at: cites 23 articlesThis article

Subscriptionshttp://jimmunol.org/subscriptions

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at:

Author Choice Author Choice option

The Journal of ImmunologyFreely available online through

Email Alertshttp://jimmunol.org/cgi/alerts/etocReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2014 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on August 4, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on August 4, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

A Novel Mouse Model for the Hyper-IgM Syndrome: ASpontaneous Activation-Induced Cytidine DeaminaseMutation Leading to Complete Loss of Ig Class Switching andReduced Somatic Hypermutation

Carin I. M. Dahlberg,*,1 Minghui He,†,1 Torkild Visnes,‡ Magda Liz Torres,*

Elena M. Cortizas,x Ramiro E. Verdun,x Lisa S. Westerberg,*,2 Eva Severinson,†,2 and

Lena Strom‡,2

We describe a spontaneously derived mouse line that completely failed to induce Ig class switching in vitro and in vivo. The mice

inherited abolished IgG serum titers in a recessive manner caused by a spontaneous G→A transition mutation in codon 112 of the

aicda gene, leading to an arginine to histidine replacement (AIDR112H). Ig class switching was completely reconstituted by

expressing wild-type AID. Mice homozygous for AIDR112H had peripheral B cell hyperplasia and large germinal centers in the

absence of Ag challenge. Immunization with SRBCs elicited an Ag-specific IgG1 response in wild-type mice, whereas AIDR112H

mice failed to produce IgG1 and had reduced somatic hypermutation. The phenotype recapitulates the human hyper-IgM (HIGM)

syndrome that is caused by point mutations in the orthologous gene in humans, and the AIDR112H mutation is frequently found in

HIGM patients. The AIDR112H mouse model for HIGM provides a powerful and more precise tool than conventional knockout

strategies. The Journal of Immunology, 2014, 193: 4732–4738.

An efficient immune response requires that the cells of theimmune system generate Igs of different classes that react toa wide range of pathogens. Three general strategies are used

to generate Ig diversity through processes that both permanently deletelarge regions of the Ig loci and alter parts of its coding sequence(reviewed in Refs. 1, 2). V(D)J recombination in differentiating pro-

and pre-B cells allows formation of the primary repertoire of Abs inthe form of IgM and IgD molecules, which are expressed on the cell

surface of naive B cells. Class-switch recombination (CSR) leads to

switch of the Ig class from IgM and IgD to IgA, IgE, or IgG, with

other various effector functions. The third strategy to diversify Abs is

induction of somatic hypermutation (SHM) in the V-region, leading

to increased affinity of the produced Abs for the specific Ag. The

latter two processes take place in Ag-activated B cells in the sec-

ondary lymphoid organs (reviewed in Ref. 1).Activation-induced cytidine deaminase (AID) is expressed in

activated B cells and is essential for both Ig CSR and SHM. AID

consists of a DNA cytidine deaminase activation domain and an

APOBEC-1–like domain with homology to RNA-editing enzymes

(3). Although the exact activity of AID is still a matter of debate, it

is clear that AID induces double-strand breaks (DSB) in the so-

called “switch regions” during CSR. These DSBs are then pro-

cessed by the classical nonhomologous end-joining or the alter-

native end-joining pathway, such that the Sm element, upstream of

the Cm region, is fused to a downstream S-region. The intervening

genomic region is looped out and subsequently excised and de-

graded, whereby one of the downstream Ig H chains is brought

within the open reading frame of the V(D)J region promoter (4, 5).Hyper-IgM (HIGM) syndrome is a group of primary immunode-

ficiency disorders that, because of a lack of Ig CSR, is characterized

by normal to elevated serum titers of IgM and no or low serum titers

of IgG, IgA, and IgE subclasses. HIGM syndrome can be caused by

mutations in the genes encoding CD40, CD40L, AID, or uracil–DNA

glycosylase (UNG) (6). Although mutations in CD40 and CD40L

cause HIGM by affecting the communication between B cells and Th

cells, mutations in AICDA and UNG lead to intrinsic B cell defi-

ciency. HIGM patients with loss-of-function mutations in AICDA and

UNG have increased B cell proliferation, as well as giant germinal

centers (GCs), and they suffer from recurrent infections and auto-

immunity that are caused by low-affinity IgM autoantibodies (6).

*Department of Microbiology, Tumor and Cell Biology, Karolinska Institutet, SE-171 77Stockholm, Sweden; †Department of Molecular Biosciences, The Wenner-Gren In-stitute, Stockholm University, SE-106 91 Stockholm, Sweden; ‡Department of Celland Molecular Biology, Karolinska Institutet, SE-171 77 Stockholm, Sweden; andxDivision of Gerontology and Geriatric Medicine, Department of Medicine, MillerSchool of Medicine, University of Miami, Miami, FL 33136

1C.I.M.D. and M.H. contributed equally to this work.

2L.S.W., E.S., and L.S. contributed equally to this work.

Received for publication May 15, 2014. Accepted for publication August 22, 2014.

This work was supported by grants from the Swedish Research Council (to E.S., L.S.W.,and L.S.); the Swedish Cancer Foundation, the Magnus Bergwall Foundation, and theAke Wiberg Foundation (to L.S.W. and L.S.); and Norwegian Research Council Grant205217 (to T.V.). C.I.M.D. and M.L.T. were recipients of Ph.D. fellowships from theKarolinska Institutet. L.S.W. was a recipient of Ph.D. fellowships from the RagnarSoderberg Foundation, the Childhood Cancer Foundation, and the Clas GroschinskyFoundation. R.E.V. was funded by the Stanley J. Glaser Foundation Research Award.

Address correspondence and reprint requests to Prof. Lisa S. Westerberg, Prof. EvaSeverinson, or Prof. Lena Strom, Department of Microbiology Tumor and Cell Bi-ology, Karolinska Institutet, SE-171 77 Stockholm, Sweden (L.S.W.), Department ofMolecular Biosciences, The Wenner-Gren Institute, Stockholm University, SE-106 91 Stockholm, Sweden (E.S.), or Department of Cell and Molecular Biology,Karolinska Institutet, SE-171 77 Stockholm, Sweden (L.S.). E-mail addresses: lisa.westerberg@ki.se (L.S.W.), eva.severinson@su.se (E.S.), or lena.strom@ki.se (L.S.)

The online version of this article contains supplemental material.

Abbreviations used in this article: AID, activation-induced cytidine deaminase; CSR,class-switch recombination; DSB, double-strand break; GC, germinal center; GL,germline; HIGM, hyper-IgM; KLH, keyhole limpet hemocyanin; NP, 4-hydroxy-3-nitrophenylacetyl; SHM, somatic hypermutation; TNP, 2,4,6-trinitrophenol; TNP-SRBC, TNP-conjugated SRBC; UNG, uracil–DNA glycosylase.

This article is distributed under The American Association of Immunologists, Inc.,Reuse Terms and Conditions for Author Choice articles.

Copyright� 2014 by TheAmericanAssociation of Immunologists, Inc. 0022-1767/14/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1401242

by guest on August 4, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

In this study, we identified a spontaneously occurring arginine112 histidine mutation in AID (AIDR112H) and characterized itsimpact on the B cell Ig class switching in vitro and in the vivoimmune response.

Materials and MethodsMice

Mice were bred in the animal facility of the Department of Cell andMolecular Biology at Karolinska Institutet or the Department of MolecularBiosciences at Stockholm University or were bought from Charles River.All experiments were approved by the animal ethical committee ofStockholm Region North.

In vitro cultures for class-switch analysis and retroviraltransduction

Spleen cells were enriched for total or small resting B cells, as describedpreviously (7). Different stimuli were used to detect various Ig isotypes:10 mg/ml LPS (Sigma-Aldrich) to measure switching to IgG2b and IgG3,10 mg/ml anti-CD40 + 8–16 ng/ml IL-4 to measure IgG1 and IgE, and LPS +5 ng/ml IL-5 + 0.5 ng/ml TGF-b to measure IgA (all cytokines were fromPeproTech). For retroviral transduction, splenic B cells from wild-type orAIDR112H mice were stimulated with LPS for 24 h and subsequently trans-duced with retroviral particles, as described (8). LPS or anti-CD40 + IL-4were added, and the cells were cultured for three additional days. CulturedB cells were harvested, washed, and fixed, as described (7), and stained withbiotinylated anti-mouse IgG1, IgG2b, IgG3, IgE, or IgA or isotype controls(all from BD Biosciences), followed by Streptavidin-FITC diluted in a balancedsalt solution with 0.1% saponin. The percentages of cells expressing differentIg isotypes were determined using a FACSCalibur (BD Biosciences).

RT-PCR

At the indicated time points after the addition of stimuli, cells were har-vested, and total RNAwas prepared using GenElute Mammalian Total RNAMiniprep Kit (Sigma-Aldrich). cDNAwas synthesized using AMV reverse-transcriptase (Roche) at 42˚C for 1 h, primed by Oligo dT (Roche). RT-PCRwas run on a Corbett Rotor-Gene 6000 and analyzed with Rotor-Gene6000 series software 1.7. The g1 germline (GL) transcripts were quanti-fied using SYBR Green (KAPA BIOSYSTEMS). The following primerswere used: g1 GL transcript, forward primer: 59-ACACCCCACAGA-CAAACCTGA-39 and reverse primer: 59-CCCTTGACCAGGCATCCCA-39 and Mb1, forward primer: 59-CTGCGGGTAGAAGGGGGTC-39 andreverse primer: 59-CCTTGCGAGGTCAGGGAGCC-39. mRNA levels ofAicda, Ung, and b-actin were quantified with a TaqMan expression assayusing Mm00507774_m1, Mm00449156_m1, and Mm00607939_s1, re-spectively (Life Technologies).

Immunizations and ELISA for serum levels of Ig

Wild-type and AIDR112H mice were immunized by i.p. injection of 2,4,6-trinitrophenol (TNP)-conjugated SRBCs (TNP-SRBCs) diluted 1:100 inPBS. TNP conjugation to SRBCs was performed as described (9). Micewere euthanized 6 d after immunization, and spleens were imbedded inTissue-Tek OCT compound (Sakura) for immunohistochemistry. To de-termine anti-TNP Ab titers in response to TNP-SRBC immunization, anti-TNP IgM and IgG1 Abs in sera were measured by standard ELISAtechniques. Total IgM and IgG were captured with purified anti-mouseIgH+L (Southern Biotech). Specific Abs against TNP were captured withTNP19-BSA (Biosearch Technologies). All Abs were detected withallophycocyanin-conjugated secondary anti-mouse IgM and IgG (SouthernBiotech). All samples were tested in duplicates.

Flow cytometry and immunohistochemistry

For flow cytometry, single-cell suspensions were labeled with fluorescentlyconjugated anti-mouse Abs, including B220, CD21, CD23, GL7, and IgM(all fromBioLegend), as well as IgG1, CD95, CD138, and IgM (all fromBDBiosciences). Data were acquired on an LSRFortessa cytometer (BDBiosciences) and analyzed using FlowJo software (TreeStar, Ashland, OR).For immunohistochemistry, 8–10-mm thin sections were cut from em-bedded spleens in a cryostat microtome. After overnight incubation atroom temperature, the slides were fixed in ice-cold acetone and blockedwith 5% goat serum (Dako) and with an Avidin/Biotin Blocking Kit(Vector Laboratories) in PBS. The slides were incubated with primary Absfor 30 min at room temperature, washed with PBS, incubated at roomtemperature for 30 min with a secondary Ab, and washed again with PBS.

The following reagents were used: allophycocyanin-conjugated anti-B220(BioLegend), streptavidin–Alexa Fluor 555, FITC-conjugated CD169(MOMA-1; AbD Serotech), and biotinylated peanut agglutinin (VectorLaboratories). Images were acquired using a Leica DM IRBE confocallaser scanning microscope (Leica Microsystems) equipped with one argonand two HeNe lasers, using an HC PL APO lens at 103/0.40 CS 90%glycerol (MP Biomedicals). Images were processed with Adobe PhotoshopCS4 Version 11.0.2 (Adobe Systems).

Western blot

Cells were harvested, and protein extracts were prepared at indicated timepoints. A total of 100 mg total protein extracts was run on 12% SDS-PAGEand immunoblotted with monoclonal rat anti-mouse AID (Active Motif) orrabbit anti a-actin Abs (Sigma-Aldrich), followed by HRP-conjugateddonkey anti-rat Ig (Jackson ImmunoResearch) or swine anti-rabbit Ig(Dako).

Sequencing and SHM analysis

Primary B cells were activated for 3 d with anti-CD40 + IL-4, and RNAwasextracted and reverse transcribed. The aicda sequence was amplified usingthe following primers: AID F1, 59-GCAAGCTTCCTTTGGCCTA-39 andAID R2, 59-TAACAAACTGTGTGGTCGTC-39. The amplified DNA waspurified and cloned into a TOPO TA cloning vector (Invitrogen). Threeseparate clones derived from HIGM mice were sequenced in both directionsand compared with one clone of the wild-type control and to the publishedmRNAsequence (accession numberNM_009645). For SHManalysis, VH186.2transcripts were amplified from cDNA of GC-enriched B cells from miceimmunized with 4-hydroxy-3-nitrophenylacetyl (NP)–keyhole limpet hemo-cyanin (KLH) using a nested PCR approach. The first amplification wasperformed with the following primers: Vh186.2, 59-GCTGTATCATGCTC-TTCTTG-39 and Cm, 59-AGGGGGCTCTCGCAGGAGACGAGG-39. Thesecond set of primers was Vh186.2, 59-GGTGTCCACTCCCAGGTCCA-39and Cm, 59-AGGGGGAAGACATTTGGGAAGGAC-39 for amplification ofIgM transcripts. Resulting PCR products were cloned into TOPO TA (Invi-trogen) and sequenced (Operon).

Genotyping of aicda R112H mutation

Exon 2 of the aicda gene was amplified from genomic DNA using PromegaGoTaq polymerase supplemented with 1.6 mM MgCl2 and the followingprimers: 59-TCTGGCTGCCACGTGGAATTGT-39 and 59-TGATCCCG-ATCTGGACCCCAGC-39, with 30 cycles of 95˚C for 30 s, 63˚C for 30 s,and 72˚C for 30 s, followed by 72˚C for 7 min. The resulting 259-bp PCRfragment was digested or not with 1 U BssHII at 37˚C for $3 h and run onan agarose gel. PCR products from wild-type alleles were cut into 175- and84-bp fragments, whereas the AIDR112H band was resistant to digestion.The nipbl gene was genotyped as described (10).

Chromatin immunoprecipitation

Chromatin immunoprecipitation was performed as previously described(11). In short, activated or nonactivated B cells were cross-linked with 1%formaldehyde for 20 min at room temperature, and the reaction wasstopped by addition of glycine to 125 mM final concentration. Cells werewashed twice with cold PBS, resuspended in RIPA buffer (150 mM NaCl,1% [v/v] IGEPAL CA-630, 0.5% [w/v] sodium deoxycholate, 0.1% [w/v]SDS, 50 mM Tris-HCl [pH 8], 5 mM EDTA, and protease and phosphataseinhibitors), and sonicated on a Bioruptor Next Gen (Diagenode). For immu-noprecipitation, 0.5 mg (2 mg/ml) protein extract was cleared for 2 h with 30 ml50% G protein–Sepharose slurry before Ab was added. Immunocomplexeswere eluted, and cross-linking was reversed by adjusting to 200 mM NaCl,1 mM EDTA, and 1 mM DTT and incubating overnight at 65˚C in the pres-ence of 5 mg proteinase K. DNA was purified using a QIAquick PCR puri-fication kit and quantified with real-time PCR using SYBR Green.

Statistics

Statistical significance between groups was assessed by the two-tailedStudent t test and ANOVA; a paired t test was used for the reconstitu-tion experiments. The p values , 0.05 were considered significant.

ResultsA mouse with complete absence of in vitro Ig class-switchresponse

During the course of a project analyzing mice heterozygous fora gene-trap mutation in nipbl (Nipped-B like) (10), we found that

The Journal of Immunology 4733

by guest on August 4, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

B cells of one control mouse were incapable of activating Ig classswitching in vitro. Purified peripheral B cells were stimulated withanti-CD40 + IL-4, LPS, LPS + anti–d-dex, or LPS + IL-5 +TGF-b. After 3–5 d of culture, cells were harvested, stained fordifferent Ig isotypes, and analyzed using flow cytometry. AlthoughB cells from the nipbl+/2 mouse mounted a seemingly normalisotype-switched response, the cells from the control mouse didnot respond (Supplemental Table I). When the cultures were ex-amined microscopically, cells appeared healthy and grew nor-mally. To exclude the possibility that the animals or the sampleswere mistakenly interchanged during the experimental procedure,genotyping was repeated on the cultured cells; the control cellswere indeed wild-type with respect to the nipbl gene. This caughtour interest, and we sought to understand the reason for the lack ofIg class switching.

Identification of inheritance pattern of the HIGM trait

The mouse with abolished Ig class switching in vitro was theoffspring of a mating between a nipbl+/2 male and two CD-1

wild-type females. This mating produced 18 pups, of which 6were heterozygous for the nipbl transgene. We hypothesized thatthis breeding serendipitously gave rise to immunodeficient off-

spring. A consequence of this might be an inability to secrete IgG

in the bloodstream. The 15 available nipbl+/2 and nipbl+/+ siblings

of the mouse in which we had detected abolished in vitro Ig class

switching were used for further analysis. We measured the IgM

and IgG serum titers and found that 3 of 15 mice from two litters

had IgG serum titers close to null but normal to slightly elevated

IgM serum titers (Fig. 1A, F0CD1 generation). We called this

phenotype HIGM, because it resembled the human hyper-IgM

immunodeficiency. Breeding was initiated to monitor the inheri-

tance of the HIGM phenotype. The two HIGM females were bred

to the single HIGM male, and all of the resulting offspring had

IgG serum titers close to the detection limit (Fig. 1A, F1CD1

generation), indicating that the HIGM trait is genetically inherit-

able. We crossed the same HIGM male to two wild-type C57BL/6

females and found that none of the 16 pups born had the HIGM

FIGURE 1. Inheritance of the HIGM phenotype. (A)

ELISA analysis of total serum titers of IgM and IgG Abs in

littermates of the founder HIGM mouse (F0CD1, n = 11)

identified three mice as having the HIGM phenotype (in-

dicated in red). Interbreeding of these mice resulted in 20

offspring (F1CD1) with very low IgG serum titers. (B) One

HIGM male was bred with C57BL/6 females to start

generating mice backcrossed to the C57BL/6 background

[F1(B6/CD1) n = 16]. Further interbreeding generated

F2(B6/CD1) mice (n = 54), of which a quarter displayed the

HIGM phenotype. (C) Pedigree of mice carrying the HIGM

trait, which was discovered in 4 of 18 littermates (F0 gen-

eration). The originally discovered F0 mouse is crossed out,

whereas breeding of the remaining male HIGM mouse

with its female HIGM siblings resulted in 20 offspring all

displaying the HIGM phenotype (F1). None of the 16

offspring from the HIGM male mated to C57/B6 females

displayed the HIGM phenotype, whereas interbreeding the

F1 offspring reproduced the HIGM phenotype in about one

quarter of the offspring (F2).

4734 A NEW SPONTANEOUS MOUSE MODEL FOR STUDIES OF HIGM SYNDROME

by guest on August 4, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

phenotype (Fig. 1B, F1B6/CD1 generation). In the F2B6/CD1 gen-

eration, we got 54 pups that were analyzed for IgM and IgG serumtiters (Fig. 1B). Roughly one fourth of these displayed very lowIgG titers at 6 wk of age, suggesting that the observed HIGMphenotype displayed a recessive Mendelian inheritance pattern(Fig. 1C).Mice from these subsequent generations were tested for the ability

to perform Ig class switching in vitro. The class-switching defectidentified in the original founder of the phenotypewas then confirmed(Fig. 2A, 2B). This might be due to defective proliferation, but wefound that the proliferative response was indistinguishable betweenHIGM and control B cells (Supplemental Fig. 1).We then investigated whether the reason for the abolished Ig class

switching was reduced expression of GL transcripts, which are nec-essary to create chromatin accessibility of S-regions in the Ig locus.After stimulation with anti-CD40 + IL-4, wild-type and HIGM Bcells expressed equal levels of GL g1 transcripts (Fig. 2C). aicdawas an obvious candidate gene that, if mutated, could cause this HIGMphenotype. However, we found comparable quantities of AID inHIGM and wild-type B cells at both the mRNA and protein levels(Fig. 2C, 2D). In addition, we tested the expression of Ung mRNAand also found that to be equally expressed (Fig. 2C).

Identification of the R112H mutation in aicda

Of the genes causing the human HIGM syndrome, CD40L was nota likely candidate, because stimulation with anti-CD40 would thenhave resulted in normal Ig class switching in vitro. In an attempt toidentify the mutation causing the HIGM phenotype, we sequenced

cDNA for aicda, as a first candidate. Compared with the publishedsequence, we found one alteration at position 427 consisting ofa G→A mutation, leading to the amino acid replacement arginine112 to histidine (Fig. 3A). This mutation is localized in a region ofthe aicda gene between the cytidine deaminase domain and theAPOBEC1-like domain. The corresponding residue in humanAICDA is frequently mutated in HIGM patients harboring AIDdeficiencies (12). This region is localized just outside a short se-quence that constitutes a recognition loop determining the sub-strate specificity toward deoxycytidines (Fig. 3B).To confirm that the HIGM phenotype was due to the mutation in

aicda, we reintroduced wild-type AID by means of retroviraltransduction using a construct in which the aicda gene was flankedby GFP, separated by an internal ribosome entry site motif. LPS-activated B cells isolated from AIDR112H or wild-type mice weretransduced with retroviral particles containing the genes for AIDand GFP or only GFP. Cells were activated with LPS or anti-CD40 + IL-4 for three additional days, and Ig class switchingwas measured among the GFP+ cells. Ectopic expression of wild-type AID in AIDR112H B cells fully rescued the Ig class-switchingresponse (Fig. 3C); therefore, we concluded that the HIGM phe-notype is due to the R112H mutation.Because AIDR112H was expressed at normal levels (Fig. 2D), we

were interested in how the mutation influenced nuclear translocationand chromatin association. To address this, we performed chromatinimmunoprecipitation in combination with real-time PCR usingprimers specific for the AID binding site in the Ig Sm locus (11).DNAwas isolated from wild-type and AIDR112H B cells at 0 or 48 h

FIGURE 2. In vitro Ig class-switch re-

sponse of HIGM B cells. (A) Ig class-switch

response in vitro on day 5. Purified B cells

were stimulated with anti-CD40 + IL-4 for

switching to IgG1 and IgE or with LPS for

production of IgM and IgG3. Thereafter, they

were harvested, fixed, and permeabilized,

and intracellular staining was performed.

Gates were set so that the corresponding

isotype controls were ,0.01%. Flow cy-

tometry plots are shown from one represen-

tative experiment. (B) Quantification of Ig

class switching to indicated isotypes. Aver-

age from three independent experiments is

shown. (C) mRNA expression of Aicda, Ung

mRNA, and GL g1 transcripts in HIGM

B cells. Average from two independent

experiments is shown. The expressions are

related to those of wild-type mice. (D) AID

protein expression on day 3 after stimulation

in vitro (left panel). Relative quantification of

AID protein expression from two indepen-

dent experiments (right panel). *p , 0.05.

The Journal of Immunology 4735

by guest on August 4, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

after stimulation with anti-CD40 + IL-4. At 0 h, no binding could bedetected in either population of cells (data not shown). At 48 h,maximum AID binding was detected in wild-type cells, whereas theAIDR112H association was 50% of that.

Naive AIDR112H mice have increased peripheral B cell numberand large GCs and fail to induce SHM

HIGM patients with loss-of-function mutations in AID have highlyproliferative IgM+IgD+ B cells and enlarged GCs (13). Likewise,gene-targeted aicda2/2 mice display expansion of B cell folliclesand large GCs in the absence of Ag challenge (14). In the bonemarrow, AIDR112H mice and littermate controls showed similarB cell development, as characterized by the Hardy classification(C.I.M. Dahlberg and L.S. Westerberg, unpublished observations).When assessing peripheral B cell development and function, be-fore and after immunization with TNP-SRBCs, AIDR112H micehad an increased number of total B cells in the spleen comparedwith wild-type littermate mice (Fig. 4A). This increased numberof B cells was evident in both the follicular and the marginal zonesubsets of B cells (C.I.M. Dahlberg and L.S. Westerberg, unpub-lished observations). Wild-type mice showed few, if any, GCsbefore immunization but formed large GCs in response to TNP-SRBCs at day 6 (Fig. 4A, 4B). AIDR112H mice already had en-larged GCs in the spleen before challenge, and they remainedlarge after challenge with TNP-SRBCs (Fig. 4A, 4B). We nextexamined the anti-TNP Ab response in serum. Wild-type miceshowed increased serum titers of TNP-specific IgG1 after immu-nization, whereas AIDR112H mice, as expected from the in vitrodefect in Ig class switching, failed to induce a TNP-specific IgG1serum response (Fig. 4C). We did not detect any differencesin total plasma cell number when comparing wild-type and

AIDR112H mice before and after TNP-SRBC immunization. Whenassessing IgG1-producing plasma cells, wild-type mice respondedto TNP-SRBCs with an increased number of IgG1+ plasma cells,whereas AIDR112H failed to do so (Fig. 4D).To investigate SHM, AIDR112H mice were immunized with the

T cell–dependent Ag NP-KLH. NP-KLH induces selection of upto 50% of NP-specific B cells that express the specific V H chain(VH) VH186.2 (15, 16). Upon immunization, wild-type mice hadincreased serum titers of NP-specific IgG1, whereas AIDR112H

mice lacked an IgG1 response (data not shown). To establish therate of SHM and introduction of the reported high-affinity muta-tion in CDR1, W33L, in response to NP-KLH, we prepared cDNAfrom GC-enriched B cells and sequenced VH186.2 combined withthe constant segment for IgM. Wild-type B cells induced SHM, asevident by many replacement mutations and presence of the high-affinity W33L mutation in 6 of 72 sequences (data not shown,Fig. 4E). In contrast, AIDR112H B cells had an overall low rateof mutations and failed to induce the W33L mutation (0/86sequences) (Fig. 4E).

DiscussionIn this article, we describe an inherited trait in mice that is causedby a point mutation in aicda and results in a failure to switch Igclass. The arginine 112 histidine mutation renders AID catalyti-cally inactive (17), thus explaining the lack of Ig class switchingand SHM in AIDR112H B cells. The AIDR112H mutation has beenobserved in several unrelated HIGM patients, and it is a frequentcause of AID deficiency (12, 18).It is interesting that the AIDR112H mutation appeared with the

nipbl+/2 genotype. The most well-characterized function forNIPBL is DNA loading of the cohesin complex. Both NIPBL and

FIGURE 3. The defect in class switching is caused by an AID arginine 112 histidine mutation. (A) Sequence of aicda cDNA, corresponding to amino

acids 111–113. Mutant (upper panel); wild-type (lower panel). (B) Schematic illustration of the AID protein; the R112H mutation and important functional

domains are indicated. (C) Primary splenic B cells were stimulated in vitro with LPS for 24 h, transduced with retroviral particles, and stimulated with either

LPS or anti-CD40 + IL-4. After a total of 5 d, the cells were stained for expression of the indicated Ig isotypes and analyzed by flow cytometry. Average of

three independent experiments is shown. *p , 0.05, **p , 0.01, ***p , 0.001, paired t test. Differences between AID-reconstituted wild-type and

AIDR112H mice were not significant. (D) Chromatin immunoprecipitation was performed to detect AID association with Sm 48 h after B cells were

stimulated with anti-CD40 + IL-4. Data are mean 6 SD of five mice/group. *p , 0.05, Student t test.

4736 A NEW SPONTANEOUS MOUSE MODEL FOR STUDIES OF HIGM SYNDROME

by guest on August 4, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

cohesin are important for chromosome segregation, gene regula-tion, and DSB repair (19–21). Thus, nipbl+/2 mice could benefitfrom lacking functional AID, because AID induces these types ofDNA lesions.The spontaneous AIDR112H mouse model described in this article

shares many similarities with aicda2/2 mice described by Mur-amatsu et al. (14), as well as the human HIGM syndrome caused bymutations in AICDA that was identified by Revy et al. (13). They all

have abolished in vitro Ig class switching, expansion of peripheralB cells, and large GCs. An important advantage of the AIDR112H

mouse is that AID is expressed normally, allowing for evaluation ofits posttranslational regulation. Moreover, the AIDR112H proteintranslocated to the nucleus and bound to the Sm region, albeit withdecreased efficiency. Possible reasons for this are deficient inter-actions with proteins important for targeting of AID to the S regionsor a poor binding capacity of AIDR112H to S regions.

FIGURE 4. AIDR112H mice have increased

peripheral B cell numbers, enlarged GCs, and

reduced SHM. Wild-type and AIDR112H mice

were immunized by i.p. injection of TNP-

SRBCs and sacrificed 6 d later. (A) Flow

cytometric analysis was performed to define

the absolute numbers of B cells in wild-type

or AIDR112H mice. Analysis of GCs by im-

munohistochemistry and flow cytometry of

spleens from either naive or day 6–immunized

wild-type and AIDR112H mice. (B) GC forma-

tion was detected by staining for B220 (blue),

PNA (red), and MOMA (green). Original

magnification 310. Photo-collages from four

representative spleens for each condition are

shown. Quantification of GC B cells (B220+

GL7+CD95+) was performed by flow cytom-

etry. Note the marked increase in GC B cells in

naive AIDR112H mice compared with wild-

type. (C) Serum from naive or immunized

wild-type and AIDR112H mice was tested for

IgM (left panel) and IgG1 (right panel) bind-

ing to TNP by ELISA. Test samples were

corrected for background binding. (D) Plasma

cells (CD138+) (left panel) and class-switched

plasma cells (CD138+IgG1+) (right panel)

from spleen were analyzed by flow cytometry.

Each group represents eight mice examined in

two experiments. Horizontal black lines rep-

resent the mean 6 SEM for the group. (E)

Frequency of the high-affinity mutation, W33L,

in CDR1 in wild-type and AIDR112H mice in

response to NP-KLH. Wild-type: 72 sequences

from five mice; AIDR112H: 86 sequences from

four mice. *p , 0.05, **p , 0.01, ***p ,0.001, ****p , 0.0001, Student t test.

The Journal of Immunology 4737

by guest on August 4, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

Although recent studies, to some extent, defined the require-ments for GC selection of B cells and showed that GC B cellscompete for help from T follicular helper cells (22), the exactcues that make B cells leave the GC remain unknown. It wassuggested that aicda2/2 mice have larger than normal GCs be-cause of decreased apoptosis (23). Why B cells without functionalAID accumulate and/or fail to leave the GCs remains to beaddressed.A recent study examined the potential role of AID as an epi-

genetic eraser and transcriptional regulator (24). The investigatorsfound that the identified alterations in the transcription profilepotentially were not related to the aicda2/2 genotype but insteadto a region surrounding the targeted aicda locus originating fromthe founding CBA mouse strain. Therefore, care should be takenwhen interpreting aicda2/2 correlated phenotypes. The AIDR112H

mouse strain described in this study could become a valuable toolin the study of AID function, as well as a model for the humanHIGM syndrome.

AcknowledgmentsWe thank Prof. Tasuku Honjo (Kyoto University, Kyoto, Japan) for the kind

gift of the pMX-AID-IRES-GFP construct.

DisclosuresThe authors have no financial conflicts of interest.

References1. Pinaud, E., M. Marquet, R. Fiancette, S. Peron, C. Vincent-Fabert, Y. Denizot,

and M. Cogne. 2011. The IgH locus 39 regulatory region: pulling the strings frombehind. Adv. Immunol. 110: 27–70.

2. Kato, L., A. Stanlie, N. A. Begum, M. Kobayashi, M. Aida, and T. Honjo. 2012.An evolutionary view of the mechanism for immune and genome diversity. J.Immunol. 188: 3559–3566.

3. Muramatsu, M., V. S. Sankaranand, S. Anant, M. Sugai, K. Kinoshita,N. O. Davidson, and T. Honjo. 1999. Specific expression of activation-inducedcytidine deaminase (AID), a novel member of the RNA-editing deaminasefamily in germinal center B cells. J. Biol. Chem. 274: 18470–18476.

4. Stavnezer, J., J. E. Guikema, and C. E. Schrader. 2008. Mechanism and regu-lation of class switch recombination. Annu. Rev. Immunol. 26: 261–292.

5. Stavnezer, J. 2011. Complex regulation and function of activation-induced cy-tidine deaminase. Trends Immunol. 32: 194–201.

6. Durandy, A., N. Taubenheim, S. Peron, and A. Fischer. 2007. Pathophysiology ofB-cell intrinsic immunoglobulin class switch recombination deficiencies. Adv.Immunol. 94: 275–306.

7. Strom, L., J. Laurencikiene, A. Miskiniene, and E. Severinson. 1999. Charac-terization of CD40-dependent immunoglobulin class switching. Scand. J.Immunol. 49: 523–532.

8. Strom, L., M. Lundgren, and E. Severinson. 2003. Binding of Ikaros to germlineIg heavy chain gamma1 and epsilon promoters. Mol. Immunol. 39: 771–782.

9. Karlsson, M. C., S. Wernersson, T. Diaz de Stahl, S. Gustavsson, andB. Heyman. 1999. Efficient IgG-mediated suppression of primary antibodyresponses in Fcgamma receptor-deficient mice. Proc. Natl. Acad. Sci. USA 96:2244–2249.

10. Kawauchi, S., A. L. Calof, R. Santos, M. E. Lopez-Burks, C. M. Young,M. P. Hoang, A. Chua, T. Lao, M. S. Lechner, J. A. Daniel, et al. 2009. Multipleorgan system defects and transcriptional dysregulation in the Nipbl(+/2) mouse,a model of Cornelia de Lange Syndrome. PLoS Genet. 5: e1000650.

11. Cortizas, E. M., A. Zahn, M. E. Hajjar, A. M. Patenaude, J. M. Di Noia, andR. E. Verdun. 2013. Alternative end-joining and classical nonhomologous end-joining pathways repair different types of double-strand breaks during class-switch recombination. J. Immunol. 191: 5751–5763.

12. Xu, Z., H. Zan, E. J. Pone, T. Mai, and P. Casali. 2012. Immunoglobulin class-switch DNA recombination: induction, targeting and beyond. Nat. Rev. Immunol.12: 517–531.

13. Revy, P., T. Muto, Y. Levy, F. Geissmann, A. Plebani, O. Sanal, N. Catalan,M. Forveille, R. Dufourcq-Labelouse, A. Gennery, et al. 2000. Activation-induced cytidine deaminase (AID) deficiency causes the autosomal recessiveform of the Hyper-IgM syndrome (HIGM2). Cell 102: 565–575.

14. Muramatsu, M., K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, andT. Honjo. 2000. Class switch recombination and hypermutation requireactivation-induced cytidine deaminase (AID), a potential RNA editing enzyme.Cell 102: 553–563.

15. Bothwell, A. L., M. Paskind, M. Reth, T. Imanishi-Kari, K. Rajewsky, andD. Baltimore. 1981. Heavy chain variable region contribution to the NPb familyof antibodies: somatic mutation evident in a gamma 2a variable region. Cell 24:625–637.

16. McHeyzer-Williams, M. G., G. J. Nossal, and P. A. Lalor. 1991. Molecularcharacterization of single memory B cells. Nature 350: 502–505.

17. Hu, Y., I. Ericsson, K. Torseth, S. P. Methot, O. Sundheim, N. B. Liabakk,G. Slupphaug, J. M. Di Noia, H. E. Krokan, and B. Kavli. 2013. A combinednuclear and nucleolar localization motif in activation-induced cytidine deami-nase (AID) controls immunoglobulin class switching. J. Mol. Biol. 425: 424–443.

18. Ta, V. T., H. Nagaoka, N. Catalan, A. Durandy, A. Fischer, K. Imai,S. Nonoyama, J. Tashiro, M. Ikegawa, S. Ito, et al. 2003. AID mutant analysesindicate requirement for class-switch-specific cofactors. Nat. Immunol. 4: 843–848.

19. Dorsett, D., and L. Strom. 2012. The ancient and evolving roles of cohesin ingene expression and DNA repair. Curr. Biol. 22: R240–250.

20. Liu, J., Z. Zhang, M. Bando, T. Itoh, M. A. Deardorff, D. Clark, M. Kaur,S. Tandy, T. Kondoh, E. Rappaport, et al. 2009. Transcriptional dysregulation inNIPBL and cohesin mutant human cells. PLoS Biol. 7: e1000119.

21. Remeseiro, S., A. Cuadrado, S. Kawauchi, A. L. Calof, A. D. Lander, andA. Losada. 2013. Reduction of Nipbl impairs cohesin loading locally and affectstranscription but not cohesion-dependent functions in a mouse model of Corneliade Lange Syndrome. Biochim. Biophys. Acta 1832: 2097–2102.

22. Victora, G. D., T. A. Schwickert, D. R. Fooksman, A. O. Kamphorst, M. Meyer-Hermann, M. L. Dustin, and M. C. Nussenzweig. 2010. Germinal center dy-namics revealed by multiphoton microscopy with a photoactivatable fluorescentreporter. Cell 143: 592–605.

23. Zaheen, A., B. Boulianne, J. Y. Parsa, S. Ramachandran, J. L. Gommerman, andA. Martin. 2009. AID constrains germinal center size by rendering B cellssusceptible to apoptosis. Blood 114: 547–554.

24. Hogenbirk, M. A., M. R. Heideman, A. Velds, P. C. van den Berk,R. M. Kerkhoven, B. van Steensel, and H. Jacobs. 2013. Differential program-ming of B cells in AID deficient mice. PLoS ONE 8: e69815.

4738 A NEW SPONTANEOUS MOUSE MODEL FOR STUDIES OF HIGM SYNDROME

by guest on August 4, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from