of March 23, 2018.This information is current as

Antinuclear AntibodiesofPure Nonlupus Mouse Strains to Producers

The MHC Haplotype H2b Converts Two

Kristian Hannestad and Helge Scott

http://www.jimmunol.org/content/183/5/3542doi: 10.4049/jimmunol.0900579August 2009;

2009; 183:3542-3550; Prepublished online 5J Immunol 

Referenceshttp://www.jimmunol.org/content/183/5/3542.full#ref-list-1

, 27 of which you can access for free at: cites 62 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive 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 © 2009 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The MHC Haplotype H2b Converts Two Pure NonlupusMouse Strains to Producers of Antinuclear Antibodies1

Kristian Hannestad2* and Helge Scott†

Studies of mouse lupus models have linked the MHC H2b haplotype with the earlier appearance of antinuclear autoantibodies andthe worsening of nephritis. However, it is unknown whether H2b by itself, in the context of pure nonlupus strains, is “silent” orsufficient with regard to loss of tolerance to chromatin (nucleosomes). In this study we show that, beginning �6–9 mo of age,H2b-congenic BALB/c (denoted BALB.B) mice, unlike BALB/c (H2d) and H2k-congenic BALB/c (denoted BALB.K) mice, developstrikingly increased serum levels of anti-chromatin Ab dominated by the IgG2a subclass, along with minor increase of Abs to DNAand moderately increased total serum IgG2a. The BALB.B mice did not have glomerulonephritis or an increased mortality rate.H2b-congenic C3H/He mice (designated C3.SW mice), unlike C3H/He (H2k) mice, showed low but measurable serum levels ofchromatin-reactive IgG2a Abs and minor but significant hypergammaglobulinemia. By immunofluorescence, IgG2a of sera fromboth H2b-congenic strains stained HEp-2 cell nuclei, confirming the presence of antinuclear autoantibodies. Thus, in the contextof two pure nonlupus genomes, the MHC H2b haplotype in homozygous form is sufficient to induce loss of tolerance tochromatin. The Journal of Immunology, 2009, 183: 3542–3550.

T he classical spontaneous mouse lupus models are the NewZealand Black (NZB),3 New Zealand White (NZW), andtheir inbred recombinant NZM2410 strains, males of the

BXSB strain, and the MRL/lpr strain. Usually, Abs to chromatin(nucleosomes) appear first (1, 2), followed by hypergammaglobu-linemia, splenomegaly, and fatal glomerulonephritis mediated atleast in part by the deposition of harmful immune complexes (IC)containing antinuclear Abs (ANA) typically directed to intactchromatin, dsDNA, and ssDNA (3). By activating the CpG DNA-binding TLR9 in B cells, DNA of chromatin acts as a built-inautoadjuvant that, in combination with BCR signals, recruits andexpands chromatin-specific B cells during the autoimmune re-sponse (reviewed in Ref. 4). This dual engagement of BCR andTLR9 likely plays a critical, but not necessarily sufficient, role forthe initiation and predominance of chromatin-reactive Abs in theearly stage of mouse lupus.

Many genes acting together cause mouse lupus, complicatingthe task of sorting out those genes that have a major impact onautoimmunity to chromatin. The complexity can be reduced bybreeding small chromosomal intervals, identified in advance bygenetic mapping of lupus traits, onto the nonlupus-prone C57BL/6(B6) background. In this way it was shown that intervals locatedon distal chromosome 1 such as Sle1 and its high penetrance sub-locus Sle1b from NZW, Nba2 from NZB, and Bxs1–4 from BXSB

mice are sufficient to induce production of ANA (reviewed in Refs.5 and 6). The Sle1b interval contains a number of polymorphiccandidate genes (7), among which the Ly108 gene has been con-sidered a strong candidate for mediating the Sle1b phenotype (8).

Another lupus susceptibility locus is the MHC of both mice(known as H2) and humans (HLA) that contains numerous genesincluding the polymorphic MHC class I and II genes (reviewedin Refs. 9 –11). The H2 region of the NZW strain also containsthe Sles1 lupus-resistance locus, which is located in an intervalthat also contains the class II genes (12). From the standpoint ofthe current report, relevant past studies of B6/lpr, BXSB, andNZB.H2bm12 lupus mice have shown that the MHC H2b andH2bm12 haplotypes are associated with more severe lupus serologyand nephritis when compared with the H2d and H2k haplotypes(13–17). Because the Ea gene of the H2b locus is defective, ex-pression of the I-E �-chain (E�-chain) and functional I-E het-erodimers are absent in H2b mice (18). Furthermore, a peptidederived from the E�-chain, E�52–68, binds to I-Ab moleculeswith high affinity, so that in mice expressing both I-Ab and I-Eb

this self-peptide is present in �12% of I-Ab molecules (19). Anal-yses of Ea-transgenic BXSB and related lupus mice have providedevidence that the E�-chain inhibits lupus autoimmunity but notimmune responses against several foreign Ags. Thus, transgene-encoded E�-chains suppress serological and renal manifestationsof lupus and prolong survival, most pronounced for mice that carryan unphysiologically high number (50–100) of transgene copies(20). These data have led to the proposal that the E�52–68 self-peptide (which is missing in H2b mice) partially blocks I-A mol-ecules of B cells from presenting putative pathogenic self-peptidesto Th cells, resulting in the lowering of autoantibody levels (10,20). According to this hypothesis, production of ANA depends onTh cells, and E� is a lupus resistance gene.

The H2b haplotype by itself, in the context of pure nonautoim-mune backgrounds, has to date not been associated with loss oftolerance to chromatin (21). In this study we demonstrate that,upon aging, H2b/b (called H2b here) haplotype-congenic BALB/cand C3H mice spontaneously produce anti-chromatin Abs domi-nated by the IgG2a isotype. These findings indicate that the H2b

*Institute of Immunology and †Institute of Pathology, University of Oslo and Rik-shospitalet University Hospital, Oslo, Norway

Received for publication February 20, 2009. Accepted for publication July 5, 2009.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 This work was supported by a grant from Helse Sør (to K.H.).2 Address correspondence and reprint requests to Dr. Kristian Hannestad, Institute forImmunology, Rikshospitalet, Sognsvannsveien 20, N-0027 Oslo, Norway. E-mail ad-dress: kristian.hannestad@rr-research.no3 Abbreviations used in this paper: NZB, New Zealand Black; AHGG, aggregatedhuman IgG; ANA, antinuclear autoantibody; AP, alkaline phosphatase; cM, centi-morgan; IC, immune complex; NZW, New Zealand White; p-NPP, p-nitrophenylphosphate.

Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00

The Journal of Immunology

www.jimmunol.org/cgi/doi/10.4049/jimmunol.0900579

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

haplotype contains one or more genes that play a major role ininducing ANA production.

Materials and MethodsMice

BALB/cOlaHsd (H2d) mice and the H2b haplotype-congenic BALB.B/OlaHsd (also known as C.B10-H2b) and H2k haplotype-congenic BALB.K/OlaHsd mice were obtained from Harlan. The H2b congenic interval of theBALB.B strain is externally delimited by the microsatellite markersD17Mit80 (at 12.9 centimorgans (cM)) and D17Mit232.1 (at 20.37 cM),implying that the size of the congenic interval is 7.47 cM (Ref. 22 andcorrection from C. Penha-Goncalves, unpublished observations). This in-terval includes the MHC class I, II, and III genes. Endotoxin sensitiveC3H/HeSnJ (H2k) mice and H2b-congenic C3.SW-H2b/SnJ mice were pur-chased from Jackson ImmunoResearch Laboratories. Sera from 9-mo-oldB6.Sle1b mice were kindly provided by Dr. E. Wakeland (University ofTexas Southwestern Medical Center, Dallas, TX). Unless indicated, allmice were female. Sentinel mice were examined according to recommen-dations from the Federation of European Laboratory Animal Science As-sociations and tested negative for parasitical, bacterial, and viral agents. Anin-house expanded virus test program revealed that approximately one-third of the sentinel mice tested positive for Abs to mouse norovirus. An-imal care was in accordance with national legislation and institutionalguidelines.

Preparation of soluble chromatin

To prepare nucleosome core particles, chicken erythrocytes were lysedand the nuclei were washed and digested with micrococcal nuclease(Worthington Biochemical) as described by Ausio et al. (23). We used100 U of enzyme per 100 OD260 U of DNA for 15 min at 37°C. Thedigestion was terminated by adding EDTA to a final concentration of 10mM, the suspension was spun down at 12,000� g in a SS-34 rotor at4°C for 5 min, and the supernatant was collected, aliquoted, and storedat �70°C. The preparation contained a single broad DNA band of �150bp, and electrophoresis in 18% polyacrylamide gel (SDS-PAGE) re-vealed the characteristic band pattern of the four core histones H2A,H2B, H3, and H4 (data not shown). The same preparation was usedthroughout the study.

ELISA for specific Abs

Anti-chromatin. Ninety-six-well Nunc Maxisorp plates (Nalge Nunc)were coated overnight with 5 �g/ml nucleosome core particles in PBS.After washing with PBS, the wells were blocked with 100 �l of PBSwith1.8% BSA for 1 h at 37°C. Sera typically diluted 1/200, 1/1000, and1/5000 in PBS, Tween 20, and 0.5% BSA were added to the wells for 1 h.Then, 1/2000 diluted affinity-purified biotinylated goat secondary Absagainst mouse Fc� (Jackson ImmunoResearch Laboratories) or the relevantIgG isotype (SouthernBiotech) were added for 30 min. Finally, bound sec-ondary Abs were detected with 1/3000 diluted streptavidin-alkaline phos-phatase (AP) conjugate (GE Healthcare catalog no. RPN1234V1) followedby p-nitrophenyl phosphate (p-NPP) substrate. After 30 min, plates wereanalyzed at 405 nm in a Victor3 V MultiLabel plate reader (PerkinElmer).The wells were washed four times with PBS with 0.05% Tween 20 be-tween each step.Anti-DNA. Plates precoated with methylated BSA (5 �g/ml) were coatedwith 10 �g/ml dsDNA or boiled ssDNA from salmon sperm. Serum froma patient with active systemic lupus erythematosus served as positive con-trol for coating.Anti-total histones. Wells were coated with calf thymus histones (RocheApplied Science catalog no. 10 223 565) at 2.5 �g/ml in 0.1 M carbonatebuffer (pH 9.6). An in-house IgG2a mAb (3F7-A10) derived from aBALB.B mouse served as positive control for histone coating.Anti-cardiolipin. Cardiolipin-precoated wells (Varelisa kit) were pur-chased from Phadia.Anti-U1RNP, -Sm, -Ro, and -La. Precoated wells (Varelisa Recombi ANAScreen; Phadebas) were purchased from Pharmacia. Other Ag-specificELISAs were performed by coating with 1 �g/ml heat-aggregated humanIgG (AHGG), 10 �g/ml boiled LPS, and 5 �g/ml chicken gizzard actinin.Unless specified, all reagents were obtained from Sigma-Aldrich.

Standard curves established by sandwich ELISA

Wells were coated with 2 �g/ml unlabeled IgG subclass-specific poly-clonal goat anti-mouse capture Abs (SouthernBiotech) that were blockedand the corresponding serially diluted purified mouse IgG subclasses

(SouthernBiotech) were added at known concentrations. After washing, thestandards were revealed with the corresponding biotinylated isotype-spe-cific secondary Abs (SouthernBiotech) diluted 1/2000, AP-streptavidin di-luted 1/3000, and p-NPP substrate. The serum Ab level of a given isotypewas calculated from the OD of a serum dilution that fell on the steep slopeof the standard curve for that isotype. The background OD was subtractedto obtain the final OD values.

Sandwich ELISA of total serum IgM, IgG1, and IgG2a

To wells coated with unlabeled, isotype-specific, goat capture Abs (2.5�g/ml) were added appropriate dilutions (determined by pilot experi-ments) of sera or known concentrations of purified mouse Ig standardsfor 1 h. After washing, bound Igs were revealed by incubation withAP-conjugated, isotype-specific, goat anti-mouse Abs (SouthernBiotech) andp-NPP substrate. The amount of bound Ig isotype was calculated fromOD values of sera dilutions that fell on the steep slope of the standardcurve.

Urine analysis

Proteinuria was measured using urinalysis dipsticks (Multistix 8 SG; BayerDiagnostics) and a Clinitek 500 reader and was graded as follows: �/�,trace; 1�, 0.3 g/L; 2�, 1 g/L; and 3�, 3 g/L.

Statistics

All statistics were calculated using GraphPad Prism version 4 for Win-dows. A two-tailed Mann-Whitney rank sum U test was applied to the datato assess statistical significance ( p � 0.05 is considered significant).

ResultsSerological analysis of BALB/c and H2 congenic BALB/c strains

To investigate the impact of the H2 haplotype in the context ofpure nonlupus strains on serum Abs to chromatin, we bled mice ofthe BALB.B (H2b) (n � 17–18), BALB/c (H2d) (n � 12–17), andBALB.K (H2k) (n � 18) strains at 3, 6, 9, 12, and 14 mo of age.ELISA showed that the serum IgG reactivity of BALB.B micewith immobilized nucleosome core particles increased gradually inan age-dependent fashion, indicating a developing autoimmune re-sponse to chromatin (Fig. 1). By contrast, the autoantibody levelsof most age- and gender-matched BALB.K and BALB/c mice re-mained low throughout. The difference was significant at all agesfor BALB.B vs BALB.K mice, and at 9 and 12 mo of age forBALB.B vs BALB/c mice (Fig. 1). At 9 mo of age, 100% ofBALB.B mice were deemed positive for IgG anti-chromatin (ODvalue 4 SD above the mean level of 18 BALB.K mice of the sameage). The average IgM anti-chromatin activity was also higher forthe BALB.B group: significant by 9 and 12 mo of age for BALB.Bvs BALB.K mice ( p � 0.0001), and by 12 mo of age for BALB.Bvs BALB/c mice ( p � 0.0003) (data not shown). These resultsindicate that BALB.B mice are susceptible to late-onset augmen-tation of chromatin-reactive IgM and IgG Abs.

Isotypes, specificity, and age and gender dependence ofautoantibodies

We assessed IgG subclass levels of chromatin-specific Abs of 14-mo-old BALB.B mice (n � 17). ELISA showed that IgG2a wasthe dominant subclass; its average concentration was 138 � 28�g/ml compared with 13 �g/ml for IgG2b, 5.4 �g/ml for IgG3,and 1.9 �g/ml for IgG1 (Fig. 2A). Another cohort of 12-mo-oldBALB.B mice (n � 14) had average anti-nucleosome levels of14.8 � 5.8 �g/ml for IgG2a, 7.2 � 3.8 �g/ml for IgG2b (notsignificantly different vs IgG2a), 0.69 � 0.2 �g/ml for IgG3 ( p �0.0006 vs IgG2a), and 0.12 � 0.02 for IgG1 ( p � 0.0001 vsIgG2a) (data not depicted). Thus, the IgG2a/IgG1 ratio of Abs tochromatin ranged from �70–120. A third sample of 12-mo-oldBALB.B mice (n � 12) had an average serum level of chromatin-specific IgG2a and IgM of 81 � 23 �g/ml and 19 � 7 �g/ml,respectively, i.e., an IgG2a/IgM ratio of �4. These results dem-onstrate that the autoantibodies of BALB.B mice have a distinct

3543The Journal of Immunology

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

profile with respect to isotype distribution: IgG � IgM, andIgG2a � IgG2b � IgG3 � IgG1. For the remaining analyses wefocus mainly on chromatin-specific IgG2a.

From 3 to 12 mo of age, the average anti-chromatin IgG2a levelof BALB.B mice (n � 17–18) increased �90-fold, from 0.49 �0.06 �g/ml at 3 mo, 1.67 � 0.47 �g/m at 6 mo ( p � 0.0001 vsBALB.K and p � 0.017 vs BALB/c), 22.1 � 13.9 �g/ml at 9 mo( p � 0.0001 vs BALB.K and BALB/c), and 45.4 � 11.6 �g/ml at12 mo ( p � 0.0001 vs BALB.K and BALB/c) (Fig. 2B). By com-parison, BALB.K and BALB/c mice were typically negative orlow for IgG2a Abs to chromatin (Fig. 2B). Thus, in these cohortsthe average serum IgG2a anti-chromatin level of 12-mo-oldBALB.B mice was �35- and �110-fold higher as compared withBALB/c and BALB.K mice, respectively. The same sera were as-sayed for ANA activity by indirect immunofluorescence withHEp-2 cells. A homogenous nuclear staining pattern was producedby IgG2a of sera from 15 of 16 BALB.B mice (Fig. 2E), two of 16BALB/c mice (the same two mice that had anti-chromatin IgG byELISA; see Fig. 1), and none of 18 BALB.K mice. Serum IgG1produced no nuclear staining, and no IgG2a staining of HEp-2cytoplasm was observed.

The average anti-dsDNA IgG2a levels of the sera from 12-mo-old mice were very low: 0.62 �g/ml for BALB.B mice, 0.09�g/ml for BALB/c mice ( p � 0.0001 vs BALB.B), and 0.02�g/ml for BALB.K mice ( p � 0.0001 vs BALB.B) (Fig. 2C).Thus, despite the low BALB.B serum levels of anti-dsDNAAbs, they were significantly higher than those of BALB/c andBALB.K mice. Of 17 sera from 14-mo-old BALB.B mice, 5(�30%) stood out with respect to elevated levels of IgG2a Absagainst both ssDNA (mean of 3.2 �g/ml) and dsDNA (mean of4.5 �g/ml); four of the five sera also contained �1.5 �g/mlLPS-reactive IgG2a (Fig. 2D). Of note, the abundance of anti-dsDNA IgG2a of the five sera was 8-, 15-, 62-, 83-, and 94-foldlower compared with anti-chromatin IgG2a. No IgG2a reactiv-ity was detected by ELISA against cardiolipin, actinin, humanIgG, U1RNP, Sm, SS-A/Ro, and SS-B/La (our unpublisheddata). Sera from another sample of 12-mo-old BALB.B mice(n � 12) were analyzed for IgG2a Abs to total histones. ELISAshowed that the average level was 0.093 �g/ml, which is 870-fold lower than the average concentration (81 �g/ml) of anti-chromatin IgG2a of the same sera (data not depicted).

Compared with males, the lupus of (NZB � NZW)F1 femalesis strongly accelerated (3), and the IgG anti-H2A/H2B/DNAserum levels are higher in 5-mo-old, but not 7- to 9-mo-old,B6.Sle1 female mice (24). We therefore assessed the impact of

gender on anti-chromatin IgG2a of BALB.B mice. By 12 mo ofage, the average IgG2a autoantibody level of the female cohort(n � 15) was 45 � 19 �g/ml, compared with 0.7 � 0.35 �g/mlfor the males (n � 8; p � 0.003 by Mann-Whitney U test) (datanot depicted). This 64-fold difference suggests a female bias. Itshould be noted that the BALB.B males were aggressive andmutilated each other, so that 12 of 20 mice either died or wereeuthanized before 12 mo of age.

Taken together, we conclude that the bulk of IgG2a autoanti-bodies of BALB.B mice is directed to intact nucleosome core par-ticles and is not polyreactive. The low level IgG2a reactivity withdsDNA, ssDNA, and LPS found in 30% of BALB.B mice couldrepresent a minor polyreactive fraction (25, 26).

Serum levels of total IgM, IgG1, and IgG2a

We also measured total serum concentrations of IgM, IgG1, andIgG2a (IgG2b and IgG3 were not measured). The most strikingfinding was that the total IgG2a levels of BALB.B mice rosesteeply from 3 mo through to14 mo of age, whereas those ofBALB/c and BALB.K mice rose more slowly and hit a lowerplateau (Fig. 3). By 14 mo of age, BALB.B mice had �2.6-foldmore total IgG2a (4.01 � 0.35 mg/ml) compared with BALB.K(1.52 � 0.11 mg/ml; p � 0.0001) and BALB/c (1.48 � 0.14mg/ml; p � 0.0001) mice (Fig. 3). As the average level ofchromatin-specific IgG2a in this sample of 14-mo-old mice was�0.14 mg/ml (Fig. 2A), the autoantibodies represent a smallfraction (3.5%) of total IgG2a. The total serum levels of IgMand IgG1 of BALB.B mice did not differ markedly from thoseof age-matched BALB/c and BALB.K mice (Fig. 3). Thus, thehypergammaglobulinemia of BALB.B mice mainly presents asa moderate elevation of IgG2a. This result is consistent withthat of a previous study showing that H2b mice with B10 orBALB/c backgrounds have higher serum levels of IgG2a thanH2k and H2d congenic strains (27).

Induced Ag-specific Ab responses of BALB.B mice

To analyze the IgG subclass of the response induced by a foreignprotein Ag, 10-wk-old BALB.B (n � 5) and BALB/c (n � 5) micewere injected once s.c. with 200 �g of AHGG, and the mice werebled 15 days later. ELISA showed that all mice of both strainsproduced similar levels of IgG1 anti-AHGG Abs, but only a singleBALB/c mouse had a good IgG2a anti-AHGG response (Fig. 4).Thus, young BALB.B mice have normal responsiveness and pro-duce predominantly IgG1 Abs to AHGG.

FIGURE 1. Effect of the H2 haplotype on production of IgG autoantibodies to chromatin. Sera from BALB.B (n � 17–18), BALB/c (n � 12–17),and BALB.K (n � 18) mice were diluted 1/100 and incubated in ELISA wells coated with 5 �g/ml avian nucleosome core particles. Bound Ab wasdetected by using biotinylated goat anti-mouse Fc�, AP-conjugated streptavidin, and p-NPP substrate. Shown are OD values at 405 nm. Each symbolrepresents one mouse. Horizontal bars indicate mean. The values were compared by the two-tailed Mann-Whitney U test. ns, Not significant.

3544 H2b HAPLOTYPE AND ANA

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Impact of increased NaCl concentration on autoantibodybinding avidity

Increasing NaCl concentrations have been used to gauge the avid-ity of Abs to dsDNA (28). We applied this method to assess theavidity of serum IgG2a from 14-mo-old BALB.B mice (n � 5) forimmobilized nucleosome core particles. The results are shown inFig. 5. The average concentration of NaCl that gave 50% of max-imal absorbance was 0.32M. In the presence of 0.5M NaCl, theamount of polyclonal IgG2a Ab bound to nucleosomes was re-duced by 75–90%, whereas that of three lupus IgG2ab mAbs de-rived from (NZB � BXSB)F1 mice with full-blown lupus (29) wasreduced by a modest 25–30%. A likely explanation for the highsensitivity of the BALB.B autoantibodies to moderately increasedNaCl concentration is that they depend heavily on electrostaticcompared with other forces for binding avidity. This interpretationis consistent with the highly cationic IgH CDR3 regions of anti-nucleosome mAbs derived from NZM2410 lupus mice (30).

Spleen weight

Compared with age-matched BALB/c mice (n � 17), the averagepostmortem spleen weight of 12- to14-mo-old BALB.B mice (n �15) was slightly, but significantly, increased (170 mg vs 140 mg;p � 0.0011).

Renal studies and survival

The Multistix scores of urine from 12- to 14-mo-old BALB.B mice(n � 12) did not exceed 1� (0.3 g/L), showing that the mice didnot have clinically significant proteinuria. By light microscopicanalysis (Fig. 6), histopathological abnormalities were not found inthe kidneys of any of 10 14-mo-old BALB.B mice of the cohortshown in Fig. 2A. The survival of BALB.B mice through to 14 moof age (95%) did not differ significantly from that of BALB/c andBALB.K mice (92% and 100%, respectively).

Analysis of C3H and C3.SW mice

The foregoing results prompted us to assess the nonautoimmuneC3H/HeSnJ (H2k) strain and its H2b-congenic C3.SW/SnJ strain.ELISA showed that the mean anti-chromatin IgG2a reactivity of1/200 diluted sera from C3.SW mice (n � 18) was significantlyhigher at all ages compared with the C3H control mice (n � 12)(Fig. 7A). The same C3.SW mice at 12 mo of age tested negativefor chromatin-specific IgG1 (Fig. 6A) and IgG2b Abs (data notshown; IgM and IgG3 were not measured). At 9 and 12 mo of age,83 and 72%, respectively, of the C3.SW mice were deemed pos-itive for anti-chromatin IgG2a (OD value 4 SD above C3H mean).The average anti-chromatin IgG2a concentration among the eight12-mo-old C3.SW mice with the highest ODs was 0.56 �g/ml, i.e.,�25–80-fold lower compared with the mean levels of 15 �g/mland 45 �g/ml found for two independent samples of age-matchedBALB.B mice (see above). The sera from 12-mo-old C3.SW micehad low, but highly significant, levels of IgG2a reactivity to totalhistones (Fig. 7B) and dsDNA (Fig. 7C) compared with sera fromthe C3H group; 39 and 27% of the C3.SW mice were consideredpositive for anti-histone and anti-dsDNA, respectively (OD value4 SD above C3H mean). The same sera were examined at 1/50dilution for reactivity of IgG2a with nuclei of HEp-2 cells.Whereas all 12 C3H sera were ANA negative, eight of 17 C3.SWsera (47%) were ANA positive. Of these, four exhibited homog-enous staining patterns with more bright fluorescence of mitoticchromatin (Fig. 7D, lower panel), indicative of specificity forDNA/histone. The other four sera showed weak speckled nuclear,but not cytoplasmic, staining (Fig. 7D, upper panel). Collectively,these results indicate that the H2b haplotype, when recombined ona C3H background, converts the mice to producers of ANA.

FIGURE 2. Analysis of IgG autoantibodies. A, Isotype profile of anti-chromatin Abs. Sera from 14-mo-old BALB.B mice (n � 17) were diluted5-fold starting at 1/200 and incubated in wells coated with core particles.Bound Ab was detected with biotin-labeled goat anti-IgG subclass Abs andAP-conjugated streptavidin. Ab levels were calculated by comparison withIgG subclass standard curves generated for each plate by sandwich ELISA.The numbers beside the horizontal bars indicate mean (�g/ml). Each sym-bol represents one mouse. B, Anti-chromatin IgG2a levels as a function ofage and H2 haplotype. See legend to A for details of the assay. Shown aremeans and SEM (n � 17–18). C, Effect of H2-haplotype on levels of IgG2aanti-dsDNA Abs at 1 year of age. See the legend to A for details of theassay. Each symbol represents one mouse. Horizontal bars representmeans. Sera from BALB.B (n � 17), BALB/c (n � 16), and BALB.K (n �17) mice were analyzed. The groups were compared by Mann-Whitney Utest. D, Serum levels of IgG2a Abs to ssDNA, dsDNA, and LPS. TheELISA was conducted with sera from 14-mo-old BALB.B mice (n � 17)as described in the legend to A. Note the low Ab levels compared withIgG2a anti-chromatin Abs of A. Each symbol represents one mouse. E,Typical nuclear staining of HEp-2 cells by IgG2a of BALB.B sera. Serumsamples from 1-year-old mice were diluted 1/100 and incubated withHEp-2 cells (Immuno Concepts), followed by biotin-labeled goat anti-mouse IgG2a (2.5 �g/ml), and FITC-labeled streptavidin (10 �g/ml).

3545The Journal of Immunology

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The total serum levels of IgM, IgG2a, and IgG1 were moder-ately (�2-fold) elevated in C3.SW compared with C3H mice at allages (IgG2b and IgG3 were not measured) (Fig. 8). Thus, at 12 moof age the C3.SW vs C3H levels were mean � SEM 0.98 � 0.08vs 0.73 � 0.04 mg/ml ( p � 0.044) for IgM, 0.8 � 0.13 vs 0.43 �0.05 mg/ml ( p � 0.0059) for IgG1, and 2.23 � 0.24 vs 1.25 �0.18 mg/ml ( p � 0.0031) for IgG2a. The average spleen weight ofC3.SW and C3H mice at 12 mo of age was 178 and 150 mg,respectively (not significantly different). All C3.SW and C3H micesurvived through to 1 year of age and appeared healthy.

Anti-chromatin IgG2a/c and IgG1 levels of BALB.B vs B6.Sle1bmice

The spontaneous autoimmunity to chromatin displayed byBALB.B female mice is not unlike that reported for B6.Sle1bmice. We therefore did a side-by-side comparison of levels of anti-chromatin (nucleosome) autoantibodies in sera from 12-mo-oldBALB.B (n � 8) and 9-mo-old B6.Sle1b (n � 8) mice. The serawere diluted 2-fold for seven serial steps starting at 1/200. ForBALB.B mice, the individual levels (�g/ml) of IgG2a Abs were153, 22.4, 13.6, 4.4, 2.1, 1.3, 1.2, and 0.64 (mean 24.8 � 18.5�g/ml). For B6.Sle1b mice, the levels of anti-chromatin IgG2c(also known as IgG2ab) Abs were 86.4, 70.4, 27.2, 10.4, 7.6, 4.8,

and 0.5 (mean 26 � 11.9 �g/ml; not significantly different vsIgG2a levels of BALB.B mice). Anti-chromatin IgG1 was belowdetection in four of eight BALB.B sera and 0.58, 0.54, 0.5, and0.26 �g/ml in the other four. For B6.Sle1 mice, anti-chromatinIgG1 was 0.32 �g/ml in one and below detection in the other sevensera. Thus, sera from these cohorts of BALB.B and B6.Sle1b micedisplayed similar levels of anti-chromatin IgG2a/c and minimalanti-chromatin IgG1.

DiscussionIn this work we demonstrated the following: 1) H2b in the contextof two nonlupus-prone genomes (BALB/c and C3H) is sufficient toinduce production of ANA; 2) the bulk of autoantibodies is spe-cific for intact nucleosome core particles; and 3) the most promi-nent IgG isotype is IgG2a, with some contribution of IgG2b inBALB.B mice. Abs with similar characteristics are conspicuousamong the major spontaneous mouse lupus models (3, 31, 32).However, BALB.B and C3.SW mice differ from these models byminimal representation of the IgG1 isotype, low levels of anti-DNA and anti-total histone reactivity, moderate hypergamma-globulinemia, mild splenomegaly, and no renal disease. The aver-age anti-chromatin Ab levels are 25- to 80-fold higher in BALB.Bcompared with C3.SW mice, indicating that the penetrance alsodepends on non-MHC loci.

Unlike the spontaneous anti-chromatin response, IgG1 Abs pre-vailed over IgG2a in the response of young BALB.B mice to im-munization with adjuvant-free AHGG. This dichotomy may reflectdifferent compositions of the Ags driving these responses. Thus,

FIGURE 3. Total serum IgM, IgG1, andIgG2a concentrations. The levels of BALB.B(n � 17–18), BALB/c (n � 12–18), andBALB.K (n � 18) mice were determined bysandwich ELISA and calculated by compar-ison with serially diluted standards. Mean �SEM are shown. The values of BALB/c andBALB.K mice were compared with those ofBALB.B mice by Mann-Whitney U test.���, p � 0.001, ��, p � 0.01; �, p � 0.05.

FIGURE 4. IgG1 and IgG2a Ab responses induced by AHGG. Tenweek-old BALB.B (n � 5) and BALB/c (n � 5) mice were injected s.c.with 200 �g of AHGG, sera were harvested just before and 15 days later,diluted 1/400, and incubated in wells coated with 1 �g/ml AHGG. BoundIgG1 or IgG2a Ab was determined with biotin-labeled goat anti-mouseIgG1 or IgG2a adsorbed with human IgG (SouthernBiotech) followed byAP-conjugated streptavidin and substrate. IgG1 and IgG2a anti-AHGG Ablevels were calculated by comparison with IgG subclass standard curvesgenerated by sandwich ELISA. Horizontal bars indicate means. The pre-immune sera did not contain anti-AHGG activity (not depicted). ND, Notdetectable.

FIGURE 5. Effect of NaCl concentration (conc.) on interaction ofIgG2a autoantibodies with chromatin. Wells coated with 5 �g/ml nucleo-some core particles were incubated with 1/800 diluted sera from 14-mo-oldBALB.B mice (n � 5) in the presence of NaCl ranging from 0.15 to 0.5M(dashed lines). For comparison, three IgG2ab (IgG2c) mAbs (60 ng/ml)derived from (NZB � BXSB)F1 mice with full-blown lupus were analyzedin parallel (solid lines). Bound Abs were detected with AP-conjugatedanti-IgG2a (for sera) or anti-IgG2c (for mAbs). Maximum OD is OD at 405nm in the presence of 0.15 M NaCl. Shown is a pool of two experiments.

3546 H2b HAPLOTYPE AND ANA

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

switching to IgG2a is promoted indirectly by Th1 cells produc-ing INF-� (33–35) or by direct stimulation of TLR9/MyD88 inB cells by DNA rich in nonmethylated CpG motifs (36, 37).However, CpG DNA can also induce Th1-like responses viaTLR9 (38) and may therefore activate both pathways. The di-rect pathway was responsible for the switching of spontaneousanti-DNA Abs from IgM to IgG2a/2b in B6.Fc�RIIB�/� micetransgenic for the high affinity IgH anti-DNA gene 3H9/56R(denoted B6.56R.Fc�RIIB�/� mice) (39) and for the inducedIgG2a response against a T-dependent nonself Ag mixed withCpG (40). Taken together with the chromatin specificity, thesefindings suggest that the Ag driving the IgG2a class switch ofautoantibodies of BALB.B and C3.SW mice contains CpG-richDNA, likely in the form of nucleosomes.

The benign humoral autoimmunity of BALB/c and C3.SW miceis not unlike that reported for B6.Sle1 and B6.Sle1b congenic micethat produce IgG autoantibodies preferentially targeted to H2A/H2B/DNA subnucleosomes, with little “spreading” of the speci-ficity to histone-free DNA (41, 42). In this study we show that twocohorts of BALB.B and B6.Sle1b female mice aged 12 and 9 mo,respectively, had similar levels of anti-chromatin (nucleosome)IgG2a/2c and minimal expression of IgG1 autoantibodies. One dif-ference between the strains appears to be the age of onset and theprogression of autoimmunity. Thus, for BALB.B mice this type ofANAs is usually clearly detectable by 9 mo of age, and the levelscontinue to increase through to 14 mo of age. B6.Sle1 female mice,by contrast, have detectable ANA as early as 5 mo of age but thelevels peak at 7–9 mo of age and wane thereafter (41).

The lupus susceptibility alleles at the Sle1b locus of NZW miceare present in many standard mouse strains, including BALB/c,C3H/He, and 129, but not B6, mice (7). Are these alleles alsoinvolved in the autoimmunity of H2b haplotype-congenic BALB.Band C3.SW mice? Because the vast majority of strains that carrythe Sle1b susceptibility alleles are not autoimmune (7), it seemslikely that these alleles must interact with polymorphic genes ofthe normal B6 genome to break tolerance to chromatin. This notionis supported by the finding that, unlike nonautoimmune 129 mice,B6 mice congenic for a 129-derived interval on distal chromosome1 (denoted 129chr1b) encompassing the Sle1b locus also produceANA (43). Hence, the genetic backgrounds of BALB.B andC3.SW mice may not be compatible with a role for Sle1 alleles inthe ANA responses of these mice. Another question raised by ourdata, whether the H2b interval of the B6 background is needed forthe loss of tolerance to chromatin occurring in B6.Sle1b mice, hasnot yet been studied.

The lupus of another model, that of Fc�RIIB-null B6 mice, doesdepend on the H2b haplotype. The null mutation of these mice wasgenerated in 129 embryonic stem cells (44) and then backcrossedto B6. The fact that Fc�r2 is tightly linked to Sle1b (45) and thata Sle1b-containing chromosome 1 segment derived from the 129

FIGURE 6. Representative photomicrograph of kidney sections. Kid-neys from 14-mo-old BALB.B mice were fixed in 4% formaldehyde, em-bedded in paraffin, sections were cut, and slides were stained with H&Eplus safran. The image was taken at �400 total magnification.

FIGURE 7. Analysis of autoantibodies of C3.SW and C3H mice. Serum samples diluted 1/200 were incubated in ELISA wells containing immobilizedAg. Bound Ab was detected using biotinylated goat Abs to IgG2a or IgG1 and AP-conjugated streptavidin. Shown are OD405 values. Horizontal barsindicate means. Each symbol represents one mouse. A, Reactivity of IgG2a and IgG1 with nucleosome core particles; B and C, Reactivity of IgG2a withtotal histone (B) and dsDNA (C) at 1 year of age. The values of C3.SW mice (n � 18) and C3H mice (n � 12) were compared by the Mann-Whitney Utest. ��, p � 0.0025; ���, p � 0.0001; ns, not significant. D, Representative immunofluorescence staining of HEp-2 cells by IgG2a of 1/50-diluted serafrom two C3.SW mice at 1 year of age. For details, see legend to Fig. 2D. Upper panel shows speckled nuclear staining. Lower panel shows homogenousstaining with bright fluorescence of mitotic chromatin.

3547The Journal of Immunology

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

genome mediates autoimmunity to chromatin in B6 mice de-spite carrying a normal Fc�r2 gene (43, 46) complicates theinterpretation of the phenotype of Fc�RIIB-null mice. None-theless, B6.Fc�RIIB�/� (H2b), but not BALB/c.Fc�RIIB�/�

(H2d), mice develop IgG Abs to dsDNA and chromatin as wellas glomerulonephritis (44), and spontaneous production of IgGanti-DNA and anti-IgG2a rheumatoid factors was almost elim-inated in B6.Fc�RIIB�/� mice made congenic for H2d (47). AreBALB/c.Fc�RIIB�/� mice protected from lupus because of theirH2d haplotype? When this hypothesis was tested by breeding theH2b haplotype onto BALB/c.Fc�RIIB�/� mice in homozygousform, the mice were still tolerant to chromatin (45).

The inability of H2b to convert the BALB/c.Fc�RIIB�/� strainto ANA producers is contradictory to our own observations ofFc�RIIB-sufficient BALB.B mice, opening the intriguing pos-sibility that BALB.B mice actually need Fc�RIIB for the pro-duction of ANA. The role of Fc�RIIB might be to enable cells

within secondary lymphoid organs to capture chromatin-con-taining IC and present the autoantigen of the IC in a multiva-lent, membrane-bound form to BCR, leading to immunologicalsynapse formation with B cells (48). The ensuing strong acti-vation of BCR could overcome the inhibitory signals mediatedby Fc�RIIB and result in highly efficient delivery of DNA-containing IC to endosomes. In this way, DNA may reachthreshold concentrations required for TLR 9 activation (49).This idea is supported by the following: 1) the observation thatIC-loaded dendritic cells can recycle native Ag to the cell sur-face in an Fc�RIIB-dependent manner and present it to B cells(50); 2) our previous work showing that Fc�RIIB of viablesplenic B cells binds IC containing IgG2ab (IgG2c) anti-nucleosome mAbs in vivo (29); and 3) a recent report indicatingthat follicular dendritic cells depend on Fc�RIIB for IC loadingand strong B cell stimulation (51).

Another explanation for why BALB/c.Fc�RIIB�/�.H2b miceremain tolerant may be that they carry a recently discovered alleleon BALB/c chromosome12 (designated sbb2a) that eliminates an-tinuclear IgG2 production in an otherwise B6.Fc�RIIB�/� back-ground (52). Our own data showing that Fc�RIIB-sufficientBALB.B mice produce ANA despite carrying the sbb2a suppressorlocus suggest that the mechanism underlying their autoimmunity isdistinct from that operating in B6.Fc�RIIB�/�.H2b mice.

The 7.47 cM H2b-congenic interval of BALB.B mice (see theparagraph titled “Mice” in Materials and Methods) contains nu-merous genes, including genes encoding the polymorphic Ag-pre-senting MHC class I and II proteins, complements C2, C4, and B,and the cytokines TNF-�, LT�, and LT� (where LT is lympho-toxin). Thus, the gene(s) in this interval responsible for triggeringANA is unknown. Inasmuch as the present study has revealed acritical impact of the H2b haplotype on an autoantibody response,and given the normal (nonlupus) genetic background of the miceas well as the role of class II H2 allotypes in regulating Ab re-sponses to protein Ag (53), it is tempting to consider I-Ab mole-cules as critical mediators of the BALB.B and C3.SW phenotypes.In this scenario, the role of the I-Ab allotype is visualized as de-pendent on its peptide-binding properties, i.e., the higher affinity ofthe I-Ab molecules of B cells for certain peptides (as yet unchar-acterized) when compared with the MHC II molecules encoded byH2d and H2k, leading Th cells specific for those peptides to bestrongly activated and aid ANA-producing B cells. This hypothe-sis, like that advanced by Iwamoto et al. to explain the protectiveeffect of transgene-encoded E� chains in BXSB and related lupusmice (see Introduction) (20), assumes that the anti-chromatin Abresponses depend on Th cells and autoantigen presentation onI-Ab. In this context, it has been proposed that genetic defects oflupus-prone mice lower TCR activation thresholds for autoantigen(20) (54, 55). However, the normal genetic backgrounds ofBALB.B and C3.SW mice suggest that this putative mechanismdoes not apply to these mice.

With respect to the need for Th cells, a recent study of B6 micetransgenic for the high affinity IgH anti-DNA gene 56R reportedthat, except in young mice, T cells made little difference for thespontaneous development and isotype switch of anti-dsDNA Abs(56). In contrast, as cited by Tsao et al. (56), earlier studies ofmouse lupus models with more normal B cell repertoires, such asMRL/lpr, indicated that T cells play important roles in the pro-duction of anti-DNA Abs. In line with this conclusion, clinicallyhealthy MHC II-deficient B6.Sle1 and B6.Nba2 mice showed min-imal levels of IgG Abs to chromatin, histones, and dsDNA, indi-cating the need for Ag presentation to Th cells (57). Furthermore,

FIGURE 8. Total serum Ig levels of C3.SW and C3H mice. Top panel,IgM level. Middle panel, IgG1 level. Bottom panel, IgG2a level. The levelswere determined by sandwich ELISA and calculated by comparison withserially diluted standards. The data points represent mean � SEM. Thevalues of C3.SW (n � 18) and C3H (n � 12) mice were compared by theMann-Whitney U test. ��, p � 0.01; �, p � 0.05.

3548 H2b HAPLOTYPE AND ANA

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

although TCR��-deficient B6.Sle1 mice produced a high abun-dance of IgM Abs to ssDNA and chromatin, the amount of chro-matin-reactive IgG was minimal, in sharp contrast to TCR��-suf-ficient B6.Sle1 mice (58). Thus, despite the importance of TLR9signals for eliciting production of anti-chromatin IgG in vivo (59–61), it seems that in the presence of a normal B cell repertoire withlow-avidity anti-DNA BCRs, Th cells are also required.

Previous work has shown that inoculation of 3- to 4-wk-oldmice with the Moloney leukemia virus induced a higher proportionof BALB.B mice to become ANA positive by indirect immuno-fluorescence and to develop higher ANA titers when comparedwith 5-mo-old BALB/c mice. Because this difference was not ob-served in nonviremic mice, it was concluded that high viral pro-duction was required for ANA triggering (62). In the present study,spontaneous anti-chromatin IgG became prominent beyond 6 mo,which likely explains why 5-mo-old virus-nonreplicating BALB.Bmice tested negative for ANA. Nonetheless, the findings by Piatieret al. highlight the link between environmental provocation, theH2b haplotype, and ANA responses (62). In this context it has beenreported that, after LPS stimulation, B cells from BALB.B micedisplay higher cell activation, cell proliferation, and IgM secretionwhen compared with B cells from BALB/c mice, suggesting thatgenetic variation at the MHC region contributes to the control of Bcell responsiveness to LPS stimulation (22).

Autoimmunity to chromatin has been suggested to involve sev-eral mechanisms, including altered apoptosis, reduced clearance ofapoptotic material or IC, altered cytokine environment, disruptionof the balance between activation and inhibitory signaling path-ways, inefficient BCR editing, and TLR9 activation. Our observa-tion that the H2b haplotype triggers ANA even in the context ofgenetic backgrounds not prone to lupus indicates that the H2b hap-lotype contains one or more major lupus susceptibility genes or islacking genes that confer resistance to ANA responses. Thus,BALB.B and C3.SW mouse strains may be important tools for thedissection of yet other mechanisms involved in loss of tolerance tochromatin.

AcknowledgmentsWe thank Edward Wakeland (University of Texas Southwestern MedicalCenter) for providing sera from B6.Sle1b mice.

DisclosuresThe authors have no financial conflict of interest.

References1. Burlingame, R. W., R. L. Rubin, R. S. Balderas, and A. N. Theofilopoulos. 1993.

Genesis and evolution of anti-chromatin autoantibodies in murine lupus impli-cates T-dependent immunization with self antigen. J. Clin. Invest. 91:1687–1696.

2. Amoura, Z., H. Chabre, S. Koutouzov, C. Lotton, A. Cabrespines, J. F. Bach, andL. Jacob. 1994. Nucleosome-restricted antibodies are detected before anti-dsDNA and/or antihistone antibodies in serum of MRL-Mp lpr/lpr and �/�mice, and are present in kidney eluates of lupus mice with proteinuria. ArthritisRheum. 37: 1684–1688.

3. Theofilopoulos, A. N. 1985. Murine models of systemic lupus erythematosus.Adv. Immunol. 37: 269–390.

4. Marshak-Rothstein, A., and I. R. Rifkin. 2007. Immunologically active autoan-tigens: the role of Toll-like receptors in the development of chronic inflammatorydisease. Annu. Rev. Immunol. 25: 419–441.

5. Fairhurst, A. M., A. E. Wandstrat, and E. K. Wakeland. 2006. Systemic lupuserythematosus: multiple immunological phenotypes in a complex genetic disease.Adv. Immunol. 92: 1–69.

6. Kono, D. H., and A. N. Theofilopoulos. 2006. Genetics of SLE in mice. SpringerSemin. Immunopathol. 28: 83–96.

7. Wandstrat, A. E., C. Nguyen, N. Limaye, A. Y. Chan, S. Subramanian,X. H. Tian, Y. S. Yim, A. Pertsemlidis, H. R. Garner, Jr., L. Morel, andE. K. Wakeland. 2004. Association of extensive polymorphisms in the SLAM/CD2 gene cluster with murine lupus. Immunity 21: 769–780.

8. Kumar, K. R., L. Li, M. Yan, M. Bhaskarabhatla, A. B. Mobley, C. Nguyen,J. M. Mooney, J. D. Schatzle, E. K. Wakeland, and C. Mohan. 2006. Regulationof B cell tolerance by the lupus susceptibility gene Ly108. Science 312:1665–1669.

9. Vyse, T. J., and B. L. Kotzin. 1998. Genetic susceptibility to systemic lupuserythematosus. Annu. Rev. Immunol. 16: 261–292.

10. Izui, S., N. Ibnou-Zekri, L. Fossati-Jimack, and M. Iwamoto. 2000. Lessons fromBXSB and related mouse models. Int. Rev. Immunol. 19: 447–472.

11. Nath, S. K., J. Kilpatrick, and J. B. Harley. 2004. Genetics of human systemiclupus erythematosus: the emerging picture. Curr. Opin. Immunol. 16: 794–800.

12. Subramanian, S., Y. S. Yim, K. Liu, K. Tus, X. J. Zhou, and E. K. Wakeland.2005. Epistatic suppression of systemic lupus erythematosus: fine mapping ofSles1 to less than 1 mb. J. Immunol. 175: 1062–1072.

13. Cohen, P. L., E. Creech, D. Nakul-Aquaronne, R. McDaniel, S. Ackler,R. G. Rapoport, E. S. Sobel, and R. A. Eisenberg. 1993. Antigen nonspecificeffect of major histocompatibility complex haplotype on autoantibody levels insystemic lupus erythematosus-prone lpr mice. J. Clin. Invest. 91: 2761–2768.

14. Ibnou-Zekri, N., M. Iwamoto, L. Fossati, P. J. McConahey, and S. Izui. 1997.Role of the major histocompatibility complex class II Ea gene in lupus suscep-tibility in mice. Proc. Natl. Acad. Sci. USA 94: 14654–14659.

15. Merino, R., L. Fossati, M. Lacour, R. Lemoine, M. Higaki, and S. Izui. 1992.H-2-linked control of the Yaa gene-induced acceleration of lupus-like autoim-mune disease in BXSB mice. Eur. J. Immunol. 22: 295–299.

16. Merino, R., M. Iwamoto, L. Fossati, P. Muniesa, K. Araki, S. Takahashi,J. Huarte, K. Yamamura, J. D. Vassalli, and S. Izui. 1993. Prevention of systemiclupus erythematosus in autoimmune BXSB mice by a transgene encoding I-E �chain. J. Exp. Med. 178: 1189–1197.

17. Chang, B. L., E. Bearer, A. Ansari, K. Dorshkind, and M. E. Gershwin. 1990. TheBm12 mutation and autoantibodies to Dsdna in Nzb.H-2Bm12 mice. J. Immunol.145: 94–101.

18. Mathis, D. J., C. Benoist, V. E. Williams, M. Kanter, and H. O. McDevitt. 1983.Several mechanisms can account for defective E� gene expression in differentmouse haplotypes. Proc. Natl. Acad. Sci. USA 80: 273–277.

19. Rudensky, A. Y., S. Rath, P. Preston-Hurlburt, D. B. Murphy, and C. A. Janeway,Jr. 1991. On the complexity of self. Nature 353: 660–662.

20. Iwamoto, M., N. Ibnou-Zekri, K. Araki, and S. Izui. 1996. Prevention of murinelupus by an I-E � chain transgene: protective role of I-E � chain-derived peptideswith a high affinity to I-Ab molecules. Eur. J. Immunol. 26: 307–314.

21. Jorgensen, T. N., M. R. Gubbels, and B. L. Kotzin. 2004. New insights intodisease pathogenesis from mouse lupus genetics. Curr. Opin. Immunol. 16:787–793.

22. Rodo, J., L. A. Goncalves, J. Demengeot, A. Coutinho, and C. Penha-Goncalves.2006. MHC class II molecules control murine B cell responsiveness to lipopoly-saccharide stimulation. J. Immunol. 177: 4620–4626.

23. Ausio, J., F. Dong, and K. E. Van Holde. 1989. Use of selectively trypsinizednucleosome core particles to analyze the role of the histone “tails” in the stabi-lization of the nucleosome. J. Mol. Biol. 206: 451–463.

24. Mohan, C., E. Alas, L. Morel, P. Yang, and E. K. Wakeland. 1998. Geneticdissection of SLE pathogenesis. Sle1 on murine chromosome 1 leads to a selec-tive loss of tolerance to H2A/H2B/DNA subnucleosomes. J. Clin. Invest. 101:1362–1372.

25. Hannestad, K. 1969. �M rheumatoid factors reacting with nitrophenyl groups anddenatured DNA. Ann. NY Acad. Sci. 168: 63–75.

26. Wardemann, H., S. Yurasov, A. Schaefer, J. W. Young, E. Meffre, andM. C. Nussenzweig. 2003. Predominant autoantibody production by early humanB cell precursors. Science 301: 1374–1377.

27. Matthews, V. B., F. T. Christiansen, and P. Price. 2000. A novel locus affectingserum levels of IgG2a maps to the murine H2 region. Eur. J. Immunogenet. 27:135–139.

28. Xie, C., Z. Liang, S. Chang, and C. Mohan. 2003. Use of a novel elution regimenreveals the dominance of polyreactive antinuclear autoantibodies in lupus kid-neys. Arthritis Rheum. 48: 2343–2352.

29. Li, X., E. Egorina, E. L. Bertelsen, H. Dahlen, and K. Hannestad. 2004. Anti-nucleosome autoantibodies bind directly to cell lines in vitro and via the Fc�RIIBreceptor to B lymphocytes in vivo: a role for immune complexes in interactionsbetween antinucleosome IgG2a and B cells of BXSB lupus mice. Scand. J. Im-munol. 60: 121–133.

30. Liang, Z., C. Xie, C. Chen, D. Kreska, K. Hsu, L. Li, X. J. Zhou, and C. Mohan.2004. Pathogenic profiles and molecular signatures of antinuclear autoantibodiesrescued from NZM2410 lupus mice. J. Exp. Med. 199: 381–398.

31. Fisher, C. L., R. A. Eisenberg, and P. L. Cohen. 1988. Quantitation and IgGsubclass distribution of anti-chromatin autoantibodies in SLE mice. Clin. Immu-nol. Immunopathol. 46: 205–213.

32. Slack, J. H., L. Hang, J. Barkley, R. J. Fulton, L. D’Hoostelaere, A. Robinson,and F. J. Dixon. 1984. Isotypes of spontaneous and mitogen-induced autoanti-bodies in SLE-prone mice. J. Immunol. 132: 1271–1275.

33. Stevens, T. L., A. Bossie, V. M. Sanders, R. Fernandez-Botran, R. L. Coffman,T. R. Mosmann, and E. S. Vitetta. 1988. Regulation of antibody isotype secretionby subsets of antigen-specific helper T cells. Nature 334: 255–258.

34. Snapper, C. M., and W. E. Paul. 1987. Interferon-� and B cell stimulatory fac-tor-1 reciprocally regulate Ig isotype production. Science 236: 944–947.

35. Finkelman, F. D., I. M. Katona, T. R. Mosmann, and R. L. Coffman. 1988. IFN-�regulates the isotypes of Ig secreted during in vivo humoral immune responses.J. Immunol. 140: 1022–1027.

36. Liu, N., N. Ohnishi, L. Ni, S. Akira, and K. B. Bacon. 2003. CpG directly inducesT-bet expression and inhibits IgG1 and IgE switching in B cells. Nat. Immunol.4: 687–693.

37. Lin, L., A. J. Gerth, and S. L. Peng. 2004. CpG DNA redirects class-switchingtowards “Th1-like” Ig isotype production via TLR9 and MyD88. Eur. J. Immu-nol. 34: 1483–1487.

3549The Journal of Immunology

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

38. Hemmi, H., O. Takeuchi, T. Kawai, T. Kaisho, S. Sato, H. Sanjo, M. Matsumoto,K. Hoshino, H. Wagner, K. Takeda, and S. Akira. 2000. A Toll-like receptorrecognizes bacterial DNA. Nature 408: 740–745.

39. Ehlers, M., H. Fukuyama, T. L. McGaha, A. Aderem, and J. V. Ravetch. 2006.TLR9/MyD88 signaling is required for class switching to pathogenic IgG2a and2b autoantibodies in SLE. J. Exp. Med. 203: 553–561.

40. Jegerlehner, A., P. Maurer, J. Bessa, H. J. Hinton, M. Kopf, and M. F. Bachmann.2007. TLR9 signaling in B cells determines class switch recombination to IgG2a.J. Immunol. 178: 2415–2420.

41. Mohan, C., L. Morel, P. Yang, H. Watanabe, B. Croker, G. Gilkeson, andE. K. Wakeland. 1999. Genetic dissection of lupus pathogenesis: a recipe fornephrophilic autoantibodies. J. Clin. Invest. 103: 1685–1695.

42. Morel, L., K. R. Blenman, B. P. Croker, and E. K. Wakeland. 2001. The majormurine systemic lupus erythematosus susceptibility locus, Sle1, is a cluster offunctionally related genes. Proc. Natl. Acad. Sci. USA 98: 1787–1792.

43. Carlucci, F., J. Cortes-Hernandez, L. Fossati-Jimack, A. E. Bygrave,M. J. Walport, T. J. Vyse, H. T. Cook, and M. Botto. 2007. Genetic dissection ofspontaneous autoimmunity driven by 129-derived chromosome 1 loci when ex-pressed on C57BL/6 mice. J. Immunol. 178: 2352–2360.

44. Bolland, S., and J. V. Ravetch. 2000. Spontaneous autoimmune disease inFc�RIIB-deficient mice results from strain-specific epistasis. Immunity 13:277–285.

45. Bolland, S., Y. S. Yim, K. Tus, E. K. Wakeland, and J. V. Ravetch. 2002. Geneticmodifiers of systemic lupus erythematosus in Fc�RIIB�/� mice. J. Exp. Med.195: 1167–1174.

46. Bygrave, A. E., K. L. Rose, J. Cortes-Hernandez, J. Warren, R. J. Rigby,H. T. Cook, M. J. Walport, T. J. Vyse, and M. Botto. 2004. Spontaneous auto-immunity in 129 and C57BL/6 mice–implications for autoimmunity described ingene-targeted mice. PLoS Biol. 2: E243.

47. Moll, T., L. Nitschke, M. Carroll, J. V. Ravetch, and S. Izui. 2004. A critical rolefor Fc�RIIB in the induction of rheumatoid factors. J. Immunol. 173:4724–4728.

48. Batista, F. D., D. Iber, and M. S. Neuberger. 2001. B cells acquire antigen fromtarget cells after synapse formation. Nature 411: 489–494.

49. Yasuda, K., P. Yu, C. J. Kirschning, B. Schlatter, F. Schmitz, A. Heit, S. Bauer,H. Hochrein, and H. Wagner. 2005. Endosomal translocation of vertebrate DNAactivates dendritic cells via TLR9-dependent and -independent pathways. J. Im-munol. 174: 6129–6136.

50. Bergtold, A., D. D. Desai, A. Gavhane, and R. Clynes. 2005. Cell surface recy-cling of internalized antigen permits dendritic cell priming of B cells. Immunity23: 503–514.

51. El Shikh, M. E., R. M. El Sayed, A. K. Szakal, and J. G. Tew. 2009. T-inde-pendent antibody responses to T-dependent antigens: a novel follicular dendriticcell-dependent activity. J. Immunol. 182: 3482–3491.

52. Tarasenko, T., H. K. Kole, and S. Bolland. 2008. A lupus-suppressor BALB/clocus restricts IgG2 autoantibodies without altering intrinsic B cell-tolerancemechanisms. J. Immunol. 180: 3807–3814.

53. Schwartz, R. H. 1985. T-lymphocyte recognition of antigen in association withgene products of the major histocompatibility complex. Annu. Rev. Immunol. 3:237–261.

54. Vratsanos, G. S., S. Jung, Y. M. Park, and J. Craft. 2001. CD4� T cells fromlupus-prone mice are hyperresponsive to T cell receptor engagement with lowand high affinity peptide antigens: a model to explain spontaneous T cell activa-tion in lupus. J. Exp. Med. 193: 329–337.

55. Martinez-Soria, E., N. Ibnou-Zekri, M. Iwamoto, M. L. Santiago-Raber,S. Kikuchi, M. Kosco-Vilbois, and S. Izui. 2004. Epitope-dependent inhibition ofT cell activation by the Ea transgene: an explanation for transgene-mediatedprotection from murine lupus. J. Immunol. 173: 2842–2848.

56. Tsao, P. Y., J. Jiao, M. Q. Ji, P. L. Cohen, and R. A. Eisenberg. 2008. T cell-independent spontaneous loss of tolerance by anti-double-stranded DNA B cellsin C57BL/6 mice. J. Immunol. 181: 7770–7777.

57. Stohl, W., D. Xu, T. E. Metzger, K. S. Kim, L. Morel, and B. L. Kotzin. 2004.Dichotomous effects of complete versus partial class II major histocompatibilitycomplex deficiency on circulating autoantibody levels in autoimmune-pronemice. Arthritis Rheum. 50: 2227–2239.

58. Sobel, E. S., M. Satoh, Y. Chen, E. K. Wakeland, and L. Morel. 2002. The majormurine systemic lupus erythematosus susceptibility locus Sle1 results in abnor-mal functions of both B and T cells. J. Immunol. 169: 2694–2700.

59. Lartigue, A., P. Courville, I. Auquit, A. Francois, C. Arnoult, F. Tron, D. Gilbert,and P. Musette. 2006. Role of TLR9 in anti-nucleosome and anti-DNA antibodyProduction in lpr mutation-induced murine lupus. J. Immunol. 177: 1349–1354.

60. Christensen, S. R., M. Kashgarian, L. Alexopoulou, R. A. Flavell, S. Akira, andM. J. Shlomchik. 2005. Toll-like receptor 9 controls anti-DNA autoantibody pro-duction in murine lupus. J. Exp. Med. 202: 321–331.

61. Christensen, S. R., J. Shupe, K. Nickerson, M. Kashgarian, R. A. Flavell, andM. J. Shlomchik. 2006. Toll-like receptor 7 and TLR9 dictate autoantibody spec-ificity and have opposing inflammatory and regulatory roles in a murine model oflupus. Immunity 25: 417–428.

62. Piatier, D., P. Debre, P. S. Mach, and S. Gisselbrecht. 1981. Genetic control ofantinuclear antibodies in mice inoculated with the Moloney leukaemia virus.J. Immunogenet. 8: 177–184.

3550 H2b HAPLOTYPE AND ANA

by guest on March 23, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from