Cell Activation and Systemic Autoimmunity Mutant Mice Triggers

The Journal of Immunology

The Th2 Lymphoproliferation Developing in LatY136F Mutant Mice Triggers Polyclonal B Cell Activation and Systemic Autoimmunity1

Ce´line Genton,* Ying Wang,† Shozo Izui,‡ Bernard Malissen,† Georges Delsol,§¶ Gilbert J. Fournie´,ሻ Marie Malissen,2† and Hans Acha-Orbea2, 3*

LatY136F knock-in mice harbor a point mutation in Tyr136 of the linker for activation of T cells and show accumulation of Th2 effector cells and IgG1 and IgE hypergammaglobulinemia. B cell activation is not a direct effect of the mutation on B cells since in the absence of T cells, mutant B cells do not show an activated phenotype. After adoptive transfer of linker for activation of T cell mutant T cells into wild-type, T cell-deficient recipients, recipient B cells become activated. We show in vivo and in vitro that the LatY136F mutation promotes T cell-dependent B cell activation leading to germinal center, memory, and plasma cell formation even in an MHC class II-independent manner. All the plasma and memory B cell populations found in physiological T cell-dependent B cell responses are found. Characterization of the abundant plasmablasts found in secondary lymphoid organs of LatY136F mice revealed the presence of a previously uncharacterized CD93-expressing subpopulation, whose presence was confirmed in wild-type mice after immunization. In LatY136F mice, B cell activation was polyclonal and not Ag-driven because the increase in serum IgG1 and IgE concentrations involved Abs and autoantibodies with different specificities equally. Although the noncomplement-fixing IgG1 and IgE are the only isotypes significantly increased in LatY136F serum, we observed early-onset systemic autoimmunity with nephritis showing IgE autoantibody deposits and severe proteinuria. These results show that Th2 cells developing in LatY136F mice can trigger polyclonal B cell activation and thereby lead to systemic autoimmune disease. The Journal of Immunology, 2006, 177: 2285–2293.

ystemic autoimmune disease is often associated with au- deposition of immune complexes in kidney glomeruli is part of the toantibody production. Systemic lupus erythematosus, for disease process (for review, see Ref. 3). Abs with (cross)-reactivity S example, is a chronic autoimmune condition that results in to glomerular membrane proteins have been suggested to be of inflammation and kidney damage. Many disease susceptibility importance (4, 5). Several excellent mouse models of spontaneous genes are involved, and several of these affect survival and acti- nephritis exist. Models with germinal center (GC)4 formation, af- vation of B lymphocytes resulting in increased autoantibody pro- finity maturation, and preferential amplification of autoantibody- duction (1, 2). Autoantibodies directed toward cellular and extra- producing plasma cells, as well as models with polyclonal activa- cellular components such as DNA, chromatin, cell membrane, and tion of B cells, resulting in similar disease patterns, have been matrix play an important role in the development of nephritis, and described previously (6–8). During T-dependent immune responses, Ag-specific, naive T cells are primed by Ag-presenting dendritic cells (DC) (9). B cells *Department of Biochemistry, University of Lausanne, Epalinges, Switzerland; †Cen- recognizing native Ag through their surface Ig are activated in tre d’Immunologie de Marseille-Luminy, Institut National de la SanteétdelaRe- parallel in the follicles and preferentially localize to the interfol- cherche Me´dicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)- Universite´delaMe´dite´rraneé, Marseille, France; ‡Department of Pathology and licular regions of secondary lymphoid organs (10–12). There, they Immunology, University of Geneva, Geneva, Switzerland; §INSERM Unite´ 563, Cen- receive T cell help if they are able to present to the primed T cells tre de Physiopathologie de Toulouse-Purpa, Toulouse, F-31300 France; ¶ De´parte- the same peptides that were presented by DC during T cell priming ment Oncogeńe`se et Signalisation dans les Cellules He´matopoie´tiques, Institut Fe´d- e´ratif de Recherche 30 and Laboratoire d’Anatomie Pathologique, Purpan’s Hospital, (11, 13). Interaction between T and B cells constitutes a key event Toulouse F-31300, France; ࿣INSERM Unite´ 563, Centre de Physiopathologie de Tou- in both the formation of extrafollicular plasmablasts producing # louse-Purpan, Toulouse, F-31300 France; and De´partement de Geńe´tique Fonction- IgM or switched isotypes and in the initiation of the GC reaction nelle des Maladies des E´ pithe´liums, Institut Fe´de´ratif de Recherche 30, Purpan’s Hospital, Toulouse F-31300, France in the follicular regions of secondary lymphoid organs (14). In GC, Received for publication November 28, 2005. Accepted for publication May isotype switching, affinity maturation, and differentiation into 23, 2006. short- or long-lived plasma cells or memory B cells are achieved The costs of publication of this article were defrayed in part by the payment of page with the help of follicular DC (FDC). After having recaptured the charges. This article must therefore be hereby marked advertisement in accordance (same) priming Ag from FDC, apoptosis-sensitive centrocytes re- with 18 U.S.C. Section 1734 solely to indicate this fact. ceive T cell help from follicular T cells (14). These steps ensure 1 This work was supported by the Swiss National Science Foundation (to H.A.-O. and S.I.), INSERM, CNRS, Ministe`re de l’Education Nationale et de la Recherche (Plate- survival and maturation of B cells that have maintained Ag spec- forme Rassemblement Inter Organsimes), the European Communities (MUGEN Net- ificity and optimized their affinity for Ag (15). work of Excellence) (to B.M. and M.M.), and Poˆles Alliances des Recherches sur le Cancer (to B.M. and G.D.), as well as the Experimental Histopathology platform of IFR30. 4 Abbreviations used in this paper: GC, germinal center; BAFF, B cell-activating 2 M.M. and H.A.-O. contributed equally to this work. factor belonging to the TNF family; DC, dendritic cell; FDC, follicular dendritic cell; 3 Address correspondence and reprint requests to Dr. Hans Acha-Orbea, Department LAT, linker for activation of T cell; PNA, peanut lectin (agglutinin); WT, wild type; of Biochemistry, University of Lausanne, Chemin Des Boveresses 155, CH-1066 NP, 4-hydroxy-3-nitrophenylacetate; PALS, periarteriolar lymphoid sheet; TNP, Epalinges, Switzerland. E-mail address: [email protected] 2,4,6-trinitrophenyl.

Physiological T cell-B cell interactions are sometimes not as mice were maintained at the Swiss Institute for Experimental Cancer Re- specific as might be expected from the above description. It has search. All animal experiments were done in agreement with Institutional been observed that only ϳ10% of plasma cells secrete Abs specific and Swiss regulations. for the encountered Ags during an immune response induced by Immunohistological analysis different protein Ags such as sheep RBC (16). Both Ag-indepen- Studies were performed on acetone-fixed frozen sections and stained using dent polyclonal B cell activation by Ag-primed T cells and Ag- standard methods. The following reagents were used: biotinylated anti- dependent T cell help to B cells that have captured Ag through B220 (RA3-6B2; Caltag Laboratories), anti-CD4 (H129) culture superna- other mechanisms than uptake via surface Ig have been described tant, biotinylated peanut lectin (agglutinin) (PNA) (Vector Laboratories), previously (17). In this latter mode of “unspecific” T-B cell help, biotinylated anti-IgG1 (Caltag Laboratories), biotinylated anti-IgE (Uni- ϳ versity of Louvain), and goat anti-C3 (Cappel Laboratories). Biotinylated the amounts of Ag required are 10,000 times higher than those Abs were visualized with HRP-conjugated streptavidin (Jackson Immu- required for T-B interactions involving B cells expressing an Ag- noResearch Europe), except for the double staining where anti-B220 was specific Ig (17, 18). detected with alkaline phosphatase-conjugated streptavidin (Boehringer Linker for activation of T cell (LAT) is an adaptor protein that Mannheim). Nonconjugated supernatant was detected using HRP-conju- constitutes a major substrate for the Zap70 protein tyrosine kinase gated goat anti-rat (BioSource International). Goat Abs were detected using HRP-conjugated anti-goat Ab (Chemicon International). The substrate in T cells (19). Once phosphorylated, LAT coordinates the assem- used for HRP was 3,3Ј-diaminobenzidine/urea tablets (Sigma-Aldrich) bly of a multiprotein signaling complex that links the TCR to the complemented with hydrogen peroxide and for alkaline phosphatase Fast main intracellular pathways regulating T cell development and Blue BB salt (Sigma-Aldrich). Counterstaining was done using Mayer’s function. LatY136F knock-in mice carry a mutation in LAT that hematoxylin. For immunofluorescent detection of kidney deposits, anti- IgG1 was detected with Cy3-conjugated streptavidin (Jackson ImmunoRe- replaces Tyr at position 136 with phenylalanine and principally search Europe), anti-IgE with Alexa 488-conjugated anti-rat Ab (Invitro- 136 eliminates binding of phospholipase C-␥1. Mutation of Tyr re- gen Life Technologies), and anti-C3 with Alexa 488-conjugated anti-goat sults in a partial block at the two developmental checkpoints that Ab (Invitrogen Life Technologies). Histopathologic examination was per- Y136F punctuate intrathymic ␣␤ T cell development (20, 21). However, formed on 12 Lat mice and 4 wild-type (WT) mice. All major organs beginning at ϳ2–3 wk of age, LatY136F mice develop a lympho- (spleen, thymus, lymph node, lung, and kidney) were fixed in neutral buff- ered 10% Formalin. Paraffin-embedded sections were stained with H&E proliferative disorder involving polyclonal CD4 T cells that pro- and with Giemsa. Immunohistochemistry was performed with CNA.42 Ab duce high amounts of Th2 cytokines. This exaggerated Th2 dif- directed against FDC (24) After incubation with the aforementioned Abs, ferentiation is most likely responsible for the massive activation of sections were immunostained by the streptavidin-biotin-peroxidase com- B cells and the hypergammaglobulinemia of IgE and IgG1 iso- plex method using DakoCytomation StrepABComplex/HRP Duet (Mouse/ Rabbit) Kit (code K0492; DakoCytomation) as described elsewhere (25). types that ensues. Immunostaining on paraffin sections was performed using the method de- In this study, we show that the polyclonal B cell activation in scribed by Shi et al. (26) with some modifications (25). LatY136F mice is not due to a direct (intrinsic) effect of the LAT Flow cytometry mutation on B cells but indirectly results from the presence of CD4 Th2 effector cells. This polyclonal B cell activation involves the Single-cell suspensions were stained using standard techniques. mAbs used formation of abundant GC and leads to the formation of abnor- for flow cytometry were B220 (RA3-6B2; eBioscience), MHC class II (2G9; BD Pharmingen), syndecan-1 (281-2; BD Pharmingen), GL-7 mally large numbers of memory B cells, plasmablasts, and plasma Y136F (Ly77; BD Pharmingen), PNA (Sigma-Aldrich), CD4 (GK1.5; BD Pharm- cells. The exaggerated B cell differentiation occurring in Lat ingen), CD8 (53-6.7; Biolegend), CD93 (PB493; BD Pharmingen), and mice allowed the characterization of a novel CD93-positive plas- IgD (11-26c.2a). Biotinylated Abs were visualized with streptavidin-allo- mablast subpopulation that was also found in normal B cell re- phycocyanin (Caltag Laboratories) and streptavidin/PE/Cy5.5 (eBio- ␥ sponses. The GC and plasma cell populations found in LatY136F science). Culture supernatant 2.4G2 has been used to block Fc Rs. Flow cytometry was performed on a FACSCalibur flow cytometer. Analysis was mice are phenotypically comparable to those found in physiolog- performed on a live gate based on forward and side scatter characteristics ical immune responses. Surprisingly, the B cell help provided by using CellQuest software 3.2.1f1 (BD Pharmingen) and FlowJo 6.2.1. the LatY136F CD4 T cells was independent of TCR-MHC interac- Analysis of Ab titers tion, as documented by the fact that LatY136F CD4 T cells induced MHC class II-negative B cells to secrete Ab in vitro as efficiently Serum levels of total IgM-, IgG1-, IgG2a-, IgE-, and 4-hydroxy-3-nitro- as MHC class II-expressing B cells. In vivo, adoptive transfer of phenylacetate (NP)-specific IgG1 Abs were quantified by ELISA using Y136F polyclonal goat Abs specific for mouse Ig isotypes for detection (Caltag Lat T cells resulted in reduced, but clearly detectable, MHC Laboratories) and o-phenylenediamine developing reagents (Sigma- class II-independent GC formation and polyclonal Ab secretion. Aldrich). To detect anti-NP Abs, plates were precoated with NP (23)-BSA The Th2-mediated polyclonal B cell activation observed in (Biosearch Technologies). For quantifying serum urea, the kit from Sigma- LatY136F mice resulted in early-onset systemic autoimmune disease Aldrich was used. with nephritis. The increase in autoreactive Abs observed in In vitro coculture assay Y136F Lat mice was not Ag driven but was proportional to the in- ϩ crease in total Ig. Based on their IgG1 and IgE isotypes, the au- CD4 cells were first enriched by MACS (Miltenyi Biotec) using anti-CD4 Y136F beads and then sorted by FACS for maximal purity (at least 99%). A total toantibodies present in Lat mice were expected to be less of 3 ϫ 104 CD4ϩ cells was cultured with 9 ϫ 104 WT or MHC class II Ϫ/Ϫ pathogenic than complement-fixing autoantibodies. However, ne- splenocytes in 200 ␮l of DMEM (Invitrogen Life Technologies) with 10% phritis with IgE and in ϳ30% of cases complement deposits in the FCS (Brunschwig), 10 mM HEPES buffer (Invitrogen Life Technologies), ␮ ␮ glomeruli, increased serum urea levels, as well as proteinuria, were 200 g/ml gentamicin (Invitrogen Life Technologies), and 50 M 2-ME Y136F (Invitrogen Life Technologies). IgG1 was quantified in culture superna- readily observed in young Lat mice. tants by ELISA at different time points. Concentration of IgG1 was determined by comparing a test sample dilution series with that using a control Materials and Methods IgG1 standard (BD Pharmingen). Mice Adoptive transfer LatY136F mice (20), Cd3-␧⌬5/⌬55 mice (22), and Cd3-e⌬5/⌬5 ϫ LatY136F Single-cell suspensions of spleen and lymph nodes from 5- to 8-wk-old mice have been obtained from the Centre d’Immunologie de Marseille- donor mice were enriched for CD4 T cells using magnetic beads conju- Luminy and were backcrossed for more than six generations to the gated with CD4 (Miltenyi Biotec) on an AutoMACS. A total of 3 ϫ 106 C57BL/6 background. MHC class IIϪ/Ϫ mice ((I-A␣Ϫ/Ϫ) were a gift from CD4ϩ T cells (purity Ͼ 95%) from the specified mice was injected i.v. into W. Reith (Centre Me´dical Universitaire, Geneva, Switzerland) (23). All 4- to 6-wk-old CD3-␧⌬5/⌬5 mice or CD3- ␧⌬5/⌬5 ϫ IA␤Ϫ/Ϫ mice. Mice The Journal of Immunology 2287

FIGURE 1. B cell activation in LatY136F mice. A,HϩE staining of naive lymphocytes in WT lymph node. B, Ac- cumulation of plasma cells in PALS from 6-wk-old LatY136F mice. C, Immunohistochemical analysis of lymph node sections from 6-wk-old WT and 2-, 4-, and 6-wk-old LatY136F mice. Anti-B220 (blue) and anti-CD4 (brown) staining shows the T and B cell distribution. PNA (brown) staining indicates GC formation. Nuclei are counterstained with Mayer’s hematoxylin (blue). D, T cell, GC, and plasmablast/plasma cell content per lymph node were plotted using logarithmic scale. All experiments are representative of two individuals per condition.

were treated with Bactrim and maintained in specific pathogen-free con- Statistical analysis ditions for 8 wk before analysis. Statistical analysis was performed with the Mann-Whitney rank-sum test to Autoimmune disease analysis compare WT and LatY136F mouse autoantibody level, Ig fold increase, kidney disease index, and for in vitro assay. The Fisher exact test has been Serum levels of IgG and IgE autoantibodies against chromatin and DNA used to determine significant differences between WT and LatY136F pro- were determined by ELISA as described previously (27, 28). Results are teinuria analysis. Values of p Ͻ 0.05 were determined as significant for expressed either as units per milliliter, in reference to a standard curve each of the tests. lpr derived from a serum pool of MRL-Fas mice, or as OD405. A flow cytometric assay was used to detect Coombs’ anti-erythrocyte autoanti- ␬ Results bodies using PE-labeled anti-mouse L chain Abs (BD Pharmingen), as Y136F described previously (29). The results are expressed as mean fluorescence Status of the B cell compartment in Lat mice intensity. The presence of proteinuria was measured in a semiquantitative Adult LatY136F mice show serum IgG1 and IgE concentrations that way using Albustix (Bayer). Mice were scored as positive for proteinuria when protein levels exceeded 300 mg/dl defined by the colorimetric test for are elevated 200 and 10,000 times, respectively, compared with two measures at an interval of at least 1 wk. Paraformaldehyde-fixed kid- WT mice (20, 21). Histopathologic examination of the spleen ney histological sections were stained with periodic acid-Schiff reagent. showed white pulp hyperplasia. Periarteriolar lymphoid sheaths The extent of glomerulonephritis was graded on a scale of 0–4 based on (PALS) were pale with a striking accumulation of plasma cells the intensity and extent of histological changes, as described previously associated with plasmablasts and large immunoblast-like cells (30). The scoring was performed in a double-blind mode. Ab deposits in Y136F the kidney were detected using anti-IgG1 and anti-IgE staining of kidney (Fig. 1, A (control) and B (Lat )). Variable numbers of eosin- sections. ophils and basophils/mast cells were found. Lymph nodes were

FIGURE 2. LatY136F B cell status in the absence of LatY136F T cells. A, Immunohistochemical analysis of lymph node sections from 3-mo-old WT, LatY136F, Cd3- ␧⌬5/⌬5, and Cd3-␧⌬5/⌬5 KO LatY136F mice. B cells were stained with anti-B220 (brown), and nuclei were counterstained with Mayer’s hematoxylin (blue). B, Flow cytometry analysis was performed on lymph nodes to quantify B cell populations. GC B cells defined as GL-7ϩ PNA high are shown as percent of B220ϩIgDϪ cells. The percentage of plasma cells (B220Ϫsyndecan-1ϩ) is indicated. C, IgG1 serum levels were measured by ELISA. The concentrations are plotted on a logarithmic scale. All experiments are representative of two individuals per condition. 2288 Th2-MEDIATED SYSTEMIC AUTOIMMUNITY IN LatY136F MICE enlarged and showed variable hyperplasia of lymphoid follicles affected later stages of B cell activation and differentiation, we and an accumulation of cells with morphologic features compara- generated double-mutant mice deficient for both CD3␧ polypep- ble to that observed in the spleen. tides (Cd3-␧⌬5/⌬5) (22) and the LatY136F mutation. In these double- Analysis of the development of this pathologic phenotype mutant mice, direct activation of B cells via T cell help can be showed that until 2 wk after birth, most B lymphocytes had a naive excluded because they lack both ␣␤ and ␥␦ T cells. FACS analysis phenotype (data not shown). At this age, we observed a normal of developing B cells in the bone marrow and of the maturation distribution of B cells in lymphoid follicles but almost a complete and activation status of mature B cells in the spleen and lymph absence of CD4 T cells in the paracortex and PALS (Fig. 1C and nodes of Cd3-␧⌬5/⌬5 ϫ LatY136F double-mutant mice showed no data not shown). FACS analysis confirmed these results and differences compared with double heterozygous or WT littermate showed that in 2-wk-old LatY136F mice there were eight times controls (data not shown). Histological analysis of lymph nodes fewer CD4 T cells in lymph nodes than in littermate controls (Fig. from double-mutant mice showed no effect on B cell localization 1D). As previously described (20, 21), there were few CD8 T cells or activation, and no GC developed in 3-mo-old mice (Fig. 2, A in LatY136F mice at all ages examined. and B). FACS analysis showed that most B220ϩ cells had a naive In secondary lymphoid organs of 4-wk-old LatY136F mice, CD4 phenotype as in WT mice (Fig. 2B). When Cd3-␧⌬5/⌬5 ϫ LatY136F T cell numbers increased and were mostly localized to the para- mice were compared with WT and Cd3-␧⌬5/⌬5 mice, they showed cortex and the PALS. Coincident with this increase in CD4 T cells, no increase in either GC B cell populations or in Ab-secreting cells large PNAϩ GC appeared in every lymph node and spleen B cell (Fig. 2B). Moreover, there was no B cell lymphoproliferation in follicle (Fig. 1, C and D, and data not shown). In addition to ac- Cd3-e⌬5/⌬5 ϫ LatY136F mice; their lymph node cellularity was sim- tivated B and T cells, both lymph nodes and spleen contained large ilar to that of Cd3-e⌬5/⌬5 mice. Their serum levels of IgG1 and IgE quantities of Ab-secreting (B220low, MHC class IIint, syndecan- were also comparable to those of Cd3-e⌬5/⌬5 mice (Fig. 2C). 1ϩ) plasmablasts (Fig. 1D). In 6-wk-old LatY136 mice, Ab-secret- Therefore, the massive B cell activation observed in young ing plasmablasts/plasma cells (as defined by detecting B220 down- LatY136F mice was not due to a direct effect of the LatY136F point regulation and syndecan-1 expression) reached 5% of total lymph mutation on B cells but was clearly T cell dependent. node cells. Considering that lymph nodes of 6-wk-old LatY136F 6 mice are five times bigger than those of WT controls, 0.6 ϫ 10 Phenotype of GC and Ab-secreting cells in LatY136F mice Ab-secreting cells, 2.3 ϫ 106 GC cells, and 6 ϫ 106 T cells could To compare the phenotype of activated and differentiated B cells in be isolated from a single LatY136 lymph node (Fig. 1D). This rep- LatY136F mice, we analyzed their surface phenotype. All memory resents an ϳ500-fold increase in plasmablasts/plasma cell num- and effector B cell populations that characterize a primary carrier- bers compared with either age-matched, heterozygous littermate hapten-specific immune response were found in LatY136F mice controls or to WT mice. Analysis of LatY136F mice showed that (Fig. 3, A and B, and data not shown). Surprisingly, a prominent lymph node activation became more severe with time, leading to a proportion of plasmablasts/plasma cells expressed CD93, a previ- complete disruption of lymphoid follicle architecture by 3 mo of ously unrecognized marker of this population (Fig. 3C). The same age (data not shown). Thus, there was a clear correlation between population can also be found in physiological immune responses the increase in splenic and lymph node CD4 T cell number, Th2 (data not shown). Therefore, the B cell subpopulations seen in phenotype, and the activation status of B cells. LatY136F mice are comparable to those found in physiological im- The LatY136F mutation acts indirectly on B cells mune responses, differing in the strongly increased number in the LatY136F mice. LAT is expressed during early stages of B cell differentiation at levels, however, strikingly lower than in T cells (31, 32), becoming Y136F undetectable at later stages of B cell differentiation (31, 32). It has Polyclonal B cell activation in Lat mice is detected in the been suggested that LAT links the pre-BCR to calcium signaling. absence of MHC class II However, its absence has no measurable effect on either B cell To examine whether the aberrant CD4 T cells developing in differentiation or activation (31–33). To exclude that expression of LatY136F mice activate B cells polyclonally or primarily activate mutant LATY136F molecules at early stages of B cell development autoreactive B cells, we compared the relative titers of anti-NP,

FIGURE 3. Plasma cell compartment in LatY136F lymph node. A, Flow cytometric analysis of plasma cells in lymph node from 6-wk-old WT and 2-, 4-, and 6-wk- old LatY136F mice. Syndecan-1ϩ cells are shown as black dots on the B220/MHC class II graph. B, Percentage of syndecan-1ϩ among B220Ϫ MHC class IIϩ cells in the lymph node. WT mice (ﬁlled) are compared with LatY136F mice (black line). C, CD93 expression by syndecan-1ϩ cells in lymph node from 2-mo-old LatY136F mice. The Journal of Immunology 2289

Table I. Increase in Ab in LatY136F mice Abs were easily detectable after 6 days, and levels further increased with time. Purified WT B cells together with WT CD4 T Isotype/Specificity Fold Increasea pb or sorted CD4 T cells from LatY136F mice cells did not induce production of detectable amounts of IgG1. Based on our observa- IgM 6 0.026 ␤ IgG1 158 Ͻ0.001 tion that blocking Abs to 2 integrin completely prevented IgG1 IgG2a 0.08 variable! NS secretion, we conclude that B cell activation is contact dependent. IgG3 5 Ͻ0.01 In addition, blocking IL-4 with mAbs prevented IgG1 secretion IgE 225Ј864 Ͻ0.001 (data not shown). When B cells lacking MHC class II expression Anti-NP IgG1 20 0.001 (I-A␣Ϫ/Ϫ) were incubated in the presence of LatY136F CD4 T cells, Anti-DNA IgG1 100 Ͻ0.001 Anti-laminin IgG 17 0.019 they produced levels of IgG1 similar to WT B cells (Fig. 4B). In control experiments with WT CD4 T cells and WT B cells, no a Serum from seven WT mice have been compared with serum from seven LatY136F mice. Median ratio of serum Ab titers have been used to determine Ig fold induction of Ab secretion was detected during the coculture period increase from LatY136F mice compared with WT mice. Total Ig is increased by (data not shown). Therefore, consistent with the poor signaling 30-fold. Y136F b properties of the TCR expressed on Lat CD4 T cells (20, 21), Values have been determined by the Mann-Whitney rank-sum test. All Ig levels Y136F measured from LatY136F mice, except IgG2a, were statistically different from WT the fact that help delivered to B cells by Lat CD4 T cells does mice. not rely on interaction between the TCR and MHC class II molecules likely accounts for the production of an unbiased IgG1 and IgE repertoire. anti-2,4,6-trinitrophenyl (TNP) Abs, and autoantibodies in nonim- Y136F To analyze this point further, we adoptively transferred purified munized Lat mice with those of heterozygous littermate con- ⌬ ⌬ LatY136F CD4 T cells into Cd3-e 5/ 5 mice that were MHC class trols (see Table I). II expressing or deficient. In both instances, clear induction of GC Total Ig serum titers were increased 30-fold in 3-mo-old (Fig. 4C) and Ab secretion was detected (data not shown). Al- LatY136F mice. Among the predominant IgG1, which were in- though CD4 T cells from LatY136F mice do provide B cell help creased 158-fold, anti-NP-Abs were increased 20-fold and anti- MHC class II independently in vitro, the GC induced after adop- TNP 100-fold. Autoantibodies directed against DNA were in- tive transfer into MHC class II- negative recipients were smaller creased 100-fold. Other autoantibodies directed at chromatin and than those observed following transfer into MHC class II-positive laminin, as well as Abs to control proteins such as chicken gam- recipients. However, increases in Ab titers were only slightly de- maglobulin, showed low to undetectable reactivity in WT mice layed in MHC class II-deficient recipients. Taken together, these but, when detectable in controls, were increased proportional to the results suggest that the B cells found in LatY136F mice are activated total Ig increase. Proportionality for IgE could not be measured as by polyclonal T cell help and that MHC class II expression is not normal mice had very low serum IgE levels. Taken together, these required for induction of B cell activation, formation of GC cen- results indicate that LatY136F mice develop polyclonal B cell acti- ters, isotype switch, or Ab secretion. vation without a bias for B cells secreting autoantibodies. Y136F We next tested whether the help provided to B cells by Lat Y136F CD4 T cells required interaction between TCR and MHC class II Lat mice develop autoantibodies and a systemic molecules. For this purpose, we first mixed FACS-sorted LatY136F autoimmune syndrome with involvement of kidney, lung, and CD4 T cells with WT B cells and measured in vitro the induction liver of IgG1 production after 2–18 days of culture. The kinetics of LatY136F mice have elevated total IgG1 and IgE serum titers and IgG1 production by WT B cells is shown in Fig. 4A. When WT B anti-nuclear Abs (20, 21). Results shown in Table I and Fig. 5, A cells were cultured in the presence of LatY136F CD4 T cells, IgG1 and B, illustrate that B cells undergo polyclonal activation without

FIGURE 4. Independence on TCR-MHC interaction for B cell activation by LatY136F CD4ϩ T cells. A, Total WT splenocytes were cultured in the presence or absence of FACS-sorted CD4ϩ LatY136F cells. B cell activation and differentiation was monitored by measuring IgG1 production in culture supernatant by ELISA after 2, 6, 12, and 18 days of culture. Purified CD4ϩ LatY136F cells or WT splenocytes alone were used as negative controls. B, Splenocytes from MHC class II-deficient mice were used in the same assay in the presence (f) or absence (u)of CD4ϩ LatY136F cells. Data are representative of at least three independent experiments. C, CD4ϩ cells were transferred into a Cd3-␧⌬5/⌬5 recipient. Histological analyses were performed on spleen sections. Anti-B220 (blue) and anti-CD4 (brown) staining shows the T and B cell distribution. PNA (brown) staining indicates GC formation. Nuclei are counterstained with Mayer’s hematoxylin (blue). Significant differences as calculated by the Mann- .p Ͻ 0.01 ,ءء .Whitney rank-sum test are indicated 2290 Th2-MEDIATED SYSTEMIC AUTOIMMUNITY IN LatY136F MICE

works observed in the lymph nodes and spleen were also observed in these bronchial and perivascular lung infiltrates which were stained with the FDC-specific CNA.42 Ab (Fig. 7, B and C). Lym- phoid infiltrates with plasma cells comparable to those found in lungs were also observed in portal spaces of the liver (Fig. 7D). We did not detect basophil or eosinophil infiltrates in the kidneys, but eosinophil-infiltrates were detected readily in the spleen. In addition to lymphoid cell infiltrates, variable numbers of basophils/ mast cells and eosinophils were also found in liver and lung lesions. These data indicate that LatY136F mice developed chronic systemic autoimmune disease, which was apparent already at the age of 2–3 mo.

Discussion We have characterized the B cell response induced by the polyclonally activated T cell population that expands in LatY136F mutant mice in the absence of any intentional immunization and in- FIGURE 5. Generation of high levels of autoantibodies by LatY136F vestigated the consequences for autoimmunity. We show here that mice. A, Flow cytometry on RBC stained with anti-␬ L chain. Data are naive B cells are activated polyclonally and form massive GC, as represented as mean fluorescent intensity (MFI). B, Serum level of anti- well as plasma cells and memory B cells. A new plasmablast sub- chromatin Abs. C, Serum level of anti-DNA Abs from the specified iso- set expressing CD93 was identified in LatY136F mutant mice, and types. Significant differences as calculated by the Mann-Whitney rank-sum its presence was confirmed in antihapten or antiviral responses Ͻ ءءء Ͻ ءء Ͻ ء test are indicated. , p 0.05; , p 0.01; and , p 0.001. (data not shown). The effector Th2 bias of the proliferating CD4 T cells induced noncomplement-fixing IgG1 and IgE isotypes in the serum as previously shown (20, 21). This B cell response did not any preferential Ag-driven expansion. IgG1 and IgE isotypes require Ag/TCR-dependent interaction between Th2 cells and B showed increased autoantibody titers whereas IgG2a were not in- cells. The response was clearly polyclonal without any bias for creased (Fig. 5C). By measuring the titers of anti-erythrocyte Abs autoantibodies, leading to severe Th2-mediated systemic autoim- in a direct Coombs’ test (Fig. 5A), anti-laminin (Table I), anti- munity with nephritis, IgE, and C3 in the kidneys, proteinuria, and chromatin (Fig. 5B), and anti-DNA Abs (Fig. 5C), we observed increased serum urea levels. that LatY136F mice had already a significant increase in autoantibody levels at 2 mo of age. This increase was augmented at 3 mo, consistent with a proportional increase in total Ig. In correlation Polyclonal B cell activation by LatY136F T cells with the positive Coombs’ test mice developed anemia. Four WT Y136F Y136F In Lat mice, the very few double-positive thymocytes that mice and five age-matched Lat mice showed significant dif- ϩ ϩ ferences in hematocrit (control mice: hematocrit 50; Coombs’ test- successfully matured to CD4 and CD8 single-positive stages positive mice, hematocrit 30). had been selected on MHC class II and class I molecules, respec- At 2 as well as 3 mo of age, we observed signs of severe glo- tively (20). To explain the presence of a lymphoproliferative dis- Y136F merulonephritis and up to 60% of mice displayed proteinuria at the ease and of autoantibodies in Lat mutant mice, it was origi- later time point (Fig. 6E). In all the mice with albuminuria, over 1 nally hypothesized that this mutation resulted in incomplete mg/ml as determined by ELISA was found. Albuminuria in control negative selection (21). According to this hypothesis, potentially mice was 8 Ϯ 5 ␮g/ml. An increase of blood urea from Ͻ25 ␮g/ml autoreactive T cells could then migrate to peripheral lymphoid in control mice to 80 ␮g/ml in LatY136F mice developed. A clear organs and activate B cells, which would be expressing self pep- relationship between albuminuria, serum urea levels, and prema- tide-MHC class II complexes to which the T cell repertoire had not Y136F ture death was observed. In correlation with the increased inci- been negatively selected. Using Lat mice expressing the HY dence of proteinuria and serum urea levels, mice developed sig- TCR, an MHC class I-restricted TCR originally calibrated in a nificant glomerular lesions, which showed features of slightly LAT-proficient context, it has been shown recently that the “win- proliferative mesangial glomerulonephritis characterized in partic- dow” for positive and negative selection has been shifted in LAT ular by periodic acid-Schiff-positive deposits in the mesangium, mutant mice (34). Therefore, it is likely that CD4 T cells devel- accompanied by tubular cast formation (Fig. 6A). Extracapillary oping in LatY136F mice are nevertheless appropriately selected in proliferation was observed occasionally. the context of the crippled LAT molecules (19). In the periphery, Immunofluorescence analysis revealed linear deposits of IgE LatY136F mutant T cells express very low levels of surface TCR along glomerular capillary walls, together with granular IgE de- and show undetectable calcium-influx after TCR and CD28 cross- posits in mesangium of diseased glomeruli in all the mice ana- linking (34). Consistent with these observations, we showed that lyzed, while no detectable deposits of other Ig isotypes (IgG1, LatY136F mutant T cells did not require MHC class II expression by IgG2a, and IgM) were observed (Fig. 6C). In addition, a fraction WT polyclonal B cells in order for them to induce IgG1 secretion. of mice (two of seven) displayed significant deposits of C3 in In addition, in LatY136F mutant mice, titers of anti-DNA, anti- mesangium (Fig. 6B). laminin, anti-chromatin, anti-NP, and anti-TNP Abs were in- In addition to renal lesions, the lungs and liver of LatY136F mice creased proportionally to the increase in total Ig. Therefore, we can showed striking alterations. The most interesting changes were ob- exclude preferential activation of autoantibody-producing B cells. served in lungs, which showed dense bronchiovascular lymphoid The presence of autoantibodies in LatY136F mice may reflect the infiltrates consisting of small lymphoid cells, plasma cells, and fact that the LatY136F CD4 effector T cells help B cells in a “quasi- scattered large immunoblasts (Fig. 7A). Strikingly, the FDC net- mitogenic” mode, thereby inducing a massive polyclonal B cell The Journal of Immunology 2291

FIGURE 6. Glomerulonephritis developed by LatY136F mice. A, H&E staining of kidney sections of littermate or 3-mo-old LatY136F mice. B, Staining of kidney sections with anti-C3 to detect complement deposits. Top panels, WT; bottom panels, LatY136F; left panels, small enlargement (ϫ10); and right panels, larger magnification (ϫ40). C, Analysis of Ig deposits by fluorescence staining with anti-IgG1 and anti-IgE (left panels, anti-IgG1, ϫ10; mid- dle panels, anti-IgE (small enlargement, ϫ10); and right panels, anti-IgE (large magnification, ϫ40). D, Quantifi- cation of kidney lesion. E, Proportion of LatY136F mice positive for proteinuria. Significant differences as calculated by the Mann-Whitney rank-sum test for disease index and by Fisher’s exact test for proteinuria analysis are .p Ͻ 0.01 ,ءء p Ͻ 0.05; and ,ء .indicated

activation that is accompanied by the production of Abs to multi- of autoreactive plasma cells is observed, and the resultant Abs ple specificities, including autoantigens. It has been shown previ- show signs of affinity maturation with variable increases in affinity ously that polyclonally activated T cells can induce B cell prolif- for abundant autoantigens such as chromatin and DNA (for review, eration and Ab production in naive polyclonal B cells and that this see Ref. 41). Alternatively, polyclonal B cell activation models effect is not strictly dependent on expression of MHC class II mol- leading to nephritis have been described previously. B cell-acti- ecules by the B cells (35, 36). This is highlighted by the fact that vating factor of the TNF family (BAFF) transgenic mice, which MHC class II-deficient mice do form GC after adoptive transfer of display polyclonal hypergammaglobulinemia and systemic auto- LatY136F T cells. The observation that, after adoptive transfer of immunity with lupus-like features, are the most similar to LatY136F LatY136F mutant T cells into MHC class II-deficient mice, GC were mice (6). However, in our mice, no increase in serum BAFF levels smaller than those in MHC class II-expressing recipients never- was observed (data not shown). In classical murine lupus, how- theless shows that MHC class II molecules do play a role in the GC ever, pathogenic IgG2a Ab isotypes, which efficiently activate reaction. We currently have no explanation why in LatY136F mice. complement and interacts with both Fc␥RI and Fc␥RIII, have been MHC class II interactions may be required for optimal GC forma- implicated in the disease. A recently described new FcR, Fc␥RIV, tion/maintenance. So far, we were not able to detect an Ag-driven which does not fix IgG1 isotypes, might further explain the higher induction of the Ab response after immunization of LatY136F mice disease potential of IgG2a Abs (42). In LatY136F mutant mice, the (data not shown). Alternatively, the homeostasis of B cells and T Th2-induced, noncomplement-fixing IgE Abs form IgE deposits in cells may be altered in MHC class II-deficient mice.

Polyclonal B cell response and Th2 autoimmunity As described above, the B cell response in LatY136F mice is truly polyclonal without any obvious bias for autoantigens. Since the activating T cells are Th2 effector cells, the hyperimmunoglobulinemia is restricted to IgG1 (100-fold) and IgE (225,000-fold). Total Ig increases by a factor of 30-fold in 2- to 3-mo-old LatY136F mice. Similar observations were made for Th1 responses in lpr or gld mice (37). The association of lupus nephritis with Th1 responses has become clear (38). Introducing a Th2 bias results in protection from lupus (39, 40). To exclude a direct effect of IL-4 on nephritis, we crossed LatY136F mice with CD40 knockout mice abolishing lupus susceptibility despite strong Th2 accumulation. These double- knockout mice show delayed increase in IgG1 and IgE Ab titers (our unpublished observations).

Autoimmune nephritis

Most classical human and mouse autoimmune diseases, including FIGURE 7. Lung and liver pathology of LatY136F mice. A, H&E staining lupus, are clearly multigenic (1, 2). Increased total Ig has been of LatY136F lung showing peribronchial and perivascular lymphoid infil- associated frequently with lupus, leading to immune deposits in trates (ϫ50). B and C, Low and high power magnification of FDC staining renal glomeruli. In the classical murine models of lupus, (NZB ϫ in lung infiltrates (ϫ50 and ϫ200). D, lymphoid infiltrates in the portal lpr ϫ NZW)F1, MRL-Fas , and BXSB, autoantigen-driven expansion space of the liver at 200. 2292 Th2-MEDIATED SYSTEMIC AUTOIMMUNITY IN LatY136F MICE the glomeruli and induce early-onset nephritis. Immunohistology 2. Wakeland, E. K., K. Liu, R. R. Graham, and T. W. Behrens. 2001. Delineating supports a predominant role for IgE in glomerular injury in asso- the genetic basis of systemic lupus erythematosus. Immunity 15: 397–394. 3. Shlomchik, M. J., and M. P. Madaio. 2003. The role of antibodies and B cells in ciation with complement deposition in two of seven mice. the pathogenesis of lupus nephritis. Springer Semin. Immunopathol. 24: 363–375. Importantly, for our study, several models exist where single 4. Madaio, M. P., J. Carlson, J. Cataldo, A. Ucci, P. Migliorini, and O. Pankewycz. 1987. Murine monoclonal anti-DNA antibodies bind directly to glomerular an- recessive mutations result in spontaneously activated T cells and tigens and form immune deposits. J. Immunol. 138: 2883–2889. the subsequent production of autoantibodies. Two examples are 5. Mostoslavsky, G., R. Fischel, N. Yachimovich, Y. Yarkoni, E. Rosenmann, the Zap mutation or the use of autoreactive TCR transgenic mice M. Monestier, M. Baniyash, and D. Eilat. 2001. Lupus anti-DNA autoantibodies cross-react with a glomerular structural protein: a case for tissue injury by mo- (43, 44). Forward genetics has revealed a mutation in a RING-type lecular mimicry. Eur. J. Immunol. 31: 1221–1227. ubiquitin ligase that increases the activity of follicular Th cells, 6. Khare, S. D., I. Sarosi, X. Z. Xia, S. McCabe, K. Miner, I. Solovyev, N. Hawkins, resulting in autoimmune nephritis (45) via abnormally strong help M. Kelley, D. Chang, G. Van, et al. 2000. Severe B cell hyperplasia and autoimmune disease in TALL-1 transgenic mice. Proc. Natl. Acad. Sci. USA 97: to autoreactive B cells. A dominant gain of function mutant of 3370–3375. phospholipase C␥2 leads to increased inflammation and preferen- 7. Shlomchik, M. J., A. Marshak-Rothstein, C. B. Wolfowicz, T. L. Rothstein, and tial activation of autoimmune B cells due to the increased activity M. G. Weigert. 1987. The role of clonal selection and somatic mutation in autoimmunity. Nature 328: 805–804. of follicular T cells (46). In this model, on a genetic background 8. Klinman, D. M., and A. D. Steinberg. 1987. Systemic autoimmune disease arises susceptible to lupus, an autoantigen-driven process resembling from polyclonal B cell activation. J. Exp. Med. 165: 1755–1760. 9. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of classical lupus is observed (41, 47). immunity. Nature 392: 245–244. Anti-nuclear Abs contribute but are not absolutely required for 10. Catron, D. M., A. A. Itano, K. A. Pape, D. L. Mueller, and M. K. Jenkins. 2004. disease development (1, 3, 48). In addition, disease is sometimes Visualizing the first 50 hr of the primary immune response to a soluble antigen. Immunity 21: 341–347. not correlated with autoantibody titers, but evidence for specific 11. Garside, P., E. Ingulli, R. R. Merica, J. G. Johnson, R. J. Noelle, and features of autoimmunity-inducing IgG autoantibodies such as M. K. Jenkins. 1998. Visualization of specific B and T lymphocyte interactions cross-reactivity to kidney Ags have been described previously (4, in the lymph node. Science 281: 96–99. 12. Okada, T., M. J. Miller, I. Parker, M. F. Krummel, M. Neighbors, S. B. Hartley, 5, 7, 41). Nevertheless, Ag-driven expansion of autoantibody-pro- A. O’Garra, M. D. Cahalan, and J. G. Cyster. 2005. Antigen-engaged B cells ducing B cells is considered a hallmark of autoimmune kidney undergo chemotaxis toward the T zone and form motile conjugates with helper T diseases such as lupus. cells. PLoS Biol. 3: e150. 13. McHeyzer-Williams, L. J., and M. G. McHeyzer-Williams. 2005. Antigen-spe- Lupus-prone mice spontaneously develop perivascular lym- cific memory B cell development. Annu. Rev. Immunol. 23: 487–484. phoid hyperplasia with plasmocytes and lymphocytes (49). 14. MacLennan, I. C., A. Gulbranson-Judge, K. M. Toellner, M. Casamayor-Palleja, E. Chan, D. M. Sze, S. A. Luther, and H. A. Acha-Orbea. 1997. The changing Perivascular infiltrates of liver are a recurrent features of lupus preference of T and B cells for partners as T-dependent antibody responses de- in man and in mouse (50–55). These features are observed in velop. Immunol. Rev. 156: 53–66. LatY136F mutant mice. In addition, FDC are readily detectable in 15. Takahashi, Y., P. R. Dutta, D. M. Cerasoli, and G. Kelsoe. 1998. In situ studies of the primary immune response to (4-hydroxy-3-nitrophenyl)acetyl. V. Affinity the perivascular lymphocyte infiltrates in the lung. Such neofor- maturation develops in two stages of clonal selection. J. Exp. Med. 187: 885–895. mation of GC-like structures is seen repeatedly in rheumatoid 16. Boyd, W. C., and H. Bernard. 1937. Quantitative changes in antibodies and arthritis and Sjogren syndrome (56, 57). globulin fractions in sera of rabbits injected with several antigens. J. Immunol. 33: 111–122. Currently, we have neither proof nor an explanation for a patho- 17. Hunziker, L., M. Recher, A. J. Macpherson, A. Ciurea, S. Freigang, genic role of IgE in the kidneys of LatY136F mice. Several Th2- H. Hengartner, and R. M. Zinkernagel. 2003. Hypergammaglobulinemia and au- mediated autoimmune diseases with IgE deposits have been de- toantibody induction mechanisms in viral infections. Nat. Immunol. 4: 343–349. 18. Lanzavecchia, A. 1985. Antigen-specific interaction between T and B cells. Na- scribed previously. Interestingly, IgE deposits have been described ture 314: 537–534. in human diseases. Increased IgE levels were detected in some 19. Malissen, B., E. Aguado, and M. Malissen. 2005. Role of the LAT adaptor in T ϳ cell development and Th2 differentiation. Adv. Immunol. 87: 1–25. lupus patients (58), and IgE deposits found in 10% of cases in 20. Aguado, E., S. Richelme, S. Nunez-Cruz, A. Miazek, A. M. Mura, M. Richelme, human renal disease were correlated with bad prognosis (59–63). X. J. Guo, D. Sainty, H. T. He, B. Malissen, and M. Malissen. 2002. Induction Similarly, diseases related to parasite infections might involve IgE of T helper type 2 immunity by a point mutation in the LAT adaptor. Science 296: 2036–2040. immune deposits (64, 65). HgCl2-induced allergic reactions have 21. Sommers, C. L., C. S. Park, J. Lee, C. Feng, C. L. Fuller, A. Grinberg, been associated with increased IgE production (66) In addition, J. A. Hildebrand, E. Lacana, R. K. Menon, E. W. Shores, et al. 2002. A LAT there is a role for IgE deposits in Wegener’s granulomatosis, as mutation that inhibits T cell development yet induces lymphoproliferation. Sci- ence 296: 2040–2043. well as IgE myelomas (67, 68). For most of these diseases, a direct 22. Malissen, M., A. Gillet, L. Ardouin, G. Bouvier, J. Trucy, P. Ferrier, E. Vivier, role of IgE in disease has been suggested but not yet been shown and B. Malissen. 1995. Altered T cell development in mice with a targeted mu- ⑀ conclusively. tation of the CD3- gene. EMBO J. 14: 4641–4653. Y136F 23. Cosgrove, D., D. Gray, A. Dierich, J. Kaufman, M. Lemeur, C. Benoist, and Taken together, our results show that Lat T cells induce an D. Mathis. 1991. Mice lacking MHC class II molecules. Cell 66: 1051–1054. MHC class II-independent polyclonal B cell response with all the 24. Raymond, I., T. Al Saati, J. Tkaczuk, S. Chittal, and G. Delsol. 1997. CNA.42, a new monoclonal antibody directed against a fixative-resistant antigen of fol- B cell populations found in a physiological immune response. This licular dendritic reticulum cells. Am. J. Pathol. 151: 1577–1585. Th2-mediated disease leads to increases in polyclonal IgG1 and 25. al Saati, T., S. Clamens, E. Cohen-Knafo, J. C. Faye, H. Prats, J. M. Coindre, IgE, a proportional increase in autoantibodies and severe systemic J. Wafflart, P. Caveriviere, F. Bayard, and G. Delsol. 1993. Production of monoclonal antibodies to human estrogen-receptor protein (ER) using recombinant ER disease, including nephritis. (RER). Int. J. Cancer 55: 651–654. 26. Shi, S. R., M. E. Key, and K. L. Kalra. 1991. Antigen retrieval in formalin-fixed, Acknowledgments paraffin-embedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of tissue sections. J. Histochem. Cy- We thank Rod Ceredig, Daniel Muruve, and Anne Wilson for critical read- tochem. 39: 741–748. ing of the manuscript, Simon Lattmann for statistical analysis, as well as 27. Rozzo, S. J., J. D. Allard, D. Choubey, T. J. Vyse, S. Izui, G. Peltz, and Dominique Lagrange and Estelle Saüberli for excellent technical help. B. L. Kotzin. 2001. Evidence for an interferon-inducible gene, Ifi202,inthe susceptibility to systemic lupus. Immunity 15: 435–443. 28. Luzuy, S., J. Merino, H. Engers, S. Izui, and P. H. Lambert. 1986. Autoimmunity Disclosures after induction of neonatal tolerance to alloantigens: role of B cell chimerism and The authors have no financial conflict of interest. F1 donor B cell activation. J. Immunol. 136: 4420–4426. 29. Fossati-Jimack, L., L. Reininger, Y. Chicheportiche, R. Clynes, J. V. Ravetch, T. Honjo, and S. Izui. 1999. High pathogenic potential of low-affinity autoanti- References bodies in experimental autoimmune hemolytic anemia. J. Exp. Med. 190: 1. Jorgensen, T. N., M. R. Gubbels, and B. L. Kotzin. 2004. New insights into 1689–1696. disease pathogenesis from mouse lupus genetics. Curr. Opin. Immunol. 16: 30. Izui, S., M. Higaki, D. Morrow, and R. Merino. 1988. The Y chromosome from 787–793. autoimmune BXSB/MpJ mice induces a lupus-like syndrome in (NZW ϫ The Journal of Immunology 2293

C57BL/6)F1 male mice, but not in C57BL/6 male mice. Eur. J. Immunol. 18: 49. Andrews, B. S., R. A. Eisenberg, A. N. Theofilopoulos, S. Izui, C. B. Wilson, 911–915. P. J. McConahey, E. D. Murphy, J. B. Roths, and F. J. Dixon. 1978. Spontaneous 31. Oya, K., J. Wang, Y. Watanabe, R. Koga, and T. Watanabe. 2003. Appearance murine lupus-like syndromes: clinical and immunopathological manifestations in of the LAT protein at an early stage of B cell development and its possible role. several strains. J. Exp. Med. 148: 1198–1215. Immunology 109: 351–359. 50. Tanasescu, C., and S. Purice. 1992. Liver disease in systemic lupus erythemato- 32. Su, Y. W., and H. Jumaa. 2003. LAT links the pre-BCR to calcium signaling. sus. Rom. J. Intern. Med. 30: 169–164. Immunity 19: 295–294. 51. Kushimoto, K., K. Nagasawa, A. Ueda, T. Mayumi, Y. Ishii, Y. Yamauchi, 33. Wang, Y., O. Horvath, A. Hamm-Baarke, M. Richelme, C. Gregoire, Y. Tada, H. Tsukamoto, T. Kusaba, and Y. Niho. 1989. Liver abnormalities and R. Guinamard, V. Horejsi, P. Angelisova, J. Spicka, B. Schraven, et al. 2005. liver membrane autoantibodies in systemic lupus erythematosus. Ann. Rheum. Single and combined deletions of the NTAL/LAB and LAT adaptors minimally Dis. 48: 946–952. affect B cell development and function. Mol. Cell. Biol. 25: 4455–4465. 52. Moyer, C. F., J. D. Strandberg, and C. L. Reinisch. 1987. Systemic mononuclear- 34. Sommers, C. L., J. Lee, K. L. Steiner, J. M. Gurson, C. L. Depersis, cell vasculitis in MRL/Mp-lpr/lpr mice: a histologic and immunocytochemical D. El-Khoury, C. L. Fuller, E. W. Shores, P. E. Love, and L. E. Samelson. 2005. analysis. Am. J. Pathol. 127: 229–242. ␥ Mutation of the phospholipase C- 1-binding site of LAT affects both positive and 53. Cerny, A., M. Kimoto, A. W. Hugin, R. Merino, and S. Izui. 1989. Anti-IgM negative thymocyte selection. J. Exp. Med. 201: 1125–1134. treatment of C57BL/6–1pr/1pr mice: depletion of B cells reduces 1pr gene-in- 35. Hodgkin, P. D., L. C. Yamashita, R. L. Coffman, and M. R. Kehry. 1990. Sep- duced lymphoproliferation and mononuclear cell vasculitis. Clin. Exp. Immunol. aration of events mediating B cell proliferation and Ig production by using T cell 77: 124–129. membranes and lymphokines. J. Immunol. 145: 2025–2034. 54. Berden, J. H., L. Hang, P. J. McConahey, and F. J. Dixon. 1983. Analysis of 36. Noelle, R. J., J. Daum, W. C. Bartlett, J. McCann, and D. M. Shepherd. 1991. vascular lesions in murine SLE. I. Association with serologic abnormalities. Cognate interactions between helper T cells and B cells. V. Reconstitution of T J. Immunol. 130: 1699–1705. helper cell function using purified plasma membranes from activated Th1 and Th2 T helper cells and lymphokines. J. Immunol. 146: 1118–1124. 55. Alexander, E. L., C. Moyer, G. S. Travlos, J. B. Roths, and E. D. Murphy. 1985. 37. Ishigatsubo, Y., A. D. Steinberg, and D. M. Klinman. 1988. Autoantibody pro- Two histopathologic types of inflammatory vascular disease in MRL/Mp auto- duction is associated with polyclonal B cell activation in autoimmune mice which immune mice: model for human vasculitis in connective tissue disease. Arthritis express the lpr or gld genes. Eur. J. Immunol. 18: 1089–1093. Rheum. 28: 1146–1155. 38. Takahashi, S., L. Fossati, M. Iwamoto, R. Merino, R. Motta, T. Kobayakawa, and 56. Kim, H. J., V. Krenn, G. Steinhauser, and C. Berek. 1999. Plasma cell develop- S. Izui. 1996. Imbalance towards Th1 predominance is associated with acceler- ment in synovial germinal centers in patients with rheumatoid and reactive aration of lupus-like autoimmune syndrome in MRL mice. J. Clin. Invest. 97: thritis. J. Immunol. 162: 3053–3062. 1597–1604. 57. Weyand, C. M., and J. J. Goronzy. 2003. Ectopic germinal center formation in 39. Santiago, M. L., L. Fossati, C. Jacquet, W. Muller, S. Izui, and L. Reininger. rheumatoid synovitis. Ann. NY Acad. Sci. 987: 140–149. 1997. Interleukin-4 protects against a genetically linked lupus-like autoimmune 58. Atta, A. M., C. P. Sousa, E. M. Carvalho, and M. L. Sousa-Atta. 2004. Immu- syndrome. J. Exp. Med. 185: 65–70. noglobulin E and systemic lupus erythematosus. Braz J. Med. Biol. Res. 37: 40. Yoh, K., K. Shibuya, N. Morito, T. Nakano, K. Ishizaki, H. Shimohata, M. Nose, 1497–1501. S. Izui, A. Shibuya, A. Koyama, et al. 2003. Transgenic overexpression of 59. Hyun, J., and M. A. Galen. 1981. Acute interstitial nephritis: a case characterized GATA-3 in T lymphocytes improves autoimmune glomerulonephritis in mice by increase in serum IgG, IgM, and IgE concentrations—Eosinophilia, and IgE with a BXSB/MpJ-Yaa genetic background. J. Am. Soc. Nephrol. 14: 2494–2502. deposition in renal tubules. Arch. Intern. Med. 141: 679–681. 41. Radic, M. Z., and M. Weigert. 1994. Genetic and structural evidence for antigen 60. Juncos, L. I., J. C. Muino, N. H. Garcia, C. I. Ferrer, M. Romero, R. H. Sambuelli, selection of anti-DNA antibodies. Annu. Rev. Immunol. 12: 487–520. and D. Beltramo. 2000. Renal tubular acidosis and vasculitis associated with IgE 42. Nimmerjahn, F., P. Bruhns, K. Horiuchi, and J. V. Ravetch. 2005. Fc␥RIV: a deposits in the kidney and small vessels. Am. J. Kidney Dis. 35: 941–949. novel FcR with distinct IgG subclass specificity. Immunity 23: 41–44. 61. Moriber-Katz, S., J. Loose, A. S. Miller, and A. B. Schwartz. 1988. IgE and 43. Korganow, A. S., H. Ji, S. Mangialaio, V. Duchatelle, R. Pelanda, T. Martin, immune complex glomerulonephritis. Ann. Clin. Lab. Sci. 18: 34–34. C. Degott, H. Kikutani, K. Rajewsky, J. L. Pasquali, et al. 1999. From systemic 62. Robertson, M. R., E. V. Potter, M. L. Roberts, and R. Patterson. 1976. Immu- T cell self-reactivity to organ-specific autoimmune disease via immunoglobulins. noglobulin E in renal disease. Nephron 16: 256–254. Immunity 10: 451–461. 63. Tuma, S. N., F. Llach, S. Sostrin, E. L. Dubois, and S. G. Massry. 1981. Glo- 44. Sakaguchi, N., T. Takahashi, H. Hata, T. Nomura, T. Tagami, S. Yamazaki, merular IgE deposits in patients with lupus nephritis. Am. J. Nephrol. 1: 31–34. T. Sakihama, T. Matsutani, I. Negishi, S. Nakatsuru, and S. Sakaguchi. 2003. Altered thymic T cell selection due to a mutation of the Zap70 gene causes 64. Maeno, Y., P. Perlmann, H. Perlmann, Y. Kusuhara, K. Taniguchi, autoimmune arthritis in mice. Nature 426: 454–460. T. Nakabayashi, K. Win, S. Looareesuwan, and M. Aikawa. 2000. IgE deposition 45. Vinuesa, C. G., M. C. Cook, C. Angelucci, V. Athanasopoulos, L. Rui, in brain microvessels and on parasitized erythrocytes from cerebral malaria pa- K. M. Hill, D. Yu, H. Domaschenz, B. Whittle, T. Lambe, et al. 2005. A RING- tients. Am. J. Trop. Med. Hyg. 63: 128–132. type ubiquitin ligase family member required to repress follicular helper T cells 65. Adu, D., D. G. Williams, I. A. Quakyi, A. Voller, Y. Anim-Addo, and autoimmunity. Nature 435: 452–458. A. A. Bruce-Tagoe, G. D. Johnson, and E. J. Holborow. 1982. Anti-ssDNA and 46. Yu, P., R. Constien, N. Dear, M. Katan, P. Hanke, T. D. Bunney, S. Kunder, antinuclear antibodies in human malaria. Clin. Exp. Immunol. 49: 310–314. L. Quintanilla-Martinez, U. Huffstadt, A. Schroder, et al. 2005. Autoimmunity 66. Ochel, M., H. W. Vohr, C. Pfeiffer, and E. Gleichmann. 1991. IL-4 is required for and inflammation due to a gain-of-function mutation in phospholipase C ␥2 that the IgE and IgG1 increase and IgG1 autoantibody formation in mice treated with specifically increases external Ca2ϩ entry. Immunity. 22: 451–465. mercuric chloride. J. Immunol. 146: 3006–3011. 47. Rahman, A. 2004. Autoantibodies, lupus and the science of sabotage. Rheuma- 67. Kamali, S., E. Kasapoglu, F. Akturk, A. Gul, M. Inanc, L. Ocal, O. Aral, and tology 43: 1326–1336. M. Konice. 2003. Eosinophilia and hyperimmunoglobulinemia E as the present- 48. Vyse, T. J., C. G. Drake, S. J. Rozzo, E. Roper, S. Izui, and B. L. Kotzin. 1996. ing manifestations of Wegener’s granulomatosis. Clin. Rheumatol. 22: 333–335. Genetic linkage of IgG autoantibody production in relation to lupus nephritis in 68. Vaerman, J. P. 1979. A new case of IgE myeloma (Des) ending with renal failure. New Zealand hybrid mice. J. Clin. Invest. 98: 1762–1772. J. Clin. Lab. Immunol. 2: 343–348.