Role of Complement-Binding CD21/CD19/CD81 in Enhancing Human Protection from Fas-Mediated Apoptosis

This information is current as Patricia K. A. Mongini, Anna E. Jackson, Sonia Tolani, of October 3, 2021. Rasem J. Fattah and John K. Inman J Immunol 2003; 171:5244-5254; ; doi: 10.4049/jimmunol.171.10.5244 http://www.jimmunol.org/content/171/10/5244 Downloaded from

References This article cites 83 articles, 48 of which you can access for free at: http://www.jimmunol.org/content/171/10/5244.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on October 3, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2003 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Role of Complement-Binding CD21/CD19/CD81 in Enhancing Human B Cell Protection from Fas-Mediated Apoptosis1

Patricia K. A. Mongini,2*† Anna E. Jackson,* Sonia Tolani,* Rasem J. Fattah,‡ and John K. Inman‡

Defective expression of Fas leads to B cell autoimmunity, indicating the importance of this apoptotic pathway in eliminating autoreactive B cells. However, B cells with anti-self specificities occasionally escape such regulation in individuals with intact Fas, suggesting ways of precluding this apoptosis. Here, we examine whether coligation of the B cell Ag (BCR) with the complement (C3)-binding CD21/CD19/CD81 costimulatory complex can enhance the escape of human B cells from Fas-induced death. This was warranted given that BCR-initiated signals induce resistance to Fas apoptosis, some (albeit not all) BCR-triggered events are amplified by coligation of BCR and the co-stimulatory complex, and several self Ags targeted in autoimmune diseases

effectively activate complement. Using a set of affinity-diverse surrogate Ags (receptor-specific mAb:dextran conjugates) with Downloaded from varying capacity to engage CD21, it was established that BCR:CD21 coligation lowers the BCR engagement necessary for inducing protection from Fas apoptosis. Enhanced protection was associated with altered expression of several molecules known to regulate Fas apoptosis, suggesting a unique molecular model for how BCR:CD21 coligation augments protection. BCR:CD21 coligation impairs the generation of active fragments of caspase-8 via dampened expression of membrane Fas and augmented expression of

FLIPL. This, in turn, diminishes the generation of cells that would be directly triggered to apoptosis via caspase-8 cleavage of

caspase 3 (type I cells). Any attempt to use the mitochondrial apoptotic protease-activating factor 1 (Apaf-1)-dependent pathway http://www.jimmunol.org/ for apoptosis (as type II cells) is further blocked because BCR:CD21 coligation promotes up-regulation of the mitochondrial antiapoptotic molecule, Bcl-2 . The Journal of Immunology, 2003, 171: 5244–5254.

he survival of B is determined in part by tinued viability of activated cells as long as foreign Ag is present. whether the Fas (CD95) death-inducing pathway is suc- On the other hand, the impaired BCR signaling that characterizes T cessfully executed. Not expressed on resting B cells, Fas most mature B cell engagements with self Ag probably promotes is significantly up-regulated on CD40-activated B cells (1Ð3), Fas-mediated elimination of autoreactive B cells (11Ð13). Anti-self making the latter vulnerable to apoptosis upon contact with Fas cells that escape deletion while immature or that are generated as

3 by guest on October 3, 2021 ligand (FasL) -bearing T lymphocytes (4). B cell Ag receptor a result of later somatic typically engage self Ag weakly (BCR) signaling does not up-regulate Fas, but, rather, protects and/or are in an anergic state. This is thought to preclude induction CD40-activated B cells from Fas-mediated death (5, 6). Protection of BCR-mediated protective signals and to encourage the demise has been linked to the heightened expression of several antiapop- of potentially pathogenic clones activated upon contact with CD40 totic molecules, i.e., mitochondria-associated Bcl-x (7, 8), Fas L ligand (CD40L)-bearing T cells (11Ð13). Consistent with this in- death effector domain-associating FLIP (9), and Fas apoptosis L terpretation, deficiencies in Fas or FasL expression result in in- inhibitory molecule, FAIM (10). creased numbers of activated autoimmune B cells in both mice and The extent to which protective signals are generated controls patients with autoimmune lymphoproliferative disease (14Ð16). whether responses to foreign Ag are prolonged and responses to self Ag are aborted (11, 12). Thus, BCR signals ensure the con- Additionally, the germinal centers of Fas-deficient mice harbor an elevated number of highly mutated B cells. These may represent B cells with lowered affinity for the and increased reac- tivity for self (17, 18). *Department of Rheumatology, Hospital for Joint Diseases, New York, NY 10003; The B cell costimulatory complex, comprised of complement †Department of Pathology, Kaplan Comprehensive Cancer Center, New York Uni- versity School of Medicine, New York, NY 10016; and ‡Laboratory of Immunology, (C3)-binding CD21(CR2), CD19, and CD81, significantly lowers National Institutes of Health, Bethesda, MD 20892 the Ag:BCR binding threshold for inducing several (albeit not all) Received for publication December 23, 2002. Accepted for publication September BCR-triggered phenomena (19Ð23). This raises the important is- 16, 2003. sue of whether protection against Fas apoptosis is more effectively The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance achieved by complement-bound Ags. Viruses, bacteria, and Ag:Ab with 18 U.S.C. Section 1734 solely to indicate this fact. complexes promote deposition of C3 fragments via either the clas- 1 This work was supported by a research award from the Arthritis Foundation and in sic or alternative complement activation pathways (24). An ability part by a grant from the U.S. Public Health Service (National Institutes of Health of such C3d(g)-bearing BCR ligands to lower the BCR threshold Grant R01AI052189). for inducing protection might be important in sustaining B cell 2 Address correspondence and reprint requests to Dr. Patricia Mongini, Department of Rheumatology, Hospital for Joint Diseases, New York University Medical Center, viability as concentrations of foreign Ag become limiting. Addi- 301 East 17th Street, New York, NY 10003. E-mail address: patricia.mongini@med. tionally, the complement-activating potential of certain atypical nyu.edu self Ags (25Ð29) might promote the escape of autoreactive B cells 3 Abbreviations used in this paper: FasL, ; BCR, B cell receptor; CD40L, CD40 ligand; dex, high molecular weight dextran; DISC, death-inducing from Fas-triggered apoptosis despite the low affinity of the signal complex; FADD, Fas-associated death domain. latter’s BCR.

Copyright © 2003 by The American Association of Immunologists, Inc. 0022-1767/03/$02.00 The Journal of Immunology 5245

The present study directly examines whether coengagement of Flow cytometric assays for B cell expression of cell surface and BCR and the C3d-binding costimulatory complex on CD40-acti- intracellular molecules vated human follicular B lymphocytes influences their resistance For analysis of surface molecules, cells were stained with FITC-mAb XG4 to Fas-mediated apoptosis. The effects of coligation were investi- anti-IgM (31), FITC-anti-CD3/PE-anti-CD16 (Immunotech, Westbrook, gated with surrogate C3dg-free and C3dg-bearing Ags, i.e., a set of ME), PE-anti-CD23 (Coulter, Miami, FL), FITC-anti-Fas(CD95) (BD previously described Ab:high molecular weight dextran (dex) con- PharMingen, San Diego, CA), and FITC- and PE-isotype controls (BD jugates bearing affinity-diverse anti-human IgM mAb, with and PharMingen). Staining intensity was quantitatively measured with a FACScan and associated software (BD Biosciences, Mountain View, CA). without coconjugated anti-CD21 mAb (20, 30). As discussed pre- For analysis of intracellular molecules, Ficoll-Hypaque-isolated viable viously (20, 21), the moderate multivalency of these ligands sim- cells (1Ð2 ϫ 105) from day 3 of culture (and before culture) were fixed with ulates the organized presentation of epitopes and C3 fragments on 2% paraformaldehyde, pH 7.2, in PBS for 10 min at room temperature, genuine foreign and self antigenic substrates such as bacteria, vi- washed, and made permeable with 0.1% saponin in 50% heat-inactivated human AB serum diluted in PBS/HEPES (assay buffer). Cells were then ruses, apoptotic cells, and Ag:Ab complexes. We here report that incubated with 0.5 ␮g of Ab: mouse anti-human Bcl-2 (DAKO, Carpin- under conditions of limiting receptor engagement, a ligand’s abil- teria, CA), mouse anti-human Bcl-xL (H-5, Santa Cruz Biotechnology, ity to coengage both BCR and CD21 does indeed significantly Santa Cruz, CA), or mouse IgG control mAb (MOPC-21) in 50 ␮lofthe enhance resistance to Fas-mediated death. A molecular explana- above saponin-containing assay buffer for 30 min at room temperature, washed, and then stained with R-PE-conjugated goat F(ab) anti-mouse tion for augmented protection is presented based on the degree to 2 IgG (HϩL) (Southern Biotechnology, Birmingham, AL) for an additional which BCR:CD21 coligation influences the expression of several 30 min at 4¡C. After washes, cells were refixed in 1% paraformaldehyde anti- and proapoptotic . The findings suggest that attempts and analyzed with a FACScan (same settings for day 0 and day 3 cells). to promote BCR:CD21 coligation during vaccine administration Downloaded from or, alternatively, to reduce BCR:CD21 coligation during the exac- Assay for Fas-mediated apoptosis erbating flares of B cell might help in opti- After 66- to 72-h incubation with CD40L and mAb:dex ligands, cell cul- mally managing the Fas apoptosis-inducing pathway. tures were pulsed with CH11 anti-Fas mAb (Oncor (Gaithersburg, MD) and MBL (Watertown, MA)) or mouse IgM control (TEPC-183) at a final Materials and Methods concentration of 0.5 ␮g/ml. Cells were harvested after 6Ð12 h and assayed for frequency of apoptotic cells by staining with FITC-annexin as previ- Monoclonal Ab:dex conjugates

ously described (36). Apoptotic cells exhibit both an increase in FITC- http://www.jimmunol.org/ A previous report has described the synthesis and binding properties of the annexin binding and a decrease in forward scatter (36). A maximum mAb:dex conjugates employed in this study, i.e., high m.w. dex covalently change value for specific anti-Fas-induced apoptosis under each culture linked to two distinct mAb, i.e., anti-IgM mAb (or IgG1 isotype control condition was calculated by subtracting background apoptosis (i.e., per- mAb) and THB-5 anti-CD21 mAb (or IgG2a isotype control mAb) (20). centage of FITC-annexin positive cells in culture with control IgM) from The BCR binding avidity of these ligands was varied by constructing con- the value for percentage of apoptosis in a replicate culture containing anti- jugates with three different affinity-diverse murine mAb specific for the Fas. Percent protection from anti-Fas-mediated apoptosis was calculated Ϫ same (or proximal) epitope on the C␮2 domain of human IgM: mAb HB57 using the following formula: (1 (% specific Fas-induced apoptosis in Ј ϭ ϫ 8 Ϫ1 ϭ ϫ 7 Ϫ1 CD40L-stimulated cultures containing mAb:dex/% specific Fas-induced (DA4.4), Fab (Ka 5 10 M ), mAb Mu53 (Ka 2 10 M ), and ϭϳ ϫ 6 Ϫ1 apoptosis in CD40L-stimulated cultures containing medium)) ϫ 100. mAb P24 (Ka 2 10 M ) (31, 32). The soluble conjugates have 10Ð12 anti-IgM mAb (or control mAb) and 10Ð12 anti-CD21 mAb (or control mAb) per dex molecule (20). For clarity of discussion we here Assay for cell proteins by Western blotting by guest on October 3, 2021 occasionally refer to anti-IgM:anti-CD21:dex conjugates as C3d(g)-bound ϩ Ag and anti-IgM:isotype control:dex conjugates as C3d(g)-free Ag. How- High density tonsil B cells depleted of CD27 cells were stimulated with ever, it is acknowledged that we have not eliminated the possibility that the CD40L in the presence of intermediate affinity Mu53 anti-IgM:dex, Mu53 ␮ later mAb:dex conjugates might not be fully free of C3d(g) and might have anti-IgM:anti-CD21:dex, anti-CD21:dex, or IgG control:dex (0.01 g/ml) ϫ 6 acquired some C3d(g) as the result of C3 synthesis from a low number of in 24-well plates at a density of 3Ð5 10 cells/1.5 ml/well. In all but one contaminating monocytes/dendritic cells in culture. experiment, B cells were derived from a single donor. After3dofculture, cells were washed twice with cold PBS and lysed in 50 ␮l of lysis buffer B cells and culture conditions (150 mM NaCl, 1 mM EDTA, 50 mM HEPES (pH 7.6), and 1% Nonidet P-40) containing protease inhibitor cocktail (Roche, Indianapolis, IN). Af- Human tonsil tissue from 3- to 15-year-old donors was obtained from New ter 15 min on ice, lysates were cleared of debris by centrifugation and were York Eye and Ear Infirmary (New York, NY) with the cooperation of Dr. frozen at Ϫ70¡C. For each analysis an equivalent amount of (range, S. McCormick and the surgery and pathology staffs. Tissue was used ac- 5Ð20 ␮g for Bcl-2 and Bcl-x assays, 40Ð50 ␮g for FLIP assays, and 5Ð40 cording to the guidelines of the institutional review board of New York ␮g for caspase-8, Fas-associated death domain (FADD), and Fas assays) University Medical Center. High density, -depleted B cells were ob- was resolved on an SDS-PAGE gel (12.5%), transferred to nitrocellulose, tained by rosetting and Percoll gradient density centrifugation as previ- blocked with 5% nonfat dry milk, and probed with 1) mouse anti-human ously described (32Ð34). Naive follicular B cells were further enriched by Bcl-2 (clone 124; 1/500 dilution; DAKO), 2) rabbit anti-Bcl-x (S-18; 1/500 depletion of CD27ϩ B cells (memory, , and subepithelial dilution; Santa Cruz Biotechnology), or 3) mouse anti-FLIP (mAb NF6; marginal zone B cells) (35) with anti-human CD27 magnetic beads (Milte- 1/10 dilution of culture supernatant; gift from Drs. P. Krammer and A. nyi Biotec, Auburn, CA). These were, on the average, 88% IgMϩ, 83% Krueger, DKFS, Heidelberg, Germany) (37). Abs bound to protein bands ϩ Յ ϩ CD23 , and 1% CD27 . Where indicated, some experiments were per- at 26 kDa (Bcl-2), 30 kDa (Bcl-xL), and 55 kDa (FLIPL) were detected with formed with high density B cells without CD27 depletion (on the average, HRP-conjugated goat anti-mouse IgG ( Technology, Bev- 65% IgMϩ, 64% IgDϩ, 27% CD27ϩ, and 97% CD21). Additionally, hu- erly, MA) or HRP-conjugated goat anti-rabbit Ig (BD Biosciences) and man spleen B cells (National Disease Research Interchange and Coopera- ECL detection reagent (Amersham Pharmacia Biotech, Indianapolis, IN).

tive Human Tissue Network) were tested. Cells were cultured in microtiter Detection of FLIPS (p25) was best achieved by use of Super Signal West wells at 1Ð2 ϫ 105/200 ␮l in RPMI 1640, 15% FCS, 5 ϫ 10Ϫ5 M 2-ME, Femto Maximum Sensitivity Substrate (Pierce, Rockford, IL). Blots were and gentamicin. For lysate preparation, cells were cultured in 24-well stripped and reprobed with rabbit anti-catalase (gift from Dr. P. Lazarow, plates at 3Ð5 ϫ 106 cells/2 ml. Generally this was in the form of culture Mount Sinai School of Medicine, New York, NY) as a means of standard- supernatants of COS cells transiently transfected with CD8-hgp39 (human izing the amount of protein loaded using an intrinsic protein. Additional CD40L; used at a final concentration of 3Ð5%; provided by Dr. R. Noelle, stripping and reprobing with 1) mouse anti-caspase-8 (1C12; 1/2000; Cell Dartmouth Medical School, Lebanon, NH). In these supernatants CD40L Signaling Technology), 2) mouse anti-FADD (clone 1; 1/250; BD Bio- was present as a trimer, and no additional cross-linking agent was added. sciences), and 3) mouse anti-Fas (clone B-10; 1/100; Santa Cruz Biotech- Some experiments used a CD40L activation (Oncogene Research Prod- nology) revealed levels of the proapoptotic molecules in the same lysates. ucts, Cambridge, MA) or Alexis Biochemicals (San Diego, CA); recom- For each lane a ratio was obtained of the densitometric intensity of the binant CD40L at 0.5 ␮g/ml and enhancer at 1 ␮g/ml) with essentially Bcl-2, Bcl-x, FLIP, caspase-8, FADD, or Fas band relative to the loading similar results. In all experiments B cells were exposed to mAb:dex ligand control band, as described previously (21). To facilitate the pooling of data and CD40L (plus enhancer) from the initiation of culture. from separate experiments, the lane with the highest ratio in each blot was 5246 BCR:CD21 COLIGATION AUGMENTS PROTECTION FROM FAS APOPTOSIS

arbitrarily given a maximal value of 100%, and the ratios of the other lanes binding requirements for eliciting protection from Fas-mediated were expressed as a percentage of this maximal value. apoptosis. Heightened protection from Fas-mediated death at very low Results concentrations of the affinity-diverse BCR:CD21 coengaging li- Analysis of BCR binding thresholds for inducing protection from gands was associated with heightened B cell S phase entry com- Fas-mediated death with and without recruitment of the CD21/ pared with cultures containing CD40L plus the respective mono- CD19/CD81 costimulatory complex specific BCR ligands (data not shown). Synergy between low High density, CD27-depleted tonsil B lymphocytes cultured for 3 d concentrations of multivalent BCR ligands and CD40L in promot- with CD40L and high affinity HB57 anti-IgM:dex were signifi- ing B cell proliferation has been previously described (39, 40), and cantly protected from Fas-mediated apoptosis compared with B the latter findings indicate that this synergy becomes even more cells cocultured with CD40L and IgG control:dex (Fig. 1). BCR- pronounced with multivalent BCR ligands that recruit the CD21/ mediated elicitation of a protective state was indicated by dimin- CD19/CD81 costimulatory complex. ished incidence of apoptotic cells upon treatment with a surrogate for FasL, i.e., CH11 anti-Fas Ab (2, 7). The decision to assay for Fas-mediated apoptosis on day 3 of culture was based on the re- Relative expression of antiapoptotic molecules, Bcl-2, Bcl-xL, sults of several other studies with CD40-activated human B cells and FLIPL in B cells stimulated with CD40L and limiting BCR (2, 3, 7). Although maximal levels of Fas had been noted by 48 h ligand with or without CD21 binding sites of culture, B cells became most susceptible to Fas-mediated apo- Previous studies have established that high density resting tonsil B ptosis at 72 h. cells express high levels of Bcl-2, low levels of Bcl-x, and nearly Downloaded from To investigate whether the capacity of BCR ligand to coengage undetectable levels of the two c-FLIP isoforms, FLIPL and FLIPS, the CD21/CD19/CD81 complex might facilitate protection, B cells and Fas (7, 41). CD40 engagement transiently up-regulates levels

were stimulated with CD40L and various concentrations of a set of of Bcl-x, FLIPL, and FLIPS (but not Bcl-2) to a maximum between affinity-diverse anti-IgM:dex ligands with or without coconjugated 24Ð48 h of culture. While Fas expression is also rapidly induced anti-CD21 (Fig. 2A). Alternatively, cells were cocultured with and remains elevated through day 3, all the above antiapoptotic mAb:dex conjugates bearing isotype control mAb with or without proteins significantly decline by day 3 (7, 41). If BCR ligand is http://www.jimmunol.org/ anti-CD21 mAb (Fig. 2A, left). The results indicate that BCR: concomitantly present however, increases in Bcl-xL and FLIPL are CD21 coengagement significantly augments protection against sustained (7, 9, 41). The latter is thought to contribute to BCR- Fas-mediated death, provided that BCR engagement is limiting. mediated protection from death following Fas ligation. Enhancement was quite notable at 0.001Ð0.01 ␮g/ml concentra- To examine whether BCR:CD21 coengagement promotes tions of ligands with intermediate to high BCR binding sites (FabЈ greater expression of these or other antiapoptotic molecules, B ϭ ϫ 7 ϫ 8 Ϫ1 ␮ Ka 2 10 to 5 10 M ) and at 0.01Ð0.1 g/ml concen- cells were cultured with CD40L plus intermediate affinity Mu53 Ј ϭϳ ϫ 6 trations of the low affinity BCR ligand (Fab Ka 2 10 anti-IgM:dex with or without anti-CD21 (or control mAb:dex with MϪ1). At high mAb:dex concentrations (1 ␮g/ml), the ligand that or without anti-CD21) at a concentration manifesting the greatest ␮

engaged CD21 alone also induced substantial protection (Fig. 2A, protective effects of BCR:CD21 coligation (0.005Ð0.01 g/ml). by guest on October 3, 2021 left, F). Whether this is mediated independently of the BCR sig- Day 3 lysates from five experiments were assessed for the expres-

naling complex or whether this reflects the low level intrinsic as- sion of Bcl-2, Bcl-xL, FLIPL, and FLIPS by immunoblotting (Fig. sociation between CD19 and BCR (38) is unclear. Observations 3). Consistent with other studies (7, 8, 41), cultures with CD40

similar to all the above were made with high density B cells from alone exhibited detectable levels of Bcl-xL, FLIPL, and Bcl-2 de- human spleens (data not shown). Taken together, results with these spite susceptibility to Fas apoptosis (Fig. 3A). The additional pres- polyclonally activating, surrogate C3d-free and C3d-bearing Ags ence of a limiting concentration of BCR:CD21 coengaging ligand

indicate that BCR:CD21 coligation significantly reduces the BCR significantly up-regulated the expression of Bcl-2 and FLIPL,

FIGURE 1. High affinity anti-IgM:dextran can protect CD40-activated human follicular B cells from Fas-induced apoptosis. High density, CD27- depleted tonsil B cells were cocultured with CD40L and IgG control:dex or moderately multivalent HB57 anti-IgM:dex (1 ␮g/ml, a dose found optimal for the induction of several BCR-triggered phenomena (20, 21)). On day 3 cells were pulsed with apoptosis-inducing CH11 anti-Fas or IgM control (0.1 ␮g) and after 10 h were assessed for FITC-annexin staining (A) and light scatter (B)byflow cytometry. The degree of apoptosis is indicated as the percentage of the total population analyzed that exhibited high FITC-annexin binding (A) or low forward scatter (B). The Journal of Immunology 5247 Downloaded from http://www.jimmunol.org/ by guest on October 3, 2021

FIGURE 2. Coligation of BCR and the CD21/CD19/CD81 costimulatory complex reduces the level of BCR engagement required for protecting CD40-activated B cells from Fas-mediated apoptosis. B cells were cocultured with CD40L plus varying concentrations of mAb:dex ligands in which one mAb of the pair of conjugated mAbs varied in binding site affinity for IgM (high, mid, low, or none, i.e., IgG1 control), and the other conjugated mAb did or did not have specificity for CD21 (THB5 anti-CD21 or IgG2a control). Further description of the binding site affinity of the anti-IgM mAbs and the coupling ratio of the conjugates can be found in Materials and Methods. Apoptosis was assessed on day 3 as described in Fig. 1. A, The data represent the percent protection from Fas apoptosis attained in the various cultures, as determined by the formula shown in Materials and Methods (mean Ϯ SEM). This calculation helped adjust for the varying level of background apoptosis observed in the multiple B cell populations tested. Three of the six experiments used high density B cells that did not undergo further CD27 depletion; three used high density populations specifically depleted of CD27ϩ cells. The results were similar in both sets and have been pooled. High affinity anti-IgM:dex was tested in only four of six experiments. Paired Student’s t test was used to compare the values for percent protection obtained by a given concentration of a mAb:dex ligand bearing CD21 binding sites from the percent protection obtained by the respective mAb:dex ligand not bearing CD21 binding sites. The p (probability) values indicating statistical significance are shown. B, Histograms indicating the degree of apoptosis, i.e., high level binding to FITC-annexin, in a representative experiment are shown. while monospecific BCR ligand generally did not (Fig. 3, A and B). IgM:anti-CD21:dex, respectively. The mean Ϯ SEM values for Ϯ Ϯ Ϯ The alternative FLIP isoform, i.e., FLIPS (42), was weakly detect- maximal FLIPL expression were 37 20, 51 17, 76 14, and Ϯ able in two of the five experiments. In these two cases, FLIPS was 97 3%, respectively. BCR:CD21 coligation did not augment the augmented upon exposure to either BCR ligand or BCR:CD21 expression of Bcl-xL above that seen with monospecific anti-IgM: coengaging ligand, but the levels were not statistically different dex (Fig. 3, A–C). Indeed, in two of five experiments, the levels of from that in control cultures. In three subsequent experiments (not Bcl-xL in cultures with CD40L and anti-IgM:anti-CD21:dex were shown in Fig. 3) FLIPL and FLIPS were assessed by immunoblot- less than those observed in cultures with CD40L alone. This was ting with a more sensitive ECL substrate. The mean Ϯ SEM values never observed in cultures receiving signals from CD40L and for maximal expression of FLIPS per constant protein loaded were monospecific BCR ligand. 19 Ϯ 1, 9 Ϯ 9, 80 Ϯ 2, and 100 Ϯ 2% for CD40-activated cultures Interestingly, in three of five experiments (one shown in Fig. containing Ig control:dex, anti-CD21:dex, anti-IgM:dex, and anti- 3A) CD21 engagement alone augmented Bcl-2 expression above 5248 BCR:CD21 COLIGATION AUGMENTS PROTECTION FROM FAS APOPTOSIS Downloaded from

FIGURE 3. Effect of BCR:CD21 coligation on the expression of antiapoptotic molecules, Bcl-2, Bcl-xL, and FLIPL, in lysates of CD40-activated human follicular B cells. High density, CD27-depleted B cells were cocultured with CD40L and double-Ig control:dex, anti-CD21:dex, anti-IgM:dex, anti-IgM: anti-CD21:dex (0.005Ð0.01 ␮g/ml). On day 3 lysates were prepared and analyzed for levels of Bcl-2, Bcl-x , FLIP , FLIP , and catalase (loading control)

L L S http://www.jimmunol.org/ protein by immunoblotting. A, Representative blots with ratios of the densitometric intensity of antiapoptotic protein/catalase to control for loading differences between lanes; B and C, assessment of the level of antiapoptotic proteins in five pooled experiments. B, Data are calculated on the basis of densitometric intensity per constant amount of protein loaded. Ratios were expressed as the percentage of maximal expression (see Materials and Methods; mean Ϯ SEM from the pooled experiments). C, Data represent the expression of antiapoptotic proteins per 106 recovered B cells and reflect an adjustment Expression of the ,ء .of densitometric intensity values to account for the greater protein recovery per cell in cultures stimulated by BCR-engaging ligands antiapoptotic molecule under the given cell treatment is significantly greater (p Յ 0.05) than that seen in control cultures stimulated by CD40L and IgG control:dex. In the group of experiments used for lysate preparation, parallel microcultures were assessed for the percent protection from Fas-induced apoptosis. The mean Ϯ SEM (range) of such values were as follows: IgG control:dex, 2.4 Ϯ 3.3% (range, Ϫ7 to 13); anti-CD21:dex, 18.5 Ϯ 3.6 (range, 10Ð30); anti-IgM:dex, 46.4 Ϯ 10.8% (range, 15Ð77); and anti-IgM:anti-CD21:dex, 84.1 Ϯ 3.4% (range, 73Ð94). by guest on October 3, 2021 that in both control cultures and cultures with anti-IgM:dex. How- alyzed by Student’s t test, only the level of Bcl-2 in B cells ex- ever, in none of the experiments did CD21 engagement alone re- posed to BCR:CD21 coengaging ligand was significantly different ϭ sult in levels of Bcl-xL, FLIPL, or FLIPS that exceeded those in from that in control cultures ( p 0.03). Although the level of control cultures or those in cultures with anti-IgM:dex (Fig. 3; data Bcl-x was occasionally elevated by BCR engagement (Fig. 4F), not shown). when the results of seven experiments were pooled (Fig. 4 and data The above analysis does not account for the fact that the amount not shown), Bcl-x levels were not significantly greater in cells of protein recovered from 106 viable cells in cultures containing receiving additional signals via BCR with or without CD21 ligand the BCR ligands was, on the average, 1.7- to 2.3-fold greater than vs those receiving CD40 signals alone. This in part reflected the the amount of protein recovered from an equal number of viable fact that in four additional experiments in which the sensitivity for cells in cultures containing CD40L plus Ig control:dex or anti- detecting Bcl-x was higher (due to use of a polyclonal rabbit anti- CD21:dex (viability, as determined by trypan blue exclusion, was Bcl-x Ab) cultures receiving additional signals via BCR or BCR: almost identical, i.e., the range in the four treatment groups was CD21 exhibited a major subpopulation of cells with Bcl-x levels 80Ð84%). To obtain what may be a better assessment of the rel- less than those seen in cultures containing IgG control:dex. ative level of Bcl-2, Bcl-xL, and FLIPL per cell, the data are ad- Whether this truly reflected a decline in Bcl-x expression or, al- ditionally presented as normalized expression per 106 cells (Fig. ternatively, a change in the compartmentalization or intermolecu- 3C). These normalized values also indicate that at a low ligand lar association of Bcl-x with other molecules (with a resulting concentration of 0.005Ð0.01 ␮g/ml, BCR:CD21 coligation can diminished accessibility to anti-Bcl-x Ab) is not clear. In the above significantly augment Bcl-2 and FLIPL expression in CD40-acti- intracellular staining analyses, no significant level of anti-FLIP vated cells, while BCR engagement alone does not. Although by staining above isotype control background was noted with any of this analysis the levels of FLIPS were up-regulated by BCR:CD21 the cells analyzed (data not shown). coligation, the significance of this increase is questionable due to Taken together, the above results suggest that at limiting doses the small number of positive samples and the weak density of the of Ag, BCR:CD21 coengagement may protect against Fas-medi- protein bands. ated death by augmenting B cell expression of the antiapoptotic

Additional intracellular staining analyses (Fig. 4) showed that molecules, Bcl-2, FLIPL, and possibly FLIPS. Since signaling via Bcl-2 levels were most effectively sustained in CD40-activated B BCR alone does not readily promote Bcl-2 up-regulation (7, 23) cells receiving additional signals from BCR:CD21 coengaging li- (present results), it appears that the net effect of B cell engagement gand or CD21-engaging ligand (Fig. 4, C and D). When the results with C3d-bound Ag is the expression of a wider spectrum of an- of seven such experiments (Fig. 4 and data not shown) were an- tiapoptotic molecules. The Journal of Immunology 5249 Downloaded from http://www.jimmunol.org/

FIGURE 4. Effect of BCR:CD21 coligation on intracellular expression of antiapoptotic molecules in CD40-activated B cells. High density, CD27- depleted tonsil B cells were cultured with CD40L and the mAb:dex ligands described in Fig. 3 (at 0.01 ␮g/ml). Prior to culture (day 0) and on day 3 of culture, viable cells were analyzed for intracellular levels of Bcl-2 and Bcl-x. Shown are the flow cytometric profiles from two representative experiments (A, C, and E (experiment I) and B, D, and F (experiment II)). A and B, Background staining with mouse IgG1 control Ab; C and D, staining with anti-Bcl-2 (IgG1); E and F, staining with anti-Bcl-x (IgG1). The slightly higher background staining consistently observed on cells cultured with CD21- and BCR-specific dex conjugates probably represents residual cell-bound mAb:dex detected by the R-PE-conjugated anti-mouse IgG probe. ⌬ MFI (mean fluorescence intensity values above background) for Bcl-2 and Bcl-x are shown to the left of the respective histograms. In these and five additional experiments, the ⌬ MFI values for cultured cells were expressed as a percentage of the maximum within each experiment, and a comparison of these latter by guest on October 3, 2021 values (n ϭ 7) for each of the culture conditions was made. Cultures containing IgG control:dex, anti-CD21:dex, anti-IgM:dex, and anti-IgM:anti-CD21:dex exhibited 78 Ϯ3, 83 Ϯ 5, 73 Ϯ 7, and 94 Ϯ 6%, respectively, of maximum values for Bcl-2 expression; for Bcl-x expression they exhibited values of 67 Ϯ 7, 86 Ϯ 7, 69 Ϯ 11, and 53 Ϯ 11%, respectively. In these pooled experiments, while Bcl-2 expression in cells exposed to BCR:CD21 coengaging ligand was significantly different (p Ͻ 0.01, by Student’s t test) from that in control cultures (or cultures exposed to monospecific BCR ligand), the values for Bcl-x expression under the diverse culture conditions were not significantly different.

Relative expression of proapoptotic molecules, i.e., caspase-8, Because the level of membrane Fas is more physiologically rel- FADD, and Fas, in B cells stimulated with CD40L and a evant than total protein levels, flow cytometry was used to inves- limiting concentration of BCR ligand with or without CD21 tigate whether recruitment of the costimulatory complex affects B binding sites cell expression of membrane Fas. Surface Fas was assessed both before culture (Fig. 5A) and after 3 days of culture with CD40L Additional immunoblotting experiments were performed to inves- and intermediate affinity anti-IgM:dex with or without CD21 bind- tigate whether BCR:CD21 coengagement might reduce the expres- ing sites, or Ig control:dex with or without CD21 binding sites sion of molecules needed for death-inducing signal complex (0.01 ␮g/ml; Fig. 5B). While membrane Fas was augmented in all (DISC) formation, i.e., Fas, FADD, and caspase-8 (43). These lat- CD40L-containing cultures, cells exposed to BCR:CD21 coengag- ter studies showed that BCR engagement, particularly with CD21 ing ligand exhibited approximately half as much Fas as B cells coligation, up-regulated, rather than down-regulated, the day 3 ex- triggered via CD40 alone. The mean Ϯ SEM values for the max- pression of caspase-8 and FADD (data not shown). The magnitude imal Fas expression in a total of five such experiments were 100 Ϯ Ͻ of the increase was 2-fold. Whether this up-regulation might be 0, 78 Ϯ 3, 85 Ϯ 8, and 59 Ϯ 8% for cells exposed to isotype related to the recently described additional roles of caspase-8 and control:dex, anti-CD21:dex, anti-IgM:dex, and anti-IgM:anti- FADD in promoting proliferation (44Ð47) is not CD21:dex, respectively. The difference in Fas expression between clear. Unlike caspase-8 and FADD, the expression of the protein cells cultured with Ig control:dex and anti-IgM:anti-CD21:dex was Fas was not enhanced upon BCR:CD21 coligation. Indeed, in three highly significant ( p ϭ 0.006, as assessed by paired Student’s t of five experiments, Fas levels within the lysates of B cells stim- test). It was observed with concentrations of staining Ab that were ulated for 3 days with CD40L and BCR:CD21 coengaging ligand clearly saturating, i.e., 4 times that recommended by the supplier were almost 2-fold lower than those in lysates of B cells stimulated (data not shown). Other previous studies had indicated that BCR with CD40L and BCR ligand or B cells stimulated with CD40L signaling can significantly reduce Fas expression in germinal cen- alone (data not shown). ter B cells, but not other tonsil B cell populations (6, 48). It is 5250 BCR:CD21 COLIGATION AUGMENTS PROTECTION FROM FAS APOPTOSIS

therefore important to note that the high density, CD27-depleted B cells placed in culture have the phenotype of follicular B cells, but not germinal center B cells (35, 49, 50) i.e., they are almost uni- formly IgMϩ, CD23ϩ, and CD27Ϫ (Fig. 5A). Interestingly, down-regulated Fas expression was also observed in B cells stimulated with CD40L and high concentrations of monospecific anti-IgM:dex (Fig. 6, B and C-c). Under these cir- cumstances as well, the reduction in Fas correlated with an aug- mented protection from Fas-mediated apoptosis (Fig. 6, A and C-a). Thus, it would appear that Fas down-regulation is not de- pendent upon an Ag’s capacity to recruit CD21/CD19, but can be induced by C3d(g)-free Ags of moderate multivalency provided the Ag concentration is sufficiently high. B cells stimulated by CD40 and BCR:CD21 coengaging ligand exhibit a lesser level of caspase-8 fragmentation upon acute exposure to Fas ligand, i.e., CH11 anti-Fas mAb A decline in membrane Fas, taken together with a rise in FLIP (which binds FADD with higher affinity than procaspase-8 and can FIGURE 5. BCR:CD21 coligation results in a diminished level of mem- compete with the latter for binding to FADD) (37, 42, 43), might Downloaded from brane Fas on B cells costimulated by CD40L. High density, CD27-depleted be expected to impair DISC formation and cleavage of pro- tonsil B cells were analyzed by flow cytometry for the expression of var- caspase-8 into its active fragments. To investigate whether B cells ious surface molecules prior to culture (A) and after3dofculture with activated by CD40L and a limiting concentration of BCR:CD21 CD40L and mAb:dex (B; as in Fig. 3; dose of 0.01 ␮g/ml). Background staining with FITC-labeled IgG control is indicated by the light histogram coengaging ligand do indeed exhibit diminished caspase-8 activa- to the left of each plot, and the relative levels of Fas are shown as the mean tion, cultures costimulated with CD40L and anti-IgM:dex with or fluorescence intensity (MFI). Percent protection from anti-Fas-induced ap- without anti-CD21, or Ig control:dex with or without anti-CD21 http://www.jimmunol.org/ optosis in other replicate cultures (assayed as described in Fig. 3) was as (0.005 ␮g/ml) were exposed to Fas-cross-linking mAb (or control follows: IgG control:dex, 13%; anti-CD21:dex, 30%; anti-IgM:dex, 33%; mAb) on day 3. Lysates prepared after6hofanti-Fas exposure and anti-IgM:anti-CD21:dex, 73%. by guest on October 3, 2021

FIGURE 6. At high ligand concentrations, BCR-specific ligand devoid of CD21 binding sites can down-regulate Fas expression. High density CD27- depleted tonsil B cells were cultured with CD40L and mAb:dex ligands (as described in Fig. 3). Dex-conjugated mAb were present at a concentration of either 1 or 0.01 ␮g/ml. After 3 days one replicate set of cultures was analyzed for the expression of Fas by flow cytometry with FITC-anti-Fas (B, C-c, and C-d). CH11 anti-Fas or IgM control was added to other sets of replicate cultures during the last 8Ð10 h of culture, and the percent protection from Fas-mediated death was determined (A, C-a, and C-b). The Journal of Immunology 5251 were analyzed for the expression of uncleaved and processed of the p18 fragments of caspase-8 (important in downstream caspase forms of caspase-8. As shown in Fig. 7, B cells receiving concom- 3 activation) is thought to occur only under conditions where mole- itant activation signals from CD40L and BCR:CD21 coengaging cules of procaspase-8 or its p43/41 fragments lie adjacent to one an- ligand were notably resistant to anti-Fas-induced cleavage of other, i.e., when levels of competing FLIP molecules are low, and caspase-8 into its intermediate (p41/43) and active p18 fragments. transproteolytic cleavage is possible (42, 51, 52). When compared to the levels in anti-Fas-treated cultures stimu- lated with CD40L alone, cleavage to p41/43 was less affected than Discussion cleavage to p18 (65 vs 84% inhibition, respectively). Anti-Fas- Using soluble receptor-specific mAb:dextran conjugates as surro- induced cleavage of the pro form of FLIPL into its p43 fragment gate Ags, the present study has noted that coengagement of BCR (which results from the proximity of FLIPL and caspase-8 within and C3 fragment-binding CD21 lowers the Ag concentration the DISC) (42, 51) was also reduced in B cells activated by CD40L threshold for eliciting protection from Fas apoptosis by 2Ð3 orders and BCR:CD21 coengaging ligand (Fig. 7). Generation of the p43 of magnitude. This protection was evident over a wide range of

FLIPL fragment, like generation of p43/41 caspase-8, was sup- BCR:ligand binding affinities. Because the in vitro B cell-activat- pressed to a lesser degree by BCR:CD21 coligation than genera- ing properties of these moderately multivalent anti-IgM:dex li- tion of p18 caspase-8. These differences might be linked to the gands (34) appear comparable with those of moderately multiva- stepwise manner in which DISC-associated proteins are cleaved lent Ags such as bacteriophage (53), we consider that these and to the relative degree that FLIPL inhibits generation of the findings have some parallels in vivo. Thus, with the caveat that B p43/41 and p18 fragments of caspase-8 (41, 51, 52). Following Fas cell responses to certain Ag:C3d(g) complexes in vivo may be cross-linking in cells bearing moderate levels of FLIPL, both influenced by parameters not present in this model system, the Downloaded from caspase-8 and FLIPL are recruited to the DISC. Indeed, at such present study strongly suggests that the innate ’s levels FLIPL can promote caspase-8 recruitment and activation. capacity to tag Ags with fragments of complement (C3) contrib- The initially recruited caspase-8 initiates cleavage of FLIPL to its utes significantly to the escape of CD40-activated B cells from p43 fragment as well as autocatalytically cleaves itself, generating Fas-mediated death. Because the deposition of complement is the proteolytically active p43/41 fragment (42, 51, 52). Generation greatly augmented on the surfaces of microorganisms, compared with most body cells (24), T cell-dependent B cell responses to microorganisms can be selectively amplified over those to most http://www.jimmunol.org/ self Ags. Problems may arise, however, with occasional self anti- genic substrates that also effectively activate complement, e.g., ap- optotic cells and IgG in Ab:Ag complexes. Coligation of BCR and CD21 on autoreactive B cells specific for these Ags may under some circumstances drive autoantibody production. The possibility that the coligation of BCR and the C3d(g)-bind- ing costimulatory complex might prevent Fas-mediated elimina- tion of B cells during an immune response was indeed previously by guest on October 3, 2021 suggested from the in vivo murine studies by Carroll and col- leagues (54). These investigators found that follicular survival of moderate affinity B cells with defective expression of the Cr2 gene products, i.e., CD21 and CD35, was significantly reduced after immunization with a T cell-dependent Ag compared with that of wild-type B cells of the same affinity (54). However, no difference in the follicular survival of Cr2Ϫ/Ϫ and Cr2ϩ/ϩ B cells was noted if the B cells were deficient in Faslpr/lpr (55). In the above-cited study, B cells of high affinity for Ag were not dependent upon Cr2 for follicular survival, but were dependent on Cr2 for germinal center entry and/or survival (54). Whether the diminished recovery of high affinity Cr2Ϫ/Ϫ cells within this site reflected impaired resistance to Fas-mediated apoptosis and/or an impaired receipt of viability signals from Ag:C3dg complexes on germinal center den- FIGURE 7. B cells stimulated by CD40L and BCR:CD21 coengaging dritic cells was not determined. ligand exhibit diminished production of active caspase-8 fragments fol- lowing cross-linking of Fas on day 3 of culture. B cells stimulated with Several factors argue that the CD21/CD19 costimulatory com- CD40L and mAb:dex (0.005 ␮g/ml), as described in Fig. 3, were pulsed on plex might be of relevance in protecting high affinity B cells from day 3 for 6 h with CH11 anti-Fas (or control IgM) at a final concentration germinal center Fas-mediated apoptosis. Firstly, Fas apoptosis ap- of 0.5 ␮g/ml prior to cell lysis. Lysates (30 ␮g for the representative ex- pears to play a role in B cell selection within this site (50, 56). periment shown; 40 and 5.3 ␮g for two additional experiments) were an- Secondly, within germinal centers the dose of Ag, i.e., the avail- alyzed for the uncleaved (p55/57) form of caspase-8 as well as the inter- ability of free antigenic epitopes, is probably quite limiting. Fi- mediate processed form of p41/43 and the active fragment of p18 by nally, from the present study it is clear that when BCR:ligand immunoblotting with an anti-caspase-8 mAb specific for the carboxyl-ter- affinity is high, low doses of ligand are needed to observe depen- minal sequence of p18 (1C12). Following stripping, blots were reprobed dency upon the costimulatory complex. with anti-catalase and anti-FLIP (NF6). The relative amount of a given The fact that BCR coligation with CD21 failed to increase the protein is shown as a ratio of the densitometric intensity of the relevant band/catalase (loading control). Results from three caspase-8 immunoblots protection elicited by high doses of the higher affinity ligands is (and the two FLIP immunoblots with 30 and 40 ␮g of protein/lane) were consistent with past observations that in vivo Ab responses elicited pooled and assessed as a percentage of the maximum ratio for each under conditions of near-optimal BCR engagement (high doses of ء Ϯ caspase-8 band or each FLIPL band (mean SEM). , The indicated Ag or high antigenic valency) are relatively independent of C3 and densitometric analysis of that band was unreliable due to a blemish. CD21/CD19 (19, 22, 57). However, it was somewhat unexpected 5252 BCR:CD21 COLIGATION AUGMENTS PROTECTION FROM FAS APOPTOSIS that recruitment of the costimulatory complex did not enhance the gagement of CD21 on human germinal center B cells promoted suboptimal protection achieved at high doses of the low affinity cell viability via a Bcl-2-dependent mechanism (70). Additionally, BCR ligand. Given that the mAbs conjugated to dex were intact Roberts and Snow (23) noted that engagement of CD19 on LPS- IgG Abs, it is possible that at high concentrations, the inhibitory activated murine B cells promoted both increased Bcl-2 expression influences of Fc␥RIIB engagement (58, 59) over-ride the positive and increased proliferation. In another parallel with the present influences of recruiting CD21/CD19/CD81. However, in two ex- study, the latter investigators noted that CD19 engagement up- periments in which 100 ␮g/ml of Fc␥RII-blocking mAb (AT10) regulated Bcl-2, but not Bcl-x, and conversely, BCR engagement (30) was added to cultures 30 min before addition of CD40L and up-regulated Bcl-x, but not Bcl-2. Additionally, coligation of BCR the mAb:dex conjugates, little to no augmentation in protection and CD19 was found to result in Bcl-2 expression greater than that was noted (P.K.A.M., unpublished observations). observed with CD19 engagement alone (23). The present study’s examination of how BCR:CD21 coligation Although prior studies found that BCR induced down-regulation affects the expression of various anti- and proapoptotic molecules of Fas only in germinal center B cells (48), the present report provides important insights into probable mechanisms for aug- clearly indicates that this phenomenon can occur in CD40-acti- mented protection. The reduced density of membrane Fas in B vated follicular cells as well. There are several reasons why this cells exposed to BCR:CD21 coengaging ligand is probably a major may not have been observed previously (5, 6, 9, 48). Firstly, it factor responsible for diminished susceptibility, since 1) formation appears that without recruitment of the costimulatory complex, of the active DISC is dependent upon aggregation of preformed down-modulation of Fas requires high concentrations of moder- Fas oligomers (43, 60), and 2) 2-fold changes in the density of ately multivalent BCR ligand. This level of BCR engagement other receptors can have notable effects on cell behavior (61Ð63). might not have been reached in other studies with nongerminal Additionally, the slight, but statistically significant, augmentation center B cells. Secondly, B cell heterogeneity, e.g., incidence of Downloaded from in FLIPL expression upon BCR:CD21 coligation may contribute to cells binding the BCR ligand, may affect the degree to which Fas the diminished susceptibility to Fas apoptosis. FLIPL bears the down-regulation is detected. Thirdly, the time at which BCR sig- FADD binding domain of caspase-8, but lacks the latter’s catalytic nals are delivered relative to CD40 signals may affect the degree to activity (42, 43). In a manner quite dependent upon the expression which Fas expression is modulated. Of relevance, the addition of level, FLIP dampens Fas-mediated apoptosis by competing with anti-IgM:dex or anti-IgM:anti-CD21:dex (0.01 ␮g/ml) late after caspase-8 for binding to FADD (42, 43, 51, 52). Importantly, CD40 activation (an approach used in some studies (5, 8)) resulted http://www.jimmunol.org/ FLIP’s higher affinity for FADD (37) may promote this interfer- in no down-modulation of Fas. Although protection was less pro- ence despite the higher intracellular levels of caspase-8. A major nounced than when mAb:dex ligands were added at the beginning finding, which was consistent with a diminished formation of Fas: of culture, the BCR:CD21 coengaging ligand remained more ef- FADD:caspase-8 DISC complexes in B cells costimulated with fective than the ligand that engaged solely BCR (P.K.A.M., un- CD40L and BCR:CD21 coengaging ligand, is the observation that published observations). Thus, upon delayed exposure to BCR: such B cells were significantly compromised in the cleavage of CD21 coengaging ligand, CD40-activated B cells are protected by caspase-8 into its p43/41 and p18 fragments following cross-link- a mechanism(s) independent of a reduction in membrane Fas. Pro- ing of Fas. tection might involve modulation of the above antiapoptotic mol- by guest on October 3, 2021 Diminished cleavage of caspase-8 is expected to affect the two ecules or others not presently explored, e.g., Fas apoptosis inhib- cellular programs available for executing Fas apoptosis, i.e., type itory molecule (10). Alternatively, (Akt) I and type II, to differing degrees. As discussed previously (3, 64, and inactivation of proapoptotic BAD (71, 72) 65), Fas apoptosis in type I cells requires extensive caspase-8 ac- and caspase-9 (73) might be important, since the costimulatory tivation at the DISC to initiate direct cleavage and activation of complex is known to amplify signaling pathways that lead to aug- downstream caspase-3. It is unaffected by antiapoptotic molecules mented protein kinase B (Akt) function (19, 74, 75). within the mitochondria, i.e., Bcl-2 and Bcl-x. In contrast, Fas Importantly from a therapeutic perspective, several lines of ev- apoptosis in type II cells requires low levels of DISC activation. In idence suggest that the C3-binding costimulatory complex could such cells, although the level of caspase-8 activation is not suffi- be a relevant target for blocking the activation of mature autore- cient for cleavage and activation of caspase-3, it is sufficient for active B cells. Firstly, as shown here, the costimulatory complex cleavage of the cytosolic proapoptotic molecule, BID. Truncated plays an important role in protecting CD40-activated B cells from BID translocates to the mitochondria, where it induces the release Fas-mediated death when BCR is suboptimal, of cytochrome c and the activation of an alternative apoptotic pro- a situation expected when autoreactive escapees from immature B tease-activating factor 1 (Apaf1)-dependent apoptotic program cell deletion confront self Ag in the periphery (76). Secondly, (43, 64, 65). Thus, although the level of caspase-8 activation fol- Foote et al. (77) showed that anergic B cells are not incapable of lowing Fas aggregation in cells exposed to BCR:CD21 coengaging signaling a Fas-resistant state, but require higher levels of BCR ligand might be too low to activate direct type I apoptosis, it might cross-linking to do so. BCR engagement with CD21/CD19/CD81 be sufficient to initiate type II apoptosis were it not for the height- may lower the degree of BCR engagement required to convert ened level of the mitochondrial antiapoptotic molecule, Bcl-2. Im- anergic B cells into a Fas-resistant state. Thirdly, although exten- portantly in this regard, deliberate overexpression of Bcl-2 in a B sive deposition of C3 on most body cells is prevented by specific cell line was shown to reduce Fas-mediated apoptosis (66). Al- complement regulatory mechanisms and the presence of high lev- though the increases in Bcl-2 noted here are generally Յ2-fold, els of sialic acid, two self Ag substrates known to be targeted in they may very well be relevant to cell survival. It has been noted autoimmune diseases, i.e., apoptotic cells expressing DNA and that only a 30% increase in Bcl-2 expression in vivo can effectively nucleoproteins on surface blebs (25Ð29) and complexes of Ag and rescue B cells from apoptosis (67). IgG Ab, have a proclivity to bind C3 fragments. Interestingly, mice The present study’s correlation between CD21-promoted resis- and humans deficient in Fas characteristically exhibit augmented tance to Fas apoptosis and Bcl-2 expression is consistent with production of autoantibodies specific for these self Ags (14, 78Ð other findings linking the CD21/CD19 complex to increased B cell 81). Indeed, mice deficient in Fas accumulate rheumatoid-factor viability and Bcl-2 expression (68Ð70). Of greatest relevance to producing cells within the T cell-rich regions of lymphatic tissue the present work, Bonnefoy et al. (69) reported that in vitro en- (81). Thus, deliberate blocking of C3d:CD21 interactions might be The Journal of Immunology 5253 expected to enhance Fas-mediated elimination of mature autore- 15. Cohen, P. L., and R. A. Eisenberg. 1991. Lpr and gld: single gene models of active B cells. systemic autoimmunity and lymphoproliferative disease. Annu. Rev. Immunol. 9:243. Paradoxically, on certain genetic backgrounds, mice with defec- 16. Drappa, J., A. Vaishnaw, K. Sullivan, J. Chu, and K. Elkon. 1997. Fas gene tive expression of both Cr2 (CD21/CD35) and Fas have greater in Canale-Smith syndrome, an inherited lymphoproliferative disorder associated with autoimmunity. N. Engl. J. Med. 335:1643. levels of lupus-like autoantibodies than do mice with either defect 17. Hoch, S., M. Boyd, B. Malone, G. Gonye, J. Schwaber, and J. Schwaber. 2000. alone (82, 83). Part of the explanation may lie in the fact that the Fas-mediated apoptosis eliminates B cells that acquire self-reactivity during the murine Cr2 gene encodes both CR1(CD35) and CR2(CD21), two germinal center response to NP. Cell. Immunol. 203:103. 18. Takahashi, Y., H. Ohta, and T. Takemori. 2001. Fas is required for clonal selec- molecules with differing expression patterns and functions (83, tion in germinal centers and the subsequent establishment of the 84). Additionally, coengagement of BCR and the CD21/CD19/ repertoire. Immunity 14:181. CD81 costimulatory complex might have differing functional ef- 19. Fearon, D. T., and M. C. Carroll. 2000. Regulation of B lymphocyte responses to foreign and self by the CD19/CD21 complex. Annu. Rev. Immunol. fects depending upon the state of the target autoreactive B cell 18:393. (83). Thus, such receptor coligation might augment the tolerance 20. Mongini, P. K. A., M. A. Vilensky, P. F. Highet, and J. K. Inman. 1997. The of immature B cells that reside in or are recent emigrants from the affinity threshold for human B cell activation via the antigen receptor complex is reduced upon co-ligation of the antigen receptor with CD21 (CR2). J. Immunol. bone marrow. In mature B cells that have not received strong co- 159:3782. stimuli such as CD40L, coligation might help promote anergy. 21. Mongini, P. K. A., S. Tolani, R. J. Fattah, and J. K. Inman. 2002. Antigen re- ceptor triggered upregulation of CD86 and CD80 in human B cells: augmenting Finally, as suggested by the present study, if mature autoreactive B role of the CD21/CD19 co-stimulatory complex and IL-4. Cell. Immunol. 216:50. cells receive signals from a BCR:CD21 coengaging ligand con- 22. Chen, Z., S. B. Karalov, M. Gendelman, M. C. Carroll, and G. Kelsoe. 2000. Ϫ Ϫ comitantly with CD40 signals, the result will be augmented growth Humoral immune responses in Cr2 / mice: enhanced affinity maturation but impaired persistence. J. Immunol. 164:4522. and protection from Fas apoptosis. In developing interventional

23. Roberts, T., and E. C. Snow. 1999. Cutting edge: recruitment of the CD19/CD21 Downloaded from therapy for B cell autoimmune diseases, it will be important to coreceptor to the B cell antigen receptor is required for antigen-mediated expres- consider ways of targeting costimulatory complex-blocking agents sion of bcl-2 by resting and cycling hen egg lysozyme transgenic B cells. J. Im- munol. 162:4377. to the latter cell population. 24. Janeway, C. A., P. Travers, M. Walport, and M. Schlomchik. 2001. The com- plement system and innate immunity. In Immunobiology, 5th Ed. Garland, New Acknowledgments York, p. 43Ð64. 25. Casciola-Rosen, L. A., G. Anhalt, and A. Rosen. 1994. Autoantigens targeted in We thank Drs. Randy Noelle and Peter Krammer for their contributions of systemic lupus erythematosus are clustered in two populations of surface struc- reagents, John Brunda for technical assistance, Drs. Mandar Dave and tures on apoptotic keratinocytes. J. Exp. Med. 179:1317. http://www.jimmunol.org/ Mukundan Attur for advice in performing the immunoblots, Drs. Robert 26. Miranda, M. E., C. E. Tseng, W. Rashbaum, R. L. Ochs, C. A. Casiano, Clancy and Ashok Amin for useful discussions, and Sonali Trivedi for F. DiDonato, E. K. L. Chan, and J. P. Buyon. 1998. Accessibility of SSA/Ro and SSB/La antigens to maternal autoantibodies in apoptotic human fetal cardiac assistance with figure preparation. myocytes. J. Immunol. 161:5061. 27. Rauterberg, A., E. G. Jung, and E. W. Rauterberg. 1993. Complement deposits in References epidermal cells after ultraviolet B exposure. Photodermatol. Photoimmunol. Pho- 1. Schattner, E. J., K. B. Elkon, D. H. Yoo, J. Tumang, P. H. Krammer, M. K. Crow, tomed. 9:135. and S. M. Friedman. 1995. CD40 ligation induces Apo-1/Fas expression on hu- 28. Mevorach, D., J. O. Mascarenhas, D. Gershov, and K. B. Elkon. 1998. Comple- man B lymphocytes and facilitates apoptosis through the Apo-1/Fas pathway. ment-dependent clearance of apoptotic cells by human . J. Exp. J. Exp. Med. 182:1557. Med. 188:2313. 29. Korb, L. C., and J. M. Ahearn. 1997. C1q binds directly and specifically to

2. Garrone, P., E. M. Neidhardt, E. Garcia, L. Galibert, C. van Kooten, and by guest on October 3, 2021 J. Banchereau. 1995. Fas ligation induces apoptosis of CD40-activated human B surface blebs of apoptotic human keratinocytes. J. Immunol. 158:4525. lymphocytes. J. Exp. Med. 182:1265. 30. Mongini, P. K. A., and J. K. Inman. 2001. dependency of human B cell 3. Daniel, P. T., and P. H. Krammer. 1994. Activation induces sensitivity to APO-1 cycle progression elicited by ligands which co-engage BCR and the CD21/CD19/ (CD95)-mediated apoptosis in human B cells. J. Immunol. 152:5624. CD81 costimulatory complex. Cell. Immunol. 207:127. 4. Ju, S.-T., H. Cui, D. J. Panka, R. Ettinger, and A. Marshak-Rothstein. 1994. 31. Rudich, S. M., E. Mihaesco, R. Winchester, and P. K. A. Mongini. 1987. Anal- Participation of target Fas protein in apoptosis pathway induced by CD4ϩ Th1 ysis of the domain specificity of various murine anti-human IgM monoclonal and CD8ϩ cytotoxic T cells. Proc. Natl. Acad. Sci. USA 91:4185. differing in human B lymphocyte signaling activity. Mol. Immunol. 5. Rothstein, T. L., J. K. M. Wang, D. J. Panka, L. C. Foote, Z. Wang, B. Stanger, 24:809. H. Cui, S.-T. Ju, and A. Marshak-Rothstein. 1995. Protection against Fas-depen- 32. Rudich, S. M., K. H. Roux, R. J. Winchester, and P. K. A. Mongini. 1988. dent Th1-mediated apoptosis by antigen receptor engagement in B cells. Nature Anti-IgM-mediated B cell signaling: molecular analysis of ligand binding requi- 374:163. sites for human B cell clonal expansion and tolerance. J. Exp. Med. 168:247. 6. Lagresle, C., P. Mondiere, C. Bella, P. H. Krammer, and T. Defrance. 1996. 33. Mongini, P. K. A., C. A. Blessinger, and J. P. Dalton. 1991. Affinity requirements Concurrent engagement of CD40 and the antigen receptor protects naive and for induction of sequential phases of human B cell activation by membrane IgM- memory human B cells from APO-1/Fas-mediated apoptosis. J. Exp. Med. cross-linking ligands. J. Immunol. 146:1791. 183:1377. 34. Mongini, P. K. A., C. A. Blessinger, P. F. Highet, and J. K. Inman. 1992. Mem- 7. Zhang, X., L. Li, C. Jongseon, S. Krajewski, J. C. Reed, C. Thompson, and brane IgM-mediated signaling of human B cells: effect of increased ligand bind- ing site valency on the affinity and concentration requirements for inducing di- Y. S. Choi. 1996. Up-regulation of Bcl-xL expression protects CD40-activated human B cells from Fas-mediated apoptosis. Cell. Immunol. 173:149. verse stages of activation. J. Immunol. 148:3892. 8. Schneider, T. J., D. Grillot, L. C. Foote, G. E. Nunez, and T. L. Rothstein. 1997. 35. Dono, M., S. Zupo, N. Leanza, G. Melioli, M. Fogli, A. Melagrana, N. Chiorazzi, Bcl-x protects primary B cells against Fas-mediated apoptosis. J. Immunol. and M. Ferrarini. 2000. Heterogeneity of tonsillar epithelial B lymphocytes, the 159:4834. splenic marginal zone equivalents. J. Immunol. 164:5596. 9. Wang, J., A. A. Lobito, F. Shen, F. Hornung, A. Winoto, and M. J. Lenardo. 36. Mongini, P. K. A., M. A. Vilensky, P. F. Highet, and J. K. Inman. 1998. Mem- 2000. Inhibition of Fas-mediated apoptosis by the B cell antigen receptor through brane IgM-stimulated human B lymphocytes succumb to activation-related apo- 3 c-FLIP. Eur. J. Immunol. 30:155. ptosis at a G1 S transition: influence of ligand affinity and valency. Cell. Im- 10. Schneider, T. J., G. M. Fischer, T. J. Donohoe, T. P. Colarusso, and munol. 188:137. T. L. Rothstein. 1999. A novel gene coding for a Fas apoptosis inhibitory mol- 37. Scaffidi, C., I. Schmitz, P. H. Krammer, and M. E. Peter. 1999. The role of c-FLIP ecule (FAIM) isolated from inducibly Fas-resistant B lymphocytes. J. Exp. Med. in modulation of CD95-induced apoptosis. J. Biol. Chem. 274:1541. 189:949. 38. Carter, R. H., G. M. Doody, J. B. Bolen, and D. T. Fearon. 1997. Membrane 11. Rothstein, T. L., X. Zhong, B. R. Schram, R. S. Negm, T. J. Donohoe, IgM-induced phosphorylation of CD19 requires a CD19 domain that D. S. Cabral, L. C. Foote, and T. J. Schneider. 2000. Receptor-specific regulation mediates association with components of the B cell antigen receptor complex. of B-cell susceptibility to Fas-mediated apoptosis and a novel Fas apoptosis in- J. Immunol. 158:3062. hibitory molecule. Immunol. Rev. 176:116. 39. Snapper, C. M., M. R. Kehry, B. E. Castle, and J. J. Mond. 1995. Multivalent, but 12. Rathmell, J. C., M. P. Cooke, W. Y. Ho, J. Grein, S. E. Townsend, M. M. Davis, not divalent, antigen receptor cross-linkers synergize with CD40 ligand for in- and C. C. Goodnow. 1995. CD95 (Fas)-dependent elimination of self-reactive B duction of Ig synthesis and class switching in normal murine B cells: a redefi- cells upon interaction with CD4ϩ T cells. Nature 376:181. nition of the TI-2 vs T cell-dependent antigen dichotomy. J. Immunol. 154:1177. 13. Rathmell, J. C., S. E. Townsend, J. C. Xu, F. A. Flavell, and C. C. Goodnow. 40. Wheeler, K., J. D. Pound, J. Gordon, and R. Jefferis. 1993. Engagement of CD40 1996. Expansion or elimination of B cells in vivo: dual roles for CD40- and Fas lowers the threshold for activation of resting B cells via antigen receptor. Eur. (CD95)-ligands modulated by the B cell antigen receptor. Cell 87:319. J. Immunol. 23:1165. 14. Izui, S., V. E. Kelley, K. Masuda, H. Yoshida, J. B. Roths, and E. D. Murphy. 41. Hennino, A., M. Berard, M. Casamayor-Palleja`, P. H. Krammer, and T. Defrance. 1984. Induction of various autoantibodies by mutant gene lpr in several strains of 2000. Regulation of the Fas death pathway by FLICE-inhibitory protein in pri- mice. J. Immunol. 133:227. mary human B cells. J. Immunol. 165:3023. 5254 BCR:CD21 COLIGATION AUGMENTS PROTECTION FROM FAS APOPTOSIS

42. Krueger, A., I. Schmitz, S. Baumann, P. H. Krammer, and S. Kirchhoff. 2001. 63. George, J., S. J. Penner, J. Weber, J. Berry, and J. L. Claflin. 1993. Influence of Cellular FLICE-inhibitory protein splice variants inhibit different steps of membrane Ig receptor density and affinity on B cell signaling by antigen. Impli- caspase-8 activation at the CD95 death-inducing signaling complex. J. Biol. cations for affinity maturation. J. Immunol. 151:5955. Chem. 276:20633. 64. Scaffidi, C., S. Fulda, A. Srinivasan, C. Friesen, F. Li, K. J. Tomaselli, 43. Schmitz, I., S. Kirchhoff, and P. H. Krammer. 2000. Regulation of death receptor- K. M. Debatin, P. H. Krammer, and M. E. Peter. 1998. Two CD95 (APO-1/Fas) mediated apoptosis pathways. Int. J. Biochem. Cell Biol. 32:1123. signaling pathways. EMBO J. 17:1675. 44. Chun, N. J., L. Zheng, M. Ahmad, J. Wang, C. K. Speirs, R. M. Siegel, J. K. Dale, 65. Schmitz, I., H. Walczak, P. H. Krammer, and M. E. Peter. 1999. Differences J. Puck, J. Davis, C. G. Hall, et al. 2002. Pleiotropic defects in lymphocyte between CD95 type I and type II cells detected with the CD95 ligand. Cell Death activation caused by caspase-8 mutations lead to human immunodeficiency. Na- Differ. 6:821. ture 419:395. 66. Alam, M. K., S. Davison, N. Siddiqui, J. D. Norton, and J. J. Murphy. 1997. 45. Mack, A., and G. Hacker. 2002. Inhibition of caspase or FADD function blocks Ectopic expression of Bcl-2, but not Bcl-xL rescues Ramos B cells from Fas proliferation but not MAP kinase-activation and interleukin-2-production during mediated apoptosis. Eur. J. Immunol. 27:3485. primary stimulation of T cells. Eur. J. Immunol. 32:1986. 67. Grimaldi, C. M., J. Cleary, A. S. Dagtas, D. Moussai, and B. Diamond. 2002. 46. Zhang, J., D. Cado, A. Chen, N. H. Kabra and A. Winoto. 1998. Fas-mediated Estrogen alters thresholds for B cell apoptosis and activation. J. Clin. Invest. apoptosis and activation-induced T-cell proliferation are defective in mice lack- 109:1625. ing FADD/Mort1. Nature 392:296. 68. Kozono, Y., R. C. Duke, M. S. Schleicher, and V. M. Holers. 1995. Co-ligation 47. Kennedy, N. J., T. Kataoka, J. Tschopp, and R. C. Budd. 1999. Caspase activation of mouse complement receptors 1 and 2 with surface IgM rescues splenic B cells is required for T cell proliferation. J. Exp. Med. 190:1891. and WEHI-231 cells from anti-surface IgM-induced apoptosis. Eur. J. Immunol. 48. Choe, J., K. Han-Soo, X. Zhang, R. J. Armitage, and Y. S. Choi. 1996. Cellular 25:1013. and molecular factors that regulate the differentiation and apoptosis of germinal center B cells: anti-Ig down-regulates Fas expression on CD40 ligand-stimulated 69. Bonnefoy J.-Y., S. Henchoz, D. Hardie, M. J. Holder, and J. Gordon. 1993. A germinal center B cells and inhibits Fas-mediated apoptosis. J. Immunol. subset of anti-CD21 antibodies promote the rescue of germinal center B cells 157:1006. from apoptosis. Eur. J. Immunol. 23:969. 49. Jung, J., J. Choe, L. Li, and Y. S. Choi. 2000. Regulation of CD27 expression in 70. Liu Y.-J., D. Y. Mason, G. D. Johnson, S. Abbot, C. D. Gregory, D. L. Hardie, the course of germinal center B cell differentiation: the pivotal role of IL-10. Eur. J. Gordon, and I. C. M. MacLennan. 1991. Germinal center cells express bcl-2 I Immunol. 30:2437. protein after activation by signals which prevent their entry into apoptosis. Eur. 50. Hennino, A., M. Be«rard, P. H. Krammer, and T. Defrance. 2001. FLICE-inhib- J. Immunol. 21:1905. Downloaded from itory protein is a key regulator of germinal center B cell apoptosis. J. Exp. Med. 71. Datta, S. R., H. Dudek, X. Tao, S. Masters, H. Fu, Y. Gotoh, and 193:447. M. E. Greenberg. 1997. Akt phosphorylation of BAD couples survival signals to 51. Micheau O., M. Thome, P. Schneider, N. Holler, J. Tschopp, D. W. Nicholson, the cell-intrinsic death machinery. Cell 91:231. C. Briand, and M. G. Gra«tter. 2002. The long form of FLIP is an activator of 72. del Peso, L., M. Gonzalez-Garcia, C. Page, R. Herrera, and G. Nunez. 1997. caspase-8 at the Fas death-inducing signaling complex. J. Biol. Chem. Interleukin-3-induced phosphorylation of BAD through the protein kinase Akt. 277:45162. Science 278:687. 52. Chang, D. W., Z. Zing, Y. Pan, A. Algeciras-Schimnich, B. C. Barnhart, 73. Cardone, M. H., N. Roy, H. R. Stennicke, G. S. Salvesen, T. F. Franke,

S. Yaish-Ohad, M. E. Peter, and X. Yang. 2002. c-FLIPL is a dual function E. Stanbridge, S. Frisch, and J. C. Reed. 1998. Regulation of cell death protease http://www.jimmunol.org/ regulator for caspase-8 activation and CD95-mediated apoptosis. EMBO J. caspase-9 by phosphorylation. Science 282:1318. 21:3704. 74. Buhl, A. M., C. M. Pleiman, R. C. Richert, and J. C. Cambier. 1997. Qualitative 53. Kouskoff, V., S. Famiglietti, G. Lacaud, P. Lang, J. E. Rider, B. K. Kay, regulation of B cell antigen receptor signaling by CD19: selective requirement for J. C. Cambier, and D. Nemazee. 1998. Antigens varying in affinity for the B cell PI3-kinase activation, inositol-1,4,5-triphosphate production and Ca2ϩ mobiliza- receptor induce differential B lymphocyte responses. J. Exp. Med. 188:1453. tion. J. Exp. Med. 186:1897. 54. Fischer, M. B., S. Goerg, S. LiMing, A. P. Prodeus, C. C. Goodnow, G. Kelsoe, 75. Astoul, E., S. Watton, and D. Cantrell. 1999. The dynamics of protein kinase B and M. C. Carroll. 1998. Dependence of germinal center B cells on expression of regulation during B cell antigen receptor engagement. J. Cell Biol. 145:1511. CD21/CD35 for survival. Science 280:582. 76. Shlomchik, M. J., J. E. Craft, and M. J. Mamula. 2001. From T to B and back 55. Barrington, R., M. Zhang, M. Fischer, and M. C. Carroll. 2001. The role of again: positive feedback in systemic autoimmune disease. Nat. Rev. Immunol. complement in inflammation and adaptive immunity. Immunol. Rev. 180:5. 1:147. 56. Van Eijk, M., T. Defrance, A. Hennino, and C. de Groot. 2001. Death-receptor 77. Foote, L. C., A. Marshak-Rothstein, and T. L. Rothstein. 1998. Tolerant B lym- contribution to the germinal-center reaction. Trends Immunol. 22:677. phocytes acquire resistance to Fas-mediated apoptosis after treatment with inter- by guest on October 3, 2021 57. Heyman, B., E. J. Wiersma, and T. Kinoshita. 1990. In vivo inhibition of the leukin 4 but not after treatment with specific antigen unless a surface immuno- antibody response by a complement receptor-specific . globulin threshold is exceeded. J. Exp. Med. 187:847. J. Exp. Med. 172:665. 78. Pisetsky, D. S., S. A. Caster, J. B. Roths, and E. D. Murphy. 1982. Lpr gene 58. Fong, D. C., A. Brauweiler, S. A. Minskoff, P. Bruhns, I. Tamir, I. Mellman, control of the anti-DNA antibody response. J. Immunol. 128:2322. M. Daeron, and J. C. Cambier. 2000. Mutational analysis reveals multiple distinct 79. Wang, H., and M. J. Shlomchik. 1999. Autoantigen-specific B cell activation in sites within Fc␥ receptor IIB that function in inhibitory signaling. J. Immunol. Fas-deficient rheumatoid factor immunoglobulin transgenic mice. J. Exp. Med. 165:4453. 190:639. 59. Kiener, P. A., M. N. Lioubin, L. R. Rohrschneider, J. A. Ledbetter, J. G. Nadler, and M. L. Diegel. 1997. Co-ligation of the antigen and Fc receptors gives rise to 80. Brard, F., M. Shannon, E. L. Prak, S. Litwin, and M. Weigert. 1999. Somatic the selective modulation of intracellular signaling in B cells: regulation of the mutation and light chain rearrangement generate autoimmunity in anti-single- association of phosphatidylinositol 3-kinase and inositol 5-phosphatase with the stranded DNA transgenic MRL/lpr mice. J. Exp. Med. 190:691. antigen receptor complex. J. Biol. Chem. 272:3838. 81. Jacobson, B. A., D. J. Panka, K. A. Nguyen, J. Erikson, A. K. Abbas, and 60. Siegel, R. M., J. K. Frederiksen, D. A. Zacharias, F. K.-M. Chan, M. Johnson, A. Marshak-Rothstein. 1995. Anatomy of autoantibody production: dominant D. Lynch, R. Y. Tsien., and M. J. Lenardo. 2000. Fas preassociation required for localization of antibody-producing cells to T cell zones in Fas-deficient mice. apoptosis signaling and dominant inhibition by pathogenic mutations. Science Immunity 3:509. 288:2354. 82. Prodeus, A. P., S. Goerg, L. M. Shen, O. O. Pozdnyakova, L. Chu, E. M. Alicot, 61. Molina, H., V. M. Holers, B. Li, Y. Fung, S. Mariathasan, J. Goellner, C. C. Goodnow, and M. C. Carroll. 1998. A critical role for complement in J. Strauss-Schoenberger, R. W. Karr, and D. D. Chaplin. 1996. Markedly im- maintenance of self-tolerance. Immunity 9:721. paired humoral immune response in mice deficient in complement receptors 1 and 83. Wu, X., N. Jiang, C. Deppong, J. Singh, G. Dolecki, D. Mao, L. Morel, and 2. Proc. Natl. Acad. Sci. USA 93:3357. H. D. Molina. 2002. A role for the Cr2 gene in modifying autoantibody produc- 62. Sato, S., M. Hasegawa, M. Fujimoto, T. F. Tedder, and K. Takehara. 2000. tion in systemic lupus erythematosus. J. Immunol. 169:1587. Quantitative genetic variation in CD19 expression correlates with autoimmunity. 84. Chen, Z., S. B. Koralov, and G. Kelsoe. 2000. Regulation of humoral immune J. Immunol. 165:6635. responses by CD21/CD35. Immunol. Rev. 176:194.