Expression, Regulation, and Function of -Expressed CD154 in Germinal Centers Amrie C. Grammer, Richard D. McFarland, Jonathan Heaney, Bonnie F. Darnell and Peter E. Lipsky This information is current as of September 25, 2021. J Immunol 1999; 163:4150-4159; ; http://www.jimmunol.org/content/163/8/4150 Downloaded from References This article cites 74 articles, 33 of which you can access for free at: http://www.jimmunol.org/content/163/8/4150.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/ • 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 September 25, 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 © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Expression, Regulation, and Function of B Cell-Expressed CD154 in Germinal Centers1

Amrie C. Grammer,* Richard D. McFarland,† Jonathan Heaney,* Bonnie F. Darnell,† and Peter E. Lipsky2*

Activated B cells and T cells express CD154/CD40 ligand in vitro. The in vivo expression and function of B cell CD154 remain unclear and therefore were examined. Tonsillar B and T cells expressed CD154 at a similar density both in situ and immediately ex vivo, whereas a significantly higher percentage of the former expressed CD154. CD154-expressing B cells were most frequent in the CD38positiveIgD؉ pre- (GC)/GC founder, CD38positive GC and CD38؊IgD؊ memory populations, and were .also found in the CD38؊IgD؉ naive and CD38brightIgD؉ plasmablast subsets, but not in the CD38brightIgD؊ subset B cell expression of CD154 was induced by engaging surface Ig or CD40 by signals that predominantly involved activation of

AP-1/NF-AT and NF-␬B, respectively. The functional importance of CD154-mediated homotypic B cell interactions in vivo was Downloaded from .indicated by the finding that mAb to CD154 inhibited differentiation of CD38positiveIgD؊ GC B cells to CD38؊IgD؊ memory cells In addition, mAb to CD154 inhibited proliferation induced by engaging sIg or CD40, indicating the role of up-regulation of this molecule in facilitating B cell responsiveness. Of note, CD154 itself not only functioned as a ligand but also as a direct signaling molecule as anti-CD154-conjugated Sepharose beads costimulated B cell responses induced by engaging surface Ig. These results indicate that CD154 is expressed by human B cells in vivo and plays an important role in mediating B cell responses. The Journal of Immunology, 1999, 163: 4150–4159. http://www.jimmunol.org/

cells express CD154/CD40 ligand following activation in cient in CD40 (12–14) or its ligand (CD154) (14–16) do not form vitro by ionomycin and phorbol ester (1–3) as well as by functional GCs following immunization with T-dependent (TD) B the polyclonal B cell activators Staphyloccus aureus Ag. Additionally, humans with X-linked hyper-IgM syndrome Cowan I (1, 2) or LPS (4, 5). The finding that CD154 expressed by (HIgMXL) syndrome, who cannot express a functional CD154 (re- activated B cells mediates homotypic CD154-CD40 interactions viewed in Refs. 17 and 18) do not form functional GCs in response resulting in aggregation, clonal expansion, and differentiation into to TD Ag (19, 20). Ig-secreting cells (1, 2, 4, 5) suggests that up-regulation of this The GC is one of the structures in which maturation of the by guest on September 25, 2021 molecule might play an essential role in propagating humoral re- humoral response to Ag occurs, fostering somatic hypermutation, sponses in vivo. Further support for this possibility is the finding selection, and isotype switching of activated B cells (reviewed in that circulating B cells from patients with autoimmune diseases, Refs. 21–23). Although CD154-CD40 interactions are essential for such as systemic lupus erythematosus, characterized by polyclonal initiation and propagation of the GC reaction, there are stages of B cell activation express CD154 (5–7). Moreover, certain B cell the GC reaction that appear to proceed in the absence of CD154ϩ lymphomas express CD154 in a functional manner (8–11). These T cells. CD154ϩ T cells are absent from the dark zone (DZ), where considerations prompted an examination of whether CD154 is ex- rapid B cell proliferation and somatic hypermutation occur, and are pressed physiologically in vivo at sites of B cell activation in sec- found infrequently if at all in the basal zone (LZ) of the GC ondary lymphoid tissue as well as a delineation of the regulation of (24–28), where high avidity Ag-binding B cells are rescued from B cell CD154 expression and its functional activity. apoptosis (29, 30). Despite the paucity of CD154-expressing T Signaling through CD40 is important in the formation and per- cells, an established GC rapidly disassembles when CD154-CD40 petuation of the germinal center (GC)3 reaction since mice defi- interactions are blocked (31). One possible explanation for this finding is that cells other than T cells express CD154 in the DZ and LZ of GCs in secondary lymphoid tissues. Activated CD154-ex- *Harold C. Simmons Arthritis Research Center and Departments of Internal Medicine pressing B cells are prime candidates to provide the essential † and Pathology, University of Texas Southwestern Medical Center, Dallas, TX 75235 CD40-mediated signals in these regions. Received for publication February 24, 1999. Accepted for publication July 29, 1999. To test this hypothesis requires an examination of CD154 ex- The costs of publication of this article were defrayed in part by the payment of page pression by B cells in secondary lymphoid organs. In humans, charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. tonsils have been employed extensively to understand GC behav- 1 This research was supported by National Institutes of Health (NIH) Grant AI-31229. ior, despite their chronic inflammatory and often infected character A.C.G. was supported in part by NIH Postdoctoral Training Grant AR-18550. (reviewed in Ref. 32). Therefore, the current experiments were 2 Address correspondence and reprint requests to Dr. Peter E. Lipsky, Harold C. undertaken to determine whether CD154 is expressed in situ by Simmons Arthritis Research Center, University of Texas Southwestern Medical Cen- tonsillar B cells, to examine the nature of signals that regulate B ter, 5323 Harry Hines Boulevard, Dallas, TX 75235-8884. E-mail address: [email protected] cell CD154 expression, and to investigate the functional activity of 3 Abbreviations used in this paper: GC, germinal center; MFI, mean fluorescence CD154 expressed by human B cells. The data clearly indicate that intensity; TD, T-dependent; DZ, dark zone; LZ, light zone; sIg, surface Ig; hIg, human CD154 is expressed by tonsillar B cells, is up-regulated by en- Ig; Cy, cyclosporine; MNC, mononuclear cell; MKK1, mitogen-activated ki- nase kinase 1; LAC, lactacystein; HIgMXL syndrome, X-linked hyper-IgM syn- gagement of surface Ig (sIg) or CD40 itself, and is likely to be of drome. great importance in propagating GC reactions.

Copyright © 1999 by The American Association of Immunologists 0022-1767/99/$02.00 The Journal of Immunology 4151

Materials and Methods Ridgefield, CT) or biotinylated h2431 followed by either streptavidin Cells conjugated to PE (Becton Dickinson), 613, or 670 (Life Technologies). Alternatively, CD154 expression was analyzed with 8976-PE (Becton Tonsils were minced and digested in RPMI medium (Life Technologies, Dickinson) or with 2443 (kind gift of Dr. Randolph J. Noelle) or uncon- Grand Island, NY) containing 210 U/ml collagenase type I (Worthington jugated 8976, followed by biotinylated rat anti-mouse Ig and streptavidin- Biochemical, Lakewood, NJ) and 90 KU/ml DNase (Sigma, St. Louis, PE. In some cases, B cells were acid washed as described (1) to remove MO) for 30 min at 37°C. Following filtration through a wire mesh, the cells bound CD40 before staining with anti-CD154 mAb. Subsets of tonsillar B were washed twice in 20% NHS-RPMI and once with 10% NHS-RPMI. cells were delineated with Ab specific for CD19 (PerCP, APC, Becton Mononuclear cells (MNC) were obtained by centrifugation of heparinized Dickinson; biotinylated, Coulter), IgD (FITC, Caltag; PE, Southern Bio- venous blood or digested tonsil tissue over diatrizoate/Ficoll gradients technology, Birmingham, AL), CD38 (HB7, APC or PE, Becton Dickin- (Sigma). Blood was obtained from healthy adult volunteers or an HIgMXL son; HIT2, FITC, Caltag), CD23 (biotinylated, The Binding Site, San Di- patient previously demonstrated to lack functional CD154 expression. The ego, CA; FITC, Becton Dickinson), TdT (FITC, Supertech, Rockville, mutations in CD154 of this HIgMXL patient have been reported (Refs. 1, MD), Ki67 (Biogenics, Ramon, CA) followed by anti-mouse IgG1 (FITC, 33; EMBL accession number X96710). In some cases, MNCs were de- The Binding Site), CD44 (A3D8, biotinylated, ATCC), and CD77 (Biode- Ј pleted of NK cells and monocytes and separated into -enriched and sign, Kennebunk, ME) followed by goat F(ab )2 anti-rat IgM FITC (Jack- B cell-enriched populations as described (1). The B cell population was son ImmunoResearch). Isotype-matched mAb were used as controls. Anal- further purified by positive selection on a Ceprate streptavidin column ysis was performed using CellQuest and Paint-a-Gate Software (Becton (CellPro, Bothell, WA) following staining of cells with biotinylated anti- Dickinson). CD19 mAb (Coulter, Hialeah, FL). Alternatively, the B cell population was purified by negative selection on a magnetic column (StemCell Technol- RNA extraction, cDNA synthesis, and PCR analysis ogies, Vancouver, Canada), following staining of cells with a mixture of RNA and cDNA were prepared as previously described (1). PCR analysis dextran cross-linked mAb specific for , CD2, CD3, CD14, was made semiquantitative by varying the number of amplification cycles CD16, and CD56, followed by exposure to a magnetic colloid covalently and performing dilutional analysis so that there was a linear relationship Downloaded from linked to anti-dextran mAb. The resultant population of B cells was ana- ϩ between the amount of cDNA used for each reaction and the intensity of lyzed by flow cytometry and found to be Ͼ97% sIg (FITC-conjugated the band obtained by Southern hybridization of the PCR product. PCR anti-polyvalent Ig Ab, Caltag, South San Francisco, CA; FITC-conjugated reactions and Southern blotting were performed as described (1). MgCl2 anti-␬ mAb, Becton Dickinson, San Jose, CA; and PE-conjugated anti-␭ ␣ ϩ ϩ was used at a concentration of 2 mM for G6PD and TCR and 2.5 mM for mAb, Coulter) and Ͻ3% CD3 or CD4 (FITC; PerCP, Becton Dickin- CD154. The primers used for the PCR reactions were: CD154: L5 and III3 son). To eliminate remaining T cells from the purified B cell population, (1); G6PD: 5Ј-ACC TAC AAG TGG GTG AAC CC-3Ј and 5Ј-CTT GGC where indicated, cells were stained with FITC-conjugated anti-CD3 mAb AGC TGA GGA ATG TAG C-3Ј; and TCR␣:5Ј-GAA CCC TGA CCC http://www.jimmunol.org/ Plus (Sigma) and sorted for the CD3-negative population using the FACStar TGC CGT GTA CC-3Ј and 5Ј-ATC ATA AAT TCG GGT AGG ATC (Becton Dickinson). C-3Ј. In some cases, PCR products were purified using the QIAquick-spin Culture conditions PCR purification (Qiagen, Chatsworth, CA) following the manufactur- er’s instructions and subjected to nested PCR using primers II5Ј and II3Ј B cells were cultured (1 ϫ 105/well) in U-bottom 96-well microtiter plates (1). The following probes were used for Southern blotting: CD154, Ia3Ј or (Costar, Cambridge, MA) in RPMI medium supplemented with penicillin III5Ј (1); G6PD, 5Ј-ATT GAC CTC AGC TGC ACA TTC C-3Ј; and G (200 U/ml), gentamicin (10 ␮g/ml), and 10% FCS. In some cases, cells TCR␣,5Ј-GTC ACT GGA TTT AGA GTC TCT C-3Ј. were activated with 10 ng/ml PMA (Sigma) and 1.34 ␮M ionomycin (Cal- biochem, La Jolla, CA). sIg was cross-linked with anti-IgM, anti-IgD (Bio- Analysis of B cell function source International, Camarillo, CA), or glycine-conjugated Sepharose 3

Proliferation was analyzed by [ H]thymidine incorporation as previously by guest on September 25, 2021 ␮ Ј beads. Alternatively, sIg was engaged with 10 g/ml of a F(ab )2 of anti- described (1). The percentage of live and apoptotic cells in a given pop- human IgM or polyvalent Ig Ab (Jackson ImmunoResearch, West Grove, ulation was analyzed by hypotonic propidium iodide staining as previously PA) or whole mouse anti-human IgM mAb (DA4.4, American Type Cul- described (36). ture Collection (ATCC), Manassas, VA). CD40 was engaged with membranes from Sf9 cells infected with re- Results combinant baculovirus encoding murine CD154 and prepared as previ- ously described (34, 35). Specificity of the effects of this reagent was dem- Freshly isolated tonsillar B cells express a higher density of onstrated by incubation with or without 5–10 ␮g/ml of either MR1, a CD154 than in vitro-activated peripheral B cells hamster anti-mouse CD154 mAb (kind gift of Dr. Randolph Noelle, Dart- Initial experiments demonstrated that tonsillar B and T cells in situ mouth Medical School, Lebanon, NH) or 2C11, a control hamster anti- mouse CD3 mAb (ATCC) that has no reactivity with murine CD154 or and freshly isolated ex vivo express CD154 at equivalent densities human lymphocytes. without in vitro stimulation (Figs. 1 and 2). The density of CD154 To analyze the effects of CD154, cells were incubated in the presence expression by tonsillar T and B cells is lower than that expressed ␮ of 10 g/ml of a mouse anti-hCD154 mAb (5c8; Ref. 25; kind gift of by peripheral blood T cells activated in vitro and higher than that Biogen, Boston, MA), or P1.17 (ATCC), an isotype-matched control mouse mAb. In some cases, cells were incubated with intact humanized expressed by in vitro-activated peripheral blood B cells as assessed mouse anti-hCD154 mAb (h24-31; kind gift of the IDEC, San Diego, CA), by binding of a CD40.Ig construct (1) or the anti-CD154 mAb F(ab) fragments of h24-31 generated by pepsin digestion, or h24-31 con- 8976 (Fig. 1A), 2431, and 2443 (data not shown). The specificity jugated to Sepharose beads. of anti-CD154 mAb was demonstrated by the finding that prein- Pooled human Ig (hIg, Sandoglobulin, Novartis, East Hanover, NJ), cubation with excess CD40.Ig decreased both the mean fluores- F(ab) fragments of hIg, glycine conjugated to Sepharose beads, or mem- branes from Sf9 cells infected with wild-type baculovirus were used as cence intensity (MFI) (Fig. 1A) and the percentage of cells stained controls. with these mAb (data not shown). Additionally, both FACS anal- Inhibitors ysis and immunohistochemistry demonstrated that the anti-CD154 mAb employed in these studies did not stain T or B cells isolated Calcineurin-dependent NF-AT activation was inhibited with cyclosporine from a donor with HIgMXL syndrome (data not shown) lacking (Cy, 100 ng/ml; Novartis), MKK1 activation was inhibited with PD98059 functional CD154 expression (1, 33). It is important to note that (200 ␮M; Calbiochem), and NF-␬B activity was blocked with an inhibitor of proteosome-mediated I␬B degradation, lactacystein (100 ␮M; Calbio- tonsillar B cells not only expressed the CD154 epitopes recognized chem). New protein synthesis was inhibited with cycloheximide (1 ␮g/ml; by the various mAb, but had functionally active CD154 in that it Sigma). bound a CD40.Ig construct (Fig. 1A). Finally, in agreement with Flow cytometric analysis the staining data, RT-PCR and Southern blotting determined that activated peripheral blood T cells expressed the greatest CD154 Cells were stained with mAb as previously described (1) and analyzed with a FACScan, a FACScalibur, a FACStarPlus, or a FACS Vantage flow cy- mRNA, tonsillar T and B cells expressed moderate CD154 mRNA, tometer (Becton Dickinson). CD154 expression was analyzed with biotin- and activated peripheral blood B cells expressed the least CD154 ylated CD40.Ig (kind gift of Dr. Marilyn Kehry, Boehringer Ingelheim, mRNA (Fig. 1B). Importantly, T cell contamination of the B cells 4152 B CELL-EXPRESSED CD154 IN GERMINAL CENTERS

munohistochemical analysis of serial sections of tonsil tissue con- firmed that T cells expressing CD154 were localized to the inter- follicular regions (25) and that B cells expressing CD154 were found in the IgD-positive mantle zone and the GC itself. In addi- tion, CD154 was expressed by IgD-positive plasmablasts in the interfollicular and lymphoepithelial regions (data not shown). Expression of CD154 by tonsillar B cell subsets Initial analysis of tonsillar B cell staining by FACS confirmed that they could be contained in the previously described (22, 23, 37) CD38ϪIgDϩ naive, CD38ϪIgDϪ memory, CD38ϩIgDϩ mantle zone/pre-GC/GC founder, CD38ϩIgDϪ GC, and CD38bright plasma cell subsets (Fig. 3). Furthermore, immunohistochemical analysis of frozen tonsil sections confirmed that the mantle zone and GC contained CD38ϩ cells that were IgDϩ and IgDϪ respec- tively, whereas CD38bright cells were found in the interfollicular and lymphoepithelial regions (data not shown). Further analysis of the FACS data revealed eight populations of

B cells defined by expression of IgD and CD38. It should be noted Downloaded from that there was considerable heterogeneity in the distribution of B cell subsets in individual tonsils. Thus, some tonsils contained all eight B cell populations (n ϭ 2), whereas other tonsils contained only some (n ϭ 8). A plot of a tonsil containing all eight subsets is shown in Fig. 3A, along with the distribution of CD19ϩ tonsillar

B cells into those subsets in ten different tonsils (Fig. 3B). Novel http://www.jimmunol.org/ subpopulations of CD38lowIgDϩ and CD38lowIgDϪ tonsillar B cells were observed following staining with the HB7 anti-CD38 (Fig. 3A) but not the HIT2 anti-CD38 mAb (data not shown). Analysis of tonsillar B cell subpopulations for expression of CD154 revealed that all tonsillar B cell subsets displayed CD154 to varying degrees, with the exception of CD38bright/ϩϩϩIgDϪ cells (Fig. 4A). Further analysis revealed there was no significant difference in the percentage of CD154-expressing cells observed in the IgDϩ and IgDϪ subsets of tonsillar B cells, 35.4 Ϯ 28.4% by guest on September 25, 2021 (range 3.8–93.6%) and 64.7 Ϯ 28.4% (range 6.4–96.2%), respec- tively ( p ϭ 0.08, n ϭ 10). However, a significantly higher per- centage of B cells expressing CD154 was found in the FIGURE 1. Freshly isolated tonsillar B cells express a higher density of CD38ϩIgDϪ subpopulation (27.3 Ϯ 17.2%; range 0.7–48.7%) CD154 than in vitro-activated peripheral blood B cells. A, CD154 expres- ϭ ϩ ϩ when compared with all other tonsillar B cell subsets ( p 0.02, sion on CD3 T cells and CD19 B cells was assessed by FACS analysis ϩ n ϭ 10). Moreover, a significantly lower percentage of CD154 B following incubation with or without CD40. Ig and staining with FITC- Ϫ Ϫ Ϯ conjugated mAb to CD19 or CD3 and PE-conjugated anti-CD154 mAb cells was found in the CD38 IgD memory (17.4 11.0%; range 0.1–39.3%), CD38ϩIgDϩ (14.1 Ϯ 8.0%; range 1.9–26.3%), (89-76). Tonsillar MNCs were stained immediately following isolation. ϩϩ ϩ ϩϩ Ϫ Highly purified peripheral blood T and B cells were stained following CD38 IgD (12.4 Ϯ 8.6%; range 0–53.6%), and CD38 IgD culture for 18 h with ionomycin and PMA. The results of one of two subpopulations (14.8 Ϯ 8.7%; range 0.6–28.2%) when compared ϩ Ϫ experiments with similar findings are shown as density of CD154 expres- with the CD38 IgD subpopulation ( p Ͻ 0.05, n ϭ 10). The sion by the respective populations (MFI). B, Analysis of CD154 mRNA lowest percentage of CD154ϩ B cells was observed in the was performed by RT-PCR using varying amounts of cDNA from freshly CD38ϪIgDϩ naive (5.9 Ϯ 3.8%; range 0–11.6%) and isolated tonsillar T and B cells as well as from highly purified peripheral T CD38ϩϩϩIgDϩ (6.2 Ϯ 10.0%; range 0.1–25.9%) subpopulations and B cells following stimulation for 18 h with ionomycin and PMA. (p Ͻ 0.05, n ϭ 10). Finally, CD154-expressing B cells were never Following Southern blotting analysis, the intensity of the CD154 PCR ϩϩϩ Ϫ found in the CD38 IgD subpopulation (0.5 Ϯ 0.5%; range product band was quantitated by AMBIS analysis and plotted as the rela- ϭ tive density of PCR product as a function of the amount of initial cDNA. 0–1.7%; n 10). The greatest CD154 density was observed in activated naive B cells, mantle zone/pre-GC/GC founder (CD38ϩ/ ϩ ϩ Ϫ ϩϩϩ ϩ ϩϩIgD ), as well as in the CD38 IgD and CD38 IgD pop- was ruled out since mRNA for the ␣-chain of the TCR (TCR␣) ulations. Moreover, the average density of CD154 was signifi- could not be amplified from these B cell samples (data not shown). cantly greater on CD38ϪIgDϩ naive B cells and CD38ϩIgDϪ B All tonsils studied contained both T and B cells, although the cells when compared with CD38ϪIgDϪ memory B cells and percentages of these subsets varied with the donor (Fig. 2A). Rou- CD38ϩϩIgDϪ GC B cells (data not shown). tinely, tonsil MNCs contained a significantly higher percentage of CD23 expression by tonsillar B cell subpopulations defined by B cells compared with T cells (Fig. 2A), and a greater percentage IgD and CD38 was also examined, since previous studies had de- of tonsillar B cells expressed CD154 compared with tonsillar T fined this Ag as a marker of mantle zone/pre-GC/GC founder cells cells (Fig. 2B), although the density of CD154 expressed was (reviewed in Ref. 22). Although CD23-expressing B cells were equivalent (Fig. 1C). Furthermore, all tonsils contained CD154- found in all tonsillar subpopulations defined by IgD and CD38 expressing B cells, but somewhat surprisingly only some tonsils (data not shown), a significantly higher percentage ( p ϭ 0.02, n ϭ contained CD154-expressing T cells (data not shown). Finally, im- 5) of the CD38ϩ and CD38ϩϩ subsets expressed CD23 (11.5 Ϯ The Journal of Immunology 4153

FIGURE 2. Comparative density of CD154 expressed by tonsillar B and T cells. Freshly isolated MNCs from eight individual tonsils were assessed for the percentage of T and B cells (A), the percentage of these lymphocyte populations that express CD154 (B), and the relative density of CD154 expression (% maximum MFI of either cell type determined as a fraction) by tonsillar CD19Ϫ T cells and CD19ϩ B cells (C). FACS analysis was performed after staining with PE-conjugated anti-CD154 mAb (8976) and FITC-conjugated anti-CD19 mAb. Similar results were found when CD154 expression by the CD3ϩ and CD3Ϫ populations were analyzed (data not shown). Data are expressed as the mean Ϯ SEM of results from eight tonsils. Significance was determined by paired two-sample Student’s t test. Downloaded from

7.5%, range 1.3–20.5%, and 6.4 Ϯ 6.2%, range 0.6–18.1, respec- these receptors. It is important to emphasize that engagement of tively) when compared with the CD38Ϫ and CD38ϩϩϩ subpopu- CD40 was accomplished with Sf9 membranes expressing recom- lations (2.5 Ϯ 1.6%, range 0.4–4.4%, and 1.4 Ϯ 2.6%, range binant murine CD154 that is not recognized by the anti-human 0–6.5%, respectively). This trend was observed regardless of IgD CD154 mAb used for detection. Furthermore, experiments were expression. In addition, there was no significant difference in the conducted using an amount of mCD154 previously shown to in- overall percentage of cells expressing CD23 when the IgDϩ and duce a variety of functional outcomes (34–36). Analysis by FACS http://www.jimmunol.org/ IgDϪ B cells were compared (12.6 Ϯ 12.6%, range 0–30.3% vs (Fig. 5) and fluorescence microscopy (data not shown) demon- 13.9 Ϯ 11.8%, range 0.4–29.1%; p ϭ 0.26, n ϭ 5). strated that engagement of CD40 or sIg on B cells induced ex- Analysis of CD154 expression by tonsillar B cell subpopula- pression of CD154. Moreover, there was a relationship between tions defined by CD38, IgD, and CD23 (Fig. 4B) revealed that a the amount of mCD154 used for stimulation and the level of sub- significantly greater percentage ( p ϭ 0.05, n ϭ 5) of CD154- sequent CD154 expression (Fig. 5B). expressing cells was observed in CD23ϩCD38ϩ tonsillar B cells (32.0 Ϯ 13.6%; range 10.8–52.2%) when compared with CD23ϪCD38ϩ tonsillar B cells (11.6 Ϯ 6.6%; range 0.3–18.9%). Induction of CD154 requires new protein synthesis and is by guest on September 25, 2021 Moreover, within CD23-expressing cells, a significantly ( p Ͻ inhibited by cyclosporine, PD98059, or lactacystein 0.03, n ϭ 5) higher percentage of CD154-expressing cells was To investigate potential signaling mechanisms involved in regu- ϩ ϩϩ found in the CD38 and CD38 subsets (32.0 Ϯ 13.6%, range lating CD154 expression in human B cells, highly purified periph- 10.8–52.2%, and 19.0 Ϯ 13.3%, range 5.5–42.4%, respectively) eral blood B cells were activated with Sf9 membranes expressing Ϫ ϩϩϩ when compared with the CD38 and CD38 subpopulations recombinant mCD154 or with anti-Ig in the presence or absence of (6.6 Ϯ 4.0%, range 1.6–10%, and 5.2 Ϯ 8.2%, range 0–21.1%, various inhibitors (Fig. 6). Engagement of CD40 or sIg induced respectively). By contrast, examination of CD154 expression in CD154 protein expression and mRNA (Fig. 6). Importantly, T cell Ϫ CD23 tonsillar B cells revealed that a significantly higher per- contamination of the peripheral B cells was ruled out since TCR␣ Ϫ ϩ ϩϩ centage ( p Ͻ 0.02, n ϭ 5) were CD38 , CD38 , and CD38 , mRNA could not be amplified from these samples (Fig. 6B). More- (11.1 Ϯ 6.8%, range 0–19.1%; 11.6 Ϯ 6.6%, range 0.3–18.9%; over, specificity for the effect of mCD154-expressing Sf9 mem- ϩϩϩ 14.6 Ϯ 8.5%, range 2.6–25.4%), than were CD38 (0.3 Ϯ branes was documented, since all effects were blocked by an anti- 0.4%, range 0–0.5%). mCD154 mAb (Fig. 6B, data not shown). Cycloheximide inhibited induction of surface CD154 ( p Ͻ 0.02) following engagement of Engagement of CD40 or sIg on human B cells up-regulates sIg or CD40, demonstrating that expression requires new protein CD154 expression synthesis (Fig. 6A). Additionally, CD154 induced by engaging sIg To examine whether engagement of CD40 or sIg induces CD154 or CD40 was blocked with lactacystein ( p Ͻ 0.01; Fig. 6, A and expression, B cells were initially analyzed following ligation of B), an inhibitor of proteosome-mediated degradation of I␬B. By

FIGURE 3. Subsets of B cells from multiple ton- sils characterized by expression of CD38 and IgD. Freshly isolated MNCs or negatively selected B cells from ten individual tonsils were assessed for CD19ϩ subpopulations by FACS analysis following staining with FITC-conjugated anti-IgD, PerCP-conjugated anti-CD19, and APC-conjugated anti-CD38. Naive, memory, and GC subsets are indicated as previously described (22, 23, 37). The mean Ϯ SEM percent- ages of CD19ϩ cells in each subset defined by CD38 and IgD expression from ten tonsils are depicted in B. A plot of a tonsil containing all eight subpopula- tions is shown in A. 4154 B CELL-EXPRESSED CD154 IN GERMINAL CENTERS

FIGURE 4. Expression of CD154 by tonsil- lar B cell subsets. Freshly isolated MNCs or negatively selected B cells from individual tonsils were assessed for CD154 expression by FACS analysis following staining with PE- conjugated anti-CD154 (8976), PerCP-conju- gated anti-CD19, APC-conjugated anti-CD38, and FITC-conjugated anti-IgD (A) or FITC- conjugated anti-CD23 (B). The mean Ϯ SEM percentages of CD154-expressing cells in each CD19ϩ B cell subset defined by CD38 and IgD expression from ten tonsils (A) or CD38 and CD23 expression from five tonsils (B) are

depicted. Downloaded from http://www.jimmunol.org/ contrast, inhibiting calcineurin with cyclosporine or MKK1 activ- GC B cells were cultured in vitro in the presence of a saturating ity with PD98059 interfered with induction of CD154 on periph- concentration of anti-CD154 mAb or an isotype-matched control eral B cells following engagement of sIg ( p ϭ 0.05) but not CD40 mAb. GC B cells spontaneously differentiated into memory B cells ( p ϭ 0.29) (Fig. 6, A and B). Of note, similar results were observed during a 3-day incubation, with 48% of CD38ϩIgDϪ and 8% of when sIg was engaged with either soluble anti-IgM, anti-IgD, or CD38ϩϩIgDϪ cells becoming CD38ϪIgDϪ during this time pe- polyvalent Ig alone or conjugated to Sepharose beads (data not riod. Of note, the emergence of CD38ϪIgDϪ memory B cells was shown). partially inhibited by blocking homotypic CD154-CD40 interac- by guest on September 25, 2021 tions with an anti-CD154 mAb ( p Ͻ 0.05; Fig. 7). As an additional Homotypic CD154-CD40 interactions between GC tonsillar B control, the impact of the anti-CD154 mAb on the spontaneous in cells play a role in differentiation to the memory phenotype vitro differentiation of CD38ϪIgDϩ naive tonsillar B cells into ϩ ϩ To examine the contribution of homotypic CD154-CD40 interac- CD38 IgD pre-GC/GC founder B cells was examined. No inhi- tions between tonsillar B cells during differentiation to the memory bition was noted (data not shown). To ensure that the impact of B cell phenotype, highly purified CD38ϩϩIgDϪ and CD38ϩIgDϪ anti-CD154 mAb on differentiation was not sec- ondary to an effect on B cell viability or apoptosis, CD38ϪIgDϪ memory B cells were analyzed only in the nonapoptotic population detected by staining with propidium iodide, and a similar effect was noted.

CD40- and sIg-induced CD154 plays a functional role in B cell responses The final experiments examined whether CD154 expressed by B cells was involved in functional responses in vitro. Whereas whole anti-CD154 mAb inhibited anti-Ig- or CD154-induced DNA syn- thesis of highly purified peripheral B cells obtained from some normal donors (Fig. 8A), anti-CD154 mAb had no effect on anti- Ig- or CD154-induced DNA synthesis of peripheral B cells from a donor with HIgMXL syndrome (Fig. 8B) previously documented not to express functional CD154 (1, 33). It should be noted that similar results were observed when sIg was engaged with either FIGURE 5. Engagement of CD40 or sIg on B cells induces surface soluble anti-IgM or anti-IgD alone or either anti-Ig conjugated to ϫ 5 expression of CD154. Highly purified peripheral blood B cells (1 10 ) Sepharose beads (data not shown). Specificity for the effect of were cultured for 18 h with medium alone, 10 ␮g/ml anti-Ig, or 2 ␮l mCD154-expressing Sf9 membranes was documented, since all membranes from Sf9 cells expressing baculovirus-encoded recombinant mCD154 (left). Right, Highly purified peripheral blood B cells were cul- effects were blocked by an anti-mCD154 mAb (data not shown). ϭ tured for 18 h with medium alone or with 2 or 6 ␮l of membranes from Sf9 Although anti-CD154 inhibited anti-Ig (n 4 of 16) and CD154 cells expressing baculovirus-encoded recombinant mCD154. Human (n ϭ 4 of 4) induced proliferation of B cells from some donors, in CD154 expression was assessed by FACS after staining with PE-conju- other donors (n ϭ 9 of 16), intact anti-CD154 costimulated anti- gated anti-CD154 (8976) and FITC-conjugated anti-CD19. Ig-induced DNA synthesis of highly purified normal peripheral B The Journal of Immunology 4155

FIGURE 6. Role of AP-1/NF-AT and NF-␬B in regulating CD154 expression by ac- tivated peripheral B cells. Highly purified pe- ripheral blood B cells (1 ϫ 105) were cultured with medium alone, 10 ␮g/ml anti-Ig, or mem- branes from Sf9 cells expressing recombinant mCD154 in the presence or absence of the 100 ng/ml Cy, 200 ␮M PD98059, 30 ␮M lactacys- tein, or 1 ␮g/ml cycloheximide. Human CD154 expression was assessed by (A) FACS analysis after staining cells with FITC-conju- Downloaded from gated anti-CD19 and PE-conjugated anti- hCD154 (8976) or by (B) RT-PCR and South- ern blotting of isolated mRNA. Data in A are expressed as the mean of two to seven exper- iments Ϯ SEM. Significance was determined by paired two-sample Student’s t test. The blot http://www.jimmunol.org/ in B is representative of three experiments with similar results. by guest on September 25, 2021

cells and in others (n ϭ 3 of 16) had little positive or negative ference in CD154 expression by B cells isolated from the tonsil impact (data not shown). To analyze this apparent functional het- compared with that by peripheral B cells stimulated in vitro sug- erogeneity further, additional experiments were conducted with gests that there may be signaling molecules present in inflamed F(ab) fragments of anti-CD154 as well as anti-CD154 conjugated secondary lymphoid tissue that costimulate up-regulation of to Sepharose beads. Whereas F(ab) fragments of anti-CD154 con- CD154 in tonsillar B cells (38–43). It is also possible that tonsillar sistently blocked DNA synthesis of normal peripheral B cells fol- B cells have differentiated sufficiently so that signals provided by lowing engagement of sIg or CD40 (Fig. 8, C and D), engaging ligation of sIg or CD40 directly induce exaggerated expression of CD154 with anti-CD154 conjugated to Sepharose beads signifi- CD154. Whether enhanced expression of CD154 by tonsillar B cantly costimulated anti-Ig-induced, but not CD154-induced, DNA cells reflects intrinsic features of the B cells themselves and/or a synthesis of normal peripheral B cells (Fig. 8C). facilitating influence of the tonsillar milieu remains to be determined. Discussion Previous studies had defined tonsil B cell subsets by their ex- pression of IgD and CD38 (22, 23, 37, 44). This study extends Tonsillar B cells expressed CD154 at a similar density to tonsillar these findings by demonstrating that CD38ϩ tonsillar B cells can T cells both in situ and ex vivo. B cell CD154 expression could be be divided into novel populations of CD38low and CD38intermediate induced following engagement of either CD40 or sIg in a manner cells by flow cytometry (Fig. 3A) with the HB7 anti-CD38 mAb. that required new protein synthesis and signaling pathways that Follicular mantle zone/pre-GC/GC founder cells are known to be predominantly involved AP-1/NF-AT and NF-␬B. Furthermore, CD38ϩIgDϩ cells (22, 45). In the current study, these cells were CD154-mediated B cell homotypic interactions played a role in found within CD38ϩIgDϩ and CD38ϩϩIgDϩ tonsillar B cell pop- promoting differentiation of GC cells to memory B cells. Finally, ulations (Figs. 3 and 4) since the highest percentage of cells ex- B cell-expressed CD154 costimulated functional responses by act- pressing the previously defined marker of mantle zone pre-GC/GC ing as a ligand for CD40 as well as a direct signaling molecule. founder cells, CD23 (22, 45, 46), was found in both subsets (Fig. Freshly isolated tonsillar B cells express a level of CD154 4B). With regard to expression of CD154 by tonsillar B cell sub- mRNA and protein equivalent to that expressed by tonsillar T cells sets, it should be noted that the majority of CD154-expressing cells and a much higher level of CD154 mRNA and protein than pe- observed in the CD38ϩIgDϩ and CD38ϩϩIgDϩ populations were ripheral B cells after in vitro stimulation (Figs. 1 and 2). The dif- CD23ϩ, with the greatest expression noted in the former (Fig. 5), 4156 B CELL-EXPRESSED CD154 IN GERMINAL CENTERS

FIGURE 7. Homotypic CD154-CD40-mediated in- teractions between GC B cells play a role in survival and differentiation to the memory subset. Following staining with APC-conjugated anti-CD19 or bio- tinylated anti-CD19 and strepavidin-670, PE-conju- gated anti-CD38, and FITC-conjugated anti-IgD, CD38ϩϩIgDϪ and CD38ϩIgDϪ GC B cells were sorted using a FACStarPlus or FACS Vantage flow cy- tometer. Representative pre- and post-sort analysis of tonsillar B cells is shown in the left panel. Tonsillar CD38ϩϩIgDϪ or CD38ϩIgDϪ B cells (1 ϫ 105; pu- rity of 98–99%), analyzed immediately (initiation) for the presence of CD38ϪIgDϪ memory B cells, were cultured in the presence of 10 ␮g/ml of the anti- Downloaded from CD154 mAb, 5c8, or an isotype-matched control mAb, P1.17 (control). Following a 3-day incubation, the presence of CD38ϪIgDϪ B cells in the nonapop- totic cells, determined by propidium iodide staining, was analyzed for the presence of CD38ϪIgDϪ mem- ory B cells. Data are expressed as the mean Ϯ SEM.

Significance was determined by paired two-sample http://www.jimmunol.org/ Student’s t test.

consistent with their status as activated pre-GC/GC founder cells. actions. Costimulation via CD154 engagement was not observed by guest on September 25, 2021 Moreover, the highest percentage of B cells expressing CD154 was either because the initial signaling was insufficient to induce this found in the CD38ϩIgDϪ population when compared with all pathway of costimulation or because the CD40 and CD154 sig- other tonsillar B cell subsets (Fig. 4A). This CD38ϩIgDϪ subset naling pathways are redundant. By contrast, engagement of sIg- was also Ki67Ϫ, TdTϪ, CD77low, and CD44high (data not shown) induced CD154 expression and also CD154-mediated bidirectional consistent with the designation of this population as costimulation via CD40 or CD154 or both. The effect of the intact (22, 23, 47). Finally, a higher percentage of centrocytes expressed anti-CD154 mAb was likely to reflect the dominance of the CD40 CD154 than was noted for the CD38ϩϩIgDϪ population (Fig. 4A) or CD154 signaling pathway and could depend on the intensity of that was Ki67ϩ, TdTϩ, CD77high, and CD44low (data not shown), the signal generated from ligation of sIg or the density of CD154 consistent with the designation of this population as centroblasts expressed or both. Although preliminary work has documented the (22, 23, 45, 47). capacity of CD154 engagement to induce a variety of proximal Engagement of CD40 or sIg on B cells induced CD154, and the signaling events in other cell types (48–53), this is the first exam- expression of this molecule clearly contributed to ongoing re- ple of the ability of CD154 to mediate B directly. sponses. Whereas intact anti-CD154 mAb consistently had no ef- The potential role of CD154 as a signaling receptor on B cells is fect on anti-Ig- or CD154-induced proliferative responses from an consistent with the previous finding that injection of a CD40.Ig HIgMXL donor (Fig. 8B), the results with normal donors sug- construct into a mouse genetically deficient in CD40 induced gested that anti-CD154 may have exerted multiple effects on small, but quantifiable, GCs after immunization and also enhanced activated B cells, blocking CD154-CD40-mediated costimulation in vivo production of IgM following immunization of normal mice in some experiments and functioning as a direct costimulator in (54, 55). others. The results using F(ab) fragments of anti-CD154 or CD154 After engagement of sIg or CD40, B cell CD154 expression was coupled to Sepharose beads provide a potential explanation for induced by specific pathways of transcriptional regulation. Of im- these findings. Whereas the anti-CD154 24-31 F(ab) fragment con- portance, engagement of both sIg and CD40 induced de novo syn- sistently blocked anti-Ig-induced proliferation, the 24-31 anti- thesis of CD154 and not reexpression of preformed protein from CD154 mAb conjugated to Sepharose beads consistently costimu- intracellular stores, as has been reported for tonsillar (27) and sy- lated. It should be noted that somewhat different from the results novial (56) T cells and anti-Ig-stimulated murine splenic B cells noted with anti-Ig stimulation, proliferation following CD40 en- (4). Utilization of specific inhibitors demonstrated that engagement gagement was blocked with both the whole anti-CD154 mAb and of sIg on resting, peripheral B cells induced CD154 expression by the F(ab) of the anti-CD154 mAb; costimulation was not observed, means of signaling pathways involving calcineurin and therefore even with anti-CD154-conjugated Sepharose beads. Thus, CD40 likely leading to nuclear translocation of NF-ATc. This finding ligation-induced proliferation appears to be dependent on the en- parallels the effects noted in T cells (38, 57–60). Moreover, dogenous expression of CD154 and ongoing CD154-CD40 inter- NF-AT motifs in the 5Ј promoter region of the CD154 gene in both The Journal of Immunology 4157 Downloaded from http://www.jimmunol.org/

FIGURE 8. Functional activity of B cell-expressed CD154. Highly purified peripheral blood B cells (1 ϫ 105) from a normal donor (A, C, and D)or a donor with HIgMXL syndrome (B) were cultured for 3 days with medium alone, 10 ␮g/ml anti-Ig (A, B, and C), or membranes from Sf9 cells expressing recombinant mCD154 (A, B, and D) in the presence or absence of intact anti-human CD154 mAb (A and B), a F(ab) fragment of an anti-CD154 mAb (C and D), or Sepharose beads conjugated with anti-CD154 mAb (C and D). Proliferation was assessed by [3H]thymidine incorporation. Data are expressed by guest on September 25, 2021 as the mean Ϯ SEM. Representative experiments are shown. The results of one of four experiments are shown for A, two experiments for B, and three experiments for C and D. Experiments with intact 24-31 anti-CD154 mAb were performed in the presence of anti-Ig 16 times and in the presence of membranes from Sf9 cells expressing recombinant mCD154 4 times.

the mouse (61) and human (62) have been shown to bind NF-ATp/ The current data suggest that CD154 expression observed by c-NF-ATn/AP-1 complexes in nuclear extracts derived from acti- tonsillar B cells may be the result of in vivo signaling through vated T cells and to control transcription of the gene. Although CD40 or the sIg complex during an immune response initially NF-AT has been thought to be a T cell-specific transcription factor, induced by T-dependent Ags. Ligation of CD40 on naive B cells there is a growing body of evidence that B cells can also be in- in the interfollicular zone of tonsils by T cells expressing CD154 duced to activate NF-AT in a Cy-sensitive manner by a variety of may induce CD154 expression on the tonsillar B cells themselves. stimuli, including engagement of sIg (63–68). The current data Engagement of CD40 on naive B cells by CD154-expressing T indicate that this pathway plays an essential role in induction of B cells has been shown to lead to expression of CD23 (71) and CD38 cell CD154 expression following engagement of sIg, and perhaps (35). Of interest, in the current study, the highest percentage of in permitting anti-Ig-activated B cells to employ CD154 as a co- CD23-expressing cells was observed in the CD38ϩ and CD38ϩϩ stimulatory molecule. subsets (data not shown). Moreover, a significantly greater per- In addition to the apparent role of NF-AT in regulating B cell centage of CD154-expressing cells was observed in the CD23ϩ CD154 expression following engagement of sIg, NF-␬B activation portion of CD38ϩ B cells when compared with those that were played an important role in up-regulating the expression of CD154 CD23Ϫ (Fig. 4B). These observations suggest the possibility that following ligation of either CD40 or sIg. In this regard, examina- induction of these molecules might be induced coordinately fol- tion of the published sequence of the human CD154 promoter lowing CD40 ligation and the initiation of GC reactions. (GenBank/EMBL accession number L47983; Ref. 62) reveals the Previous studies have demonstrated that maintenance of GC re- presence of five potential NF-␬B binding sites (69), including at actions requires ongoing signaling through the CD154-CD40 co- least one within the proximal CD154 promoter necessary for PMA receptors. Specifically, the entire GC, including the DZ, rapidly and Con A-driven transcription of the CD154 gene in Jurkat T disassembles following administration of an anti-CD154 mAb to cells (62). Importantly, ligation of both sIg and CD40 is known to an immunized mouse (31), even though T cells are largely absent activate NF-␬B in B cells (reviewed in Ref. 69). Moreover, CD40- from the DZ of GCs in the mouse or human (24–28). One expla- mediated induction of CD154 mRNA in Daudi B cells was nation for this finding is that CD154 expressed by GC B cells may blocked by the src kinase inhibitor, herbimycin A (3), previously sustain clonal expansion in the absence of T cells. This hypothesis shown to interfere with CD40-induced activation of NF-␬B (70). is strengthened by our previous finding that CD154 expression on 4158 B CELL-EXPRESSED CD154 IN GERMINAL CENTERS

B cells leads to homotypic CD154-CD40 interactions and DNA References synthesis (1, 2) and the current finding that CD40-induced prolif- 1. Grammer, A. C., M. C. Bergman, Y. Miura, K. Fujita, L. S. Davis, and eration of B cells is partially mediated by subsequent interactions P. E. Lipsky. 1995. The CD40 ligand expressed by human B cells costimulates involving the CD154-CD40 coreceptors (Fig. 8). In addition, since B cell responses. J. Immunol. 154:4996. 2. Lipsky, P. E., J. F. Attrep, A. C. Grammer, M. J. McIlraith, and Y. Nishoka. 1997. ligation of CD40 on B cells induces CD154 expression (Figs. 5, 6), Analysis of CD40-CD40 ligand interactions in the regulation of human B cell homotypic CD154-CD40 interactions between B cells may sustain function. Ann. N. Y. Acad. Sci. 815:372. CD154 expression on B cells in an autocrine or paracrine manner 3. Pinchuk, L. M., S. J. Klaus, D. M. Magaletti, G. V. Pinchuk, J. P. Norsen, and E. A. Clark. 1996. Functional CD40 ligand expressed by human blood dendritic independent of T cells after initial activation. Further evidence that cells is up-regulated by CD40 ligation. J. Immunol. 157:4363. CD154-CD40 interactions between tonsillar B cells propagate GC 4. Wykes, M., J. Poudrier, R. Lindstedt, and D. Gray. 1998. Regulation of cyto- plasmic, surface and soluble forms of CD40 ligand in mouse B cells. Eur. J. Im- reactions is provided by the finding that in vitro differentiation of munol. 28:548. highly purified GC B cells to those with a memory phenotype was 5. Blossom, S., E. B. Chu, W. O. Weigle, and K. M. Gilbert. 1997. CD40 ligand partially blocked by an anti-CD154 mAb (Fig. 7). It should be expressed on B cells in the BXSB mouse model of systemic lupus erythematosus. J. Immunol. 157:4580. noted that the impact of anti-CD154 was documented only by phe- 6. Desai-Mehta A., L. Lu, R. Ramsey-Goldman, and S. K. Datta. 1996. Hyperex- notypic analysis. However, there is currently no other means to pression of CD40 ligand by B and T cells in human lupus and its role in patho- dissect centrocytes from memory cells, and, therefore, it is likely, genic autoantibody production. J. Clin. Invest. 97:2063. 7. Devi, B. S., S. van Noordin, T. Krausz, and K. A. Davies. 1998. Peripheral blood but not definitely proven, that this phenotypic change implies an lymphocytes in SLE: hyperexpression of CD154 on T and B lymphocytes and impact of anti-CD154 on memory cell differentiation. In addition, increased number of double negative T cells. J. Autoimmun. 11:471. 8. Voorzanger, N., R. Touitou, E. Garcia, H. J. Delecluse, F. Rousset, I. Joab, it should be emphasized that the effect of homotypic B cell inter- M. C. Javrot, and J. Y. Blay. 1996. (IL)-10 and IL-6 are produced in Downloaded from actions mediated by CD154-CD40 was only partial, implying other vivo by non-Hodgkin’s lymphoma cells and act as cooperative growth factors. receptor-ligand pairs are likely to be involved in this process. De- Cancer Res. 56:5499. 9. Moses, A. V., S. E. Williams, J. G. Strussenberg, M. L. Heneveld, R. A. Ruhl, spite these caveats, the data are consistent with previous evidence A. C. Bakke, G. C. Bagby, and J. A. Nelson. 1997. HIV-1 induction of CD40 on that CD40 engagement on IgDϪ cells in the absence of purposeful endothelial cells promotes the outgrowth of AIDS-associated B-cell lymphomas. Nat. Med. 3:1242. CD38 ligation has been shown previously to down-regulate CD38 10. Trentin, L., R. Zambello, R. Sancetta, M. Facco, A. Cerruti, A. Perin, M. Siviero, expression (35). These results, therefore, imply that homotypic B U. Basso, M. Bortolin, F. Adami, C. Agostini, and G. Semenzato. 1997. B lym- cell interactions mediated by CD154-CD40 interactions play a phocytes from patients with chronic lymphoproliferative disorders are equipped http://www.jimmunol.org/ ϩϩ Ϫ ϩ Ϫ with different costimulatory molecules. Cancer Res. 57:4940. central role in the maturation of CD38 IgD and CD38 IgD 11. Schattner, E. J., J. Mascarenhas, I. Reyfman, M. Koshy, C. Woo, S. M. Friedman, Ϫ Ϫ GC cells to CD38 IgD memory cells beyond any direct involve- and M. K. Crow. 1998. Chronic lymphocytic leukemia B cells can express CD40 ment of CD38 engagement. This conclusion is consistent with the ligand and demonstrate T-cell type costimulatory capacity. Blood 91:2689. 12. Kawabe, T., T. Naka, K. Yoshida, T. Tanaka, H. Fujiwara, S. Suematsu, previous observation that blocking CD154-CD40 interactions in N. Yoshia, T. Kishimoto, and H. Kikutani. 1994. The immune responses in vivo abolished Ag-stimulated clonal expansion of B cells in GCs CD40-deficient mice: impaired immunoglobulin class switching and germinal center formation. Immunity 1:167. (72, 73) and decreased the development of memory B cells (55). 13. Castigli, E., F. W. Alt, L. Davidson, A. Bottaro, E. Mizoguchi, A. K. Bhan, Furthermore, these findings extend previous reports demonstrating R. S. Geha. 1994. CD40-deficient mice generated by recombination-activating gene-2-deficient blastocyst complementation. Proc. Natl. Acad. Sci. USA that CD40 ligation preferentially induced differentiation of by guest on September 25, 2021 ϩϩ Ϫ ϩ Ϫ 91:12135. CD38 IgD and CD38 IgD GC B cells to memory cells de- 14. Oxenius, A., K. A. Campbell, C. R. Maliszewski, T. Kishimoto, H. Kikutani, fined by phenotype as well as function (74), by suggesting that the H. Hengartner, R. M. Zinkernagel, and M. F. Bachmann. 1996. CD40-CD40 ligand interactions are critical in T-B cooperation but not for other anti-viral source of CD154 in this phenomena may be the tonsillar B cell CD4ϩ T cell functions. J. Exp. Med. 183:2209. itself. 15. Xu, J., T. M. Foy, J. D. Laman, E. A. Elliott, J. J. Dunn, T. J. Waldschmidt, A final issue relates to the role of engagement of sIg during B J. Elsemore, R. J. Noelle, and R. A. Flavell. 1994. Mice deficient for the CD40 ligand. Immunity 1:423. cell maturation events in secondary lymphoid tissues that include 16. Renshaw, B. R., W. C. Fanslow, R. J. Armitage, K. A. Campbell, D. Liggitt, initial activation of naive or recirculating memory B cells in the B. Wright, B. L. Davidson, and C. R. Maliszewski. 1994. Humoral responses in CD40 ligand-deficient mice. J. Exp. Med. 180:1889. interfollicular or lymphoepithelial regions or following challenge 17. Levy, J., T. Espanol-Boren, C. Thomas, A. Fischer, P. Tovo, P. Bordigoni, with a TD or T-independent Ag, selection of B cells in the basal I. Resnick, A. Fasth, M. Baer, L. Gomez, E. A. M. Sanders, M.-D. Tabone, LZ of GCs, and maintenance of the memory B cell subset during D. Plantaz, A. Etzioni, V. Monafo, M. Abinun, L. Hammarstrom, T. Abrahamsen, bright ϩ A. Jones, A. Finn, T. Klemola, E. deVries, O. Sanal, M. C. Peitsch, and a TD response or CD38 IgD plasmablasts during a T-inde- L. D. Notarangelo. 1997. Clinical spectrum of X-linked hyper-IgM syndrome. pendent response (reviewed in Refs. 21–24, 75). The current data J. Pediatr. 131:47. provide a mechanism by which engagement of sIg without T cell 18. Seyama, K., S. Nonoyama, I. Gangsaas, D. Hollenbaugh, H. F. Pabst, A. Aruffo, and H. D. Ochs. 1998. Mutations of the CD40 ligand gene and its effect on CD40 stimulation may influence B cell function in a manner that simu- ligand expression in patients with X-linked hyper IgM syndrome. Blood 92:2421. lates some aspects of TD responses (Figs. 5, 6, and 8). Thus, 19. Ochs, H. D., S. Nonoyama, M. L. Farrington, S. H. Fischer, and A. Aruffo. 1993. The role of adhesion molecules in the regulation of responses. Semin. CD154 expression induced following sIg engagement in tonsillar Hematol. 30:72. B cell subsets may provide T cell equivalent stimulation by en- 20. Facchetti, F., C. Appiani, L. Salvi, J. Levy, and L. D. Notarangelo. 1995. Im- gaging CD40 on bystander B cells. In addition, CD154 expression munohistologic analysis of ineffective CD40-CD40 ligand interaction in lym- phoid tissues from patients with X-linked immunodeficiency with hyper-IgM. by B cells may provide a bidirectional signaling mechanism that J. Immunol. 154:6624. substitutes for the influence of activated T cells. 21. MacLennan, I. C., A. Gulbranson-Judge, K. M. Toellner, M. Casamayor-Palleja, E. Chan, D. M. Sze, S. A. Luther, and H. A. Orbea. 1997. The changing pref- erence of T and B cells for partners as T-dependent antibody responses develop. Immunol. Rev. 156:53. 22. Liu, Y.J., and C. Arpin. 1997. Germinal center development. Immunol. Rev. Acknowledgments 156:111. We thank Rehana Hussain and Christine Pavlovitch for excellent technical 23. Choi, Y.S. 1997. Differentiation and apoptosis of human germinal center B-lym- phocytes. Immunol. Res. 16:161. assistance, Angelia Jones, Mela Wierzchowski and Ricardo Olivarez for 24. Hardie, D. L., G. D. Johnson, M. Khan, and I. C. M. MacLennan. 1993. Quan- assistance with tonsil tissue acquisition, Ellis Lightfoot and Dr. Saba Jamal titative analysis of molecules which distinguish functional compartments within for histological analysis of tonsil specimens, Dr. Kate Luby-Phelps for germinal centers. Eur. J. Immunol. 23:997. advice and assistance with the fluorescence microscopy experiments, Dr. 25. Lederman, S., M. J. Yellin, G. Inghirami, J. J. Lee, D. M. Knowles, and L. Chess. 1992. Molecular interactions mediating T-B lymphocyte collaboration in human Louis Picker for advice regarding flow cytometry, and Dr. Richard Was- lymphoid follicles: roles of T cell-B cell-activating molecule (5c8 ) and serman for providing peripheral blood from an HIgMXL patient. CD40 in contact-dependent help. J. Immunol. 149:3817. The Journal of Immunology 4159

26. Van den Eertwegh, A. J. M., R. J. Noelle, M. Roy, D. M. Shepherd, A. Aruffo, 51. Brenner, B., U. Koppenhoefer, A. Lepple-Wienhues, H. Grassme, C. Muller, J. A. Ledbetter, W. J. A. Boersma, and E. Claassen. 1993. In vivo CD40-gp39 C. P. Speer, F. Lang, and E. Gulbins. 1997. The CD40 ligand directly activates interactions are essential for thymus-dependent humoral immunity. I. In vivo T-lymphocytes via tyrosine phosphorylation dependent PKC activation. Biochem. expression of CD40 ligand, , and antibody production delineates sites of Biophys. Res. Comm. 239:11. cognate T-B interactions. J. Exp. Med. 178:1555. 52. Brenner, B., U. Koppenhoefer, H. Grassme, J. Kun, F. Lang, and E. Gulbins. 27. Casamayor-Palleja, M., M. Khan, and I. C. M. MacLennan. 1995. A subset of 1997. Evidence for a novel function of the CD40 ligand as a signalling molecule CD4ϩ memory T cells contains preformed CD40 ligand that is rapidly but tran- in T-lymphocytes. FEBS Lett. 417:301. siently expressed on their surface after activation through the T cell receptor 53. Koppenhoefer, U., B. Brenner, F. Lang, and E. Gulbins. 1997. The CD40 ligand complex. J. Exp. Med. 181:1293. stimulates T-lymphocytes via the neutral sphingomyelinase: a novel function of 28. Vyth-Dreese, F. A., T. A. M. Dellemijn, D. Majoor, and D. de Jong. 1996. the CD40-ligand as signalling molecule. FEBS Lett. 414:444. Localization in situ of the co-stimulatory molecules .1, B7.2, CD40, and their 54. Van Essen, D., H. Kikutani, and D. Gray. 1995. CD40-ligand transduced co- ligands in normal lymphoid tissue. Eur. J. Immunol. 25:3023. stimulation of T cells in the development of helper function. Nature 378:620. 29. Liu, Y.-J., D. E. Joshua, G. T. Williams, C. A. Smith, J. Gordon, and 55. Gray, D., P. Dullforce, and S. Jainandunsing. 1994. Memory B cell development I. C. M. MacLennan. 1989. Mechanisms of antigen-driven selection in germinal but not germinal center formation is impaired by in vivo blockade of CD40-CD40 centres. Nature 342:929. ligand interaction. J. Exp. Med. 180:141. 30. Liu, Y.-J., D. Y. Mason, G. D. Johnson, S. Abbot, C. D. Gregory, D. L. Hardie, 56. MacDonald, K. P., Y. Nishioka, P. E. Lipsky, and R. Thomas. 1997. Functional J. Gordon, and I. C. M. MacLennan. 1991. Germinal center cells express bcl-2 CD40 ligand is expressed by T cells in rheumatoid arthritis. J. Clin. Invest. 100: protein after activation by signals that prevent their entry into apoptosis. Eur. 2404. J. Immunol. 21:1905. 57. Hodge, M. R., A. M. Ranger, F. C. de la Brousse, T. Hoey, M. J. Grusby, and 31. Han, S., K. Hathcock, B. Zheng, T. Kepler, R. Hodes, and G. Kelsoe. 1995. L. H. Glimcher. 1996. Hyperproliferation and dysregulation of IL4 expression in Cellular interactions in germinal centers: roles of CD40 ligand and B7-2 in es- NF-ATp-deficient mice. Immunity 4:397. tablished germinal centers. J. Immunol. 155:556. 58. Timmerman, L. A., J. I. Healy, S. N. Ho, L. Chen, C. C. Goodnow, and 32. Perry, M., and A. Whyte. 1998. Immunology of the tonsils. Immunol. Today G. R. Crabtree. 1997. Redundant expression but selective utilization of nuclear 19:414. factor of activated T cells family members. J. Immunol. 159:2735. 33. Lin, Q., J. Rohrer, R. C. Allen, M. Larche, J. M. Greene, A. O. Shigeoka, 59. Bischof, F., and A. Melms. 1998. Glucocorticoids inhibit CD40 ligand expression Downloaded from R. A. Gatti, D. C. Derauf, J. W. Belmont, and M. E. Conley. 1996. A single strand of peripheral CD4ϩ lymphocytes. Cell. Immunol. 187:38. conformation polymorphism study of CD40 ligand: efficient mutation analysis 60. Lobo, F. M., R. Zanjani, N. Ho, T. A. Chatila, and R. L. Fuleihan. 1999. Calcium- and carrier detection for X-linked hyper IgM syndrome. J. Clin. Invest. 97:196. dependent activation of TNF family by Ca2ϩ/calmodulin type 34. Jumper, M. D., Y. Nishioka, L. S. Davis, P. E. Lipsky, and K. Meek. 1994. IV/Gr and calcineurin. J. Immunol. 162:2057. Regulation of human B cell function by recombinant CD40 ligand and other 61. Tsytsykova, A. V., E. N. Tsitsikov, and R. S. Geha. 1996. The CD154 promoter TNF-related ligands. J. Immunol. 155:2369. contains nuclear factor of activated T cells-binding motifs which require AP-1 35. Miyashita, T., M. J. McIlraith, A. C. Grammer, Y. Miura, J. F. Attrep, binding for activation of transcription. J. Biol. Chem. 271:3763. Y. Shimaoka, and P. E. Lipsky. 1997. Bidirectional regulation of human B cell 62. Schubert, L. A., G. King, R. Q. Cron, D. B. Lewis, A. Aruffo, and

responses by CD40-CD40 ligand interactions. J. Immunol. 158:4620. http://www.jimmunol.org/ D. Hollenbaugh. 1995. The human gp39 promoter: two distinct nuclear factors of 36. Grammer, A. C., J. L. Swantek, R. D. McFarland, Y. Miura, T. Geppert, and activated T cell protein-binding elements contribute indepedently to transcrip- P. E. Lipsky. 1998. TNF receptor-associated factor-3 signaling mediates activa- tional activation. J. Biol. Chem. 270:29624. tion of p38 and Jun N-terminal kinase, secretion, and Ig production 63. Verweij, C. L., C. Guidos, and G. R. Crabtree. 1990. Cell type specificity and following ligation of CD40 on human B cells. J. Immunol. 161:1183. actuation requirements for NFAT-1 (Nuclear Factor of Activated T-cells) tran- 37. Mattila, P. S., and J. Tarkkanen. 1997. Age-associated changes in the cellular scriptional activity determined by a new method using transgenic mice to assay composition of the human adenoid. Scan. J. Immunol. 45:423. transcriptional activity of an individual nuclear factor. J. Biol. Chem. 265:15788. 38. Splawski, J. B., J. Nishioka, Y. Nishioka, and P. E. Lipsky. 1996. CD40 ligand 64. Yaseen, N. R., A. L. Abby, F. Wang, and S. Sharma. 1993. Comparative analysis is expressed and functional on activated neonatal T cells. J. Immunol. 156:119. of NFAT (Nuclear Factor of Activated T cells) complex in human T and B 39. Peng, X., J. E. Remacle, A. Kasran, D. Huylebroeck, and J. L. Ceuppens. 1998. lymphocytes. J. Biol. Chem. 268:14285. IL-12 up-regulates CD40 ligand (CD154) expression on human T cells. J. Im- munol. 160:1166. 65. Venkataraman, L., D. A. Francis, Z. Wang, J. Liu, T. L. Rothstein, and R. Sen. 1994. Cyclosporin-A sensitive induction of NF-AT in murine B cells. Immunity

40. Hirohata, S. Human Th1 responses driven by IL-12 are associated with enhanced by guest on September 25, 2021 expression of CD40 ligand. 1999. Clin. Exp. Immunol. 115:78. 1:189. 41. Sakata-Kaneko, Y. Wakatsuki, T. Usui, Y. Matsunaga, T. Itoh, E. Nishi, 66. Choi, M. S. K, R. D. Brines, M. J. Holman, and G. G. B. Klaus. 1994. Induction N. Kume, and T. Kita. 1998. Lysophophatidylcholine up-regulated CD40 ligand of NF-AT in normal B lymphocytes by anti-immunoglobulin or CD40 ligand in expression in newly activated human CD4ϩ T cells. FEBS Lett. 433:161. conjunction with IL-4. Immunity 1:179. 42. Roy, M., A. Aruffo, J. Ledbetter, P. Linsley, M. Kehry, and R. Noelle. 1995. 67. Francis, D. A., J. G. Karras, X. Ke, R. Sen and T. Rothstein. 1995. Induction of ␬ Studies on the interdependence of gp39 and B7 expression and function during the transcription factors NF- B, AP-1, and NF-AT during B cell stimulation antigen-specific immune responses. Eur. J. Immunol. 25:596. through the CD40 receptor. Int. Immunol. 7:151. ␬ Ј 43. Lund, F. E., D. A. Cockayne, T. D. Randall, N. Solvason, F. Schuber, and 68. Meyer, K. B., and J. Ireland. 1998. PMA/ionomycin induces Ig 3 enhancer M. C. Howard. 1998. CD38: a new paradigm in lymphocyte activation and signal activity which is in part mediated by a unique NFAT transcription complex. Eur. transduction. Immunol. Rev. 161:79. J. Immunol. 28:1467. 44. Dono, M., V. L. Burgio, C. Tacchetti, A. Favre, A. Augliera, S. Zupo, 69. Gerondakis, S., R. Grumont, I. Rourke, and M. Grossman. 1998. The regulation G. Taborelli, N. Chiorazzi, C. E. Grossi, and M. Ferrarini. 1996. Subepithelial B and roles of Rel/NF-kappa B transcription factors during lymphocyte activation. cells in the human palatine tonsil. I. Morphologic, cytochemical, and phenotypic Curr. Opin. Immunol. 10:353. characterization. Eur. J. Immunol. 26:2035. 70. Lapointe, R., R. Lemieux, M. Olivier, and A. Darveau. 1996. Tyrosine kinase and 45. Billian, G., C. Bella, P. Mondiere, and T. Defrance. 1996. Identification of a cAMP-dependent protein kinase activities in CD40-activated human B lympho- tonsil IgDϩ B cell subset with phenotypical characteristics of germinal center B cytes. Eur. J. Immunol. 26:2376. cells. Eur. J. Immunol. 26:1712. 71. Richards, M. L., and D. H. Katz. 1997. Analysis of the promoter elements nec- 46. Pascual, V., Y. J. Liu, A. Magalski, O. de Bouteiller, J. Banchereau, and essary for IL-4 and anti-CD40 antibody induction of murine Fc⑀RII (CD23): J. D. Capra. Analysis of somatic mutation in five B cell subsets of human tonsil. comparison with the germline epsilon promoter. J. Immunol. 158:263. J. Exp. Med. 180:329. 72. Vora, K. A., K. M. Tumas-Brundage, and T. Manser. 1998. A periarteriolar 47. Feuillard, J., D. Taylor, M. Casamayor-Palleja, G. D. Johnson, and lymphoid sheath-associated B cell focus response is not observed during the I. C. MacLennan. 1995. Isolation and characteristics of tonsil centroblasts with development of the anti-arsenate germinal center reaction. J. Immunol. 160:728. reference to Ig class switching. Int. Immunol. 7:121. 73. Garside, P., E. Ingulli, R. R. Merica, J. G. Johnson, R. J. Noelle, and 48. Cayabyab, M., J. H. Phillips, and L. L. Lanier. 1994. CD40 preferentially co- M. K. Jenkins. 1998. Visualization of specific B and T lymphocyte interactions stimulates activation of CD4ϩ T lymphocytes. J. Immunol. 152:1523. in the lymph node. Science 281:96. 49. Blotta, M. H., J. D. Marshall, R. H. DeKruyff, and D. T. Umetsu. 1996. Cross- 74. Arpin C., J. Dechanet, C. van Kooten, P. Merville, G. Grouard, F. Briere, linking of the CD40 ligand on human CD4ϩ T lymphocytes generates a costimu- J. Banchereau, and Y.-J. Liu. 1995. Generation of memory B cells and plasma latory signal that up-regulates IL-4 synthesis. J. Immunol. 156:3133. cells in vitro. Science 268:720. 50. Macaulay, A. E., R. H. DeKruyff, C. C. Goodnow, and D. T. Umetsu. 1997. 75. Tew, J. G., J. Wu, D. Qin, S. Helm, G. F. Burton, and A. K. Szakal. 1997. Antigen-specific B cells preferentially induce CD4ϩ T cells to produce IL-4. Follicular dendritic cells and presentation of antigen and costimulatory signals to J. Immunol. 158:4171. B cells. Immunol. Rev. 156:39.