sign in sign up
<<
Home , Memory B cell

Human Activation by Autologous NK Cells Is Regulated by CD40-CD40 Ligand Interaction: Role of Memory B Cells and CD5 + B Cells This information is current as of October 2, 2021. Isaac R. Blanca, Earl W. Bere, Howard A. Young and John R. Ortaldo J Immunol 2001; 167:6132-6139; ; doi: 10.4049/jimmunol.167.11.6132 http://www.jimmunol.org/content/167/11/6132 Downloaded from

References This article cites 45 articles, 17 of which you can access for free at: http://www.jimmunol.org/content/167/11/6132.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 by guest on October 2, 2021 *average

Subscription Information about subscribing to The Journal of 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 © 2001 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Human B Cell Activation by Autologous NK Cells Is Regulated by CD40-CD40 Ligand Interaction: Role of Memory B Cells and CD5؉ B Cells

Isaac R. Blanca,*† Earl W. Bere,* Howard A. Young,* and John R. Ortaldo1*

NK cells are a subpopulation of characterized primarily by their cytolytic activity. They are recognized as an important component of the against virus infection and tumors. In addition to their cytolytic activity, NK cells also participate either directly or indirectly in the regulation of the ongoing Ab response. More recently, it has been suggested that NK cells have an important role in the outcome of autoimmune diseases. Here, we demonstrate that human NK cells can induce autologous resting B cells to synthesize Ig, including switching to IgG and IgA, reminiscent of a secondary Ab response. B cell activation by the NK cell is contact-dependent and rapid, suggesting an autocrine B cell-regulated process. This NK cell function is -independent, requires an active cytoplasmic membrane, and is blocked by anti-CD40 ligand (anti-CD154) or CD40-mIg Downloaded from fusion protein, indicating a critical role for CD40-CD40 ligand interaction. Depletion studies also demonstrate that CD5؉ B cells autoreactive B-1 cells) and a heterogeneous population of CD27؉ memory B cells play a critical role in the Ig response induced) by NK cells. The existence of this novel mechanism of B cell activation has important implications in innate , B cell- mediated , and B cell neoplasia. The Journal of Immunology, 2001, 167: 6132–6139.

lassically, NK cells are characterized by their ability to man NK cells can enhance the B cell proliferative responses to the http://www.jimmunol.org/ lyse transformed cells and virally infected cells without surface Ig cross-linking agents anti-IgM or Staphylococcus aureus previous . Phenotypically mature circulating Cowan strain (6). C Ϫ ϩ ϩ dim NK cells express the CD3 CD56 CD16 CD2 surface markers By direct interaction between B cells and NK cells, it has been and are distinguishable from T cells by the lack of rearranged TCR shown that B cells are able to stimulate the production of IFN-␥ by genes. Unlike B cells, NK cells do not express surface Ig (1). NK cells and activation of the NK cells (11). Thus, both B cells In addition to their cytotoxic activity, NK cells participate either and NK cells are capable of interacting in a spontaneous manner directly or indirectly in multiple developmental and regulatory as- leading to a costimulatory effect. pects of the immune system characterized by rapid response to The evidence described above suggests that NK cells are in-

exogenous and endogenous signals by producing a variety of cy- volved in B cell maturation, Ig secretion, and switching, by guest on October 2, 2021 tokines and (2–4). pathways well-known to be regulated by CD40-CD154 interaction The ability of NK cells to produce hemopoietic cell growth fac- (12). The absence, or blockade, of the CD40-CD154 interaction re- tors, IFN, IL, TNF ␣ and ␤, TGF, and other growth factors, com- sults in gross impairment of the B cell physiological and molecular bined with their ability to respond rapidly to exogenous signals by pathways that seem to be unique to the T-dependent Ab response. up-regulating mRNA expression for various within minutes Whether the CD40-CD154 engagement plays a role in B cell demonstrates the importance of NK cells as mediators or effectors of regulation by NK cells remains to be clarified. The expression of the intercellular communication network (reviewed in Ref. 2). mRNA for CD40 ligand (CD40L)2 on purified NK cells has been Under activated conditions, NK cells express and up-regulate reported previously (13); more recently, a role for CD40-CD154 in the receptors for a variety of chemotactic factors, cytokines, NK cell interaction with other cells has been described (14). NK growth factors, and hormones including the expression of the li- cell clones expressing CD154 were able to kill target cells express- gand of the CD40 receptor (CD154) thus enhancing NK cell in- ing CD40 receptors. Although freshly isolated human NK cells teractions within the immune system. For example, it has been were unable to lyse CD40ϩ targets, when activated with rIL-2 they shown that NK cells are involved in the regulation of B lympho- were able to kill the CD40-transfected cells. This study also re- cyte functions (5–8). Donor type-activated NK cells promote mar- ported that coexpression of CD40 and MHC class I Ag on the row engraftment and B cell development during allogeneic bone target cell inhibited the lysis by the NK cells, suggesting a regu- marrow transplantation in mice (9) and in humans (10). Also, hu- latory role of MHC class I in the CD40-triggered killing. Direct interaction between B and NK cells under activated con- ditions are supported by several studies, particularly in mouse *Laboratory of Experimental Immunology, Division of Basic Sciences, National Can- models (15–18). Evidence from human studies suggests that NK cer Institute, Frederick, MD 21702; and †Instituto de Inmunologia, Facultad de Me- dicina, Universidad Central de Venezuela, Caracas, Venezuela cells and B cells can spontaneously interact in vitro, as demon- Received for publication May 11, 2001. Accepted for publication September strated by conjugate formation and activation of the interacting 21, 2001. cells (11). However, this activating function is poorly understood. The costs of publication of this article were defrayed in part by the payment of page NK cells are potent regulatory cells in the innate immune sys- charges. This article must therefore be hereby marked advertisement in accordance tem, characterized by their spontaneous interaction with immune with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Address correspondence and reprint requests to Dr. John R. Ortaldo, Laboratory of Experimental Immunology, Center for Cancer Research, Frederick Cancer Research 2 Abbreviations used in this paper: CD40L, CD40 ligand; RT, room temperature; and Development Center, National Cancer Institute, Building 560, Room 31-93, Fred- BsNK, B cells stimulated with NK cells; BsT, B cells stimulated with T cells; BMT, erick, MD 21702-1201. E-mail address: [email protected] bone marrow transplantation.

Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00 The Journal of Immunology 6133

and nonimmune cells (11, 19). In addition, growing evidence sug- microcentrifuge tubes (Fisher Scientific, Pittsburgh, PA). The cell suspen- gests that NK cells might be involved in the development of au- sion was centrifuged at 200 ϫ g for 10 s in a microcentrifuge (Capsule; toimmune diseases (20). However, the mechanisms by which NK Tomy, Tokyo, Japan) and incubated for 5 min at 37°C in a water bath. Then, cell mixtures were immediately transferred to an ice bath. These cells modulate these responses are not entirely clear. In this study, conditions were previously determined as the minimum time necessary to we tested the hypothesis that spontaneous interaction between hu- activate the B cells with the NK cells as evaluated by tyrosine-phospho- man B cells and autologous NK cells might be important in initi- rylation studies (data not shown). In some experiments, the cell mixture 5 ating the innate B cell response which, under the appropriate con- was only desegregated and the B cell concentration was adjusted to 5 ϫ 10 cells/ml in RPMI 1640 plus 10% FCS (complete medium) and cultured for ditions, might provide an early protective advantage to the host or 6 days at 37°C in an atmosphere of 5% CO2. In other experiments, the contribute to the development of autoimmunity. stimulating T cells or NK cells were removed from the mixture by mag- netic columns (as described above) after labeling at 4°C with biotinylated Materials and Methods anti-CD3 plus streptavidin microbeads or anti-CD56 microbeads, respec- Reagents tively. The remaining B cells stimulated with T cells (BsT cells) or with NK cells (BsNK cells) contained Ͻ2% of T cells or NK cells as determined RPMI 1640 culture medium and Dulbecco’s PBS, L-glutamine, and peni- by flow cytometry using double-labeling with anti-CD5 FITC/anti-CD19 cillin-streptomycin were purchased from BioWhittaker (Walkersville, PE and anti-CD19 FITC/CD16 PE, respectively. Both BsT cells and BsNK MD); FBSЈ was purchased from Biofluids (Rockville, MD). cells were adjusted to 5 ϫ 105 cells/ml and cultured for 6 days in the same separation medium and BSA were obtained from ICN Pharmaceuticals conditions as above. Each culture was performed in triplicate in 1 ml of (Aurora, OH). Percoll was obtained from Amersham Pharmacia Biotech medium using 12 ϫ 75 round-bottom sterile culture tubes (BD Bio- (Upsala, Sweden); polyoxyethylene-sorbitan monolaurate (Tween 20) was sciences). Finally, the cell cultures were centrifuged at 1200 ϫ g for 10 min obtained from Sigma-Aldrich (St. Louis, MO). ELISA kits for human and the cell-free supernatants were harvested, filtered through a 0.22 ␮m TNF␣, IL-6, IL-4, IFN-␥, and IL-10 quantification were obtained from Millipore filter (Bedford, MA), and tested for secreted Ig. R&D Systems (Minneapolis, MN). Downloaded from Monoclonal and polyclonal Abs Fixation of NK cells Anti-CD3 FITC, anti-CD19 PE, anti-CD14 FITC, anti-CD56 PE, anti-CD5 In some experiments, the NK cells were fixed for 5 min with 1% glutar- FITC, anti-CD40 PE, anti-CD40L (anti-CD154) PE, anti-CD27 FITC, anti- aldehyde, washed three times with PBS, rested for1hatroom temperature IgM FITC, anti-IgG, IgD, and anti-CD27 FITC mAbs were purchased from (RT) to achieve polymerization of the fixative, and suspended in RPMI BD Biosciences (San Jose, CA). Anti-CD56 microbeads (CD56 beads) and 1640 plus 10% FCS before use for B cell stimulation.

streptavidin microbeads were obtained from Miltenyi Biotec (Oslo, Swe- http://www.jimmunol.org/ den). CD40-muIg fusion protein was obtained from Ancell (Bayport, MN). Ig analysis Lymphocyte preparation Quantitation of IgM, IgG, and IgA in cell-free supernatants was performed PBMC were isolated from the buffy coats of healthy donors (obtained from by an ELISA specific for human IgM, IgG, and IgA (Bethyl Laboratories, the National Institutes of Health bank) after centrifugation on a lym- Montgomery, TX). Nun Maxisorp C bottom-well plates (Nunc, Naperville, ␮ phocyte separation medium. Cells were washed twice with Dulbecco’s IL) were coated with isotype-specific capture Abs at 1 g/well in 0.1 ml of PBS and suspended in RPMI 1640 medium supplemented with 2 mM of 0.5 M sodium carbonate (pH 9.6) for1hatRT.Theplates were washed L-glutamine, 100 IU/ml penicillin, 50 ␮g/ml streptomycin, and 10% FCS. twice with a wash solution containing 50 mM of Tris (pH 8.0), 0.1 M of Adherent cells were removed by incubation in plastic flasks for1hat37°C NaCl, and 0.05% Tween 20 and incubated for 30 min at RT with a postcoat solution (1% BSA in 50 mM of Tris (pH 8.0), 0.15 M of NaCl) to block

and the nonadherent cells were recovered by being gently washed with by guest on October 2, 2021 warmed medium and incubated on nylon wool columns for 1 h at 37°C. nonspecific binding. Subsequently, cell-free supernatants and standards The nylon nonadherent cells (mostly T cells and NK cells) were eluted with (reference serum or calibrator) were added in duplicate to the plate wells ␮ prewarmed RPMI 1640 medium and the nylon adherent cells (enriched B (100 l/well). Dilution of the standards was made in a blocking solution to cell fraction) were recovered by washing and soaking the nylon wool with avoid binding of serum components to the wells. All plates were incubated cold PBS plus 1 mM of EDTA. for1hatRT,washed three times with wash solution, and incubated1hat RT with 100 ␮l/well optimal concentration of isotype-specific (anti-human Cell fractionation by Percoll gradient IgM, IgG, and IgA) Abs conjugated with HRP. After washing three times, the enzyme substrate hydrogen peroxide plus 3,3Ј,5,5Ј-tetramethylbenzi- Both nylon adherent (B enriched fraction) and nonadherent cells (T cells dine was added for 20 min at RT in the dark and the reaction was stopped plus NK cells) were fractionated separately on a seven-step Percoll gradi- with 50 ␮lof2MofHSO . Plates were read in a MRX microplate reader ent as previously described (21). High-density fractions (F4 to F6) from the 2 4 (Dynatech Laboratories, Chantilly, VA) and isotype concentrations were nylon adherent cell gradients containing 60–80% resting B cells were used extrapolated from a reference curve (range: 500–7.5 ng/ml of the corre- for further purification of B cells and the low density fraction 2 from the sponding isotype). A revelation program incorporated with the microplate nonadherent cell gradients (40–60% NK cells) was used to purify NK ϩ reader (revelation program, Dynatech) calculated the Ig concentration. cells. Fraction 5 from the NK cell gradient (96 Ϯ 2% CD3 T cells) was used as the source of T cells. Purification of B cells and NK cells by magnetic columns Results B cells and NK cells obtained by Percoll gradient centrifugation were fur- NK cell activation of autologous B cells ther depleted from the remaining T lymphocytes and monocytes by nega- Previous reports (11, 22) have hypothesized that contact interac- tive selection with anti-CD3 and anti-CD14 mAbs. The cells were labeled tion between B cells and NK cells induces two-way activation for 30 min on ice with biotinylated anti-CD3 and anti-CD14. After remov- ing the unbound Abs by washing with cold PBS plus 1% BSA, the cells signals. These may be important in the regulation of B cell func- were incubated for 15 min with streptavidin microbeads (Miltenyi Biotec) tion by the NK cells, especially if occurring in the absence of and the positive cells (CD3ϩ and CD14ϩ) were removed with a magnetic exogenous stimuli. To test this hypothesis, we evaluated Ig syn- column (MACS; Miltenyi Biotec). Final purity was 98 Ϯ 1% CD19-pos- Ϯ ϩ Ϫ thesis by highly purified human resting B cells cocultured with itive cells for B cells and 97 2% CD56 /CD5 cells for NK cells as ϫ 5 determined by flow cytometry analysis (FacsSort, BD Biosciences) with autologous NK cells. A fixed number of 5 10 B cells were anti-CD19 and anti-CD56/anti-CD5, respectively. cocultured for the indicated days at 37°C with NK cells, using Phenotypically, the B cell population was composed of 85 Ϯ 3% different B-NK cell ratios and the B cell function was monitored by ϩ ϩ IgM IgD ;2Ϯ 0.5% IgM alone, 3 Ϯ 0.6% IgD alone, 6 Ϯ 2% Ig production in cell-free supernatants as described in Materials IgMϪIgDϪIgGϩ;13Ϯ 3% CD5ϩ (B1), 21 Ϯ 3% IgMϩIgDϩ, CD27ϩ, ϩ and Methods. As shown in Fig. 1, cocultures of B cells with NK cells and Ͻ3 Ϯ 2% IgA cells as determined by flow cytometry. at 1:1 and 1:2 B-NK ratios induced consistent activation and differ- In vitro assay for B cell activation with NK cells entiation of the B cells into Ab-producing cells. The peak for each Ig Highly purified B lymphocytes (1 ϫ 106 cells/ml) were mixed at different isotype varied from a 1:1 B-NK cell ratio for IgG to a 1:4 B-NK cell ratios with autologous NK cells or T lymphocytes in 1.5-ml polystyrene ratio for IgA. In general, B cell and NK cell cocultures at B-NK ratios 6134 B-LYMPHOCYTE ACTIVATION BY NK CELLS

The time curve for Ig synthesis induced by NK cells (Fig. 2) follows kinetics similar to that reported for T cell-dependent hu- man B cell differentiation into IgM-, IgG-, and IgA-producing cells (23). This Ig response reaches its maximum between 7 and 8 days with a plateau between 8 and 10 days of culture. Cocultures for 12 days or more showed a decrease in the accumulated Ig levels in comparison with those with 8–10 days of incubation, due probably to an increase in the protease activity released by dead cells or generation of some other factors interfering with the ELISA co- cultures of B cells with unstimulated T cells that failed to activate the B cells (data not shown). This stimulatory activity of the NK cells on the B cells was selective and specific because T cells failed to induce the B cell activation. B cell activation by NK cells is contact-dependent To evaluate whether activation of B cells by NK cells was contact- dependent or contact-independent, we tested the activation of B cells by NK cells using transwell plates containing two chambers separated by a semipermeable membrane (polycarbonate, 0.4-um Downloaded from pore membrane). This membrane permits the circulation of soluble FIGURE 1. Ig synthesis by B cells cocultured with NK cells. Highly purified B cells (5 ϫ 105 B cells/0.5 ml) were mixed with autologous NK factors between both chambers but prevents B cell and NK cell ϫ 5 cells at the indicated B-NK cell ratio in 1 ml total volume of RPMI 1640 contact. A fixed number of 2.5 10 B cells were seeded in the 5 medium containing 10% FCS and cocultured for 6 days at 37°C. Ig syn- lower chamber and cocultured with 5 ϫ 10 NK cells seeded in the thesis was measured in cell-free supernatants using a human isotype-spe- upper chamber (1:2, B-NK cell ratio) for 7 days at 37°C in RPMI cific ELISA test. Data represent mean Ϯ SD of triplicate determinations. 1640 medium plus 10% FCS. As shown in Fig. 3, NK cells failed Data represent one of two experiments with similar results. to activate the B cells when cocultured in the transwell plate in http://www.jimmunol.org/ comparison to those cocultured in the same plate at an identical B-NK cell ratio. These results show that activation of resting B higher than 1:4 were not effective in inducing the activation and dif- cells by unstimulated NK cells is a cell contact-dependent process ferentiation of the B cells, suggesting that this NK cell function is and, in contrast to -activated B cells which are induced to regulated for a restricted range of interaction with the B cells. Ab production by NK soluble factor(s) (7), resting B cells are not To place this response in the context of a mitogen-stimulated B inducible by NK cell soluble factors alone. cell response, we have also evaluated the activation of the B cells by NK cells in combination with PWM or IL-2 (Table I). The data B cells primed for 5 min with NK cells progress into Ig- by guest on October 2, 2021 demonstrate that the response of the B cells to stimulation with producing cells NKϩ PWM is over 40-fold for IgM and IgG in both unfractionated To evaluate whether the B cell activation required the continuous B cells and high-density B cells. However, when we compare the presence of the NK cells to become activated, we designed exper- Ig synthesis induced by NK cells alone with respect to B cells plus iments of 5-min interaction between B cells and NK cells at 37°C medium, PWM, and IL-2 alone, the NK-induced Ig is over 10-fold as described in Materials and Methods. B cells were mixed at a 1:1 higher. Although the isotype distribution is similar between un- ratio with autologous NK cells or T cells inducing a close cell-cell fractionated B cells and high-density B cells, the background level contact for 5 min at 37°C. Then, the mixture was desegregated by of Ig showed a higher variability in comparison to the high-density vortexing in 1 mM of EDTA and suspended in RPMI 1640 plus fraction. The NK-induced Ig response by unstimulated NK cells is 10% FCS or T cells and NK cells were depleted by magnetic similar to the B cell response reported by Cocks et al. (13) that was columns using anti-CD3 and anti-CD56 microbeads, respectively. achieved by stimulation of purified B cells with CD40L-trans- The number of B cells in each experimental condition (BsNK, fected cells. Thus, in contrast with Gray and Horwitz (22) who BsT) and those cocultured together (B ϩ NK) and (B ϩ T), was study the amplification by IL-2 of the NK-induced Ig synthesis, we adjusted to 5 ϫ 105 B cells/ml and cultured for 6 days at 37°Cin have analyzed in greater detail this consistent B cell activation RPMI 1640 medium containing 10% FCS. Cell-free supernatants induced by unstimulated autologous NK cells to understand its were tested for Ig production. As shown in Fig. 4, the B cells possible role in the immune response. cocultured with NK cells (B ϩ NK mixed) and B cells stimulated

Table I. NK-induced Ig synthesis by autologous B cellsa

Unfractionated B Cells High-Density Resting B Cells

Stimulus IgM IgG IgA IgM IgG IgA

Medium 9.6 3.2 1.5 6.5 1 1 PWM 14.5 5.4 2.3 8.9 3.2 2.2 IL-2 10.2 8.1 5.1 9.6 3.2 2.2 NK 223.0 155.0 22.9 185.5 120.0 30.0 NK ϩ PWM 1109.0 2430.0 657.0 1105.0 1234.0 160.0 NK ϩ IL-2 439.0 677.0 367.0 355.0 150.0 45.0

a Highly purified B cells (5 ϫ 105/ml) were stimulated with PWM (dilution 1/200), IL-2 (200 U/ml), NK cells (5 ϫ 105/ml), or the indicated combination for 6 days at 37°C. Ig synthesis was measured by ELISA and expressed in nanograms per milliliter. The Journal of Immunology 6135

FIGURE 4. Activation of B cells by short-term interaction with NK cells. B cells were mixed at a 1:1 ratio with NK cells or T cells, pelleted by centrifugation, and incubated for 5 min at 37°C. The cell mixtures were FIGURE 2. Time course of Ig synthesis by B cells cocultured with NK kept together (B ϩ NK mixed) or depleted of NK cells (BsNK) or T cells 5 cells. B cells (5 ϫ 10 B cells/0.5 ml) were mixed at a 1:1 ratio with NK (BsT) with anti-CD56 and anti-CD3 microbeads, respectively. The con- Downloaded from cells in 1 ml of RPMI 1640 medium with 10% FCS and cultured for the centration of B cells in each sample was adjusted to 5 ϫ 105 B cells/ml indicated time at 37°C. Ig synthesis was measured at each time point in the before culturing for 6 days at 37°C. Production of Ig was determined in the cell-free supernatants. Results represent the median Ϯ SD of net accumu- cell-free supernatant as described in Materials and Methods. Results rep- lation of Ig synthesis (stacked curve) of triplicate determinations. Data resent the median Ϯ SD of three different assays. represent one of three experiments with similar results.

stimulation with the NK cells. IL-10 and IL-4 were not detected in http://www.jimmunol.org/ for only 5 min with NK cells (BsNK) underwent activation and these system (data not shown). differentiation into Ig-producing cells in a similar way. This data Fixation of NK cells with glutaraldehyde abolished their indicates that a 5-min interaction with NK cells is sufficient to stimulatory capacity induce B cell differentiation into Ab-producing cells. We also eval- uated the synthesis of IL-4, IL-10, IL-6, TNF␣, and IFN-␥ in these The contact-dependent cognate recognition involved in B cell ac- cultures. As shown in Fig. 5, both B ϩ NK and BsNK cocultures tivation suggested that it might require an active membrane func- produced TNF␣ and IL-6, while IFN-␥ was present only in B ϩ tion. Cognate interaction is an active process that requires cell membrane polarization, a phenomenon involved in many other

NK cocultures, suggesting that this factor (produced by the NK by guest on October 2, 2021 cells) was not critical for the differentiation of the B cells after processes, such as cell differentiation, induction of immune re- sponse, and target cell recognition and killing (24, 25). Cell po- larization is required for the conjugate formation between NK cells and their target cells (26, 27). To further evaluate whether the activation of B cells by NK cells requires an intact NK cell membrane mobility, we performed ex- periments in which the NK cell membrane was fixed with 1% glutaraldehyde before testing its capacity to stimulate B cells. Glu- taraldehyde-fixed NK cells failed to activate the B cells after a

FIGURE 3. Ig response of B cells cocultured with NK cells in transwell FIGURE 5. Production of TNF␣, IL-6, and IFN-␥ by B cells stimulated plate. B cells (2.5 ϫ 105) were seeded in the lower chamber of a transwell with NK cells. Cell-free supernatants from B cells, cocultured for 6 days plate and cocultured for 7 days with an equal number of autologous NK with NK cells at a 1:1 ratio (B ϩ NK) or stimulated for 5 min with NK cells cells seeded in the upper chamber in a total volume of 1 ml of RPMI 1640 at a 1:1 ratio (BsNK) and cultured for 6 days in RPMI 1640 plus 10% FCS, medium plus 10% FCS (B ϩ NK transwell). Mixed B and NK cells at the were evaluated using ELISA kits for human TNF␣, IL-6, and IFN-␥. BsNK same B-NK cell ratio were seeded in the same plate and cocultured for the cells contained Ͻ2% of NK cells as tested by double-labeling with anti- same period (B ϩ NK mixed). Results represent the median Ϯ SD of two CD16 FITC/CD56 PE. Data represent the mean Ϯ SD of two independent independent experiments. experiments. 6136 B-LYMPHOCYTE ACTIVATION BY NK CELLS

FIGURE 6. Glutaraldehyde-fixed NK cells lose their capacity for B cell stimulation. B cells were mixed at a 1:1 ratio with untreated NK cells or T cells or NK cells fixed for 5 min with 1% glutaraldehyde before use. Cell mixtures were pelleted by centrifugation and incubated for 5 min at 37°C. The stimulating NK cells or T cells were then removed from the cell mix- ture as described in Materials and Methods. The B cells stimulated with untreated NK cells (BsNK), fixed NK cells (BsNKf), or T cells (BsT) were adjusted to 5 ϫ 105 B cells/ml before culturing for 6 days at 37°C. BsNKf Downloaded from were also cocultured in the presence of fresh NK cells (1:1 ratio) alone or stimulated with a 1/200 dilution of PWM. Production of Ig was determined in cell-free supernatants as described in Materials and Methods. Results represent the median Ϯ SD of fold increase Ig synthesis of two indepen- dent assays. http://www.jimmunol.org/

5-min interaction (Fig. 6) or when cocultured for 6 days. These results indicate that this contact-dependent activation phenomenon is also dependent upon an active cellular membrane. Longer in- teraction (6 days coculture) failed to activate the B cells (data not shown). The possibility that the fixed NK cells can be deleterious for the B cell activity is unlikely because the B cells were still able FIGURE 7. Activation of B cells by NK cells is inhibited by anti- to respond to fresh NK cells when cocultured for 6 days or when CD154. A fixed number (5 ϫ 105) of B cells were mixed with an equal cocultured with fresh NK cells plus PWM (IgM ϭ 626 ng/ml; number of NK cells (1:1 B-NK cell ratio) in the presence of the indicated by guest on October 2, 2021 IgG ϭ 130 ng/ml; and IgA ϭ 36 ng/ml). concentrations of anti-CD154 or IgG1 of isotype-matched control. The cells were cultured for 6 days at 37°C in RPMI 1640 medium plus 10% B cell activation by NK cells is inhibited by anti-CD154 and FCS. Cell-free supernatants were evaluated for Ig production. Data is rep- CD40-mIg fusion protein resentative of two independent experiments. Numerous studies have dealt with the molecular mechanisms reg- ulating B cell and T cell interactions, suggesting that the ligand pair CD40/CD154 plays an important role (12). To determine as a stimulus for the B cells at 1:1 B-NK cell ratio for 5 min. As whether the capacity of the NK cells to induce B cell activation depicted in Fig. 8B, pretreatment of NK cells with anti-CD154 was dependent on the CD40-CD154 interaction, we first cocul- abolished their capacity to activate the B cells. tured both B and NK cells in the presence of different doses of ϩ ϩ anti-CD154 mAb (range: 0.1–2 ␮g/ml). As a control for nonspe- Depletion of CD27 B cells and CD5 B cell subpopulations cific binding via FcR, we included an identical concentration of modifies the B cell response to NK cell activation irrelevant mouse IgG1 that matched the anti-CD154 isotype. As As described above, the Ig response in our system is reminiscent shown in Fig. 7, anti-CD154 inhibited the activation of the B cells of a secondary response in which the IgG was the main isotype in in a dose-dependent manner as demonstrated by the inhibition of most of the donors. In addition, phenotypic analysis of our resting IgM, IgG, and IgA synthesis. B cell population obtained from the high-density Percoll gradient As a second approach, we evaluated the blocking activity of revealed a population of 21 Ϯ 6% CD27ϩ cells and 13 Ϯ 3% CD154 and the CD40-mIg fusion protein during a short-term in- CD5ϩ (B1) cells. From these, only 2.5 Ϯ 1% were CD27ϩ/CD5ϩ. teraction period. B and NK cells were cocultured for 5 min at a 1:1 To determine whether CD27ϩ B cells and CD5ϩ B cell sub- B-NK cell ratio in the presence of 1 ␮g/ml anti-CD154, 1 ␮g/ml populations were participating in the B cell response induced by irrelevant IgG1 isotype, 1 ␮g/ml anti-CD56, or 100 ng/ml CD40- NK cells, we proceeded to deplete these B cell subsets from the mIg. The B cells were separated from the NK cells as described in tested B cell populations before their interaction with the NK cells. Materials and Methods and cultured in the absence of anti-CD154 This approach was used due to the difficulty in isolating these B or CD40 mIg. As shown in Fig. 8A, the presence of anti-CD154 or cell subsets by negative selection, avoiding the presence of Ab on CD40-mIg during the B-NK interaction blocked the activation of the responder B cells. Purified B cells were labeled with biotin- the B cells, suggesting a critical role of the CD40-CD40L(CD154) conjugated anti-CD27ϩ, anti-CD5ϩ, or anti-CD27ϩ anti-CD5 interaction in the activation of the B cells by NK cells. mAbs and the positive cells were removed by magnetic columns As further proof of the role of the CD40-CD154 interaction, we using streptavidin microbeads as described in Materials and Meth- performed experiments involving pretreatment (1 h at 4°C) of the ods. The unfractionated B cell population (total), CD27Ϫ B cells, NK cells with 1 ␮g/ml anti-CD154 and anti-CD56 before their use CD5Ϫ B cells, and the double-negative (CD27ϪCD5Ϫ) B cells The Journal of Immunology 6137

FIGURE 9. Effect of depletion of CD27ϩ B cells and CD5ϩ B cell subsets in the Ig response induced by NK cell stimulation. Total B cells were depleted of CD27ϩ B cells and CD5ϩ B cells with biotin-conjugated anti-CD27 and anti-CD5 mAb, respectively, followed by streptavidin mi- Downloaded from crobeads and selection through magnetic columns. B cells, negatively se- lected as CD27Ϫ, CD5Ϫ, or CD27ϪCD5Ϫ, were cocultured at 5 ϫ 105 cells/ml with an equal number of NK cells for 6 days at 37°C. Cocultures of total B cells with NK cells at the same B-NK cell ratio were included as controls. Ig isotypes were determined as above. Results are expressed as the mean Ϯ SD of three independent experiments, except for the

CD27CD5 double-negative experiment that was performed twice. Confi- http://www.jimmunol.org/ dence levels in comparison with the total B cells obtained by Student’s t .p Ͻ 0.001 ,ءء ;p Ͻ 0.05 ,ء test for paired samples are FIGURE 8. Short-term activation of B cells by NK cells is inhibited by anti-CD154 or CD40L-mIg fusion protein. A, B cells were stimulated for 5 min with NK cells at a 1:1 ratio, in the presence of IgG1 (1 ␮g/ml), Discussion anti-CD56 (1 ␮g/ml), anti-CD154 (1 ␮g/ml), or CD40-mIg (100 ng/ml). B, The present study has investigated the interaction between human NK cells were pretreated with anti-CD154 (1 ␮g/ml) or anti-CD156(1 ␮g/ B lymphocytes and NK cells. Using highly purified human B cells ml)for1hat4°C before use as a stimulus for the B cells. After removal of the and autologous NK cells, we have shown that unstimulated NK stimulating cells, the stimulated B, containing Ͻ2% of NK, cells were adjusted

cells activate resting B cells when cocultured for 6 days at ratios of by guest on October 2, 2021 ϫ 5 to 5 10 cells/ml and cultured for 6 days at 37°C in RPMI 1640 medium plus B-NK cells Ն1:1. We also demonstrate that 5 min of close contact 10% FCS. Cell-free supernatants were evaluated for Ig production. Data rep- at 37°C between both cells, followed by removal of the NK cells, resent two of three independent experiments with similar results. was sufficient to induce the differentiation of the B cells into Ig- producing cells. This activating function is contact-dependent and requires an active NK cell membrane because coculturing both were cocultured at 5 ϫ 105 cells/ml with autologous NK cells at a cells in transwell plates or fixing the NK cell with glutaraldehyde 1:2 B-NK cell ratio for 7 days at 37°C. Cell-free supernatants were failed to activate the B cells. Indeed, cognate interaction as a result tested for Ig synthesis. of cell-cell conjugation has been shown to be an important feature As shown in Fig. 9, depletion of CD27ϩ B cells abrogated IgM of the interaction of B cells and NK cells (11), a process that synthesis induced by the NK stimulation and reduced the levels of requires the redistribution of adhesion molecules and cell mem- IgG and IgA. A similar effect was also observed when the CD5ϩ brane polarization as demonstrated in conjugate formation of NK B cell subset was removed from the responder B cells. B cells cells with target cells (27). It has been shown that many membrane depleted in both B cell subsets (CD27ϩ and CD5ϩ) did not re- proteins are freely mobile and thus present an array of independent spond to activation with autologous NK cells, suggesting a critical binding sites (30). Thus, very low affinity interactions can effec- role of these B cell subpopulations in the B cell response induced tively generate organized areas of synapses. Moreover, the capac- by NK cells. Although depletion of CD27ϩ includes switched cells ity to activate resting B cells, in the absence of exogenous stimuli, (IgGϩ B cells), the depletion of IgGϩ memory B cells alone (per- is specific for NK cells, because autologous resting T cells failed formed in separated experiments, data not shown) did not induce to induce B cell activation. It is well known that helper T cells may significant changes in the IgG synthesis, suggesting that other sub- activate and cooperate with B cells only after their activation using sets of memory B cells may be responsible for IgG switching by Ag receptors or by Ag-unlinked (cognate) interaction (31). Thus, activation with the NK cells. Recently, studies have shown that the lack of B cell activation by unstimulated T cells was expected. human IgMϩIgDϩCD27ϩ B cells switch to all IgG subclasses Although B cell and NK cell interaction has been supported by (28). These B cell subsets in our B cell population accounted for several studies, particularly in rodent models, the experimental de- 18 Ϯ 5% of the total B cells, while the IgGϩ subset was 6 Ϯ 2% signs required the preactivation of the NK cells (22) or the B cells of the total B cells. Although the IgMϩIgDϩCD27ϩ B cell subset (17). In human studies, the evidence for direct interaction of unstimu- represents nonactivated resting B cells, it has been shown to ex- lated NK cells with B cells is more limited. In a previous report, Wyatt press higher levels of Ig mRNA than naive (IgMϩIgDϩ) B cells and Dawson (11) showed that human B cells are able to con- (29). Thus, our results suggest that in addition to IgGϩ B celIs, jugate to NK cells and to induce their activation and IFN-␥ produc- IgMϩ IgDϩCD27ϩ B cells may be responsible for the switch to an tion, which was suggested as a possible mediator of the Ig synthesis IgG isotype in our experimental system. inhibition reported for NK cells (5, 32, 33). However, in this study, the 6138 B-LYMPHOCYTE ACTIVATION BY NK CELLS ability of B cells to function as Ab-producing cells was not reported. source of Ab production in young individuals and might be im- In our system, the production of IFN-␥ was detected only in B cells portant for T cell-independent Ab response. As they tend to pro- cocultured with the NK cells (B ϩ NK), but not in the supernatant of duce autoreactive Igs, they have been linked to autoimmune dis- B cells stimulated for 5 min with NK cells. This result suggests that eases. Remarkably, the malignant cells in nearly all cases of IFN-␥ may not play a critical role. human chronic lymphocytic leukemia carry the CD5 marker, sug- Although in our system the B cell activation is contact depen- gesting that this malignancy arises from the CD5ϩ B cell subset. dent, as demonstrated by transwell plate cultures, we cannot ex- Both, the CD27ϩ B cells and CD5ϩ B cell subpopulation belong clude the possibility that during the interaction (11), the NK cells to the resting B cell pool in human peripheral blood and have been produce soluble factors in the intercellular space capable of driving implicated in the generation of natural Abs and participation in the the B cells into Ab-producing cells. The synthesis of a late-acting pathophysiology of certain autoimmune diseases (20). soluble factor by the NK cell, which has an enhancing effect on Ig This population, once activated by the NK cells, is capable of synthesis by B cells, has been hypothesized (34). It is possible that driving its own differentiation by autocrine synthesis of cytokines. B-NK cell interaction may reduce the B cell threshold necessary to To test this hypothesis, we performed depletion studies of CD27ϩ respond to activating signals (35). Indeed, it has been shown that memory B cells and CD5 ϩ B-1 autoreactive B cell subsets, which, activated B cells are able to synthesize and express CD40L13 and in our B cell populations, represented 21 Ϯ 5% and 15 Ϯ 3% of to produce cytokines with B cell costimulatory activity. Among the total population, respectively. Removal of CD27ϩ memory B these cytokines, IL-6, IL-10, and TNF␣ have been shown to func- cells or CD5ϩ cells significantly reduced the Ig response, while tion as autocrine loops (36–38). Although in our system IL-10 was depletion of both B cell subsets completely abolished the Ig re- not detected in the cultures supernatant of BsNK cells, IL-6 and sponse induced by NK cell activation. This data suggests that di- Downloaded from TNF␣ were always present. rect interaction of the NK cells with these two B cell subsets may The role of CD40-CD40L (CD154) interaction on B cell acti- play an important role in the memory Ab response and B cell- vation and differentiation has been well established for T-depen- mediated autoimmunity. The substantial reduction of IgM synthe- dent Ab responses. However, the role of this interaction in the sis almost to background levels caused by the removal of CD27ϩ T-independent B cell response remains to be elucidated. In our B cells or CD5ϩ B cells suggests that the IgM are CD27ϩ/CD5ϩ. system, B cell activation by the NK cell was inhibited by the presence Although 2.5% of the B cells were CD27ϩ/CD5ϩ, further studies of either anti-CD154 mAb or CD40-mIg fusion protein during the are needed to elucidate this point. It is possible that after activation http://www.jimmunol.org/ B-NK cell interaction or by pretreatment of the NK cells with anti- with the NK cells both B cell subsets require homotypic interaction CD154 before their interaction with the B cells. This data suggests to differentiate into IgM-producing cells. that the CD40-CD154 interaction plays an important role in this T As depletion of the IgGϩ B cell subset does not significantly cell-independent NK-mediated activation of B lymphocytes. The ex- reduce the IgG synthesis, we postulate that another memory B cell pression of CD154 on NK cells has been well-documented. Purified subset with switching capacity to IgG may be involved. Recent NK cells express mRNA for CD40L (13) and CD40L surface expres- studies (29, 42) have shown that the produces sion and its regulatory role on NK cell cytotoxicity have been dem- three different types of V(D)J mutated B cells in similar proportion Ϫ ϩ ϩ ϩ

onstrated in IL-2-activated NK cells and NK cell lines (39). Thus, including the Ig-switched, IgD IgM (IgM only), and IgD IgM by guest on October 2, 2021 very low levels of CD154 may be expressed constitutively on the NK cells which together represent the CD27ϩ compartment of rested cells or during the interaction with the B cells. recirculating memory B cells. In human studies, mutated Ig se- The possibility that activated B cells may express CD40L as sug- quences are found exclusively in this CD27ϩ B cell population and gested by Grammer et al. (40) is of functional importance in our the IgDϩIgMϩCD27ϩ cells have been shown to express a high system due to its possible contribution in homotypic B cell costimu- switching capacity into most Ig isotypes and IgG subclasses (28). lation which, in combination with autocrine B cell factors, might drive In our study, this IgDϩCD27ϩ population represented 18 Ϯ 5% of the B cell into an Ab-producing cell. However, the possibility that the purified B cells. Thus, these cells represent an important can- CD40L acts as initiator of the B cell activation is very unlikely be- didate for direct interaction with the NK cells. cause purified B cells under the same interacting conditions as the Although NK cells provide a survival signal to the B cells as B-NK cell mixture never become activated. This data suggests that determined by the improvement in viability of the B cells cocul- contact-dependent NK cell activating signals are required. tured with NK cells in comparison to B cells cultured alone, this is This direct contact activation of the B cells by NK cells which, not sufficient to achieve activation of the B cells because a contact- in addition to the CD40-CD40L interaction, includes the activation dependent signal is still required. This is demonstrated by B-NK of production by the B cell capable of promoting its differ- cell cocultures in transwell plates in which the survival signals entiation into an Ab-producing cell, is extremely important in auto- provided by the NK cells are still present without activation of the immunity and oncogenesis. In addition, increasing evidence, both in B cells. A similar improvement of B cell survival is observed in B human and mouse models, has suggested an important role of NK cells cocultured with resting T cells without subsequent activation cells in autoimmunity (20, 41). Nevertheless, the link between the NK of the B cells. The addition of a survival signal such as IL-2 to cell and the autoreactive B cell or transformed B cell is still missing. resting B cells in the absence of NK cells prolongs their survival In our study, the Ig response elicited by NK-activated B cells is without accumulation and secretion of Ig. reminiscent of a secondary Ab response expressing most of the Ig The importance of this mechanism of B cell activation in vivo is isotype (IgM, IgG, and IgA). This result prompted us to hypoth- supported by different studies. For example, the presence of NK esize that the responder B cells might belong to a certain subset of cells in the germinal center of the human tonsil in proximity to B memory B cells or autoreactive B cells. One interesting B cell cells suggests a close interaction with the B cells in this lymphoid subgroup is CD27ϩ B cells, which include, in addition to classical organ (44). Secondly, NK cells could play a critical role in the isotype-switched IgGϩ, IgAϩ, and IgE memory B cells, the initiation of an autoimmune response in combination with host IgMϩIgDϩ and IgM-only B cells (42). In contrast, CD5ϩ B cells intrinsic or extrinsic factors. Studies by Shi et al. (20) have shown are long-lived cells that appear to be derived from precursor cells that NK cells can affect the outcome of the adaptive immune re- present in infant, but not in adult, bone marrow and classified as sponse. NK cells, but not NK1.1ϩ T cells, were found to partici- B1 cells (43). As they arise soon after birth, they are the major pate in the development of myasthenia gravis in mice (41). Third, The Journal of Immunology 6139 the contribution of NK cells to B cell activation and differentiation Phenotypic and functional characterization of a panel of cytotoxic murine NK cell in vivo has been suggested from human bone marrow transplan- clones that are heterogeneous in their enhancement of Ig secretion in vitro. Int. Immunol. 10:1093. tation (BMT). In recipients of T cell-depleted allogeneic bone mar- 19. Goodman, D. J., M. Von Albertini, A. Willson, M. T. Millan, and F. H. Bach. row transplants, large granular lymphocytes which contain mostly 1996. Direct activation of porcine endothelial cells by human natural killer cells. Transplantation 61:763. NK cells were found responsible for the secretion of B cell dif- 20. Shi, F. D., H. G. Ljunggren, and N. Sarvetnick. 2001. Innate immunity and ferentiation factors capable of maintaining the Ig expression and autoimmunity: from self-protection to self-destruction. Trends Immunol. 22:97. specific Ab levels in the absence of mature T cells. It appears likely 21. Timonen, T., J. R. Ortaldo, and R. B. Herberman. 1981. Characteristics of human large granular lymphocytes and relationship to natural killer and K cells. J. Exp. that NK cells play a significant role in maintaining B cell function Med. 153:569. in vivo upon T cell-depleted BMT (10, 45). Donor type-activated 22. Gray, J. D., and D. A. Horwitz. 1995. Activated human NK cells can stimulate NK cells have also been shown to promote BMT in mice (9). resting B cells to secrete Ig. J. Immunol. 154:5656. 23. Brinkmann, V., and C. H. Heusser. 1993. T cell-dependent differentiation of The establishment of a link between NK cells, autoreactive B human B cells into IgM, IgG, IgA, or IgE plasma cells: high rate of cells, and circulating memory B cells in our study also suggests a production by IgE plasma cells, but limited clonal expansion of IgE precursors. Cell. Immunol. 152:323. potential regulation by NK cells of the adaptive immune response 24. Hoshino, T., R. T. Winkler-Pickett, A. T. Mason, J. R. Ortaldo, and H. A. Young. through their direct interaction with cells and the production of a 1999. IL-13 production by NK cells: IL-13-producing NK and T cells are present number of immunoregulatory factors. NK cells might provide an in vivo in the absence of IFN-␥. J. Immunol. 162:59. 25. Mahajan, S., J. A. Gollob, J. Ritz, and D. A. Frank. 2001. CD2 stimulation leads advantage to the host before the development of specific and po- to the delayed and prolonged activation of STAT1 in T cells but not NK cells. tent helper T cells. The extension of our findings into animal mod- Exp. Hematol. 29:209. els should allow further testing of these hypotheses. 26. Takahashi, K., S. Miyake, T. Kondo, K. Terao, M. Hatakenata, S. Hashimoto, and T. Yamamura. 2001. Natural killer type-2 bias in remission of multiple sclerosis.

J. Clin. Invest. 107:R23. Downloaded from Acknowledgments 27. Kopp, E. B., and S. Ghosh. 1995. NF-␬B and rel protein in innate immunity. Adv. We thank Dr. Dan McVicar for helpful advice and review of the manu- Immunol. 58:1. script and Anna Mason for performing IL assays. 28. Werner-Favre, C., F. Bovia, P. Schneider, N. Holler, M. Barnet, V. Kindeler, J. Tschopp, and R. H. Zubler. 2001. IgG subclass switching capacity is low in switched and IgM-only, but high in IgDϩIgMϩ, postgerminal center (CD27ϩ) References human B cells. Eur. J. Immunol. 31:243. 1. Herberman, R. B., J. Djeu, H. D. Kay, J. R. Ortaldo, C. Riccardi, G. D. Bonnard, 29. Klein, U., R. Ku¨ppers, and K. Rajewsky. 1997. Evidence for large compartment H. T. Holden, R. Fagnani, A. Santoni, and P. Puccetti. 1979. Natural killer cells: of IgM-expressing memory B cells in humans. Blood 89:1288. characteristics and regulation of activity. Immunol. Rev. 44:43. 30. Dustin, M. L., D. E. Golan, D. Zhu, J. M. Miller, W. Meier, E. A. Davies, and http://www.jimmunol.org/ 2. Degliantoni, G., M. Murphy, M. Kobayashi, M. K. Francis, B. Perussia, and P. A. Van der Merwe. 1997. Low affinity interaction of human or rat T cell G. Trinchieri. 1985. Natural killer (NK) cells-derived hematopoietic colony in- adhesion molecule CD2 with its ligand aligns adhering membranes to achieve hibiting activity and NK cytotoxic factor: relationship with tumor necrosis factor high physiological affinity. J. Biol. Chem. 272:30889. and synergism with immune interferon. J. Exp. Med. 162:1512. 31. Koenig-Marrony, S., P. Soulas, S. Julien, A. M. Knapp, J. C. Garaud, T. Martin, 3. Fehniger, T. A., M. H. Shah, M. J. Turner, J. B. VanDeusen, S. P. Whitman, and J. M. Pascuali. 2001. Natural autoreactive B cells in transgenic mice repro- M. A. Cooper, K. Suzuki, M. Wechser, F. Goodsaid, and M. A. Caligiuri. 1999. duce an apparent paradox to the clonal tolerance theory. J. Immunol. 166:1470. Differential cytokine and gene expression by human NK cells fol- 32. Brieva, J. A., S. Targan, and R. H. Stevens. 1984. NK and T cell subsets regulate lowing activation with IL-18 or IL-15 in combination with IL-12: implication for antibody production by human in vivo -induced lymphoblastoid B cells. the innate immune response. J. Immunol. 162:4511. J. Immunol. 132:611. 4. Maghazachi, A. A. 2000. Chemokines, G proteins and natural killer cells. Arch. 33. Peter, H. H., W. Friedrich, R. Dopfer, W. Muller, C. Kortmann, W. J. Pichler, Immunol. Ther. Exp. (Warsz) 48:65. F. Heinz, and C. H. Rieger. 1983. NK cell function in severe combined immu- 5. Abruzzo, L. V., and D. A. Rowley. 1983. Homeostasis of the antibody response: nodeficiency (SCID): evidence of a common T and NK cell defect in some but by guest on October 2, 2021 immunoregulation by NK cells. Science 222:581. not all SCID patients. J. Immunol. 131:2332. 6. Katz, P., G. Whalen, T. R. Cupps, S. R. Mitchell, and M. Evans. 1989. Natural 34. Kimata, H., E. H. Sherr, and A. Saxon. 1988. Human natural killer (NK) cells killer cells can enhance the proliferative responses of B lymphocytes. Cell. Im- produce a late-acting B-cell differentiation activity. J. Clin. Immunol. 8:381. munol. 120:270. 35. Klaus, G. G. B., M. Holman, C. Johnson-Leger, J. R. Christenson, and M. Kehry. 7. Rodriguez, M. A., I. Blanca, M. L. Baroja, S. Arama, M. Leon-Ponte, I. Abadi, 1999. Interaction of B cells with activated T cells reduces the threshold for CD40- and N. E. Bianco. 1987. Helper activity by human large granular lymphocytes in mediated B cell activation. Int. Immunol. 11:71. in vitro immunoglobulin synthesis. J. Clin. Immunol. 7:356. 36. Burdin, N., C. Van Kooten, L. Galibert, J. Abrams, J. Banchereau, and 8. Kimata, H., F. Shanahan, M. Brogan, S. Targan, and A. Saxon. 1987. Modulation F. Rousset. 1995. Endogenous IL-6 and IL-10 contribute to the differentiation of of ongoing human immunoglobulin synthesis by natural killer cells. Cell. Immu- CD40-activated human B lymphocytes. J. Immunol. 154:2533. nol. 107:74. 37. Iwasaki, T., T. Hamano, J. Fujimoto, A. Ogata, and E. Kakishita. 1998. Regu- 9. Murphy, W. J., M. Bennett, V. Kumar, and D. L. Longo. 1992. Donor-type lation of cytokine expression by an autoreactive B cell clone derived from MRL/ activated natural killer cells promote marrow engraftment and B cell development MP-lpr/lpr mice. Clin. Exp. Immunol. 114:1. during allogeneic bone marrow transplantation. J. Immunol. 148:2953. 38. Rieckman, P., F. D’Alessandro, R. P. Nordan, A. Fauci, and J. Kehrl. 1991. IL-6 10. Brenner, M. K., J. E. Reittie, J. P. Grob, J. Z. Wimperis, S. Stephens, J. Patterson, and tumor necrosis factor-␣: autocrine and paracrine cytokines involved in B cell A. V. Hoffbrand, and H. G. Prentice. 1986. The contribution of large granular function. J. Immunol. 146:3462. lymphocytes to B cell activation and differentiation after T-cell-depleted alloge- 39. Reed, A. J., M. P. Riley, and J. H. Chace. 2000. Virus-induced maturation and neic bone marrow transplantation. Transplantation 42:257. activation of autoreactive memory B cells. J. Exp. Med. 92:1763. 11. Wyatt, R. M., and J. R. Dawson. 1991. Characterization of a subset of human B 40. Grammer, A. C., M. C. Bergman, Y. Miura, K. Fujita, L. S. Davis, and lymphocytes interacting with natural killer cells. J. Immunol. 147:3381. P. E. Lipsky. 1995. The CD40 ligand expressed by human B cells costimulates 12. Grewal, I. S., and R. A. Flavell. 1998. CD40 and CD154 in cell-mediated im- B cell responses. J. Immunol. 154:4996. munity. Annu. Rev. Immunol. 16:111. 41. Shi, F. D., H. Wang, H. Li, H. Seokmann, M. Taniguchi, H. Link, L. V. Kaer, and 13. Cocks, B. G., R. de Waal Malefyt, J. Galizzi, J. E. de Vries, and G. Aversa. 1993. H. G. Ljunggren. 2000. Natural killer cells determine the outcome of B cell- IL-13 induces proliferation and differentiation of human B cells activated by the mediated autoimmunity. Nat. Immun. 1:245. CD40 ligand. Int. Immunol. 5:657. 42. Klein, U., K. Rajewsky, and R. Ku¨ppers. 1998. Human immunoglobulin 14. Carbone, E., G. Ruggiero, G. Terrazano, C. Palomba, S. Fontana, H. Spits, K. (Ig)MϩDϩ peripheral blood B cells expressing the CD27 cell surface antigen Ka¨rre, and S. Zappacosta. 1997. A new mechanism of NK cytotoxicity activation: carry somatically mutated variable region genes: CD27 as a general marker for the CD40-CD40 ligand interaction. J. Exp. Med. 185:2053. somatically mutated (memory) B cells. J. Exp. Med. 188:1679. 15. Yuan, D., J. Wilder, T. Dang, M. Bennett, and V. Kumar. 1992. Activation of B 43. Hayakawa, K., and R. Hardy. 1988. Normal, autoimmune, and malignant CD5ϩ lymphocytes by NK cells. Int. Immunol. 4:1373. B cells: the Ly-1B lineage. Annu. Rev. Immunol. 6:197. 16. Yuan, D., C. Y. Koh, and J. A. Wilder. 1994. Interaction between B lymphocytes 44. Okada, T., T. Nishimura, M. Yagisawa, and K. Naito. 1988. The role of natural and NK cells. FASEB J. 8:1012. killer cells in human tonsillar tissue focusing on the change of tonsillar tissue with 17. Snapper, C. M., H. Yamaguchi, M. A. Moorman, R. Sneed, D. Smoot, and aging. Acta Otolaryngol. Suppl (Stockh) 454:96. J. J. Mond. 1993. NK cells induce activated murine B cells to secrete Ig. J. Im- 45. Brenner, M. K., A. Vyakarnam, J. E. Reittie, J. Z. Wimperis, J. P. Grob, munol. 151:5251. A. V. Hoffbrand, and H. G. Prentice. 1987. Human large granular lymphocytes 18. Vos, Q., J. R. Ortaldo, M. Conan-Cibotti, M. D. Vos, H. A. Young, induce immunoglobulin synthesis after bone marrow transplantation. Eur. J. Im- S. K. Anderson, K. Witherspoon, I. Prager, C. M. Snapper, and J. J. Mond. 1998. munol. 17:47.