Characterization of the Roles of TNF Receptor-Associated Factor 6 in CD40-Mediated B Lymphocyte Effector Functions¹

CD40 is a member of the TNF receptor (TNFR)³ family of surface molecules and is expressed on bone marrow B cells, mature B cells, and certain accessory cells, including bone marrow-derived and follicular dendritic cells. Activated CD4⁺ T cells express CD154, the ligand for CD40 (1, 2). Signaling through CD40 in B cells leads to B cell proliferation (3), Ig and IL-6 secretion (3, 4), isotype switching (5), and up-regulation of surface molecules such as B7-1 (6), LFA-1, ICAM-1 (7), and Fas (8). However, the downstream signaling events following CD40 ligation are not well understood. The cytoplasmic (CY) tail of CD40 does not contain any enzymatic domains, but associates with a family of signal transducer proteins, the TNF receptor-associated factors (TRAFs). TRAFs have no identified enzymatic functions and may serve as adapter molecules. TRAF protein structure is characterized by a conserved carboxyl-terminal TRAF-C domain, a coiled-coil TRAF-N domain, and, with the exception of TRAF1, amino-terminal Zn RING and finger domains. The TRAF-C domain binds to receptor cytoplasmic domains, whereas the TRAF-N domain mediates TRAF homo- and hetero-oligomerization (9, 10). The Zn-binding domains are believed to be essential for downstream signaling on the basis of mostly indirect evidence (9, 11, 12). Of the six TRAFs identified to date, TRAFs 2, 3, 5, and 6 are reported to associate directly with the cytoplasmic tail of CD40 (10, 12). TRAFs 2, 3, and 5 interact with the same region whereas TRAF6 binds to a more membrane proximal region (11). Although the function of TRAF2, which binds most of the known TNFR family of molecules, has received considerable attention, little is yet known about the function of other TRAF molecules. The developmentally early lethality of TRAF deficiencies in genetically altered mice has precluded the use of TRAF^−/− mice as an optimal model system for understanding the roles of TRAFs in mature lymphocytes. The most common experimental approach to date for examining TRAF function has been transient overexpression of TRAFs in the easily transfected human adenovirus-transformed kidney epithelial cell line 293. When overexpressed in 293 cells, TRAFs 2,3, 5, and 6 constitutively associate with TNFR family receptors, and expression of TRAFs alone is sufficient to activate the transcription factor NF-κB (9, 12, 13). However, when lymphocytes are examined, evidence has been presented that TRAFs may neither associate with TNFR family molecules in resting cells nor activate NF-κB in a constitutive fashion (14), and TRAF function may thus have cell-type-specific features.

TRAF6 consists of 522 amino acids and has a m.w. of 57,00. It shares 30% sequence identity with other TRAFs in the TRAF-C and the Zn-binding domains (11). A role for TRAF6 in activation of NF-κB has been reported for various receptors such as CD40 (12), IL-1 receptor (11), and hToll receptor (15). TRAF6 has also been reported to stimulate extracellular signal-related kinase activation in CD40 signaling along a ras-independent pathway (16). However, all of these experiments were done using transient transfection of non-B cells, and, as described above, their results may not be applicable to the functions of TRAF6 in response to CD40 signaling in B lymphocytes. We have used two alternative approaches to study the physiologic role of TRAF6 in CD40 signaling to B cells. In the first approach, we produced and stably expressed in mouse B cell lines a human CD40 molecule with two cytoplasmic domain point mutations (hCD40EEAA); this mutant fails to bind TRAF6, while showing normal association with TRAFs 2 and 3. In the second approach, we inducibly expressed in B cells a transfected dominant-negative (DN) TRAF6 molecule which contains only the C-terminal TRAF-binding domain of TRAF6. Using these complementary approaches, we find that TRAF6 association with CD40 is important for CD40-induced IL-6 and Ig secretion as well as B7-1 up-regulation. Interestingly, however, CD40-mediated activation of NF-κB and c-Jun kinase in B cells was unaffected by disrupting the CD40-TRAF6 association.

Mouse B cell lines CH12.LX (17) and M12.4.1 (18) were cultured in RPMI 1640 supplemented with 10% FCS, 10 μM 2-ME, and antibiotics [B cell medium (BCM) 10]. Transfected B cells were cultured in BCM-10 supplemented with 400 μg/ml geneticin (Life Technologies, Grand Island, NY). Chinese hamster ovary cells (CHO-K1) obtained from the American Type Culture Collection (Manassas, VA) were cultured in DMEM (high glucose) supplemented with 10% FCS, 10 μM 2-ME, 1× MEM non-essential amino acids (Sigma, St. Louis, MO), and antibiotics (MEM culture medium (MCM)-10). CHO cells transfected with a plasmid encoding the mouse CD40 ligand CD154 (19) were cultured in MCM-10 supplemented with 1 mg/ml geneticin. CHO cells transfected with a plasmid encoding human CD154 were a kind gift from Dr. Amelia Black (IDEC Pharmaceuticals, San Diego, CA) and were cultured in CHO-S-SFM II (Life Technologies) supplemented with 50 nM methotrexate. Embryonic kidney epithelial cell lines 293 and 293-T were cultivated in MCM-10.

Flag-tagged DNTRAF6 consists of the C-terminal 228 amino acids of TRAF6, corresponding to the TRAF-binding domain but lacking the Zn fingers and the RING finger. DNTRAF6 and WTTRAF6 were prepared by PCR amplification of cDNA from M12 cells. For DNTRAF6, CCGTCGACATGGAAACTATCAAACAGTTGGAGAGTC and CCTCTAGATTGAACACAAGTACATGGACGC primers and for wild-type (WT) TRAF6, GGGTCGACATGAGTCTCTTAAACTGTGAGAAC and CCGAATGGTCCGTTTGAGCTC primers were used, and cDNA from M12.4.1 was used as a template. Amplified cDNA was inserted into the mammalian expression vector pRSV.5(neo) (20) for constitutive expression or the plasmid pOPRSVI.mcs1 (19) for inducible expression of DNTRAF6. Mutant hCD40EEAA, in which two glutamic acid residues at positions 232 and 235 were substituted by alanines, was PCR amplified using oligonucleotide primers containing appropriately positioned point mutations and phCD40.neo (21) as a template. The PCR product was inserted into pRSV.5(neo). DNA sequencing confirmed that all PCR products were free of any undesired mutations. FLAG-tagged TRAF2 and hemagglutinin (HA)-tagged TRAF3 constructs have been described previously (22).

M12.4.1 and CH12.LX stable transfectants constitutively expressing lac repressor (LacI) have been described previously (19, 23). Super transfection of CH12.LAC and M12.LAC with DNTRAF6 and transfection of CH12.LX and M12.4.1 with hCD40EEAA constructs was conducted using electroporation as previously described (24). Geneticin-resistant clones were analyzed for expression of DNTRAF6 and hCD40EEAA using intracellular or surface membrane immunofluorescence staining, with analysis by flow cytometry on a FACScan (Becton Dickinson, Mountain View, CA) flow cytometer, as described (21, 22). Induction of inducible DNTRAF6 expression was accomplished by incubation of transfectants with 100 μM isopropylthio-β-d-galactoside (IPTG) for 48 h as described previously (19, 23). Generation of B cell transfectants expressing WThCD40 has been described previously (21).

A mAb specific for the FLAG epitope tag was purchased from Eastman Kodak (New Haven, CT). Anti-HA epitope tag Ab was purchased from Babco (Richmond, CA). Anti-mouse IgG-HRP was purchased from Bio-Rad (Hercules, CA). Recombinant mouse IL-6 and the mAbs 16/10A1 (fluorescein-conjugated anti-mouse B7-1, hamster IgG), G235-2356 (antitrinitrophenyl, isotype control, hamster IgG), Jo2 (anti-mouse Fas, hamster IgG) were purchased from PharMingen (San Diego, CA). Goat anti-hamster IgG-FITC was purchased from Jackson ImmunoResearch Laboratories (West Grove, PA). MOPC-21, streptavidin-HRP, and luciferin were purchased from Sigma (St. Louis, MO). Goat anti-mouse IgG1-FITC and goat anti-rat IgG-FITC were purchased from Southern Biotechnology Associates (Birmingham, AL). The following hybridomas were purchased from the American Type Culture Collection (ATCC) or were generous gifts from the indicated individuals: G28–5 (anti-human CD40, mouse IgG1; ATCC); M17/4.4.11.9 (anti-mouse LFA-1α, rat IgG2a; ATCC); YN1/1.74 (anti-mouse ICAM-1, rat IgG2a; ATCC); UC8–169 (hamster IgG, isotype control; ATCC); 20F3.11 and 32C11.4 (anti-mouse IL-6, rat IgG; ATCC); 1C10 (anti-mouse CD40, rat IgG2a) from Dr. Frances Lund (Trudeau Institute, Saranac Lake, NY), and EM95.3 (anti-mouse IgE, rat IgG2a) from Dr. Thomas Waldschmidt (University of Iowa, Iowa City, IA).

Transient transfection of 293-T cells was performed by calcium phosphate precipitation (22) with the indicated constructs. Cells were harvested 36–48 h after transfection and lysed in 400 μl of lysis buffer [1% Triton X-100, 150 mm NaCl, 20 mM HEPES (pH 7.0), 0.4 mM EDTA, and protease inhibitors; Boehringer Mannheim, Indianapolis, IN] for 30 min on ice. The lysates were centrifuged at 14, 000 × g for 15 min at 4°C, and 370 μl of the supernatants was immunoprecipitated with anti-hCD40-coated protein G-agarose beads for 1.5 h at 4°C. The beads were washed four times with lysis buffer, resolved by SDS-PAGE, and protein bands were transferred to nitrocellulose. Subsequent immunoblotting was performed with anti-FLAG mAb (for TRAFs 2 and 6) or anti-HA (for TRAF3) Ab followed by HRP-labeled goat anti-mouse IgG Ab, as previously described. Protein bands were visualized using a chemiluminescent detection system (Pierce, Rockford, IL).

M12.4.1 transfectants (10⁵) expressing DNTRAF6 (induced with 100 μM IPTG for 24 h) or hCD40EEAA were stimulated with 2 μg of anti-CD40 mAbs or isotype control Ab for 72 h in a volume of 2 ml. Surface expression of B7-1, ICAM-1, LFA-1, and Fas was determined by flow cytometry as described previously (21).

Transfected CH12.LX cells (1 × 10⁵) plus untransfected CHO-K1, CHO-mouse CD154, or CHO-hCD154 cells were cocultured at a ratio of B cells:CHO cell (4:1) with or without IPTG in BCM-10 for 48 h, and the supernatants were quantitated for secreted IL-6 by ELISA as described (19). Values given represent the mean ± SE of triplicate wells. Exogenous IL-6 (10 ng/ml) and anti-IL-6 mAbs (10 μg/ml) were used at previously determined saturating concentrations.

CH12.LX cells express surface and secreted IgM specific for phosphatidylcholine, an Ag found on the surface of sheep RBC (25). IgM secretion by CH12.LX transfectants was determined as described previously (26). Briefly, cells were preincubated with or without 100 μM IPTG for 24 h for transfectants expressing inducible DNTRAF6 followed by incubation with the indicated stimuli for 48 h. Cells expressing hCD40EEAA were stimulated for a total of 72 h. To study the role of IL-6 in Ab secretion, cells were stimulated in the presence or absence of either blocking or nonblocking IL-6 Abs or in the presence of exogenous recombinant mouse IL-6. The number of IgM-secreting cells/10⁶ viable recovered cells was enumerated as cells able to form lytic plaques on a lawn of sheep RBC as described (26). Sheep erythrocytes used as a source of phosphatidylcholine Ag were purchased from Elmira Biologicals (Iowa City, IA).

Both nuclear protein extraction and EMSA were performed as described previously (14). Briefly, 10⁷ viable cells, previously induced with 100 μM IPTG for 48 h when indicated, were stimulated with 1 μg/ml of anti-CD40 Ab or isotype control Ab for 1.5 h at a concentration of 10⁶ cells/ml. Cells were lysed and nuclear extracts were prepared as described previously (14). The extracts were stored at −70°C in the presence of protease inhibitors (Mini Complete; Boehringer Mannheim). The double-stranded NF-κB probe, previously described (14), was end labeled with [γ-³²P]ATP using T4 polynucleotide kinase. A total of 5 μg of nuclear extract was incubated with 0.25–0.5 ng of probe for 30 min at room temperature. The samples were resolved on a 5% native polyacrylamide gel at a constant current of 20 mA. The gel was dried and exposed to x-ray film overnight at −70°C.

M12.4.1 stable transfectants (induced for 24 h with 100 μM IPTG in the case of cells expressing DNTRAF6) were transiently transfected by electroporation with 5 μg of CMV-β-galactosidase (β-gal) construct and 10 μg of a luciferase reporter construct (4xIκB-Luc) under the control of four NF-κB binding sites and a minimal promoter as previously described (14). 293 cells were transiently transfected with 1 μg each of CMV-β-gal construct and 4xIκB-Luc construct, 5 μg of various TRAF6 expression vectors, and 1 μg of WT hCD40 construct by calcium phosphate precipitation. The total amount of DNA transfected into 293 cells was always adjusted to 8 μg with a control expression vector. Viable M12.4.1 and 293 cells (5 × 10⁵) were stimulated in triplicate 24 h after transfection with 1 μg/ml anti-mCD40, anti-hCD40, or isotype control Ab for 24 h at 37°C. Cells were assayed for luciferase activity, as described previously (14), with a luminometer (TD-20/20; Turner Designs, Sunnyvale, CA) immediately after the addition of 100 μl of 1 mM luciferin. β-gal activity was assayed using a Galacto-Light-Plus assay system (Tropix, Bedford, MA) and was used to normalize luciferase activity to correct for transfection efficiency.

M12.4.1 transfectants (2 × 10⁶) were stimulated with 1 μg/ml anti-mCD40, anti-hCD40, or isotype control Ab for 5 min or with 0.6 M sorbitol for 20 min at 37°C and c-Jun kinase activity was measured as described previously (27). Samples were resolved by SDS-PAGE, and phosphorylated c-Jun was visualized by autoradiography.

An inducible expression system was used for DNTRAF6 to enable comparison of various functions of the same transfected B cell clone in the presence or absence of DNTRAF6 expression, as well as to avoid any long-term toxic effects of constitutive expression of a negative signaling protein. DNTRAF6 expression was induced in the mouse B cell lines M12.LAC and CH12.LAC by incubation with IPTG for 24–48 h (Fig. 1). Expression of DNTRAF6 had no effect on expression of CD40 (data not shown). hCD40EEAA was expressed constitutively in M12.4.1 and CH12.LX transfectants. The level of expression of the mutant hCD40 molecule was determined by flow cytometry (Fig. 1). Clones were selected for expression similar to that of WThCD40 on previously transfected subclones (last panel).

The cytoplasmic domain of CD40 has an overlapping binding site for TRAFs 2, 3, and 5, and a distinct binding site for TRAF6 (10). Pullen et al. (10) have reported that an 8-mer peptide, QEPQEINF, derived from amino acids 231–238 of hCD40 binds TRAF6. To analyze the specific role of TRAF6 binding in CD40-mediated signaling to B cells, we generated an hCD40 mutant molecule in which two glutamic acid residues at position 232 and 235 were substituted with alanines. We determined the ability of this molecule to bind TRAFs 2, 3, and 6 in comparison to the WThCD40 molecule. We were unable to immunoblot for endogenous WTTRAF6 during our binding experiments because commercially available Abs to mouse TRAF6 showed no reactivity on Western blots. To circumvent this problem, we transiently expressed our DNTRAF6 construct for these experiments. We coexpressed either WT or hCD40EEAA with TRAFs 2 and 3 or DNTRAF6 in an epithelial cell line, 293-T, immunoprecipitated hCD40, and blotted for the different TRAFs. Association of TRAFs 2 and 3 with hCD40EEAA was similar to that of WThCD40. However, DNTRAF6 did not associate with hCD40EEAA (Fig. 2).

CD40 signaling has been shown to induce IL-6 secretion in B cells (4), including the mouse B cell line CH12.LX (19). Previous studies in our laboratory have found that a deletion mutant of CD40, lacking 32 amino acids from the cytoplasmic tail (hCD40Δ32), cannot bind TRAFs 2 and 3 (22). However, hCD40Δ32 induces IL-6 secretion comparable to that induced by WThCD40 (28). The TRAF6 binding site in hCD40Δ32 is still intact, but a further-truncated molecule, hCD40Δ55, cannot induce IL-6 secretion and lacks the TRAF6 binding site (28). Thus, we postulated that CD40-mediated IL-6 secretion might require TRAF6. Data presented in Fig. 3,A show that hCD40EEAA stimulated 75% less IL-6 secretion upon engagement with CHO-hCD154 cells as compared with IL-6 secreted upon engagement of the WThCD40 molecule. IL-6 production induced via WThCD40 is always considerably higher than that induced by mouse CD40, as we have previously shown (28). Two likely possibilities account for this. First, the expression of human CD40 molecules achieved via transfection is always significantly higher than that of endogenous mouse CD40 on our B cell subclones, and, second, for unknown reasons, the expression of human CD154 on the transfected CHO cells we use for a stimulus for IL-6 production is higher than the maximum expression of mouse CD154 we could achieve on CHO transfectants. Thus, a combination of lower amounts of stimulating ligand and lower receptor expression may account for the lower IL-6 production. Importantly, however, both of these transfectants secreted comparable levels of IL-6 when stimulated with CHO-mouse CD154 cells. This control indicates that the two cell lines are similar in their inherent ability to secrete IL-6 in response to a mouse CD40 signal. A similar although less pronounced defect was seen in CH12.LAC cells inducibly expressing DNTRAF6. IL-6 secretion was reproducibly ∼25% lower when expression of DNTRAF6 was induced by IPTG (Fig. 3 B). The more dramatic effect seen with hCD40EEAA compared with DNTRAF6 may reflect the complete lack of TRAF6 binding to hCD40EEAA compared with the competition for TRAF6 binding achievable using DNTRAF6. Because presently there is not a reliable Ab which detects endogenous TRAF6 in cells, we were unable to determine the relative level of expression of transfected DNTRAF6 vs endogenous WTTRAF6 in the B cells, and therefore cannot precisely quantitate the degree of competition that DNTRAF6 provides for endogenous TRAF6 binding to CD40.

CD40 signaling leads to Ab secretion in CH12.LX cells and can synergize with B cell Ag receptor (BCR) signaling to enhance Ab secretion (29). There are no previous reports on the role of TRAF6 in CD40-mediated Ab secretion. However, it has been previously shown that both CD40 signals and exogenously added IL-6 induce CH12.LX to secrete Ab, and both signals are enhanced by BCR signaling (19, 30). Thus, it was of interest to determine whether TRAF6 is also involved in CD40-mediated Ab production. CH12.LX cells expressing either WThCD40 or hCD40EEAA were stimulated through their endogenous mCD40 molecules or through the transfected hCD40 molecule in the presence or absence of Ag, and Ab-secreting cells were enumerated as plaque-forming cells on a lawn of sheep RBC. Ab secretion stimulated by hCD40EEAA was decreased ∼75% in comparison to that stimulated in cells expressing WThCD40 (Fig. 4,A). However, the ability of hCD40EEAA transfectants to secrete IgM in response to mCD40 was normal. We also saw synergy with BCR signaling by hCD40EEAA, but it was not enough to reverse the signaling defect. Similarly, when CH12.LAC cells inducibly expressing DNTRAF6 were stimulated with anti-mCD40 Ab in the presence of IPTG, Ab secretion was 75% lower (Fig. 4 B) compared with the uninduced cells. Once again, synergy with BCR signaling was seen but was not sufficient to reverse the signaling defect. Thus, TRAF6 binding is important for CD40-mediated Ab secretion in B cells.

To further pursue the potential relationship between IL-6 and Ab secretion, CH12.LX cells were incubated in the presence of CD154 or IL-6 along with either blocking or nonblocking mAbs specific for mouse IL-6. An IL-6 blocking Ab (20F3) decreased Ab secretion by B cells by ∼50% when cells were incubated with CD154 (Fig. 5,A), whereas the isotype-matched non-blocking anti-IL-6 mAb (32C11) had no effect. Addition of exogenous IL-6 alone also induced Ab secretion by B cells but it was much lower compared with cells stimulated with CD154. This suggests that CD40-mediated Ab secretion by CH12.LX cells is driven in part, although not entirely, by secreted IL-6. To test whether a causal relationship exists between TRAF6-mediated IL-6 secretion and Ab secretion, cells expressing DNTRAF6 were incubated with anti-mCD40 Ab and either exogenously added IL-6 or IL-6-blocking Ab. Induced expression of DNTRAF6 decreased Ab secretion, which was restored to normal levels when exogenous IL-6 was also included (Fig. 5 B). Furthermore, when DNTRAF6 expression was induced, concurrent addition of the IL-6-blocking mAb did not further reduce the IgM secretion stimulated through CD40. These results are consistent with the hypothesis that TRAF6 mediates its effects on CD40-stimulated Ig secretion principally through its effects on IL-6 production by the B cell.

We have previously demonstrated that CD40 stimulation of M12.4.1 leads to up-regulation of a number of B cell surface molecules including adhesion molecules and the costimulatory molecule B7-1 (21). Thus, we compared surface molecule up-regulation in B cells stimulated through either the endogenous mCD40 molecule or the transfected hCD40 molecule. CD40-mediated up-regulation of B7-1 was decreased by 50% upon stimulation through hCD40EEAA as compared with the endogenous mCD40 stimulation (Fig. 6). Up-regulation of LFA-1, ICAM-1, and Fas was unaffected (data not shown). Induced expression of DNTRAF6 in M12.4.1 transfectants did not result in any significant effect on up-regulation of these surface molecules including B7-1 (data not shown). As discussed above, this may be due to insufficient competition for endogenous WTTRAF6 by the DNTRAF6, particularly as the effect of the hCD40EEAA mutation was less drastic on B7 up-regulation than on IL-6 and Ab secretion, suggesting that B7 up-regulation may be less dependent upon TRAF6.

TRAF6 has been reported to be involved in CD40-mediated NF-κB activation (16, 31). However, these studies were performed in either 293 kidney epithelial cells or in Jurkat T cells, cell lines which normally do not express CD40. In addition, although studies in 293 cells indicated that TRAF2 also plays an important role in CD40-mediated NF-κB activation (9), subsequent experiments performed in mice (32, 33) and in B lymphocytes (14) showed that in cells which normally express CD40, TRAF2 is not required for CD40-mediated NF-κB activation. Thus, to study the role of TRAF6 in CD40-mediated NF-κB activation in B cells, we first examined nuclear translocation of NF-κB in M12.4.1 cells expressing hCD40EEAA. Nuclear translocation of NF-κB in M12.4.1 cells was similar in cells stimulated through endogenous mCD40 or the transfected hCD40EEAA molecule (Fig. 7,A). Also, no difference in NF-κB translocation was seen when M12.4.1 cells were induced to express DNTRAF6 compared with the uninduced cells (Fig. 7 B). Similar results were seen using CH12.LX cells (data not shown).

To measure CD40-mediated NF-κB-induced transcription, M12.4.1 transfectants expressing hCD40EEAA or WThCD40 were transiently transfected with a NF-κB reporter construct and then stimulated through the endogenous mCD40 or the hCD40 molecule. No difference in activation of the reporter gene was seen by hCD40 signaling vs mCD40 signaling (Fig. 8,A). Once again, similar results were obtained with M12.4.1 inducibly expressing DNTRAF6 (Fig. 8,B). This strongly suggests that TRAF6 does not have a critical role in CD40-mediated activation of NF-κB in B lymphocytes. These results are in direct contrast to what has been reported previously in non-B cells (31); therefore, we repeated our experiments in 293 epithelial cells. Transient expression of DNTRAF6 alone in 293 cells decreased NF-κB activation below the basal level. Expression of WTTRAF6 or WThCD40 increased NF-κB activation, and coexpression of both gave an additive effect (Fig. 8 C). Interestingly, although overexpression of hCD40EEAA alone gave lower activation of the NF-κB reporter than did WThCD40, there was still a cooperative effect with TRAF6, although hCD40EEAA does not detectably bind TRAF6. These findings suggest that TRAF6 function in CD40-mediated signaling differs at least partially between B lymphocytes and epithelial or T cells, and reinforces the need to study CD40 signaling in lymphocytes to completely understand its function in these cells.

CD40 is known to activate c-Jun NH₂-terminal kinase (JNK) in B cells (23). To determine the role of TRAF6 in CD40-mediated JNK activation, M12.4.1 cells expressing hCD40EEAA were stimulated through the endogenous mCD40 or through the transfected hCD40EEAA. JNK activation was comparable when cells were stimulated through either receptor showing that CD40-mediated activation of JNK was independent of TRAF6 (Fig. 9 A). Similar results were obtained in M12.4.1 cells inducibly expressing DNTRAF6 (Fig, 9B). Similar results were found in CH12.LX cells (data not shown).

TRAF6 has been reported to play a role in signaling via CD40 (16, 31), IL-1R (11), and hToll-like receptor (15). Reports studying TRAF6 have largely been restricted to studying the potential role of TRAF6 in transcription factor activation (11, 15, 16) and have largely been performed in either 293 kidney epithelial cells or Jurkat T cells, neither of which normally express CD40. However, because we are interested in understanding how CD40 delivers its important activation signals to B lymphocytes, we wished to study the role of TRAF6 in CD40-mediated signaling in B cells. We adopted two different complementary approaches to accomplish this. In the first approach, we inducibly expressed DNTRAF6 in two different B cell lines and compared various functions of the B cells with or without expression of DNTRAF6. In the second approach, we tested signaling via a transfected mutant of hCD40 which does not bind TRAF6 but binds TRAFs 2 and 3 normally, and used signaling via the endogenous mouse CD40 molecule as an internal control. The collective results obtained using both approaches showed that TRAF6 plays a role in CD40-mediated IL-6 secretion, differentiation, and up-regulation of B7 in B cells, functions previously not tested for the TRAF6 molecule.

Our results further reveal a connection between CD40-mediated IL-6 production and Ab secretion, linked by a requirement for TRAF6 binding to CD40. Previous reports, including those from our own laboratory, have linked IL-6 with B cell differentiation (30, 34, 35, 36, 37). Other reports have shown that IL-6 secretion is induced by CD40 signaling in B cells (4, 19), and have suggested a connection between CD40-mediated IL-6 production and the induction of B cell differentiation (38, 39). Data presented here show that preventing TRAF6 from binding to CD40 decreased both IL-6 secretion (Fig. 3) and Ab secretion (Figs. 4 and 5) by ∼75%. That this similarity in effect is likely to be causal rather than coincidental is demonstrated by the findings that addition of exogenous IL-6 reversed the inhibition of differentiation seen when CD40 is stimulated in cells induced to express DNTRAF6, but exogenous IL-6 did not further increase CD40-mediated Ab secretion in cells that do not express DNTRAF6. Consistent with these findings, addition of blocking Ab to IL-6 did not further decrease DNTRAF6-induced inhibition of differentiation. Data in Fig. 5 also show that CD40-mediated Ab secretion is partially TRAF6 and IL-6 independent. Addition of saturating amounts of exogenous IL-6 could not produce the same amount of Ab secretion as the CD40 signal itself, nor could inhibition of either IL-6 secretion (Fig. 5) or TRAF6 binding (Fig. 4) completely eliminate CD40-mediated Ab secretion. In addition, CD40 molecules which cannot detectably bind TRAF6 (hCD40EEAA) were still able to stimulate IL-6 production above basal levels (Fig. 3). Although this result may indicate that hCD40EEAA is able to bind an undetectable but biologically significant amount of TRAF6, it is at least as likely that CD40-stimulated IL-6 production is partially TRAF6 independent.

A major function attributed to TRAF molecules, based principally upon studies performed in transiently transfected 293 epithelial cells, has been the activation of the transcription factor NF-κB, and several studies have reported that TRAF6 plays a role in CD40-mediated activation of NF-κB in 293 cells (16, 31). Although we could reproduce these results in 293 cells, our experiments in B cells indicate the lack of a critical role for TRAF6 in CD40-mediated NF-κB activation (Figs. 7 and 8). While this manuscript was in preparation, Lomaga et al. (40) reported that CD40-mediated proliferation and NF-κB activation were impaired in splenocytes obtained from TRAF6-deficient mice. However, these mice have low perinatal and postnatal survival, and before death show enlarged spleens. The number, developmental stage, subset composition, or phenotype (including level of CD40 expression) of B cells in the TRAF6^−/− mice was not described. Thus, it is difficult to know whether their loss of NF-κB activation is a direct or indirect effect of TRAF6 deficiency. TRAF6-induced JNK activation in 293 epithelial cells has also been reported (41, 42). However, our results suggest that CD40-mediated JNK activation in B cells does not absolutely require TRAF6 binding.

We believe the most likely explanation for the partial discordance in the results of CD40-signaling assays in various cells is that different cell types have overlapping but distinct CD40-signaling pathways. This explanation is consistent with previous findings that TRAF2 appears to be very important for NF-κB activation in non-B cells (9) but not in B cells (14, 32, 33). In addition, TRAF molecules associate constitutively with TNFR family molecules in transiently transfected 293 cells, but this does not appear to be the case in B cells (43). This distinction between CD40 function on cells that normally express it and those that do not may have considerable functional significance. CD40 was originally identified as an Ag expressed on normal B cells and malignant bladder carcinoma cells (44), and CD40 expression has been reported on melanoma cells (45) as well as prostate, renal, and cervical carcinomas (46, 47, 48). This raises the intriguing possibility that when CD40 expression is induced on nonhematopoietic cells, particularly those of epithelial origin, cell type-specific consequences of CD40 signaling may contribute to cellular transformation.

The two most reproducible early signaling events documented in B cells following CD40 engagement are the activation of JNK and activation of NF-κB, neither of which were noticeably affected in our study by preventing TRAF6 from binding to CD40. However, a loss or decrease in TRAF6 binding was clearly able to inhibit CD40-mediated IL-6 secretion, differentiation, and B7 up-regulation, indicating that TRAF6 participates significantly in CD40 signaling. Other events reported to result from CD40 signaling to B cells include activation of protein tyrosine kinases (49, 50), phosphatidyl inositol 3-kinase, phospholipase Cγ2 (50), and Jak 3 (51). Determining which of these events is triggered by TRAF6 will be the subject of future investigation.

We thank Luis Ramirez for technical assistance.