Cellular responses to murine CD40 in a mouse B cell line may be TRAF dependent or independent

0014-2980/02/0101-39$17.50+.50/0© WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2002

Cellular responses to murine CD40 in a mouse Bcell line may be TRAF dependent or independent

Eric Manning1, Steven S. Pullen2, Donald J. Souza2, Marilyn Kehry2 and Randolph J.Noelle1

1 Department of Microbiology, Dartmouth Medical School, Lebanon, USA2 Department of Biology, Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, USA

Engagement of CD40 by its ligand induces IKK and mitogen-activated protein kinase(MAPK) phosphorylation and transcriptional activation, leading to activation and differentia-tion of B cells. These events are most likely transduced by adaptor molecules that arerecruited to the CD40 cytoplasmic domain, called TNF receptor-associated factors (TRAF).We have engineered a chimeric CD40 molecule using the human extracellular sequence andthe murine cytoplasmic domain to assess the contribution that specific TRAF bindingdomains provide to the cytoplasmic signaling functions of CD40. The data presented hereshow that the shared binding site for TRAF2 and TRAF3 accounts for receptor internaliza-tion, and the majority of signaling through CD40, but is redundant with the TRAF6 bindingsite for activation of p38 and NF ‹ B signaling pathways. Disruption of the TRAF2/3 bindingsite results in a delayed and diminished kinase pathway induction, but complete preclusionof all signals requires the disruption of more than the two known TRAF binding sites. Thespecific TRAF dependency of CD40-induced growth arrest, TNF- § production, and phos-phorylation of signaling molecules are shown, while p38 MAPK activation and cell surfaceantigen modulation suggest TRAF independent CD40 signaling in B cells.

Key words: CD40 / TRAF / Signaling / B cell / Murine

Received 6/8/01Revised 13/9/01Accepted 22/10/01

[I 22369]

Abbreviations: TRAF: TNF receptor-associated factorMAPK: Mitogen-activated protein kinase JNK: Jun N-terminal kinase TD: Thymus-dependent

1 Introduction

CD40 is a receptor expressed on the surface of B lym-phocytes that is essential for thymus-dependent (TD)humoral immunity (reviewed in [1]). The interactionbetween CD40 and its ligand, CD154, constitutes a cen-tral element in triggering B cell expansion and differentia-tion in response to TD antigens [2, 3]. The various biolog-ical activities induced via CD40 are mediated throughTNF receptor-associated factors (TRAF) [4–8]. Within thecytoplasmic domain of CD40 there are two definedstructural motifs that mediate the binding of eitherTRAF2 and TRAF3, or TRAF6. These adapter moleculeshave been shown, in various systems, to link to all themitogen-activated protein kinase (MAPK) signaling path-ways (reviewed in [9]). Identification of the signaling cas-cades evoked as a consequence of TRAF binding hasbeen sought in a variety of systems using a spectrum of

approaches. Originally, data from TRAF overexpressionsystems suggested possible functions for TRAF pro-teins, only some of which were confirmed in vivo [5–7,10–12]. Furthermore, data suggest that the role of TRAFin downstream activation of NF- ‹ B, JNK, and p38 maydiffer depending on the cell type studied. For example,TRAF6 has recently been implicated as the TRAF familymember responsible for activating the p38 kinase cas-cade in HEK293 cells [13], while in a human B cell line,TRAF3 filled the same role [10]. Others have alsoreported that the function of TRAF proteins vary depend-ing on the cell type studied [14]. Therefore, to understandthe function of TRAF in B cell biology, extrapolationsfrom other cell types may not be reliable.

To gain an understanding of the role of TRAF proteins inB cell activation, the signaling capacity of CD40 mole-cules that are defective in TRAF binding have been stud-ied. Mutant forms of human CD40 with selective disrup-tions in the TRAF binding sites for TRAF2/TRAF3 orTRAF6 have been expressed in murine B cells to assessthe impact of these mutations on CD40 signaling. Inthese studies, the common TRAF2 and TRAF3 bindingsite was found to be at least partially responsible forinduction of costimulatory molecules [14–17]. TRAF2/3recruitment has also been shown to be important for the

Eur. J. Immunol. 2002. 32: 39–49 Studies on TRAF-dependent CD40 signaling in a murine B cell line 39

Fig. 1. Chimeric CD40 DNA constructs used to generate stable cell lines. (A) Schematic diagrams of the CD40 gene. The extra-cellular domain is entirely human sequence, while the transmembrane and cytoplasmic domains are of murine sequence. Thereported TRAF binding domains are shown; critical residues and point substitutions are highlighted. (B) Flow cytometry histo-grams of clonal lines of M12 expressing human-murine chimeric CD40. At least three clones of every line were tested; represen-tative clones are shown stained with a human CD40 antibody. Outlines show the staining pattern of an isotype-matched controlantibody.

induction of NF- ‹ B and Ig secretion [16, 17], and inWEHI-231, the TRAF2/3 binding site has been reportedto be a major mediator of MAPK activation [18]. In CH12cells, another murine B lymphoma, it has been shownthat TRAF6 can signal antibody production [14]. Non-hematopoietic cells, mostly the HEK293 cell line, havealso been engineered to express mutant CD40, and relymore heavily on TRAF6 for CD40 signaling kinase activa-tion and cytokine production [13, 14, 19]. Taken together,studies show that the relative contributions of TRAFrecruitment to CD40 signaling vary among cell types,and species.

To gain an appreciation for TRAF function in B cells,murine B cell lines have been engineered to express achimeric human/mouse CD40 protein with defined muta-tions in the TRAF2/3 site, the TRAF6 site, and both sitestogether. The use of a chimeric molecule with a murinecytoplasmic domain ensures that any species-specificcytoplasmic protein interactions will remain intact. Stud-ies presented evaluate the causal relationships betweenTRAF recruitment and the biochemical and biologicalresponses of B cells to CD40 triggering. The data pre-sented show that the shared binding site for TRAF2 andTRAF3 accounts for the majority of signaling through

CD40, but is redundant with the TRAF6 binding site foractivation of p38 and NF- ‹ B signal pathways. Disruptionof the TRAF2/3 binding site results in a delayed anddiminished kinase pathway induction, but complete pre-clusion of those signals requires the disruption of morethan both the TRAF2/3 and TRAF6 sites.

2 Results

2.1 Generation of M12 clones bearing a human-mouse chimeric CD40

Upon engagement of CD40 by its ligand, the aggregatedreceptor recruits TRAF 1, 2, 3, 5 and 6 [8, 10, 13, 18, 20]. Inan effort to explore the distinct roles of TRAF family pro-teins, a set of chimeric CD40 molecules have been engi-neered that bear the human extracellular domain and themurine transmembrane and cytoplasmic domains, withor without point mutations in TRAF binding sites [4, 13].The mutant chimeric molecules shown in Fig. 1A, thatlack either the TRAF 6, TRAF2/3, or both TRAF2/3 andTRAF6 binding sites, and a truncated molecule were pro-duced. The murine cytoplasmic domain was used inthese constructs because of sequence disparity

40 E. Manning et al. Eur. J. Immunol. 2002. 32: 39–49

Fig. 2. JNK activation relies on the TRAF2/TRAF3 bindingsite on CD40. Western blot probed with an antibody directedagainst the dually phosphorylated form of Jun N-terminalkinase. M12 cell lines expressing chimeric CD40 weretreated with human CD154 for the times indicated, or with1c10 for 5 min as a control (C). Cells were then lysed, andsamples were prepared for Western blotting. Shown is a rep-resentative from three independent Western blots.

between human and murine CD40. Most other studiesthat have addressed the function of the CD40 cytoplas-mic domain in signaling have used heterologous sys-tems.

Chimeric CD40 constructs were engineered into thepDOI-5 plasmid for expression under control of the MHCclass II E § promoter in antigen-presenting cell types [21].Those constructs were electroporated into the M12.4.1murine B cell line along with a selectable vector to gener-ate stable cell lines. Clonal lines were generated by limit-ing dilution, under antibiotic selection, and screened bystaining for the extracellular domain of human CD40.Expression analyses of representative clones of M12 thatexpress the human/mouse chimeric CD40 protein arepresented in Fig. 1B. As shown, staining with anti-humanCD40 detected similar levels of the transfected geneproduct in each of the selected clones. Experimentsherein were performed with multiple clones of eachmutant line, which all demonstrated similar expression ofthe transfected gene.

2.2 The TRAF2/TRAF3 binding site is critical forCD40-induced JNK activation

One of the biochemical responses to CD40 engagementis the induction of the Jun N-terminal kinase (JNK) cas-cade [22]. Substantial data from both TRAF2-deficientmice and studies using a truncated CD40 in a B cell lineshowed that TRAF2 was critical for JNK activation inB cells [18, 23]. However, conflicting data surroundthe function of TRAF6 in signaling JNK from CD40. InHEK293 cells, TRAF6 overexpression induces JNK [12]and the disruption of the TRAF6 binding site blockedJNK activation by CD40 [13]. In a B cell line, however, adominant negative TRAF6 demonstrated no impact onJNK induction by CD40 [19]. We stimulated each of theM12 clones bearing a chimeric CD40 molecule using sol-uble human CD154 (CD40 ligand) for times ranging from5 to 30 min, and Western blotted with an antibodydirected at the dually phosphorylated form of JNK. Dis-ruption of the TRAF2/3 binding site ablated JNK activa-tion. In contrast, disruption of the TRAF6 site did notinfluence JNK activation across any of the time pointsmeasured (Fig. 2), indicating that CD40 induces JNKphosphorylation through TRAF2/3 recruitment. An ago-nistic mAb specific to murine CD40 was used to controlfor the ability of each clone to transduce a signal to theJNK pathway.

2.3 CD40 activates p38 through a redundantmechanism

CD40 has been shown to induce p38MAPK in a varietyof cellular models, including HEK293 cells, B cell linesRamos, Daudi, WEHI231, and human tonsillar B cells[10, 13, 24, 25]. It has also been shown that p38 is criticalfor CD40-mediated induction of IL-12 in dendritic cells,and IL-1 g in B cells [26]. Conflicting data in the existingliterature show that TRAF6 is either critical [27] or dis-pensable [10] for p38 induction by CD40. To furtheraddress this question, we examined phosphorylation ofp38 MAPK in the M12 transfectants. M12 clones werestimulated with soluble human CD40 ligand for varioustimes between 5 and 30 min, together with untreatedcontrols. Lysates were Western blotted with an antibodydirected against the dually phosphorylated form of p38.When the TRAF2/3 binding site was disrupted the kinasep38 showed a profile of phosphorylation in which thekinetics were delayed, and the intensity of signal wasdiminished (Fig. 3). In contrast with data from HEK293cell systems, disruption of the TRAF6 binding site has nodiscernable effect on p38 induction, even in the absenceof TRAF2/3 recruitment. Disruption of all known TRAFbinding domains was insufficient to completely block theCD40 signal to p38 phosphorylation, suggesting thatanother signaling domain exists in the CD40 cytoplasmictail. Treating each clone with an agonistic antibodyagainst murine CD40 showed that each clone was capa-ble of inducing p38 via the endogenous murine CD40


Fig. 3. MAPK p38 is activated by multiple pathways fromCD40. Western blots probed with antibodies directedagainst the phosphorylated and activated form of p38. M12cell lines expressing chimeric CD40 were treated withhuman CD154 for the times indicated, then samples wereprepared for Western blotting. Shown is a representativefrom three independent Western blots.

Fig. 4. I ‹ B § phosphorylation can be induced through eitherTRAF binding site. Western blots probed with an antibodydirected against the activation-specific phosphoserines onI ‹ B § . M12 cell lines expressing chimeric CD40 were treatedwith human CD154 for the times indicated, and sampleswere prepared for Western blotting. Results are representa-tive of multiple independent experiments.

(data not shown) similar to the induction through the“wild-type“ transfected chimeric CD40 molecule.

2.4 CD40 activates the NF- ‹ B pathway throughmultiple TRAF-dependent pathways

The activation of the NF- ‹ B pathway has been shown toarise as a consequence of CD40 engagement in a widerange of cell types. Contrasting data have implicated dif-ferent TRAF proteins in mediating this signal. TRAF-deficient mice show that either TRAF2 or TRAF6 is criti-cal for NF- ‹ B activation by CD40 [28, 29]. TruncatedCD40 with the TRAF6 site intact was incapable of acti-vating NF- ‹ B in M12 or CH12 cells [17], yet TRAF6 hasbeen shown to be important for NF- ‹ B activation byhuman CD40 transfected into HEK293 cells [19], but notmurine B cells [14].

M12 clones were stimulated with soluble human CD154for times varying from 5 to 30 min, lysed and Westernblotted with an antibody directed against the duallyphosphorylated form of I ‹ B § . Fig. 4 shows that mutationof the TRAF2/3 binding site on CD40 resulted in adelayed and diminished activation of I ‹ B § phosphoryla-tion. This correlated well with the activation profile of p38shown in Fig. 3. In contrast to the p38 activation profile,however, I ‹ B § phosphorylation completely relied onTRAF recruitment. The delayed and diminished activa-tion seen in the absence of TRAF2/3 recruitment iscompletely abrogated when the TRAF6 binding site wasalso disrupted. These data corroborate the delay in NF-

‹ B activation found in TRAF2-deficient mice by Mak etal. [30] and further demonstrated that there is no TRAF-independent activation of NF- ‹ B via CD40 in murine Bcells. Signaling via endogenous murine CD40 confirmedthe capacity of all of these clones to trigger I ‹ B § phos-phorylation similar to the “wild-type” transfected mole-cule (data not shown).

2.5 Disruption of TRAF recruitment to CD40 onlymarginally interferes with the up-regulationof cell surface molecules

Engagement of CD40 by its ligand induces the modula-tion of a myriad of cell surface molecules on B cells andB cell lines. Previous studies have shown that deficien-cies in TRAF2 or TRAF3 do not block the CD40-dependent up-regulation of CD23 or CD80 [23, 31], butthe existing literature neither addresses the role ofTRAF6 in these functions, nor compares the contribu-tions of the different TRAF binding domains. We evalu-ated requirements for TRAF binding in the up-regulationof CD80 and CD23 on M12 clones through CD40. Over24, 36, or 48 h, neither the TRAF2/3 nor TRAF6 siteappeared to be vital to the up-regulation of either surfacemarker. Surface marker induction at 48 h is shown inFig. 5. Clones expressing the truncated form of CD40 didnot signal a significant increase in expression of thesesurface molecules in response to human CD154, show-ing that any TRAF-independent activation is still CD40dependent. These data suggest that another TRAF-independent domain of the murine CD40 cytoplasmicdomain exists that is capable of mediating the up-regulation of surface markers.


Fig. 5. CD40-driven expression of CD23 and CD80 do notrequire TRAF recruitment. M12 cell lines expressing chime-ric CD40 were seeded in culture plates, and treated withhuman CD154 for 48 h. (A) Lines were stained with a biotiny-lated antibody against CD80, detected with streptavidin-phycoerythrin. Flow cytometry data is converted to meanfluorescence intensity of living cells. (B) Cell lines werestained with an mAb to CD23 directly conjugated to PE.Mean fluorescence intensity is graphed as before. Resultsare representative of multiple independent experiments.

Fig. 6. M12 culture growth arrest is mediated by TRAF pro-teins. Cells were cultured in the presence or absence of sol-uble human CD154 or anti-murine CD40 1c10 for 60 h.Resultant cultures were counted with a hemacytometer andviability assessed by trypan blue exclusion. Data are shownas the percent reduction in culture size as compared to anuntreated control, and represent compiled results from threeseparate experiments.

2.6 CD40 halts the growth of M12.4.1 B cells inculture through TRAF recruitment

Like many tumor lines, the M12.4.1 cell line is known toundergo growth arrest in response to engagement ofCD40. The Thr-Ala substitution that disrupts the TRAF2/TRAF3 binding site was originally identified on humanCD40 in experiments testing this biological response[32]. The role of TRAF6 in this growth arrest function hasnever been examined. To address this, each transfectedline was stimulated with human CD40 ligand, or anti-murine CD40 as a control, for 60 h. At the end of thatincubation, culture growth was determined by counting,and viability confirmed with trypan blue exclusion. Asshown in Fig. 6, neither disruption of the TRAF6 site nor

of the TRAF2/TRAF3 binding site was sufficient to blockthis response. However, when both TRAF bindingdomains were mutated, CD40 induced growth arrestwas ablated. Over this 60-h stimulation, the anti-murineCD40 control resulted in approximately 90% loss in cellnumber from each culture (data not shown). Thesedata indicate that the TRAF binding domains can actindependently to mediate CD40 dependent culturegrowth arrest.

2.7 CD40 induces the secretion of TNF- >in a TRAF-dependent manner

Engagement of CD40 on human B cells induces the pro-duction of TNF- § as an autocrine growth factor [33]. Pre-liminary experiments using RNase protection assaysshowed the induction of TNF mRNA, suggesting that theM12 clones used for these experiments might produceTNF- § in response to CD154 (data not shown). To assessthe dependency of TNF- § production on the recruitmentof specific TRAF proteins, 48-h culture supernatantsfrom M12 cells triggered via human CD40 were assayedby ELISA. Fig. 7 shows the relative induction of TNF- §secretion through the various mutant forms of chimerichuman CD40. The control supernatant was taken froman M12.4.1 culture stimulated in parallel with an agonis-tic anti-murine CD40. Disruption of the TRAF2/3 bindingsite reduced the production of TNF- § by about 50%,whereas disruption of the TRAF6 site was without effect.However, disruption of both TRAF binding domainsreduced the induction of TNF- § secretion to near back-ground levels, indicating a functional role for TRAF6.


Fig. 7. TNF- § secretion by B cells depends on TRAF recruit-ment to CD40. Cell lines were stimulated in triplicate withhuman CD154 for 48 h. Culture supernatants were recov-ered and assayed for TNF- § content by sandwich ELISA.Absorbance values were converted to ng/ml, and concen-trations were normalized against untreated controls, to yieldmeasures of fold induction. Positive controls were a cultureof an M12 clone bearing WT chimeric CD40, incubated 48 hwith an agonistic mAb against anti-murine CD40. Resultsare representative of three independent experiments.

Fig. 8. Internalization of CD40 requires recruitment of TRAF2and/or TRAF3. Cells were cooled to 4°C and stained withCD8-CD154 soluble fusion protein for 60 min. Excess fusionprotein was washed away, and cells were resuspended in37°C cRPMI for the times indicated. Internalization washalted by flooding with cold paraformaldehyde-based fixa-tive, and cell-external CD40 ligand was detected with anmAb against CD8. (A) Histograms show that intensity ofstaining decreases proportionally to the amount of time cellsare given to internalize CD40 at 37°C. (B) M12 clonesexpressing either the WT or TRAF-less versions of chimericCD40 were stained with CD8-CD154 as described above,with the addition of 5 ? g/ml anti-murine CD40 agonisticmAb. Results are representative of three or more indepen-dent experiments.

These data suggest that both the TRAF6 and TRAF2/3sites contribute to the induction of TNF- § production byB cells.

2.8 CD40 internalization is impaired bydisruption of TRAF binding sites

Recently published experiments have shown thatengagement of CD40 induces its translocation intomembrane rafts [34, 35]. Often, association of receptorswith glycolipid-rich membrane domains like rafts andcaveolae precedes their internalization in lymphocytes[36, 37]. Accordingly, CD40 has been shown to internal-ize after engagement with a staining antibody [34]. Also,TRAF2 was recently reported to interact with caveolinand couple TNF receptor 2 to caveolae [38]. We decidedto investigate whether TRAF2 specifically might beimportant in internalization of CD40. To test this function,we stained chimeric CD40 M12 clones with soluble,recombinant CD8-CD154. This fusion protein wasallowed to bind CD40 at 4°C, then unbound excess waswashed away. The cells were warmed to 37°C for 5, 10,30, or 60 min to allow internalization. The amount ofCD8-CD154 remaining on the cell surface was estimatedusing a fluorochrome-coupled anti-CD8 antibody. Flowcytometry histograms (Fig. 8a) showed that wild-typeCD40 rapidly cleared the CD8-CD154 from the mem-brane over time. However, the disruption of the TRAF2/3binding site on CD40, caused receptor bound CD8-CD154 to persist on the cell surface, indicating a failure

of receptor internalization. In contrast, loss in TRAF6binding had no appreciable effect on the ligand-inducedinternalization of CD8-CD154.

The loss in ligand-induced internalization in the absenceof TRAF2 and TRAF3 recruitment could be the result ofloss in kinase activation or in inability of the TRAF2/3


deficient receptor complex to interact with the appropri-ate cytoplasmic machinery for endocytosis. To addressthe former issue, the endogenous murine CD40 receptorwas simultaneously triggered with an anti-murine CD40mAb to elicit kinase activation, while internalization of thetransfected chimeric CD40 was assessed. As can beseen in Fig. 8B, the engagement of the endogenousmurine receptor could not rescue the ability of theTRAF2/3-deficient CD40 to mediate down-regulation.Therefore, it appears that TRAF2 or 3 may play a struc-tural function in mediating CD40 receptor internalization,perhaps through recruitment of CD40 to caveolae ormembrane rafts. Further, this result suggests that trans-fected chimeric CD40 and endogenous murine CD40 donot mix in receptor complexes, which might havemasked the effects of these mutations.

3 Discussion

This study evaluates the role of TRAF2/3 and 6 in the bio-chemical and biological responses of a murine B cell lineto CD40 engagement. Using a murine B cell line stablytransfected with a chimeric CD40 molecule, we haveshown the biological and biochemical function of CD40in its native environment under conditions of alteredTRAF recruitment. While activation of the JNK cascaderelied completely on recruitment of TRAF2/3, activationof the p38 kinase and NF- ‹ B cascades relied on a mech-anism either redundant or shared with TRAF6. There wasa delayed and diminished NF- ‹ B activation in theabsence of TRAF2 that recapitulates findings in the firstreports of TRAF2-deficient mice [30]. The residual phos-phorylation of p38 and the up-regulation of cell surfacemolecules together established that in the absence ofTRAF recruitment, some CD40 signaling events can beinduced. While TRAF recruitment appeared nonessentialfor the up-regulation of activation markers, other eventsincluding receptor internalization, TNF secretion, andculture growth arrest appeared to rely on TRAF recruit-ment.

Recruitment of TRAF2 and TRAF3, in comparison to thatof TRAF6, appeared to impart distinctive effects onmediating signal transduction and alterations in B cellbiology. With the loss of the TRAF2/3 binding site JNKsignaling is completely lost. While the recruitment ofboth TRAF2 and TRAF3 are lost, it is likely this effect isprimarily due to the lack of TRAF2 recruitment. The roleof TRAF2 in CD40-induced JNK activation was first sug-gested by overexpression of TRAF2 in the HEK293 cellline, which resulted in JNK activation [12]. These datawere reiterated by reconstitution of a CD40 signal cas-cade in HEK293 cells, which implicated TRAF2 as well asTRAF6 in JNK signaling [13]. In ex vivo B cells from

TRAF2-deficient mice, the role of TRAF2 in JNK signalingwas also shown [30]. Therefore, multiple, independentlines of evidence show definitively that TRAF2 recruit-ment is critical for JNK activation in B cells.

In contrast to the clear function of TRAF2 in JNK activa-tion, the role of TRAF2 in NF- ‹ B activation in B cells isdifficult to resolve. Initial reports from TRAF2-deficientmice showed that TRAF2 was not important for CD40signaling to NF- ‹ B [30], while subsequent work from thesame group in a TNFR1-deficient system has shown thatTRAF2 was absolutely required for the same function[28]. These observations are perplexing in light of datafrom another knockout mouse showing that TRAF6 isalso absolutely required for induction of NF- ‹ B [29].Knockout mice might indicate that TRAF recruitment isnot redundant for NF- ‹ B activation, but other work in theM12 system has indicated that both TRAF2 and TRAF6were partly responsible for NF- ‹ B induction [19]. Over-expression of the TRAF proteins has supported the ideathat TRAF2 and TRAF6 are each autonomously capableof inducing NF- ‹ B [6]. Data presented in this study showthat loss of TRAF2/3 recruitment greatly diminished anddelayed CD40-induced NF- ‹ B activation. No effect wasobserved as a result of disruption of the TRAF6 site. Arole for TRAF6 was indicated when disruption of both theTRAF2/3 and TRAF6 sites completely ablated NF- ‹ Bactivation. Therefore, NF- ‹ B activation in B cells isdependent on TRAF recruitment, though it is more effec-tively induced in the presence of TRAF2 and TRAF3.

CD40 partially activates p38 MAPK in the absence ofTRAF recruitment, indicating a TRAF-independent sig-naling function within the cytoplasmic domain of thereceptor. This is particularly interesting in light of thedemonstrated requirement for p38 in inducing a re-stricted subset of gene activation by CD40 in B cells. Inprimary B cells, p38 has been shown critical for IL-1 gand IL-10 production, as well as proliferation [24, 26].Work in vivo has also shown that p38 signaling controlsthe CD40-dependent production of IL-12 [39]. At thesame time, a number of genes have been shown torespond to CD40 independently of p38 activation [24].Proposed signaling domains in CD40, which mightexplain TRAF-independent signaling, are reported bothmembrane proximal to the TRAF binding sites, andmembrane distal [17, 40]. Our data do not distinguishwhich proposed domains are capable of signalingthrough p38 or to induce the up-regulation of B cell acti-vation markers.

Because the engagement of CD40 elicits a wide spec-trum of biological responses, the contribution of specificmotifs in the CD40 cytoplasmic domain to these eventshas been sought. We have shown that the up-regulation


of CD23 and CD80 was induced by CD40 engagement inthe absence of TRAF recruitment. These data conflictwith previous reports that show that the up-regulation ofthese molecules rely on the TRAF2/3 binding site fortheir expression on M12 cells [15–17]. One possibleexplanation for this discrepancy may lie in the speciesspecificity of proteins being recruited to the cytoplasmicdomain of CD40. All other studies to date have used full-length human CD40 in murine B cells, whereas our stud-ies have employed a chimeric human/mouse CD40 mol-ecule. Sequence alignment of the human and murineCD40 cytoplasmic domains shows disparity within theTRAF6 binding site, as well as within other proposed sig-naling motifs. It could be argued that human CD40 isinefficient at recruiting murine TRAF6 or other signalingproteins in a murine B cell. Therefore, the murine cyto-plasmic domain may have activities that the humandomain cannot manifest in the environment of a murinecell.

One of the more interesting deficits of function with theloss of TRAF2 and TRAF3 recruitment is the failure ofcells to internalize ligated receptor. Mounting data havedemonstrated the role of topographical regulation asinstrumental in receptor signaling [41]. Specifically,ligated receptors may be translocated into microdo-mains in the lipid membrane rich in signaling machinery,and perhaps then endocytosed. It has recently beenreported that CD40 is a receptor that signals throughmembrane rafts [34, 35]. Interestingly, TRAF2 wasreported to interact with caveolin, indicating that CD40may signal through the specialized membrane pits calledcaveolae [38]. Also, a novel molecule called MIP-T3 wasreported to link TRAF3 with cytoskeletal elements [42].Receptor-based internalization via clathrin includesassociation with cytoskeletal elements. It may be thatTRAF2 mediates relocalization to caveolae, and TRAF3is required for its loss from the cell surface. This wouldexplain the loss of receptor internalization with theT255A mutation.

Disruption of the TRAF6 binding site alone had little, ifany effect on any of the signaling functions we tested.However, the NF- ‹ B signaling and TNF- § production byCD40, which were not completely ablated by the loss ofTRAF2/3, were further diminished by the loss of TRAF6binding. These data suggest that the role of TRAF6 in Bcells is largely redundant with other TRAF proteins. Thispresents a contrast to data gathered in HEK293 overex-pression systems, but is not unprecedented, as TRAF6had little effect in CD40 driven NF- ‹ B reporter assays inB cells as opposed to HEK293 cells [14]. Since TRAF6also participates in signaling NF- ‹ B activation andTNF- § production from the Toll-like receptors [43, 44],it is conceivable that in the absence of TRAF2 and

TRAF3 recruitment, TRAF6 independently initiates cyto-plasmic signaling.

Because protein interactions vary between species andcell types, we have employed a system that allows forstudies of mutant forms of CD40 in its native cytoplasmicenvironment. We have shown that in murine B cells,murine CD40 functions through TRAF-dependent andTRAF-independent mechanisms. The induction of cellsurface markers on B cells and p38 MAPK appear, atleast in part, to be independent of TRAF recruitment. Incontrast, cytokine production, activation of JNK and NF-‹ B, and internalization of CD40 appear to rely on the

recruitment of TRAF to the receptor complex. Differ-ences between these data and that previously publishedcould lie in the fact that previous studies have employedthe human cytoplasmic domain expressed either inmurine or non-immunological cell types. Further studiesmust focus on CD40 and TRAF’s function within theirnative environments.

4 Materials and methods

4.1 Cell culture and transfection

Murine B cell lymphoma M12.4.1 was grown in completeRPMI 1640 supplemented with 10% fetal bovine serum,10 ? M 2-mercaptoethanol, and antibiotics. Cultures weregrown in 6.0% carbon dioxide at 37°C. Clonal lines wereselected in the same media supplemented with 400 ? g/mlG418 (Life Technologies, Rockville, MD). Five million cellswere resuspended in 0.5 ml culture medium, and transferredto a disposable electroporation cuvette (Life Technologies).Sterile PBS with 10 ? g plasmid DNA was added to thecuvette, and electroporation was performed using a Cell-Porator™ (Life Technologies) set to 800 ? F and 250 V.Selection of transfectants commenced no sooner than 12 hafter electroporation. Selected bulk cultures were thencloned by limiting dilution.

4.2 Construction of chimeric CD40

The oligonucleotides 5’-TTGGATCCATGGTTCGTCTGCC-TCTGCAGT-3’ and 5’-ACAGGAATGACCAGCAGGGCTCT-CAGCCGATCCTGGGGACCA-3’ were used to amplify thehuman CD40 ectodomain sequence using phCD40/GemT[13] as a template. The oligonucleotides 5’-CAGGATCG-GCTGAGAGCCCTGCTGGTCATTCCTGTCGT-3’ and 5’-TT-ATCAAAAGGTCAGCAAGCAGCCA-3’ were used to amplifythe mouse cytoplasmic and transmembrane domainsequence containing an N-terminal overlap with the humanCD40 ectodomain sequence from a plasmid encodingmurine CD40 (a gift from Edward Clark, University of Wash-ington, Seattle). These two PCR products were annealed,extended with Taq polymerase, and ligated into pGem-T


(Promega, Madison WI) to generate ph-mCD40/GemT. Theoligonucleotides 5’-GGTGAAAGCGAATTTCTAGACACCTG-GAAC-3’ and 5’-GTTCCAGGTGTCTAGAAATTCGCTTTCA-CC-3’ were used to create a silent mutation (underlined) thateliminated the EcoRI site in the mouse CD40 cytoplasmicdomain using the QuikChange site-directed mutagenesis kit(Stratagene, La Jolla, CA) to generate ph-mCD40/EcoKO/GemT. EcoRI sites were inserted at the 5’ and 3’ ends of thechimeric CD40 coding sequence by PCR to generate ph-mCD40/Eco+/GemT. Amino acid substitutions in themCD40 cytoplasmic domain were generated using comple-mentary primers with the desired base changes and ph-mCD40/Eco+/GemT as a template using the QuikChangesite-directed mutagenesis kit. Chimeric CD40 was sub-cloned using the EcoRI sites into vector pDOI-5 for expres-sion under an MHC class II promoter [21]. All plasmid con-structs were verified by automated DNA sequencing.

4.3 TNF ELISA

Each clonal line was cultured at 105 cells/well in a 96-wellplate in the presence or absence of soluble human or murineCD154 for 48 h. Culture supernatants were transferred toMaxi-sorp plates (Nalge-Nunc, Rochester, NY) coated andassayed according to TNF- § Duoset® ELISA kit, purchasedfrom R&D Systems (Minneapolis, MN).

4.4 Antibodies and reagents

The mAb against dual-phosphorylated I ‹ B § was purchasedfrom Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Poly-clonal rabbit antibodies against the phosphorylated form ofJNK were purchased from Promega. Polyclonal rabbit anti-bodies against the phosphorylated form of MAPK p38 werepurchased from New England Biolabs (Beverly, MA). mAbS2C6 against human CD40 was the kind gift of S. Paulie(Stockholm University, Sweden). Antibodies against CD23and CD80 were purchased from BD Pharmingen (San Diego,CA). Primary antibodies were detected with goat anti-rabbitIg horseradish peroxidase (Bio-Rad, Hercules, CA) or strep-tavidin conjugated to phycoerythrin (PharMingen, TorreyPines, CA). Soluble human CD154 fusion protein was pre-pared as described previously [13].

4.5 Western blots

Protein samples were separated by 12% PAGE with SDS.Proteins were transferred to nitrocellulose and membraneswere blocked with 1% BSA and/or 5% Milk in Tris-bufferedsaline with 0.05% Tween-20. Primary antibody incubationswere carried out over 2 h at room temperature in blockingbuffers recommended by antibody suppliers. After washing,secondary antibody incubations were carried out over30 min in the same blocking buffers. Blots were developedwith SuperSignal® West Pico chemiluminescent substrate

(Pierce, Rockford, IL). Membranes were allowed to exposeX-ray films, and autoradiographs were developed.

4.6 Internalization assays

M12 clones from exponentially growing cultures were har-vested and resuspended at 4°C in staining medium con-taining approximately 2 ? g/ml soluble human CD154-CD8fusion protein, and incubated for 30 mins. Excess CD154was washed away, and cells were resuspended at 37°C incomplete RPMI for various times. Endocytosis was haltedby fixation in 1% formaldehyde at 4°C. Extracellular CD154was detected with anti-CD8-phycoerythrin (PharMingen).Flow cytometry was performed on a Becton DickinsonFACScan.

Acknowledgements: We thank D. Everdeen for engineer-ing the chimeric CD40 template used for making all CD40constructs. We also thank Cory Ahonen for subcloning themutant constructs. This study was supported by grantAI26296 and AI42234 to R.J.N.

References

1 van Kooten, C. and Banchereau, J., CD40-CD40 ligand. J. Leu-koc. Biol. 2000. 67: 2–17.

2 Foy, T. M., Shepherd, D. M., Durie, F. H., Aruffo, A., Ledbetter,J. A. and Noelle, R. J., In vivo CD40-gp39 interactions areessential for thymus-dependent humoral immunity. II. Prolongedsuppression of the humoral immune response by an antibody tothe ligand for CD40, gp39. J. Exp. Med. 1993. 178: 1567–1575.

3 Gray, D., Dullforce, P. and Jainandunsing, S., Memory B celldevelopment but not germinal center formation is impaired byin vivo blockade of CD40-CD40 ligand interaction. J. Exp. Med.1994. 180: 141–155.

4 Hu, H. M., O’Rourke, K., Boguski, M. S. and Dixit, V. M., Anovel RING finger protein interacts with the cytoplasmic domainof CD40. J. Biol. Chem. 1994. 269: 30069–30072.

5 Rothe, M., Sarma, V., Dixit, V. M. and Goeddel, D. V., TRAF2-mediated activation of NF-kappa B by TNF receptor 2 and CD40.Science 1995. 269: 1424–1427.

6 Ishida, T., Mizushima, S., Azuma, S., Kobayashi, N., Tojo, T.,Suzuki, K., Aizawa, S., Watanabe, T., Mosialos, G., Kieff, E.,Yamamoto, T. and Inoue, J., Identification of TRAF6, a noveltumor necrosis factor receptor-associated factor protein thatmediates signaling from an amino-terminal domain of the CD40cytoplasmic region. J. Biol. Chem. 1996. 271: 28745–28748.

7 Ishida, T. K., Tojo, T., Aoki, T., Kobayashi, N., Ohishi, T., Wata-nabe, T., Yamamoto, T. and Inoue, J., TRAF5, a novel tumornecrosis factor receptor-associated factor family protein, medi-ates CD40 signaling. Proc. Natl. Acad. Sci. USA 1996. 93:9437–9442.

8 Pullen, S. S., Miller, H. G., Everdeen, D. S., Dang, T. T., Crute,J. J. and Kehry, M. R., CD40-tumor necrosis factor receptor-associated factor (TRAF) interactions: regulation of CD40 signal-ing through multiple TRAF binding sites and TRAF hetero-oligomerization. Biochemistry 1998. 37: 11836–11845.


9 Grammer, A. C. and Lipsky, P. E., CD40-mediated regulation ofimmune responses by TRAF-dependent and TRAF- independentsignaling mechanisms. Adv. Immunol. 2000. 76: 61–178.

10 Grammer, A. C., Swantek, J. L., McFarland, R. D., Miura, Y.,Geppert, T. and Lipsky, P. E., TNF receptor-associated factor-3signaling mediates activation of p38 and Jun N-terminal kinase,cytokine secretion, and Ig production following ligation of CD40on human B cells. J. Immunol. 1998. 161: 1183–1193.

11 Nakano, H., Oshima, H., Chung, W., Williams-Abbott, L.,Ware, C. F., Yagita, H. and Okumura, K., TRAF5, an activator ofNF-kappaB and putative signal transducer for the lymphotoxin-beta receptor. J. Biol. Chem. 1996. 271: 14661–14664.

12 Song, H. Y., Regnier, C. H., Kirschning, C. J., Goeddel, D. V.and Rothe, M., Tumor necrosis factor (TNF)-mediated kinasecascades: bifurcation of nuclear factor-kappaB and c-jun N-terminal kinase (JNK/SAPK) pathways at TNF receptor-associated factor 2. Proc. Natl. Acad. Sci. USA 1997. 94:9792–9796.

13 Pullen, S. S., Dang, T. T., Crute, J. J. and Kehry, M. R., CD40signaling through tumor necrosis factor receptor-associated fac-tors (TRAFs). Binding site specificity and activation of down-stream pathways by distinct TRAFs. J. Biol. Chem. 1999. 274:14246–14254.

14 Jalukar, S. V., Hostager, B. S. and Bishop, G. A., Characteriza-tion of the roles of TNF receptor-associated factor 6 in CD40-mediated B lymphocyte effector functions. J. Immunol. 2000.164: 623–630.

15 Goldstein, M. D. and Watts, T. H., Identification of distinctdomains in CD40 involved in B7–1 induction or growth inhibition.J. Immunol. 1996. 157: 2837–2843.

16 Hostager, B. S., Hsing, Y., Harms, D. E. and Bishop, G. A., Dif-ferent CD40-mediated signaling events require distinct CD40structural features. J. Immunol. 1996. 157: 1047–1053.

17 Hsing, Y., Hostager, B. S. and Bishop, G. A., Characterization ofCD40 signaling determinants regulating nuclear factor-kappa Bactivation in B lymphocytes. J. Immunol. 1997. 159: 4898–4906.

18 Sutherland, C. L., Krebs, D. L. and Gold, M. R., An 11-aminoacid sequence in the cytoplasmic domain of CD40 is sufficient foractivation of c-Jun N-terminal kinase, activation of MAPKAPkinase-2, phosphorylation of I kappa B alpha, and protection ofWEHI-231 cells from anti-IgM-induced growth arrest. J. Immunol.1999. 162: 4720–4730.

19 Leo, E., Welsh, K., Matsuzawa, S., Zapata, J. M., Kitada, S.,Mitchell, R. S., Ely, K. R. and Reed, J. C., Differential require-ments for tumor necrosis factor receptor-associated factor familyproteins in CD40-mediated induction of NF-kappaB and JunN-terminal kinase activation. J. Biol. Chem. 1999. 274:22414–22422.

20 Kosaka, Y., Calderhead, D. M., Manning, E. M., Hambor, J. E.,Black, A., Geleziunas, R., Marcu, K. B. and Noelle, R. J., Acti-vation and regulation of the IkappaB kinase in human B cells byCD40 signaling. Eur. J. Immunol. 1999. 29: 1353–1362.

21 Kouskoff, V., Fehling, H. J., Lemeur, M., Benoist, C. andMathis, D., A vector driving the expression of foreign cDNAs inthe MHC class II-positive cells of transgenic mice. J. Immunol.Methods 1993. 166: 287–291.

22 Sakata, N., Patel, H. R., Terada, N., Aruffo, A., Johnson, G. L.and Gelfand, E. W., Selective activation of c-Jun kinasemitogen-activated protein kinase by CD40 on human B cells.J. Biol. Chem. 1995. 270: 30823–30828.

23 Lee, S. Y., Reichlin, A., Santana, A., Sokol, K. A., Nussen-zweig, M. C. and Choi, Y., TRAF2 is essential for JNK but not

NF-kappaB activation and regulates lymphocyte proliferationand survival. Immunity 1997. 7: 703–713.

24 Craxton, A., Shu, G., Graves, J. D., Saklatvala, J., Krebs, E. G.and Clark, E. A., p38 MAPK is required for CD40-induced geneexpression and proliferation in B lymphocytes. J. Immunol. 1998.161: 3225–3236.

25 Salmon, R. A., Foltz, I. N., Young, P. R. and Schrader, J. W.,The p38 mitogen-activated protein kinase is activated by ligationof the T or B lymphocyte antigen receptors, Fas or CD40, butsuppression of kinase activity does not inhibit apoptosis inducedby antigen receptors. J. Immunol. 1997. 159: 5309–5317.

26 Aicher, A., Shu, G. L., Magaletti, D., Mulvania, T., Pezzutto, A.,Craxton, A. and Clark, E. A., Differential role for p38 mitogen-activated protein kinase in regulating CD40-induced geneexpression in dendritic cells and B cells. J. Immunol. 1999. 163:5786–5795.

27 Pullen, S. S., Labadia, M. E., Ingraham, R. H., McWhirter, S.M., Everdeen, D. S., Alber, T., Crute, J. J. and Kehry, M. R.,High-affinity interactions of tumor necrosis factor receptor-associated factors (TRAFs) and CD40 require TRAF trimerizationand CD40 multimerization. Biochemistry 1999. 38: 10168–10177.

28 Nguyen, L. T., Duncan, G. S., Mirtsos, C., Ng, M., Speiser, D.E., Shahinian, A., Marino, M. W., Mak, T. W., Ohashi, P. S. andYeh, W. C., TRAF2 deficiency results in hyperactivity of certainTNFR1 signals and impairment of CD40-mediated responses.Immunity 1999. 11: 379–389.

29 Lomaga, M. A., Yeh, W. C., Sarosi, I., Duncan, G. S., Furlonger,C., Ho, A., Morony, S., Capparelli, C., Van, G., Kaufman, S.,van der Heiden, A., Itie, A., Wakeham, A., Khoo, W., Sasaki, T.,Cao, Z., Penninger, J. M., Paige, C. J., Lacey, D. L., Dunstan,C. R., Boyle, W. J., Goeddel, D. V. and Mak, T. W., TRAF6 defi-ciency results in osteopetrosis and defective interleukin-1, CD40,and LPS signaling. Genes Dev. 1999. 13: 1015–1024.

30 Yeh, W. C., Shahinian, A., Speiser, D., Kraunus, J., Billia, F.,Wakeham, A., de la Pompa, J. L., Ferrick, D., Hum, B., Iscove,N., Ohashi, P., Rothe, M., Goeddel, D. V. and Mak, T. W., Earlylethality, functional NF-kappaB activation, and increased sensi-tivity to TNF-induced cell death in TRAF2-deficient mice. Immu-nity 1997. 7: 715–725.

31 Xu, Y., Cheng, G. and Baltimore, D., Targeted disruption ofTRAF3 leads to postnatal lethality and defective T-dependentimmune responses. Immunity 1996. 5: 407–415.

32 Inui, S., Kaisho, T., Kikutani, H., Stamenkovic, I., Seed, B.,Clark, E. A. and Kishimoto, T., Identification of the intracyto-plasmic region essential for signal transduction through a Bcell activation molecule, CD40. Eur. J. Immunol. 1990. 20:1747–1753.

33 Boussiotis, V. A., Nadler, L. M., Strominger, J. L. and Goldfeld,A. E., Tumor necrosis factor alpha is an autocrine growth factorfor normal human B cells Proc. Natl. Acad. Sci. USA 1994. 91:7007–7011.

34 Hostager, B. S., Catlett, I. M. and Bishop, G. A., Recruitment ofCD40 and tumor necrosis factor receptor-associated factors 2and 3 to membrane microdomains during CD40 signaling. J. Biol.Chem. 2000. 275: 15392–15398.

35 Vidalain, P. O., Azocar, O., Servet-Delprat, C., Rabourdin-Combe, C., Gerlier, D. and Manie, S., CD40 signaling in humandendritic cells is initiated within membrane rafts. EMBO J. 2000.19: 3304–3313.

36 D’Hondt, K., Heese-Peck, A. and Riezman, H., Protein and lipidrequirements for endocytosis. Annu. Rev. Genet. 2000. 34:255–295.


37 Deckert, M., Ticchioni, M. and Bernard, A., Endocytosis ofGPI-anchored proteins in human lymphocytes: role of glycolipid-based domains, actin cytoskeleton, and protein kinases. J. CellBiol. 1996. 133: 791–799.

38 Feng, X., Gaeta, M. L., Madge, L. A., Yang, J. H., Bradley, J. R.and Pober, J. S., Caveolin-1 associates with TRAF2 to form acomplex that is recruited to TNF receptors J. Biol. Chem. 2001.276: 8341–8349.

39 Lu, H. T., Yang, D. D., Wysk, M., Gatti, E., Mellman, I., Davis, R.J. and Flavell, R. A., Defective IL-12 production in mitogen-activated protein (MAP) kinase kinase 3 (Mkk3)-deficient mice.EMBO J. 1999. 18: 1845–1857.

40 Hanissian, S. H. and Geha, R. S., Jak3 is associated with CD40and is critical for CD40 induction of gene expression in B cells.Immunity 1997. 6: 379–387.

41 Di Fiore, P. P. and Gill, G. N., Endocytosis and mitogenic signal-ing. Curr. Opin. Cell. Biol. 1999. 11: 483–488.

42 Ling, L. and Goeddel, D. V., MIP-T3, a novel protein linkingtumor necrosis factor receptor-associated factor 3 to the micro-tubule network. J. Biol. Chem. 2000. 275: 23852–23860.

43 Yang, R. B., Mark, M. R., Gurney, A. L. and Godowski, P. J.,Signaling events induced by lipopolysaccharide-activated toll-like receptor 2. J. Immunol. 1999. 163: 639–643.

44 Underhill, D. M., Ozinsky, A., Smith, K. D. and Aderem, A., Toll-like receptor-2 mediates mycobacteria-induced proinflammatorysignaling in macrophages. Proc. Natl. Acad. Sci. USA 1999. 96:14459–14463.

Correspondence: Randolph J. Noelle, Dartmouth MedicalSchool, Department of Microbiology, Lebanon, NH 03756,USAFax: +1-603-650-6223e-mail: rjn — dartmouth.edu

M. Kehry’s present address: Department of cell biology,IDEC Pharmaceuticals Corp., San Diego, CA 92191, USA


Documents

Cellular responses to murine CD40 in a mouse B cell line may be TRAF dependent or independent