Engagement of CD40 on murine B cells by its ligand CD154 induces the binding of TNFR-associated factors (TRAFs) 1, 2, 3, and 6, followed by the rapid degradation of TRAFs 2 and 3. TRAF degradation occurs in response to signaling by other TNFR superfamily members, and is likely to be a normal regulatory component of signaling by this receptor family. In this study, we found that receptor-induced TRAF degradation limits TRAF2-dependent CD40 signals to murine B cells. However, TRAFs 1 and 6 are not degraded in response to CD40 engagement, despite their association with CD40. To better understand the mechanisms underlying differential TRAF degradation, mixed protein domain TRAF chimeras were analyzed in murine B cells. Chimeras containing the TRAF2 zinc (Zn) domains induced effective degradation, if attached to a TRAF domain that binds to the PXQXT motif of CD40. However, the Zn domains of TRAF3 and TRAF6 could not induce degradation in response to CD40, regardless of the TRAF domains to which they were attached. Our data indicate that TRAF2 serves as the master regulator of TRAF degradation in response to CD40 signaling, and this function is dependent upon both the TRAF Zn domains and receptor binding position.
Amember of the TNFR superfamily, CD40 is expressed on various cells of the immune system, including B cells, monocytes, macrophages, dendritic cells, and T cells in certain situations (reviewed in Refs.1 and 2). CD40 signaling is required for normal T-dependent B cell proliferation, differentiation, isotype switching, cytokine secretion, and surface molecule up-regulation (1, 2). CD40 uses several different signaling pathways, including activation of various kinases and transcription factors (3).
The most proximal signaling components of CD40 are the TNFR-associated factors (TRAFs),3 which are adapter molecules for many members of the TNFR superfamily. There are six known TRAFs, and a seventh has recently been proposed (3, 4). CD40 associates with TRAFs 1, 2, 3, and 6 to mediate downstream signals (3). TRAF5 has been implicated in CD40 signals, but has not been shown to directly bind CD40 in B cells (5). TRAFs 1, 2, and 3 have an overlapping binding motif on CD40, while TRAF6 has a unique site closer to the cell membrane (6, 7).
TRAFs have distinct and coinciding functions in activating CD40-mediated signaling events. TRAF2 mediates JNK activation, regulates NF-κB2, and has overlapping functions with TRAF6 in NF-κB1 activation and CD80 up-regulation (8, 9). The role of TRAF1 has been unclear, but recent work indicates that TRAF1 may cooperate with TRAF2 in CD40 signaling to B cells (P. Xie, B. Hostager, M. Munroe, C. Moore, and G. Bishop, Cooperation between TNF receptor associated factors (TRAFs) 1 and 2 in CD40-mediated B lymphocyte activation, submitted for publication). TRAF3 is a negative regulator of CD40 signaling, but may in certain settings activate JNK, and has recently been implicated in also regulating the NF-κB1 and NF-κB2 pathways (10, 11).
The many confusing reports of overlapping TRAF function could be partially explained by the structural homology of the TRAFs (12). This family of adaptors has a C-terminal TRAF domain used for receptor binding and a coiled-coiled domain involved in homo- and heterotrimerization (3). TRAFs 2–6 also have zinc (Zn) binding domains composed of five to seven Zn fingers and a Zn RING domain (3). The Zn domains are implicated in NF-κB and JNK activation (13, 14, 15). The Zn RING domains of TRAF2 and TRAF6 have also been linked to the ubiquitination pathway (15, 16, 17). TRAF1 lacks a Zn RING domain and has only one predicted Zn finger domain, so it may function principally by heterotrimerization with other TRAFs.
We have shown that TRAFs 2 and 3 are rapidly degraded after ligand-induced association with CD40 (17, 18). TRAF degradation may be a common regulatory mechanism used by TNFR superfamily members, as CD30, TNFRII/CD120b, B cell-activating factor receptor, and receptor activator of NF-κB (RANK) have all been reported to induce degradation of various TRAFs that bind to them (19, 20, 21, 22). It has been proposed that TRAF degradation is a method for down-regulating a receptor signal cascade, and that inability to initiate TRAF degradation could contribute to the transforming properties of latent membrane protein 1, an oncogenic viral mimic of CD40 produced by the EBV (18, 19, 23)
The degradation of TRAF2 following association with CD40 is dependent upon polyubiquitination, targeting to the proteasome, and an intact TRAF2 Zn RING domain (17). Lysosome-dependent proteolysis may also play a role (18). Interestingly, the CD40-mediated degradation of TRAF3 is inhibited in B cells that lack TRAF2 expression (8, 9). The re-expression of full-length TRAF2 in these cells can restore TRAF3 degradation, but it cannot be restored by a mutant TRAF2 lacking a Zn domain (9). As this TRAF2 mutant can act as a dominant-negative signaling molecule, it is not clear whether its failure to restore TRAF3 degradation is a direct or indirect effect. TRAF degradation could be TRAF and/or receptor specific as degradation of TRAF1 or TRAF6 has not been seen in response to CD40, although the degradation of these TRAFs has been reported in response to other receptors, in other cell types (17, 19, 24). Therefore, differential TRAF degradation mediated by different receptors could contribute to the diversity of signals delivered by members of the TNFR family.
The present study focused on understanding how TRAF structure controls whether or not association with CD40 induces degradation. Our model system was designed to test the hypothesis that CD40-induced degradation is controlled by specific TRAF Zn binding domains, and that the position of the TRAF domain binding site on CD40 might also influence degradation. Chimeric TRAF proteins were created to test this hypothesis. These molecules were stably and inducibly expressed in either wild-type (WT) or TRAF-deficient B cell lines, to assess the contribution of endogenous TRAFs to CD40-mediated degradation. Results show that the Zn binding region of TRAF2 is the master regulator of TRAF degradation in response to CD40 signaling, but must be attached to a TRAF that binds at the PXQXT TRAF-binding motif in the CD40 cytoplasmic domain. Additional experiments demonstrate that TRAF degradation circumscribes the intensity and duration of CD40 signals.
Materials and Methods
The mouse B cell lines A20.2J, A20.T2−/− (lacks TRAF2 expression), M12.hCD40, and M12.4.1 have been previously described (9, 25, 26, 27). B cell lines were maintained in RPMI 1640, 10 μM 2-ME (Invitrogen Life Technologies) with 10% heat-inactivated FCS (Atlanta Biologicals) and antibiotics (BCM-10). M12.4.1 was chosen to examine early TRAF1 degradation, as unstimulated A20.2J cells do not express endogenous TRAF1 levels detectable by Western blot (28). The M12.hCD40 cell line was maintained in 400 μg/ml G418 disulfate (Research Products). Transfected A20.2J B cell lines were maintained in 400 μg/ml hygromycin B (Calbiochem) and 600 μg/ml G418 disulfate. Hi-5 insect cells infected with WT baculovirus (WTBV) or baculovirus (BV)-expressing murine CD154 (mCD154) have been described previously (9). T-depleted splenocytes from 8- to 14-wk-old C57BL/6 mice were prepared, as described (29). The use of mice in these experiments was reviewed and approved by the University of Iowa Institutional Animal Care and Use Committee.
Rabbit anti-TRAF2 Ab and chicken anti-TRAF6 Ab (1B1-2) were purchased from Medical and Biological Laboratories. Rabbit anti-TRAF3 Ab (H122), rabbit anti-TRAF1 Ab (N-19), and rabbit anti-JNK1/2 Ab (FL) were purchased from Santa Cruz Biotechnology. Mouse anti-GAPDH Ab (6C5) was purchased from Abcam. Mouse anti-FLAG Ab (M2-peroxidase HRP conjugate) was purchased from Sigma-Aldrich. Rabbit anti-pIκBα, anti-IκBα, and anti-pJNK1/2 were purchased from Cell Signaling Technology. The goat anti-mouse IgG, goat anti-rabbit IgG, and donkey anti-chicken IgG secondary Abs were from Jackson ImmunoResearch Laboratories. For coimmunoprecipitation of the TRAF molecules with mouse CD40, we used the anti-mouse CD40 mAbs 1C10 and 4F11 (rat IgG2a) produced from hybridomas provided by F. Lund (Trudeau Institute, Saranac Lake, NY). The isotype control mAb EM95 (rat IgG2a) was provided by T. Waldschmidt (University of Iowa, Iowa City, IA). The anti-human CD40 mAb G28.5 (mouse IgG1) was prepared from a hybridoma obtained from the American Type Culture Collection, and the isotype control MOPC31c was purchased from Sigma-Aldrich.
The TRAF chimeric DNA constructs were created by PCR SOEing (30). The murine TRAFs used as templates have been described previously (17, 28, 31, 32). All chimeras contain a C-terminal 3× FLAG epitope tag and were cloned into the inducible expression vector pOPRSV1, which has been described previously (33). The expression vector contains a Lac repressor (LacR) binding site upstream of the cDNA inserts, which allows stable induced expression of the cDNA in the presence of isopropyl-β-d-thiogalactopyranoside (IPTG) (AMRESCO). Primers for the T2ZnT3 joint were 5′-cagattttggagaagaaggtttccctgct and 5′-ccttcttctccaaaatctggcaccgctgg. Primers for the T2ZnT1 joint were 5′-tccagcggtgccagattcttgagaagc and 5′-tgccagattcttgaggagaagctgcgt. Primers for the T2ZnT6 joint were 5′-tgcagcggctacgggaacacatgagactgttggcccag and 5′-ctgggccaacagtctcatgtgttcccgtagccgctgca. Primers for the T3ZnT2 joint were 5′-caactccctggagcagaagatagcaacctttg and 5′-tcttctgctccagggagttgctccactc. Primers for the T6ZnT2 joint were 5′-aagagaatacccagttgcacctagccctactgctgagc and 5′-gctcagcagtagggctaggtgcaactgggtattctctt. The T6ZnT2 has an AU1 tag at the N terminus of TRAF6; the primer for the tag was 5′-ataagattgcggccgcatggacacctatcgctacattatgagtctcttaaactgtgagaa.
Transfections and induced TRAF chimera expression
A20.2J and A20.T2−/− cells that stably express the bacterial LacR were transfected with 10 μg of linearized plasmid DNA by electroporation and selected with G418, as previously described (34). Hygromycin B is included in the medium to maintain expression of LacR. The growth-positive clones were treated with 100 μM IPTG overnight and screened by intracellular Ab staining and flow cytometry for chimeric TRAF molecule expression, using anti-FLAG-M2 mAb and FITC goat anti-mouse IgG, as previously described (17). This method was also used to induce expression of the transfected TRAF molecules in signaling experiments.
TRAF degradation assay
The TRAF degradation assay has been described previously (9, 17, 18). Briefly, 2 × 106 cells (5 × 106 for mouse splenic B cells) were washed in RPMI 1640, resuspended in 1 ml of BCM-10, and added to a 24-well tissue culture plate. The cells were stimulated with 4 × 105 (1 × 106 for mouse splenic B cells) Hi-5 insect cells infected with WTBV or BV-encoding mouse CD154. Plates were centrifuged for 1 min at 200 rpm to maximize cell to cell contact, then incubated for the indicated time periods at 37°C. Time courses for each chimera were performed. Although degradation is seen as early as 10 min, the 6-h time point was picked as optimal. Cells and medium were transferred into 1.5-ml Eppendorf tubes, and spun at 4°C for 2 min at 7000 rpm in a microcentrifuge. Whole cell lysates were prepared by removing the supernatant and adding 200 μl (70 μl to mouse splenic B samples) of 2× SDS-PAGE loading dye to the pellet. The pelleted cells were sonicated 15 pulses at 90% duty cycle, output 1.5. The samples were boiled for 5 min at 95°C, then put on ice before gel loading. To measure the effect of CD40-mediated TRAF2 degradation on a subsequent CD40 signal, 3 × 106 M12.4.1 cells stably transfected with human CD40 were stimulated with 10 μg/ml isotype control (EM95) or anti-mCD40 (1C10) mAbs at 37°C for 1 h in 6 ml of prewarmed BCM-10. Cells were washed twice in prewarmed BCM-10, resuspended in 6 ml of BCM-10, and aliquoted into 1.5-ml Eppendorf tubes in 1-ml fractions. Cells were rested for 1 h in a 37°C water bath and then stimulated for the indicated time course with 10 μg/ml prewarmed anti-human CD40 (G28.5), isotype control (MOPC31c) mAbs, or medium alone. Whole cell lysates were prepared, as described above.
JNK phosphorylation assay
A total of 5 × 105 cells were washed, resuspended in 1 ml of BCM-10 in 1.5-ml Eppendorf tubes, and rested for 1 h in a 37°C water bath. The cells were then stimulated for 5, 10, 30, or 60 min with anti-mCD40 (1C10), isotype control (EM95) mAbs, or medium alone. Whole cell lysates were prepared, as described above.
A total of 5–10 μl of sample were resolved on 10% SDS-PAGE. The proteins were transferred to Immobilon-P membranes (Millipore), and membranes were blocked with 10% nonfat dried milk in TBST for 1 h. The membranes were washed three times in TBST and incubated for 1 h with the anti-TRAF3 and anti-GAPDH Abs, 2 h with the anti-TRAF2 and anti-FLAG-HRP Abs, and 4°C overnight with the anti-TRAF1, anti-TRAF6 Abs, anti-pIκBα, anti-IκBα, anti-pJNK1/2, and anti-JNK1/2 Abs. The blots were incubated with secondary Abs for 1 h and developed with an ECL system (Supersignal West Pico; Pierce Biotechnology). The amount of TRAF degradation seen shows clonal variation, and is typically between 25 and 50%. It can be difficult for the human eye to compare and quantify these changes in a Western blot accurately. Therefore, to accurately compare and quantify the amount of degradation seen in these experiments, the Western blot chemiluminescence was measured on a low-light digital camera (LAS-1000 or LAS-3000; Fujifilm Medical Systems), using the Image Gauge program (Fujifilm Medical Systems).
A total of 5 × 106 B cells was stimulated with 2.5 × 106 Hi-5 cells infected with WTBV or mCD154-BV in 1 ml of BCM-10 in a 1.5-ml Eppendorf tube for the time points indicated in figure legends. Coimmunoprecipitation of TRAFs and CD40 from raft fractions was performed, as previously described (28).
CD40-induced degradation of specific TRAFs associating with CD40 in B cells
We previously observed that engagement of CD40 by CD154 in B cell lines induces rapid association, followed by degradation of TRAFs 2 and 3 (9, 17, 18, 35). However, it was not clear whether TRAF1 or TRAF6, which also binds to CD40 upon its engagement, undergoes CD40-mediated degradation. We found that TRAFs 2 and 3, but not TRAF6, are degraded in response to CD40, in normal splenic B cells and in the mouse B cell line M12.4.1 (Fig. 1, A and B). No degradation of endogenous TRAF1 in M12.4.1 cells at early time points was evident (Fig. 1, B and D). TRAF1 expression is very low in splenic B cells and B cell lines until markedly increased upon receptor engagement, so it is not possible to examine degradation of endogenous TRAF1 in splenic B cells or at later time points (Fig. 1, B and D) (36). To circumvent this problem, FLAG-tagged TRAF1, which is induced by addition of IPTG and not by an endogenous promoter, was stably transfected into A20.2J B cells, which were then stimulated with mCD154 (Fig. 1,C). As with endogenous TRAF1, FLAG-tagged TRAF1 was not degraded following stimulation; this pattern was unchanged up to 6 h after stimulation (Fig. 1,C). Neither transfected FLAG-tagged TRAF6 nor endogenous TRAF6 was degraded in response to CD40 stimulation for up to 6 h (our unpublished observations) (Fig. 1 C).
TRAF degradation alters CD40-mediated TRAF-dependent signaling in B cells
We hypothesized that differential TRAF degradation induced by an initial burst of CD40 signaling can limit continued or subsequent CD40-mediated activation. To test this hypothesis, we stimulated mouse B cells through endogenous CD40 and then restimulated them at a later time point through a transfected human CD40 (Fig. 2,A), to enable clear separation of the signals and preclude receptor modulation. B cells receiving a second CD40 signal 2 h after the initial signal exhibited a dramatic deficiency in phosphorylation of JNK1 and JNK2 compared with the cells receiving only one stimulus (Fig. 2 A).
CD40-induced IκBα phosphorylation and degradation are partially TRAF2 dependent (8, 9), so the effect on the NF-κB1 pathway of decreased levels of TRAF2 after sequential CD40 signals was also measured. There was a substantial increase in pIκBα after the second signal, but there was no change in the ratio of pIκBα in relation to the CD40-induced increased total levels of IκBα (Fig. 2,B). Thus, IκBα phosphorylation was less sensitive to decreases in TRAFs 2 and 3 following initial CD40 signaling. However, the rate of IκBα degradation was reduced after the second stimulation (Fig. 2 C). Thus, even a partial decrease in TRAF2 levels can markedly affect TRAF2-dependent CD40-mediated signals. We next addressed how the structural elements of different TRAFs regulate degradation and signaling.
A model system for structure-function analysis of TRAF signaling in B cells
Our previous studies have implied a requirement for the TRAF2 Zn binding domain in TRAF2 and TRAF3 degradation (9). However, it was unknown whether other structural elements of TRAF2 are also required, which would predict that other TRAF Zn domains could initiate degradation if attached to TRAF2 coiled-coil and TRAF domains. To address these questions, we created chimeric TRAF molecules that exchange the Zn binding domain of TRAF2 with that of TRAF 1, 3, or 6. We chose to exchange the entire Zn binding domains because the Zn fingers have been shown to influence Zn-RING-mediated JNK activation and could be important in degradation (15, 24, 37). The chimeric TRAF constructs were stably and inducibly expressed in the mouse A20.T2−/− B cell line to investigate the influence of the transfected TRAF molecules in the absence of an endogenous TRAF2 molecule. WT A20.2J cells were also transfected as a control. Because we saw the same degradation pattern in both normal mouse B cells and mouse B cell lines (Fig. 1) (17, 18), these cell lines provide an informative model for TRAF degradation experiments. This system also avoids the problems inherent in overexpression studies, as the chimeric TRAFs are stably and inducibly expressed at levels similar to those of endogenous TRAFs (Figs. 3, 4, and 5).
Role of the TRAF2 Zn binding domain in control of TRAF3 degradation
TRAF2-deficient B cells show defective CD40-induced TRAF3 degradation, and restored expression of TRAF2 molecules lacking the Zn binding domain cannot restore normal TRAF3 degradation (8, 9). Therefore, to determine whether the Zn binding domain of TRAF2 is not only necessary, but also sufficient to control CD40-induced TRAF degradation, we created a chimera with the Zn binding region of TRAF2 and the TRAF region of TRAF3 (T2ZnT3) (Fig. 3,A). Coimmunoprecipitation showed that the T2ZnT3 chimera bound to CD40 following CD154 stimulation of B cells (Fig. 3,D). After stimulation of B cells with membrane-bound CD154, the T2ZnT3 chimera was degraded in WT A20 and A20.T2−/− cells (Fig. 3, B and C). This indicates that this chimera was capable of inducing its own degradation and was not dependent upon endogenous TRAF2. The degradation of endogenous TRAF2 and TRAF3 was normal in the WT A20 cells in the presence of the chimera. However, as seen previously, the degradation of TRAF3 was defective in the A20.T2−/− cells (9). Interestingly, normal levels of CD40-mediated degradation of TRAF3 were restored by the expression of T2ZnT3 (Fig. 3 C). This indicates that the TRAF2 Zn binding domain is required for the optimal degradation of TRAF3 and there is not a specific requirement for the TRAF2 coiled-coil and TRAF domains of TRAF2.
Previous reports have shown that A20.T2−/− cells have a defect in JNK activation and that this function requires the Zn domains of TRAF2 (9, 15). Although the T2ZnT3 chimera was functional in mediating degradation, it could not substitute for full-length TRAF2 in restoring normal levels of JNK activation, as shown in Fig. 1,E, in TRAF2-deficient B cells (Fig. 3 E). CD40-mediated JNK activation may thus require additional TRAF2 structural elements. Alternatively, the negative regulatory role of TRAF3 in CD40 signaling may be mediated by the TRAF domain of TRAF3, and thus retained by the chimera (9, 38).
Influence of the TRAF2 Zn binding domain on TRAF1 and TRAF6 degradation
TRAF1 was not degraded in response to CD40 signaling in B cells (Fig. 1, B and C). This highlights the differential regulation of TRAFs by different receptors and cell types, as TRAF1 can be degraded, in a TRAF2-dependent manner, in response to CD30 signals in epithelial cells (19). The absence of degradation in response to CD40 could be due to the lack of a full Zn binding domain in TRAF1 or binding differences among receptors. In vitro studies have suggested that the direct binding of TRAF1 to CD40 is very weak and that TRAF1 is primarily recruited to CD40 by heterodimerization with TRAF2 (6, 39). This low affinity could prevent TRAF1 degradation. To test whether the Zn binding domain of TRAF2 is sufficient to induce degradation of TRAF1 or whether the low affinity of the TRAF1 binding domain prevents its degradation, we created a chimera that has the Zn binding domain of TRAF2 connected to the TRAF domain of TRAF1 (T2ZnT1) (Fig. 4,A). Coimmunoprecipitation showed binding of the chimera to CD40 upon stimulation in both the presence and absence of TRAF2 (Fig. 4 D). We have also seen endogenous TRAF1 binding to CD40 in TRAF2−/− B cells (P. Xie, B. Hostager, M. Munroe, C. Moore, and G. Bishop, submitted for publication). These data indicate that, in the absence of TRAF2, TRAF1 may have increased opportunity to bind to CD40 or that the Zn binding domain of TRAF2 helps target T2ZnT1 to the plasma membrane (35).
The T2ZnT1 chimera was degraded in response to CD40 ligation in the A20 and A20.T2−/− cells (Fig. 4, B and C). This indicates that the presence of the TRAF2 Zn binding domain on the same molecule allows CD40 signals to induce degradation of the normally degradation-resistant TRAF1. This result confirms the hypothesis that it is the absence of a Zn binding domain that precludes TRAF1 degradation in response to CD40. The degradation of T2ZnT1 also indicates that binding affinity of TRAFs for CD40 does not greatly influence their degradation. As in Fig. 3,C, the degradation of TRAF3 was defective in the absence of TRAF2, and was restored by the presence of a TRAF2 Zn binding domain (Fig. 4 C).
CD40-induced JNK phosphorylation was observed following expression of the T2ZnT1 chimera in TRAF2-deficient B cells (Fig. 4,E). This chimeric molecule can thus restore TRAF2-mediated JNK activation to a level comparable to that seen in WT cells (Fig. 1,E). This suggests that the failure of the T2ZnT3 chimera to activate JNK (Fig. 3 E) may be due to negative effects of the TRAF3 TRAF domain.
TRAF6 is not degraded in response to CD40 signaling to B cells (Fig. 1). However, TRAF6 can be degraded in macrophages in response to RANK signals, and TRAF6 mediates its own K63-linked ubiquitination in the TLR pathway (16, 22, 40). To test whether the TRAF2 Zn domain could induce CD40-mediated degradation of the TRAF domain of TRAF6 as it did for TRAF1, we created a chimera that has the Zn binding region of TRAF2 connected to the TRAF domain of TRAF6 (T2ZnT6) (Fig. 5,A). Coimmunoprecipitation showed binding of the chimera to CD40 upon stimulation (Fig. 5,D). The T2ZnT6 chimera was also degraded in response to CD40 signals in the A20 or A20.T2−/− cells (Fig. 5, B and C). However, this degradation was lower than that of endogenous TRAF2 and TRAF3, suggesting that there may be other factors inhibiting the degradation of TRAF6 in response to this signal that are only partially overcome by the presence of the TRAF2 Zn binding domain. The degradation of TRAF3 was not restored by the presence of a TRAF2 Zn binding domain in the context of the TRAF6 binding domain (Fig. 5,C). As the T2ZnT6 chimera binds at the membrane-proximal TRAF6 binding site, this indicates that the TRAF2 Zn binding domain might need to directly interact with TRAF3 at the TRAF1/2/3 PXQXT binding site to enhance its degradation. As with the T2ZnT1 chimera, restoration of JNK phosphorylation was observed upon induced expression of T2ZnT6 in TRAF2-deficient B cells (Fig. 5 E).
The degradation of the chimeras containing a TRAF2 Zn binding domain is not due to ubiquitination of the Zn-RING domain, as mutation of the lysines in the RING does not affect CD40-mediated degradation (17). Thus, the TRAF1, 3, and 6 domains could support CD40-induced degradation when connected to the TRAF2 Zn binding domain. These results reveal that the TRAF2 Zn binding domain is sufficient to induce degradation of TRAF molecules that associate with CD40 following ligand binding, but this degradation is only optimal if the TRAF binds at or near the PXQXT motif in CD40. These data also indicate that a TRAF2 Zn binding domain can restore JNK signaling, when expressed in the context of coiled-coil and TRAF domains of TRAF molecules that do not inhibit JNK activation.
Function of Zn binding domains of TRAFs 3 and 6 in CD40-induced TRAF degradation in B cells
Fig. 5 showed that the T2ZnT6 molecule was more effective in inducing its own degradation than that of WT TRAF6. This could be accounted for by differences in the Zn binding domains, or a less favorable environment for degradation at the TRAF6 membrane-proximal binding site, which a TRAF2 Zn domain could at least partially overcome. Although T2ZnT6 was unable to restore the degradation of TRAF3, it is possible that the small, but reproducible amount of TRAF3 degradation induced by CD40 signals in the A20.T2−/− cells could be caused by endogenous TRAF6. To examine these possibilities, we created a chimera with the Zn binding region of TRAF6 connected to the TRAF domain of TRAF2 (T6ZnT2) (Fig. 6,A). Coimmunoprecipitation showed binding of the chimera to CD40 upon stimulation (Fig. 6,D). The chimera was not degraded following CD40 signals in either the A20 or the A20.T2−/− cell lines (Fig. 6, B and C). It also could not restore CD40-induced degradation of TRAF3 in the A20.T2−/− cells (Fig. 6 C). This indicates that the Zn binding domain of TRAF6 cannot promote TRAF degradation in response to CD40 signaling in B cells. Consistent with this finding, CD40-mediated degradation of TRAFs 2 and 3 is normal in TRAF6-deficient B cells (our unpublished observation).
As the TRAF6 Zn binding domain could not enhance TRAF3 degradation, we next tested the possibility that the TRAF3 Zn binding domain could partially mediate its own degradation, by creating a chimera with the Zn binding region of TRAF3 and the TRAF region of TRAF2 (T3ZnT2) (Fig. 7,A). The T3ZnT2 chimera was degraded weakly in the presence of endogenous TRAF2, and not at all in the A20.T2−/− cells (Fig. 7, B and C). The degradation of endogenous TRAF3 was not restored by the presence of the T3ZnT2 chimera (Fig. 7 C). This indicates that the TRAF3 Zn binding domain was not effective in inducing degradation in response to CD40 engagement.
The findings of this study show that TRAF degradation plays a significant biological role in regulation of CD40 signaling pathways. The Zn binding domain of TRAF2 serves as a master regulator of the degradation of TRAFs in response to CD40 signals in B lymphocytes, but the receptor-binding position of the TRAF domain also influences the effectiveness of degradation. The Zn binding domains of neither TRAF3 nor TRAF6 were able to substitute for that of TRAF2 in this function. Therefore, the minimal requirements for TRAF degradation in response to CD40 signaling in B cells are a TRAF2 Zn binding domain and the ability to associate with CD40 at or near the PXQXT motif in its cytoplasmic domain (summarized in Table I).
|Chimeric TRAF Molecule .||Zn Binding Domains .||Coiled-Coil/TRAF Domains .||CD40 Binding Site .||Induced Self Degradation .||Induced TRAF3 Degradation .||JNK Activation .|
|T3ZnT2||TRAF3||TRAF2||TRAF1/2/3||No (dependent on TRAF2)||No||No|
|Chimeric TRAF Molecule .||Zn Binding Domains .||Coiled-Coil/TRAF Domains .||CD40 Binding Site .||Induced Self Degradation .||Induced TRAF3 Degradation .||JNK Activation .|
|T3ZnT2||TRAF3||TRAF2||TRAF1/2/3||No (dependent on TRAF2)||No||No|
Although it is clear that the Zn binding domain of TRAF2 controls TRAF degradation in response to CD40, it is not clear how this process is controlled. The degradation of TRAF2 in response to CD120b signals has been reported to be mediated by the binding of TRAF2 to the E3 ubiquitin ligase cellular inhibitor of apoptosis 1 (20). Although cellular inhibitor of apoptosis 1 and 2 have been reported to be recruited to CD40 when overexpressed in epithelial cells (41), we have been unable to reproduce this finding using endogenous CD40 in B cells (our unpublished observations). Seven in absentia homologue 2 (Siah2), another E3 ubiquitin ligase, is implicated in TRAF2 degradation during stress-induced cell death (42). However, Siah2 has been reported to mediate degradation of several other proteins as well, and Siah2−/− mice do not have any reported alterations in TRAF-mediated signaling (43, 44, 45, 46). Therefore, the Siah2-mediated degradation of TRAF2 may be unrelated to CD40-mediated TRAF degradation.
It is intriguing that the TRAF2 Zn domain also regulates the CD40-induced degradation of TRAF3, and this has interesting biological implications. TRAF3 has been shown to have a number of negative regulatory effects on CD40 signaling, including decreased JNK activation and IgM secretion, and inhibition of CD40-BCR synergy (28, 38, 47). TRAF2 may eliminate the negative effect of TRAF3 on CD40 signaling by promoting the degradation of TRAF3 while also controlling its own signaling cascades by self-induced degradation. Although the degradation of TRAF3 in response to CD40 is decreased in the absence of TRAF2, residual degradation of TRAF3 is still detected (Figs. 3–7) (9). As the TRAF3 Zn binding domain is unable to mediate degradation (Fig. 7), this suggests that TRAF3 may recruit another protein via its coiled-coil/TRAF domain to induce its degradation.
TRAF6 was unable to mediate degradation in response to CD40 in B cells despite the demonstration that RANK induces degradation of TRAF6 in a mouse macrophage cell line (16, 22). TRAF6 degradation was neither blocked by physical factors inherent in its binding site on CD40 (as the T2ZnT6 chimera was degraded), nor was the TRAF6 Zn domain capable of mediating self or TRAF3 degradation in response to CD40 (Figs. 5 and 6). This illustrates that although TNFR superfamily members bind overlapping sets of TRAF molecules, they regulate TRAF-mediated signal cascades in different ways.
Receptor-induced protein degradation is a general mechanism for controlling signaling cascades and cell functions, and thus important to understand. The results of this study indicate that only certain TRAFs are degraded in response to CD40 signals in B cells and that control of TRAF degradation is a combination of the properties of the TRAF itself and the receptor that it binds. It seems likely that TRAF degradation is receptor and cell type specific, a possibility that we are currently examining. This implies that members of the TNFR superfamily use differential TRAF degradation as a regulatory mechanism for altering and limiting specific signals, and thus enhance the signaling outcomes available to cells.
We are grateful to Drs. Bruce Hostager and Frederick Quelle for critical review of the manuscript, and to Karen Larison for expert technical assistance.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by National Institutes of Health Grants AI28847, AI49993, and CA099997, and a Department of Veterans Affairs career award to G.A.B. C.R.M. received support from T32-HL 07638-18.
Abbreviations used in this paper: TRAF, TNFR-associated factor; BV, baculovirus; IPTG, isopropyl-β-d-thiogalactopyranoside; LacR, Lac repressor; mCD, murine CD; RANK, receptor activator of NF-κB; Siah2, seven in absentia homologue 2; WT, wild type; WTBV, wild-type baculovirus.