Introduction

The adaptive immune response is initiated upon specific recognition of antigens presented on the surface of antigen-presenting cells (APC) by T lymphocytes. Productive activation and clonal expansion of T cells requires the concerted action of the T-cell receptor (TCR) and the CD28 co-stimulatory receptor. TCR/CD28 co-engagement induces the formation of the immunological synapse at the contact site of T cell and APC through clustering of lipid rafts, receptor molecules and cytosolic signaling mediators (van der Merwe, 2002). TCR/CD28-initiated signaling networks ultimately promote the activation of several transcription factors including NFAT, AP-1 and NF-B (Okamura and Rao, 2001).

NF-B plays a key role for the regulation of T-cell activation by mediating the induction of various genes that control T-cell proliferation, activation and survival (Ghosh et al, 1998). The NF-B transcription factor family comprises the Rel proteins p65 (RelA), RelB, c-Rel, NF-B1 (p105/p50) and NF-B2 (p100/p52), which can form various combinations of homo- and heterodimers. In unstimulated cells, NF-B proteins are retained in the cytoplasm by their tight association with inhibitory (IB) proteins (Hayden and Ghosh, 2004). TCR/CD28 co-ligation induces canonical NF-B signaling, which involves phosphorylation and degradation of small cytosolic IB inhibitors (IB, IB and IB), and subsequent nuclear translocation and DNA binding of NF-B. IB phosphorylation is catalyzed by the IB kinase (IKK) complex, which consists of the two catalytic subunits IKK and IKK and the essential regulatory subunit IKK/NEMO. Hence, an elucidation of signaling mechanisms that control IKK activation is critical for understanding the regulatory events involved in T-cell activation (Rawlings et al, 2006; Schulze-Luehrmann and Ghosh, 2006; Weil and Israel, 2006).

TCR/CD28 stimulation induces receptor proximal tyrosine phosphorylation, followed by an association of phospholipase C-, small G proteins (e.g., Rac and Ras) and guanine nucleotide exchange factors (e.g., SOS and Vav) to the immunological synapse (Weil and Israel, 2006). The protein kinase C (PKC) isoform PKC is recruited to the immunological synapse upon T-cell activation (Monks et al, 1997), and was found to be indispensable for TCR triggered NF-B activation (Sun et al, 2000; Bi et al, 2001). Genetic ablations in mice have revealed key molecules that couple PKC activation to IKK/NF-B signaling. These include Carma1 (CARD11), Bcl10, Malt1 and Caspase8 (Ruland et al, 2001, 2003; Egawa et al, 2003; Hara et al, 2003; Jun et al, 2003; Ruefli-Brasse et al, 2003; Su et al, 2005). The CARD-containing Carma1 protein is a member of the MAGUK (membrane-associated guanylate kinase) family of proteins (McAllister-Lucas et al, 2001). Carma1 is associated with the plasma membrane and recruited to the immunological synapse upon TCR/APC contact (Gaide et al, 2002). Bcl10 and Malt1, which were originally cloned from translocation breakpoints associated with malignant MALT (mucosa-associated lymphoid tissue) lymphomas (Akagi et al, 1999; Morgan et al, 1999; Willis et al, 1999; Zhang et al, 1999), are constitutively associated and recruited to Carma1 in a PKC-dependent manner (Matsumoto et al, 2005; Sommer et al, 2005). In T cells, PKC phosphorylates Carma1 in a central linker region, which triggers a conformational change of Carma1 and thereby promotes homotypic interaction between the N-terminal CARDs of Carma1 and Bcl10 and thus Carma1–Bcl10–Malt1 (CBM) complex formation (Matsumoto et al, 2005). PDK1 associates with PKC and Carma1 and might function as a molecular bridge to facilitate Carma1 phosphorylation (Lee et al, 2005).

The molecular mechanisms that trigger IKK activation downstream of the CBM complex on the route to NF-B are not fully understood. Carma1, Bcl10 and Malt1 were shown to promote NF-B activation by inducing IKK polyubiquitination (Zhou et al, 2004; Shambharkar et al, 2007). TAK1 (transforming growth factor--activated kinase 1) has been proposed to function as an IKK kinase in TCR signaling (Wang et al, 2001; Sun et al, 2004; Wan et al, 2006), and TAB2/3 contain ubiquitin-binding domains (UBDs) that can mediate recruitment of TAK1 to ubiquitinated IKK (Kanayama et al, 2004). Malt1 was suggested to possess an intrinsic E3 ligase activity that catalyzes IKK ubiquitination (Zhou et al, 2004). In addition, the RING ubiquitin ligase TRAF6 can either directly associate with the C-terminus of Malt1 via two conserved binding motifs (Sun et al, 2004), or is indirectly recruited to Malt1 through association with Caspase8 (Bidere et al, 2006). A role for TRAF6 in TCR-induced IKK activation is further supported by RNAi experiments (Sun et al, 2004; Bidere et al, 2006). However, T-cell-specific ablation of TRAF6 indicates that one or several other E3 ligases, for example, TRAF2, can compensate for the loss of TRAF6 (King et al, 2006). TRAF6 was shown to cooperate with the C-terminus of Malt1 to promote IKK ubiquitination (Sun et al, 2004), and might thus contribute ubiquitin ligase activity. However, abrogation of IKK ubiquitination caused only a partial reduction of PKC-dependent NF-B activation in T cells (Zhou et al, 2004), hinting that other targets for regulatory ubiquitination must be involved in directing CBM complex formation to IKK activation upon TCR/CD28 co-engagement.

We have identified Malt1 as a novel substrate for induced regulatory ubiquitination in response to TCR/CD28 co-ligation. Malt1 ubiquitin chains provide docking surfaces for the recruitment of IKK to the CBM complex. Congruently, NF-B signaling in T cells critically depends on an intact ubiquitin-binding motif in IKK. We further demonstrate that TRAF6 is an E3 ligase for Malt1 and induces the attachment of lysine 63-linked ubiquitin chains to the C-terminus of Malt1. Importantly, the replacement of ubiquitin acceptor sites on Malt1 impairs TCR induced NF-B activation, demonstrating a crucial role for regulatory Malt1 ubiquitination. Thus, our data provide evidence that Malt1 ubiquitination is directing CBM complex formation to IKK activation in response to TCR/CD28 engagement.

Results

Carma1-associated Malt1 is ubiquitinated upon T-cell activation

To identify further interactions or modifications critical for T-cell activation, we investigated the formation of the cellular CBM complex by analyzing Bcl10 immunoprecipitates after gel-filtration chromatography of Jurkat T-cell lysates (Figure 1A). In unstimulated Jurkat T cells, a peak of preformed Bcl10–Malt1 eluted in the low-molecular-weight fractions. Activation of Jurkat cells by PMA/ionomycin (P/I) induced the association of Bcl10–Malt1 with Carma1. The cellular CBM complex migrated with an apparent molecular mass of more than 1500 kDa, a size that by far exceeds the expected molecular weight of a heterotrimer (260 kDa). We noticed the appearance of higher molecular weight forms of Carma1–Bcl10-associated Malt1, indicative of a modification. To determine whether Malt1 is modified by ubiquitin, we analyzed Malt1 precipitates after gel-filtration chromatography under denaturing conditions to prevent non-covalent protein interactions (1% SDS, see below). Indeed, modified Malt1 in the high-molecular-weight fraction was also detected by an anti-ubiquitin antibody (Figure 1B). To see whether Malt1 ubiquitination is a primary event in response to T-cell activation, we determined the onset of Malt1 ubiquitination, IKK phosphorylation and IB degradation after P/I treatment (Figure 1C). After direct immunoprecipitation (IP) of Malt1 from Jurkat lysates, we found that the appearance of ubiquitin-conjugated Malt1 slightly preceded IKK phosphorylation and subsequent IB degradation, suggesting that Malt1 ubiquitination could be involved in delivering the signal to IKK/NF-B after T-cell activation. Further, Malt1 ubiquitination was transient and the disengagement of ubiquitin conjugates correlated with a decrease of IKK/NF-B signaling as monitored by reappearance of IB 45–60 min after stimulation (Supplementary Figure 1). In line with P/I activation, CD3/CD28 receptor co-ligation promoted polyubiquitin conjugation of Malt1 in Jurkat T cells and primary CD4+ T cells (Figure 1D). Since we could not detect Malt1 degradation during the course of stimulation, the data suggest that Malt1 is modified by non-degradative regulatory ubiquitination that could be involved in signal propagation.

IKK activity is recruited to Malt1 through the ubiquitin-binding domain of IKK

IKK was recently shown to contain a UBD (Ea et al, 2006; Li et al, 2006; Wu et al, 2006) and we asked whether IKK can associate with ubiquitinated Malt1 in response to T-cell activation. Indeed, we detect ubiquitin-conjugated Malt1 in IKK co-IPs from the extracts of activated Jurkat T cells (Figure 2A). An unspecific cross-reactivity that is detected in Malt1 western blots after co-IP is slightly larger than unmodified Malt1 (compare Figure 4B). Importantly, control IKK IPs from stimulated Jurkat T cells prove that there is no unspecific binding of ubiquitinated Malt1 species (Supplementary Figure 2A). To determine if TAB2 and TAK1 associate with ubiquitin chains attached to Malt1, we performed co-IPs and found that both proteins are recruited to ubiquitinated Malt1 upon P/I stimulation (Figure 2B). Thus, different UBD-containing signaling mediators can be recruited to ubiquitin-conjugated Malt1 upon T-cell activation.

The functional relevance of the IKK UBD to mediate interaction with ubiquitin-conjugated RIP1 was recently shown for TNF signaling (Ea et al, 2006; Li et al, 2006; Wu et al, 2006). IKK L329P and Y308S are two point mutants in the coiled-coil 2/leucine zipper of IKK (242–350) that impair binding of IKK to polyubiquitin chains and therefore its recruitment to ubiquitinated RIP1 (Li et al, 2006; Wu et al, 2006). To test whether the same mutants have a reduced affinity to ubiquitin-conjugated Malt1, we performed streptactin pull-down experiments from extracts of activated Jurkat cells using recombinant StrepIKK wt, L329P and Y308S (Figure 2C). StrepIKK wt preferentially associated with ubiquitin-conjugated Malt1 from stimulated Jurkat T-cell extracts, whereas the mutations L329P and Y308S reduce this interaction. Similar, transfected FlagIKK L329P and Y308S have a reduced affinity to ubiquitinated Malt1 from P/I stimulated Jurkat T cells (Supplementary Figure 2B). Next, we analyzed the ability of IKK L329P to associate with Malt1 and to rescue P/I-induced signaling in IKK deficient Jurkat T cells (Figure 2D–F). In reconstitution experiments, the L329P mutation severely impaired the association of IKK to ubiquitin-conjugated Malt1, even though the level of inducible Malt1 ubiquitination was equivalent in both cell lines (Supplementary Figure 2C). To verify the association of IKKs to Malt1, we also performed an in vitro kinase assay after Malt1 IP, using GSTIB 1–53 as substrate (Figure 2E). Anti-Malt1 IP coprecipitated inducible IKK activity that specifically phosphorylated serines 32/36 of IB (Supplementary Figure 2D). Moreover, IKK kinase activity was coprecipitated by an anti-Malt1 IP from extracts of Jurkat T cells reconstituted with IKK wt, but not IKK L329P. Mutation of IKK at position L329 did not affect association to IKK and IKK (Wu et al, 2006, and data not shown). In addition, only IKK wt was able to significantly rescue P/I-induced NF-B activation, as determined by IB degradation in IKK-deficient Jurkat T cells (Figure 2F). This shows that sensing of ubiquitin chains by IKK is crucial for TCR-dependent NF-B activation, and that Malt1 ubiquitin chains act as molecular bridges that connect CBM and IKK complexes.

TRAF6 functions as an E3 ligase for Malt1

TRAF6 was shown to interact with the C-terminus of Malt1 upon overexpression and in vitro (Sun et al, 2004), and we asked whether TRAF6 could act as an E3 ligase for Malt1. Indeed, coexpression of TRAF6 induced Malt1 ubiquitination in 293 cells (Figure 3A). Malt1 ubiquitination was dependent on the N-terminal RING domain of TRAF6 that confers ligase activity, although TRAF6 was still able to associate with Malt1 (Supplementary Figure 3A). To exclude background detection of TRAF6 autoubiquitination or other ubiquitin modifications, ubiquitination experiments were performed after cellular lysis under denaturing conditions (1% SDS). This lysis completely abolished the interactions of transfected or endogenous Malt1/TRAF6 or Malt1/Bcl10 (Supplementary Figure 3B, and data not shown), providing evidence that specifically Malt1 ubiquitination was detected. Since TRAF6 was shown to associate with the C-terminus of Malt1 (Sun et al, 2004), we asked whether it is also the C-terminus that is targeted by TRAF6 induced ubiquitination (Figure 3B). Indeed, TRAF6 induced strong ubiquitination of the C-terminal 200 amino acids (aa) (Malt1 612–813). In contrast, the very C-terminus of Malt1 (aa 684–813) was not ubiquitinated by TRAF6, suggesting the existence of potential acceptor lysines in the region between aa 612 and 683. However, reduced ubiquitination could partially result from decreased affinity due to the absence of the second TRAF6-binding site in Malt1 684–813 (data not shown and scheme Figure 5A).

To demonstrate that TRAF6 can function as an E3 ligase for Malt1 in vitro and to characterize the linkage of TRAF6 induced C-terminal Malt1 ubiquitin chains, we established a cell-free ubiquitin conjugation assay for Malt1. Detection was performed by anti-ubiquitin western blotting after boiling of ubiquitination assays in 1% SDS buffer, to abolish any non-covalent protein interaction, followed by Malt1 IP (Figure 3C). Depending on E2 (Ubc13/Uev1a) and ATP, TRAF6 mediated the assembly of K63-linked ubiquitin chains to the C-terminus of Malt1.

Next, we addressed TRAF6 function for Malt1 ubiquitination upon activation of Jurkat T cells. Co-IPs revealed inducible binding of endogenous Malt1 and TRAF6 (Figure 4A), correlating with enhanced Malt1 ubiquitination and degradation of IB. Vice versa, the majority of Malt1 associated with TRAF6 was ubiquitin-conjugated after T-cell activation (Figure 4B). Moreover, a reduction of TRAF6 protein amounts using three independent siRNAs diminished stimulus-dependent Malt1 ubiquitination (Figure 4C). Congruent with previous results, siRNA-mediated inactivation of TRAF6 impaired NF-B activation, as monitored by delayed degradation of IB (Figure 4D). These data demonstrate that TRAF6 can function as an E3 ligase mediating the attachment of K63-linked ubiquitin chains to the C-terminus of Malt1.

Multiple lysines in the C-terminus of Malt1 can serve as ubiquitin acceptor sites

For a functional analysis of Malt1 ubiquitination, we mutated C-terminal lysine residue(s) to map potential acceptor site(s) for the attachment of regulatory ubiquitin chains. We concentrated on the entire C-terminus of Malt1 (aa 612–813; 11 lysines), because we could not exclude that the reduced ubiquitination of Malt1 684–813 was at least partially due to decreased affinity to TRAF6 (scheme in Figure 5A). To further narrow the ubiquitination site, we exchanged pairs of lysines and all six lysines in the region 612–683 to arginines and also replaced all 11 lysines (11R) in the Malt1 C-terminus (612–813) (Figure 5B; Supplementary Figure 4A). None of the double mutants displayed reduced TRAF6-dependent Malt1 ubiquitination. Only combined exchange of all six lysine residues in the region 612–683 (Malt1 612–813 6R) significantly reduced Malt1 ubiquitination. As expected, ubiquitination was completely abolished when all eleven C-terminal lysines (Malt1 612–813 11R) were mutated to arginines. The affinity between Malt1 and TRAF6 was not altered by C-terminal lysine exchange (Supplementary Figure 4B and C). Malt1 ubiquitination was also reduced in an in vitro ubiquitin conjugation assay, if Malt1 6R or Malt1 11R instead of Malt1 wt was used as a substrate (Figure 5C). To determine if the presence of C-terminal lysine residues contributes to TCR-dependent Malt1 ubiquitination, we analyzed FlagMalt1 wt, 6R and 11R ubiquitination in Jurkat T cells (Figure 5D). Malt1 6R displayed a marked reduction and replacement of all C-terminal lysines in Malt1 11R nearly abolished inducible ubiquitin ligation, demonstrating the importance of the mapped acceptor lysines for stimulus-dependent Malt1 ubiquitination. Further, an in vitro kinase assay using GSTIB (1–53) as a substrate after IP of MycMalt1 wt, 6R and 11R revealed that C-terminal lysine to arginine exchange severely impaired the ability of Malt1 to pull down IKK kinase activity from the extracts (Figure 5E). Importantly, mutation of C-terminal lysines in Malt1 did not prevent constitutive association with endogenous Bcl10 or inducible interaction with TRAF6 or Carma1 (Figure 5F and G), demonstrating that lysine exchange in the C-terminus did not interfere with CBM complex formation and TRAF6 binding. Thus, TRAF6 and T-cell activation promote the assembly of ubiquitin chains to multiple lysine residues in the C-terminus of Malt1.

C-terminal Malt1 ubiquitination sites are required for NF-B activation and IL-2 production

T cells from Malt1-deficient mice are defective in NF-B activation and interleukin-2 (IL-2) production upon CD3/CD28 co-ligation or P/I stimulation (Ruefli-Brasse et al, 2003; Ruland et al, 2003). To investigate the functional consequences of C-terminal Malt1 ubiquitination for NF-B signaling and T-cell activation, we rescued naive CD4+ T cells from Malt1-deficient mice by retroviral infection using FlagMalt1 wt, 6R, 11R and 314–813 constructs (Figure 6). Retroviral vectors contained an IRES sequence for simultaneous expression of the surface protein Thy1.1 to identify infected cells by FACS. As determined by Flag/Thy1.1 costaining, Thy1.1-positive cells (+) expressed equivalent amounts of FlagMalt1 wt and mutant proteins (Supplementary Figure 5A). We measured IB protein levels to monitor NF-B activation (Figure 6A; Supplementary Figure 5B) and intracellular IL-2 production as a marker for T-cell activation (Figure 6B; Supplementary Figure 5C). Infection with empty vector or Malt1 314–813, a deletion mutant that cannot bind to Bcl10, did not rescue defective IB degradation in Malt1-/- T cells in response to P/I. In contrast, P/I treatment of Thy1.1+ Malt1 wt cells induced IB degradation, whereas Malt1 6R and 11R mutants were severely impaired in mediating IB degradation. Further, Malt1 wt, but not empty vector or Malt1 314–813, provoked a robust increase in IL-2 producing Thy1.1+ cells after CD3/CD28 costimulation. Again, mutation of C-terminal lysines (6R, 11R) significantly impaired the ability of Malt1 to rescue IL-2 production in Malt1-/- T cells. Thus, the data provide evidence that C-terminal ubiquitination of Malt1 is essential for NF-B activation and IL-2 induction in response to T-cell activation.

Discussion

In T cells recruitment of Bcl10/Malt1 to Carma1 is essential for IKK activation in response to TCR/CD28 co-ligation, but the molecular mechanisms of IKK activation downstream of the CBM have not been fully elucidated. In our study, we present several lines of evidence that ubiquitination of Malt1 functionally links CBM and IKK complexes. First, Malt1 ubiquitination coincides with the activation of the IKK/NF-B-signaling pathway. Second, IKK associates with Malt1 ubiquitin chains and the ubiquitin-binding motif of IKK is critical for NF-B signaling in response to T-cell activation. Third, TRAF6 can catalyze the assembly of K63-linked ubiquitin chains in vitro and functions as a potential ubiquitin ligase for Malt1 in vivo. Fourth, T-cell activation induces the assembly of ubiquitin chains to multiple C-terminal lysines of Malt1 and conservative replacement (lysine to arginine exchange) of ubiquitin attachment sites strongly decreases the ability of Malt1 to mediate T-cell stimulation-dependent NF-B signaling and IL-2 production. Collectively, the data demonstrate that polyubiquitination of Malt1 is essential for directing TCR signaling to the canonical NF-B pathway.

We suggest the following model of IKK activation in T cells (Figure 6C). TCR/CD28 co-engagement initiates a series of receptor proximal signaling events that lead to the activation of PKC (Sun et al, 2000). PKC phosphorylation of Carma1 promotes recruitment of Bcl10–Malt1 to Carma1 and thus CBM complex formation (Matsumoto et al, 2005; Sommer et al, 2005). The ubiquitin ligase TRAF6 is recruited to the C-terminus of Malt1 and mediates the assembly of K63-linked ubiquitin chains to lysine residues in the vicinity. IKK binds through its ubiquitin-binding moiety to ubiquitin chains on Malt1. Further, TAB2/TAK1 associate with ubiquitinated Malt1 upon T-cell stimulation, but it remains to be seen whether this is due to a direct interaction of the TAB2 UBD and Malt1-attached ubiquitin chains. TAK1 was shown to function as activating kinase for IKK kinase (Wang et al, 2001; Sun et al, 2004; Wan et al, 2006), suggesting that recruitment of TAK1 to ubiquitinated Malt1 upon TCR engagement could lead to subsequent IKK activation. However, the necessity for TAK1 at this stage has not been completely resolved and alternatively, binding of IKK to Malt1 ubiquitin chains could induce proximity and autoactivation of IKK complexes (Hayden and Ghosh, 2004). Future studies must determine whether recruitment of several UBD containing proteins is crucial for efficient signal propagation.

The function of Malt1 in TCR/CD28-induced IKK activation seems to be analogous to the role of RIP1 in TNF-triggered NF-B activation. It was shown that TNF stimulation-induced RIP1 polyubiquitination, potentially catalyzed by the E3 ligase TRAF2, provides a platform for the recruitment of IKKs to the TNF receptor complex (Ea et al, 2006; Li et al, 2006; Wu et al, 2006). Nevertheless, there is a discrepancy between Malt1 and RIP1 regarding the mode of ubiquitination. A single lysine residue (K377) was shown to serve as the attachment site for ubiquitin chains to RIP1 (Ea et al, 2006; Li et al, 2006). However, mutation of RIP1 at position K377 diminished its inducible interaction with TNF receptor complexes (Ea et al, 2006), indicating that lack of ubiquitination could be caused by disturbed recruitment rather than mutation of the substrate attachment site. Based on in vivo and in vitro evidence, we find that for the assembly of ubiquitin chains to Malt1, any lysine within an acceptable distance seems to be sufficient, which is in agreement with observations that RING E3 ligases often do not precisely position the ubiquitin chain to specific acceptor lysines (Passmore and Barford, 2004). Since the C-terminal lysine mutants of Malt1 associate with Bcl10 and TRAF6 and integrate into the CBM complex, we can exclude that gross structural alterations have been evoked and that upstream signaling is defective.

A previous study has suggested that Bcl10/Malt1-induced ubiquitination of IKK is crucial for NF-B signaling in T cells, and that Malt1 contains intrinsic ubiquitin ligase activity (Zhou et al, 2004). Although we cannot completely exclude that Malt1 is a TRAF6-dependent E3 ligase, in our experiments we did not observe that Malt1 is significantly autoubiquitinated after overexpression in cells or in vitro (see Figure 3A and C, and data not shown). Thus, Malt1 does not seem to confer sufficient E3 ligase activity. In line with these observations, a separate study suggested that TRAF6 might be the E3 ligase that mediates Malt1-dependent IKK ubiquitination (Sun et al, 2004). Carma1 and Bcl10–Malt1 can induce ubiquitination of IKK on K399; however, K399R mutation has only very little effect on inducible NF-B activation in T cells (Zhou et al, 2004; Shambharkar et al, 2007). Thus, the functional link between IKK ubiquitination and NF-B activation for T-cell activation is rather vague. It is tempting to speculate that different regulatory mechanisms are required for sustained productive T-cell activation. However, the kinetic of Malt1 ubiquitination and the mutagenesis of C-terminal Malt1 acceptor lysines suggest that attachment of ubiquitin chains to Malt1 is a key event to initialize IKK/NF-B signaling in response to TCR/CD28 co-engagement.

Recently, it was suggested that T-cell activation can trigger phosphorylation and ubiquitination of the IKK complex by two distinct mechanisms (Shambharkar et al, 2007). Although Carma1 and Bcl10 are involved in TRAF6-dependent ubiquitination of IKK, both proteins are dispensable for TAK1-dependent IKK/ phosphorylation. Mechanistically, it is unclear how these separate pathways are integrated. We find that the critical components, including Carma1, Bcl10, Malt1, TRAF6, TAB2/TAK1 and IKK, are directly or indirectly associating, suggesting that they should act in concert. However, it might be that some components (e.g., TAK1) can perform certain tasks independent of the other mediators.

Previous studies have reported impaired IKK/NF-B activation after siRNA-mediated downregulation of TRAF6 (Sun et al, 2004; Bidere et al, 2006). In line with this, we found that downregulation of TRAF6 resulted in a partial inhibition of Malt1 ubiquitination and NF-B signaling. Unexpectedly, mice that lack expression of TRAF6 in T cells have no apparent abnormalities in NF-B activation upon TCR engagement (King et al, 2006). However, the importance of regulatory K63-linked ubiquitination in TCR signaling is supported by the conditional excision of the UBC13 locus in T cells, as thymocytes from these mice are defective in IKK/NF-B activation (Yamamoto et al, 2006). Altogether, the data indicate that one or several unknown E3 ligases compensate for the loss of TRAF6 in T cells. Based on siRNA experiments, Sun et al (2004) have suggested that TRAF2 could be involved in TCR-dependent NF-B activation. Future studies must therefore determine whether TRAF2 or other E3 ligases might have a redundant function with TRAF6 in mediating TCR-induced NF-B signaling.

Recent results demonstrated a conserved function of Bcl10–Malt1 in directing antifungal responses and G-protein-coupled receptors (GPCR) to NF-B activation (Gross et al, 2006; Klemm et al, 2007; McAllister-Lucas et al, 2007; Wang et al, 2007). The Carma1 homologue Carma3 (CARD10) functions as a scaffold for GPCR-initiated NF-B activation, which is abrogated by TRAF6 deficiency (Grabiner et al, 2007). Further, the CBM complex is critical for the survival of a subset of malignant lymphomas (Ngo et al, 2006). Chromosomal translocations leading to the generation of API2–Malt1 fusion proteins are associated with aggressive MALT lymphoma, and API2–Malt1 requires the C-terminus of Malt1 for triggering NF-B activation (Zhou et al, 2005). Thus, C-terminal Malt1 ubiquitination may be relevant for activation of NF-B in various physiological and pathological settings.

Materials and methods

Reagents and antibodies

The following antibodies were used: human CD3, human CD28, mouse IgG1, mouse IgG2a, mouse IgG1a–FITC, IKK and IKK (all from BD Biosciences); IB (C21), Myc (9E10), TRAF6 (H274), Malt1 (H300, B12), Bcl10 (331.1), TAB2 (H300), TAK1 (M579) and IKK (FL419) (all from Santa Cruz Biotechnology); Carma1 (Abcam); flag M2 and flag M2–FITC (both from Sigma); ubiquitin (FK2; Biomol); IB, phospho-IKK/ and IKK (all from Cell Signaling); Thy1.1-APC, IL-2–FITC (both eBioscience) and ICN anti-hamster (MP Biomedicals). The following reagents and siRNAs (100 nM) were used: PMA (200 ng/ml) and ionomycin (300 ng/ml; both from Calbiochem); IL-2 (20 U/ml; Roche), Brefeldin A (10 ng/ml; Sigma); Dynabeads CD4 and DetachaBead mouse CD4 (Dynal Invitrogen) and Streptactin Superflow resin (IBA); si TRAF6.1: GCACAGCAGUGCAAUGGAAUUUAUA (Invitrogen); si TRAF6.2: CCAGCUCCUGUAGCGCUGUAACAAA (Invitrogen); si TRAF6.3: CCACGAAGAGAUAAUGGAU (Eurogentec) and si control: CCAUCCUGAUGUCGCAAUGCCGAAA (Invitrogen).

Plasmids

All Malt1 constructs were cloned with N-terminal Flag (pEF vector; Invitrogen) or Myc (pRK5 vector) epitopes. Mutagenesis was performed by standard PCR. Flag/Myc or Express (X) TRAF constructs were expressed from pRK5 or pcDNA4-His-Express vectors (Invitrogen), respectively. FlagIKK wt and mutants were cloned in pcDNA3 (Invitrogen). GST–Malt1 (493–824) and GST–TRAF6 were expressed from pGEX6p-1 (GE Healthcare). Retroviral FlagMalt1 constructs were cloned using the Gateway system (Invitrogen) into pMSCV-Thy1.1 that couples Thy1.1 and Malt1 expression via an IRES (Internal ribosome entry site) sequence.

Cell culture

HEK293 and Phoenix packaging cells were transfected using standard calcium phosphate precipitation protocols. Cell culture, transfection and stimulation of Jurkat T cells (P/I or CD3/CD28 antibody co-ligation) were performed as described (Scharschmidt et al, 2004). For RNA interference, Jurkat T cells were transfected with Atufect transfection reagent (Atugen, Berlin) and 100 nM siTRAF6 or control siRNA and analyzed after 72 h. Primary T cells were cultured in RPMI supplemented with 1% pen/strep, 1% glutamine, 10% FCS and 0.1% mercaptoethanol. Positive selection for CD4+ T cells was carried out with Dynabeads.

Retroviral infection of CD4-positive T cells and FACS analysis

Purified CD4+ T cells from spleen and lymph nodes of Malt1-/- mice (Ruland et al, 2003) were stimulated with plate-bound CD3/CD28 antibodies for 48 h essentially as described (Wegener et al, 2006). For retroviral infection, virus was harvested from Phoenix packaging cells 2 days after transfection and supplemented with Polybrene (4 g/ml). CD4+ T cells were incubated with retroviral supernatant for 6 h and then resuspended and cultured in medium with IL-2 (20 U/ml) for 3 days before analysis. Infection efficiencies between 25–50% were achieved. Infected T cells were then stimulated with P/I or plate-bound CD3/28 antibodies for the indicated times. For determination of IL-2 production, Brefeldin A was added 1 h after CD3/28 stimulation. After Thy1.1–APC staining, activated cells were fixed, permeabilized and stained with primary anti-IB and secondary anti-mouse IgG1a–FITC antibodies or anti-IL-2–FITC antibody for FACS analysis. Quantifications of IB degradation and IL-2 production represent statistical analysis of three independent experiments.

Co-IP, cellular ubiquitination and gel filtration

For binding studies, cells were lysed in co-IP buffer (25 mM HEPES pH 7.5, 150 mM NaCl, 0.2% NP-40, 10% glycerol, 1 mM DTT, 10 mM sodium fluoride, 8 mM -glycerophosphate, 20 M sodium vanadate and protease inhibitor cocktail). IP was carried out overnight at 4°C, and after washing precipitates were boiled and analyzed by western blotting. For detection of Malt1 ubiquitination, the lysis buffer was supplemented with 1% SDS. Before IPs, extracts were diluted 10-fold with co-IP buffer. For gel-filtration analysis, extracts from Jurkat T cells lysed in co-IP buffer without glycerol were fractionated on a Superose 6 column (GE Healthcare), followed by anti-Bcl10 IP. For detection of Malt1 ubiquitination, fractions were supplemented with 1% SDS and diluted to 0.1% SDS final concentration before anti-Malt1 IP.

Streptactin pull down

StrepIKK wt, L329P and Y308S were expressed in Escherichia coli BL21(DE3), bound to streptactin beads and incubated overnight with extracts from Jurkat T cells (lysis in co-IP buffer with 1% Triton X-100 instead of NP-40; final concentration of Triton-X100 for pull down was 0.1%). Analysis of material bound to beads was by western blotting.

In vitro ubiquitination and kinase assays

Recombinant GST–Malt1 (482–813) and GST–TRAF6 were expressed in E. coli BL21(DE3)RIL and purified using glutathione sepharose. In vitro ubiquitination reactions (30 l total volume) were performed in ubiquitination buffer (20 mM HEPES pH 7.2, 10 mM MgCl₂, 1 mM DTT, protease inhibitor cocktail) with 50 nM E1, 875 nM E2 (Ubc13/Uev1a), 150 M ubiquitin (wt, K63-only, K48-only, K63R or K48R) and energy-regenerating solution (all from Boston Biochemicals). The reactions were incubated for 2 h at 30°C, boiled in co-IP buffer containing 1% SDS and diluted 10-fold before Malt1 IP. For IKK kinase assays, untreated and P/I-stimulated Jurkat cells were lysed in co-IP buffer and Malt1 was precipitated using Malt1 (H300) or Myc antibodies. Kinase assays using GST IB (aa 1–53) as substrate were performed essentially as described previously (Scharschmidt et al, 2004).

Supplementary data

Supplementary data are available at The EMBO Journal Online (http://www.embojournal.org).

Acknowledgements

We thank Sandra Rohrmoser and Marc Schmidt-Supprian for helpful discussion, Evelyn Neve and Rudolf Dettmer for excellent technical assistance. We also thank SC Sun for the IKK-negative Jurkat T cells, J Ashwell for the gift of IKK wt and IKK L329P-reconstituted cells and A Abbas for the retroviral vector. This work was supported by DFG grants to DK (UR2306) and JR.

References

Akagi T, Motegi M, Tamura A, Suzuki R, Hosokawa Y, Suzuki H, Ota H, Nakamura S, Morishima Y, Taniwaki M, Seto M (1999) A novel gene, MALT1 at 18q21, is involved in t(11;18) (q21;q21) found in low-grade B-cell lymphoma of mucosa-associated lymphoid tissue. Oncogene 18: 5785–5794 | Article | PubMed | ISI | ChemPort |

Bi K, Tanaka Y, Coudronniere N, Sugie K, Hong S, van Stipdonk MJ, Altman A (2001) Antigen-induced translocation of PKC-theta to membrane rafts is required for T cell activation. Nat Immunol 2: 556–563 | Article | PubMed | ISI | ChemPort |

Bidere N, Snow AL, Sakai K, Zheng L, Lenardo MJ (2006) Caspase-8 regulation by direct interaction with TRAF6 in T cell receptor-induced NF-kappaB activation. Curr Biol 16: 1666–1671 | Article | PubMed | ISI | ChemPort |

Ea CK, Deng L, Xia ZP, Pineda G, Chen ZJ (2006) Activation of IKK by TNFalpha requires site-specific ubiquitination of RIP1 and polyubiquitin binding by NEMO. Mol Cell 22: 245–257 | Article | PubMed | ISI | ChemPort |

Egawa T, Albrecht B, Favier B, Sunshine MJ, Mirchandani K, O'Brien W, Thome M, Littman DR (2003) Requirement for CARMA1 in antigen receptor-induced NF-kappa B activation and lymphocyte proliferation. Curr Biol 13: 1252–1258 | Article | PubMed | ISI | ChemPort |

Gaide O, Favier B, Legler DF, Bonnet D, Brissoni B, Valitutti S, Bron C, Tschopp J, Thome M (2002) CARMA1 is a critical lipid raft-associated regulator of TCR-induced NF-kappa B activation. Nat Immunol 3: 836–843 | Article | PubMed | ISI | ChemPort |

Ghosh S, May MJ, Kopp EB (1998) NF-kappa B and Rel proteins: evolutionarily conserved mediators of immune responses. Annu Rev Immunol 16: 225–260 | Article | PubMed | ISI | ChemPort |

Grabiner BC, Blonska M, Lin PC, You Y, Wang D, Sun J, Darnay BG, Dong C, Lin X (2007) CARMA3 deficiency abrogates G protein-coupled receptor-induced NF-{kappa}B activation. Genes Dev 21: 984–996 | Article | PubMed | ISI | ChemPort |

Gross O, Gewies A, Finger K, Schafer M, Sparwasser T, Peschel C, Forster I, Ruland J (2006) Card9 controls a non-TLR signalling pathway for innate antifungal immunity. Nature 442: 651–656 | Article | PubMed | ISI | ChemPort |

Hara H, Wada T, Bakal C, Kozieradzki I, Suzuki S, Suzuki N, Nghiem M, Griffiths EK, Krawczyk C, Bauer B, D'Acquisto F, Ghosh S, Yeh WC, Baier G, Rottapel R, Penninger JM (2003) The MAGUK family protein CARD11 is essential for lymphocyte activation. Immunity 18: 763–775 | Article | PubMed | ISI | ChemPort |

Hayden MS, Ghosh S (2004) Signaling to NF-kappaB. Genes Dev 18: 2195–2224 | Article | PubMed | ISI | ChemPort |

Jun JE, Wilson LE, Vinuesa CG, Lesage S, Blery M, Miosge LA, Cook MC, Kucharska EM, Hara H, Penninger JM, Domashenz H, Hong NA, Glynne RJ, Nelms KA, Goodnow CC (2003) Identifying the MAGUK protein Carma-1 as a central regulator of humoral immune responses and atopy by genome-wide mouse mutagenesis. Immunity 18: 751–762 | Article | PubMed | ISI | ChemPort |

Kanayama A, Seth RB, Sun L, Ea CK, Hong M, Shaito A, Chiu YH, Deng L, Chen ZJ (2004) TAB2 and TAB3 activate the NF-kappaB pathway through binding to polyubiquitin chains. Mol Cell 15: 535–548 | Article | PubMed | ISI | ChemPort |

King CG, Kobayashi T, Cejas PJ, Kim T, Yoon K, Kim GK, Chiffoleau E, Hickman SP, Walsh PT, Turka LA, Choi Y (2006) TRAF6 is a T cell-intrinsic negative regulator required for the maintenance of immune homeostasis. Nat Med 12: 1088–1092 | Article | PubMed | ISI | ChemPort |

Klemm S, Zimmermann S, Peschel C, Mak TW, Ruland J (2007) Bcl10 and Malt1 control lysophosphatidic acid-induced NF-kappaB activation and cytokine production. Proc Natl Acad Sci USA 104: 134–138 | Article | PubMed | ChemPort |

Lee KY, D'Acquisto F, Hayden MS, Shim JH, Ghosh S (2005) PDK1 nucleates T cell receptor-induced signaling complex for NF-kappaB activation. Science 308: 114–118 | Article | PubMed | ISI | ChemPort |

Li H, Kobayashi M, Blonska M, You Y, Lin X (2006) Ubiquitination of RIP is required for tumor necrosis factor alpha-induced NF-kappaB activation. J Biol Chem 281: 13636–13643 | Article | PubMed | ISI | ChemPort |

Matsumoto R, Wang D, Blonska M, Li H, Kobayashi M, Pappu B, Chen Y, Wang D, Lin X (2005) Phosphorylation of CARMA1 plays a critical role in T cell receptor-mediated NF-kappaB activation. Immunity 23: 575–585 | Article | PubMed | ISI | ChemPort |

McAllister-Lucas LM, Inohara N, Lucas PC, Ruland J, Benito A, Li Q, Chen S, Chen FF, Yamaoka S, Verma IM, Mak TW, Nunez G (2001) Bimp1, a MAGUK family member linking protein kinase C activation to Bcl10-mediated NF-kappaB induction. J Biol Chem 276: 30589–30597 | Article | PubMed | ISI | ChemPort |

McAllister-Lucas LM, Ruland J, Siu K, Jin X, Gu S, Kim DS, Kuffa P, Kohrt D, Mak TW, Nunez G, Lucas PC (2007) CARMA3/Bcl10/MALT1-dependent NF-kappaB activation mediates angiotensin II-responsive inflammatory signaling in nonimmune cells. Proc Natl Acad Sci USA 104: 139–144 | Article | PubMed | ChemPort |

Monks CR, Kupfer H, Tamir I, Barlow A, Kupfer A (1997) Selective modulation of protein kinase C-theta during T-cell activation. Nature 385: 83–86 | Article | PubMed | ISI | ChemPort |

Morgan JA, Yin Y, Borowsky AD, Kuo F, Nourmand N, Koontz JI, Reynolds C, Soreng L, Griffin CA, Graeme-Cook F, Harris NL, Weisenburger D, Pinkus GS, Fletcher JA, Sklar J (1999) Breakpoints of the t(11;18)(q21;q21) in mucosa-associated lymphoid tissue (MALT) lymphoma lie within or near the previously undescribed gene MALT1 in chromosome 18. Cancer Res 59: 6205–6213 | PubMed | ISI | ChemPort |

Ngo VN, Davis RE, Lamy L, Yu X, Zhao H, Lenz G, Lam LT, Dave S, Yang L, Powell J, Staudt LM (2006) A loss-of-function RNA interference screen for molecular targets in cancer. Nature 441: 106–110 | Article | PubMed | ISI | ChemPort |

Okamura H, Rao A (2001) Transcriptional regulation in lymphocytes. Curr Opin Cell Biol 13: 239–243 | Article | PubMed | ISI | ChemPort |

Passmore LA, Barford D (2004) Getting into position: the catalytic mechanisms of protein ubiquitylation. Biochem J 379 (Part 3): 513–525 | Article |

Rawlings DJ, Sommer K, Moreno-Garcia ME (2006) The CARMA1 signalosome links the signalling machinery of adaptive and innate immunity in lymphocytes. Nat Rev Immunol 6: 799–812 | Article | PubMed | ISI | ChemPort |

Ruefli-Brasse AA, French DM, Dixit VM (2003) Regulation of NF-kappaB-dependent lymphocyte activation and development by paracaspase. Science 302: 1581–1584 | Article | PubMed | ISI | ChemPort |

Ruland J, Duncan GS, Elia A, del Barco Barrantes I, Nguyen L, Plyte S, Millar DG, Bouchard D, Wakeham A, Ohashi PS, Mak TW (2001) Bcl10 is a positive regulator of antigen receptor-induced activation of NF-kappaB and neural tube closure. Cell 104: 33–42 | Article | PubMed | ISI | ChemPort |

Ruland J, Duncan GS, Wakeham A, Mak TW (2003) Differential requirement for Malt1 in T and B cell antigen receptor signaling. Immunity 19: 749–758 | Article | PubMed | ISI | ChemPort |

Scharschmidt E, Wegener E, Heissmeyer V, Rao A, Krappmann D (2004) Degradation of Bcl10 induced by T-cell activation negatively regulates NF-kappa B signaling. Mol Cell Biol 24: 3860–3873 | Article | PubMed | ISI | ChemPort |

Schulze-Luehrmann J, Ghosh S (2006) Antigen-receptor signaling to nuclear factor kappa B. Immunity 25: 701–715 | Article | PubMed | ISI | ChemPort |

Shambharkar PB, Blonska M, Pappu BP, Li H, You Y, Sakurai H, Darnay BG, Hara H, Penninger J, Lin X (2007) Phosphorylation and ubiquitination of the IkappaB kinase complex by two distinct signaling pathways. EMBO J 26: 1794–1805 | Article | PubMed | ISI | ChemPort |

Sommer K, Guo B, Pomerantz JL, Bandaranayake AD, Moreno-García ME, Ovechkina YL, Rawlings DJ (2005) Phosphorylation of the CARMA1 linker controls NF-kappaB activation. Immunity 23: 561–574 | Article | PubMed | ISI | ChemPort |

Su H, Bidere N, Zheng L, Cubre A, Sakai K, Dale J, Salmena L, Hakem R, Straus S, Lenardo M (2005) Requirement for caspase-8 in NF-kappaB activation by antigen receptor. Science 307: 1465–1468 | Article | PubMed | ISI | ChemPort |

Sun L, Deng L, Ea CK, Xia ZP, Chen ZJ (2004) The TRAF6 ubiquitin ligase and TAK1 kinase mediate IKK activation by BCL10 and MALT1 in T lymphocytes. Mol Cell 14: 289–301 | Article | PubMed | ISI | ChemPort |

Sun Z, Arendt CW, Ellmeier W, Schaeffer EM, Sunshine MJ, Gandhi L, Annes J, Petrzilka D, Kupfer A, Schwartzberg PL, Littman DR (2000) PKC-theta is required for TCR-induced NF-kappaB activation in mature but not immature T lymphocytes. Nature 404: 402–407 | Article | PubMed | ISI | ChemPort |

van der Merwe PA (2002) Formation and function of the immunological synapse. Curr Opin Immunol 14: 293–298 | Article | PubMed | ISI | ChemPort |

Wan YY, Chi H, Xie M, Schneider MD, Flavell RA (2006) The kinase TAK1 integrates antigen and cytokine receptor signaling for T cell development, survival and function. Nat Immunol 7: 851–858 | Article | PubMed | ISI | ChemPort |

Wang C, Deng L, Hong M, Akkaraju GR, Inoue J, Chen ZJ (2001) TAK1 is a ubiquitin-dependent kinase of MKK and IKK. Nature 412: 346–351 | Article | PubMed | ISI | ChemPort |

Wang D, You Y, Lin PC, Xue L, Morris SW, Zeng H, Wen R, Lin X (2007) Bcl10 plays a critical role in NF-kappaB activation induced by G protein-coupled receptors. Proc Natl Acad Sci USA 104: 145–150 | Article | PubMed | ChemPort |

Wegener E, Oeckinghaus A, Papadopoulou N, Lavitas L, Schmidt-Supprian M, Ferch U, Mak TW, Ruland J, Heissmeyer V, Krappmann D (2006) Essential role for IkappaB kinase beta in remodeling Carma1–Bcl10–Malt1 complexes upon T cell activation. Mol Cell 23: 13–23 | Article | PubMed | ISI | ChemPort |

Weil R, Israel A (2006) Deciphering the pathway from the TCR to NF-kappaB. Cell Death Differ 13: 826–833 | Article | PubMed | ISI | ChemPort |

Willis TG, Jadayel DM, Du MQ, Peng H, Perry AR, Abdul-Rauf M, Price H, Karran L, Majekodunmi O, Wlodarska I, Pan L, Crook T, Hamoudi R, Isaacson PG, Dyer MJ (1999) Bcl10 is involved in t(1;14)(p22;q32) of MALT B cell lymphoma and mutated in multiple tumor types. Cell 96: 35–45 | Article | PubMed | ISI | ChemPort |

Wu CJ, Conze DB, Li T, Srinivasula SM, Ashwell JD (2006) Sensing of Lys 63-linked polyubiquitination by NEMO is a key event in NF-kappaB activation [corrected]. Nat Cell Biol 8: 398–406 | Article | PubMed | ISI | ChemPort |

Yamamoto M, Okamoto T, Takeda K, Sato S, Sanjo H, Uematsu S, Saitoh T, Yamamoto N, Sakurai H, Ishii KJ, Yamaoka S, Kawai T, Matsuura Y, Takeuchi O, Akira S (2006) Key function for the Ubc13 E2 ubiquitin-conjugating enzyme in immune receptor signaling. Nat Immunol 7: 962–970 | Article | PubMed | ISI | ChemPort |

Zhang Q, Siebert R, Yan M, Hinzmann B, Cui X, Xue L, Rakestraw KM, Naeve CW, Beckmann G, Weisenburger DD, Sanger WG, Nowotny H, Vesely M, Callet-Bauchu E, Salles G, Dixit VM, Rosenthal A, Schlegelberger B, Morris SW (1999) Inactivating mutations and overexpression of BCL10, a caspase recruitment domain-containing gene, in MALT lymphoma with t(1;14)(p22;q32). Nat Genet 22: 63–68 | Article | PubMed | ISI | ChemPort |

Zhou H, Du MQ, Dixit VM (2005) Constitutive NF-kappaB activation by the t(11;18)(q21;q21) product in MALT lymphoma is linked to deregulated ubiquitin ligase activity. Cancer Cell 7: 425–431 | Article | PubMed | ISI | ChemPort |

Zhou H, Wertz I, O'Rourke K, Ultsch M, Seshagiri S, Eby M, Xiao W, Dixit VM (2004) Bcl10 activates the NF-kappaB pathway through ubiquitination of NEMO. Nature 427: 167–171 | Article | PubMed | ISI | ChemPort |