Multiple protein domains mediate interaction between Bcl10 and MALT1

Abstract

Bcl10 and MALT1 are essential mediators of NF-kappaB activation in response to the triggering of a diverse array of transmembrane receptors, including antigen receptors. Additionally, both proteins are translocation targets in MALT lymphoma. Thus, a detailed understanding of the interaction between these mediators is of considerable biological importance. Previous studies have indicated that a 13-amino acid region downstream of the Bcl10 caspase recruitment domain (CARD) is responsible for interacting with the immunoglobulin-like domains of MALT1. We now provide evidence that the death domain of MALT1 and the CARD of Bcl10 also contribute to Bcl10-MALT1 interactions. Although a direct interaction between the MALT1 death domain and Bcl10 cannot be detected via immunoprecipitation, FRET data strongly suggest that the death domain of MALT1 contributes significantly to the association between Bcl10 and MALT1 in T cells in vivo. Furthermore, analysis of point mutants of conserved residues of Bcl10 shows that the Bcl10 CARD is essential for interaction with the MALT1 N terminus. Mutations that disrupt proper folding of the Bcl10 CARD strongly impair Bcl10-MALT1 interactions. Molecular modeling and functional analyses of Bcl10 point mutants suggest that residues Asp(80) and Glu(84) of helix 5 of the Bcl10 CARD directly contact MALT1. Together, these data demonstrate that the association between Bcl10 and MALT1 involves a complex interaction between multiple protein domains. Moreover, the Bcl10-MALT1 interaction is the second reported example of interactions between a CARD and a non-CARD protein region, which suggests that many signaling cascades may utilize CARD interactions with non-CARD domains.

FIGURE 1.

Deletional analysis defines the MALT1 Ig-like domains, the DD, and Bcl10 residues 1–114 as critical determinants of Bcl10-MALT1 interaction. A, diagram of the domains of full-length MALT1 and the MALT1 deletion mutants used in this study. For simplicity, the N-terminal FLAG and C-terminal YFP tags are not shown. B, diagram of the domains of full-length Bcl10 and the Bcl10 deletion mutants used in this study. For simplicity, the N-terminal HA and C-terminal GFP tags are not shown. C, co-IP analyses of interactions between Bcl10 (wild-type or Δ107–119) and the MALT1 deletion constructs shown in A. MALT1 and Bcl10 expression vectors were co-transfected into HEK293T cells. IP of MALT1 was performed with an anti-FLAG antibody, followed by Western blotting with an anti-HA antibody to detect co-immunoprecipitated Bcl10. Whole cell lysates were also probed with anti-HA and anti-FLAG to evaluate expression of the Bcl10 and MALT1 constructs. Note that co-transfections of Bcl10Δ107–119 and MALT1-ΔC consistently resulted in reduced levels of expression of Bcl10Δ107–119, in comparison to co-transfections with other Bcl10 deletion mutants. D, FRET analysis of interactions between Bcl10-CFP and the MALT1-YFP constructs shown in A, in the absence of specific antigen stimulation and at several time points post-stimulation with conalbumin-loaded CH12 B cells. FRET was performed as previously described (13). The dashed line indicates the threshold for significant FRET detection in this study. E, co-IP analyses of interactions between MALT1-ΔC and the Bcl10 constructs depicted in B. IPs and Western blots were performed as in C. F, luciferase assays in HEK293T cells measuring the ability of the indicated Bcl10 constructs to drive activation of an NF-κB reporter construct. Luciferase activity is expressed as –fold increase relative to NF-κB reporter alone, which is defined as 1.0. Error bars are S.E. Note that Δ107–119 and Δ107–233 had lower levels of luciferase activation than NF-κB reporter alone in this experiment. Caspase, caspase-like domain; TB, TRAF6-binding domain; WT, wild-type; IP, immunoprecipitation; IB, immunoblotting.

FIGURE 2.

Identification of conserved residues and structural modeling of the Bcl10 CARD and MALT1-binding region. A, Clustal W alignment of orthologous regions of Bcl10 from various mammals (mouse, human, and rat), birds (chicken), and fish (fugu) from amino acids 79–104 of the mouse sequence. An asterisk (*) indicates amino acid identity for all species, a colon (:) indicates conserved substitutions, and a period (.) indicates semi-conserved substitutions. Single and combinatorial point mutations analyzed in this study are shown below the alignment. The locations of CARD helices 5 and 6 and the connecting 5/6 loop, as determined by molecular modeling (see part C) are indicated above the alignment. B, Clustal W alignment of Bcl10 amino acids 105–119 was performed as described in A. The previously defined minimal MALT1 binding site (31) is indicated by a bar above the alignment. C, a three-dimensional model of Bcl10 amino acids 13–119 was generated using Swiss PDB Viewer, as described under “Experimental Procedures.” The six α-helices of the CARD are shown to the left, and the minimal MALT1 binding site (which was modeled as coil) is shown to the right. The helix 1–4 face is indicated by yellow, the helix 2–3 face is in red, and the helix 5–6 face is in green. The model is oriented so that the helix 5–6 face is closest to the viewer, with the 2–3 and 1–4 faces projecting behind. The predicted positions of side chains of five charged or polar residues that were modeled as solvent-exposed amino acids on the helix 5–6 face have been superimposed on the model.

FIGURE 3.

Analysis of NF-κB activation by Bcl10 point mutants. Luciferase assays of NF-κB activation by wild-type (WT) Bcl10 and each of the point mutants from Fig. 2 were performed as described in Fig. 1F. Error bars are S.E. Data are also tabulated in Table 1 as mean percent activation, relative to wild-type.

FIGURE 4.

Specific point mutations severely impair antigen-stimulated formation of Bcl10 POLKADOTS. Retroviral vectors expressing point mutants of Bcl10-GFP were used to infect D10 T cells, producing stable polyclonal cell lines. Each cell line was stimulated for 20 min with CH12 B cells that were loaded overnight with 250 μg/ml conalbumin (+Antigen) or with no antigen (No Ag) to induce Bcl10 redistribution into POLKADOTS. Cells were fixed and analyzed to detect Bcl10 POLKADOTS by digital deconvolution epifluorescence microscopy, as previously described (14). Data are also tabulated in Table 1, designating mutants that do and do not form POLKADOTS. Panels are arranged from left to right as differential interference contrat (DIC) image, fluorescence (GFP) image, and overlay image (GFP, green; DIC, blue). Each image is oriented to show a fluorescent T cell (top) and a non-fluorescent antigen presenting cell (bottom).

FIGURE 5.

Immunoprecipitation analysis of association between MALT1, CARMA1, and Bcl10 point mutants. A, FLAG-MALT1-ΔC and the indicated HA-tagged Bcl10 constructs were co-transfected into HEK293T cells. IP of MALT1 was performed with a monoclonal anti-FLAG antibody, followed by Western blotting with an anti-HA antibody to detect co-immunoprecipitated Bcl10. Whole cell lysates (WCL) were also probed with anti-HA and anti-FLAG to evaluate expression of the Bcl10 and MALT1 constructs. B, FLAG-CARMA1 and the indicated HA-tagged Bcl10 constructs were co-transfected into HEK293T cells. IP analyses were performed as in A. C, FLAG-MALT1-ΔC (top gel) and FLAG-CARMA1 (bottom gel) were cotransfected into HEK293T cells with decreasing amounts of the indicated HA-Bcl10 constructs. Total DNA was maintained at 3 μg/transfection by inclusion of pBluscript KS+ (Stratagene), as needed. IP analyses were performed as in A. IB, immunoblot.

FIGURE 6.

FRET analysis of association between MALT1 and Bcl10 mutants. A, FACS analysis of YFP fluorescence in D10 T cells co-expressing MALT1-ΔC-CFP and the indicated Bcl10-YFP constructs. B, FACS overlays, showing the FRET channel histograms from the indicated control or experimental cell lines (solid gray) overlaid onto the FRET histogram from the cell line co-expressing MALT1-ΔC-CFP and Bcl10-WT-YFP (black line). C, mean fluorescence intensity (MFI) of the FRET channel from the indicated D10 T cell lines as determined by FACS FRET. Dotted lines indicate the FRET MFI of the negative control exhibiting the highest FRET signal (considered the FRET background value) and the FRET MFI of cells co-expressing MALT1-ΔC-CFP and Bcl10-WT-YFP. D, FRET efficiency as determined by FLIM analysis of the indicated cell lines. Dotted lines indicate the FRET efficiency of the CFP+YFP negative control (indicating the frequency of FRET due to random CFP-YFP collisions) and the FRET efficiency of co-expressed MALT1-ΔC-CFP and Bcl10-WT-YFP. Error bars are S.E. *, p < 0.0001 versus Bcl10-WT-YFP; †, p = 0.0013 versus Bcl10-WT-YFP; all other mutants not significant.