Cytoplasmic aggregation of TRAF2 and TRAF5 proteins in the Hodgkin-Reed-Sternberg cells

Abstract

We previously reported that ligand-independent signaling by highly expressed CD30 in Hodgkin-Reed-Sternberg (H-RS) cells is responsible for constitutive activation of NF-kappa B. In the present study, we characterize the intracellular localization of tumor necrosis factor (TNF) receptor associated factor (TRAF) proteins in H-RS cells. Confocal immunofluorescence microscopy of cell lines derived from H-RS cells and HEK293 transformants highly expressing CD30 revealed aggregation of TRAF2 and TRAF5 in the cytoplasm as well as clustering near the cell membrane. In contrast, TRAF proteins were diffusely distributed in the cytoplasm in cell lines unrelated to Hodgkin's disease (HD) and control HEK293 cells. Furthermore, the same intracellular distribution of TRAF proteins was demonstrated in H-RS cells of lymph nodes of HD, but not in lymphoma cells in lymph nodes of non-Hodgkin's lymphoma. Dominant-negative TRAF2 and TRAF5 suppressed cytoplasmic aggregation along with constitutive NF-kappa B activation in H-RS cell lines. Confocal immunofluorescence microscopy also revealed co-localization of IKK alpha, NIK, and I kappa B alpha with aggregated TRAF proteins in H-RS cell lines. These results suggest involvement of TRAF protein aggregation in the signaling process of highly expressed CD30 and suggest they function as scaffolding proteins. Thus, cytoplasmic aggregation of TRAF proteins appears to reflect constitutive CD30 signaling which is characteristic of H-RS cells.

Figure 1.

A: Laser confocal immunofluorescence microscopy for intracellular distribution of TRAF2 and TRAF5 proteins. H-RS derived cell lines. Aggregation of TRAF2 and TRAF5 in the cytoplasm is clearly shown. Original magnification, ×400. B: Hematopoietic and lymphoid cell lines unrelated to Hodgkin’s disease. Both TRAF proteins are diffusely distributed in the cytoplasm. Original magnification, ×400. For A and B, following antibodies are used. First antibodies: anti-TRAF2 (C-20) rabbit antibody (Santa Cruz) and anti-TRAF5 (C-19) goat antibody. Secondary antibodies: FITC-labeled anti-rabbit donkey antibody (Amersham Pharmacia Biotech), FITC-labeled anti-goat donkey antibody (Santa Cruz). C: Immunoblot analysis of TRAF2 and TRAF5 showing little difference in the amounts of expressed TRAF proteins. D: Laser confocal immunofluorescence microscopy of biopsied lymph nodes for intracellular distribution of TRAF2 and TRAF5 proteins. HL, a lymph node of Hodgkin’s lymphoma; NHL, a lymph node of non-Hodgkin’s lymphoma (diffuse large B cell type). Instead of secondary antibodies, streptavidin-FITC or streptavidin-Texas Red was used in modified TSA system (NEN Life Science). Original magnification, ×200.

Figure 2.

Blocking CD30 signaling by dominant-negative TRAF proteins in H-RS cell lines. A: Left, aggregation of TRAF2 and TRAF5 in HEK293 cells highly expressing CD30. TRAF2 and TRAF5 were aggregated in HEK293 cells highly expressing CD30 and retaining constitutive activation of NF-κB (293CD30), but not in HEK293 cells with an empty vector (293vec). Immunofluorescence studies were done as described in Figure 1, A and B ▶ . Right, immunoblot analysis of TRAF2 and TRAF5. The levels of TRAF protein expression are almost the same between 293CD30 and 293vec cells. B: Left, electrophoretic mobility shift analysis (EMSA) with an NF-κB probe. H-RS cell derived cell lines show constitutive activation of NF-κB. Jurkat cells treated with TNF-α were used as a positive control. Right, down-regulation of basal activities of NF-κB-driven luciferase by dominant TRAF proteins in H-RS cell lines, but not in Jurkat cells. Luciferase activities are expressed as relative levels of triplicated experiments where those co-transfected with a vacant expression vector are expressed as 100%. Transfection efficiencies were corrected by dual luciferase assays using pRL-TK-Luc. C: Expression of a dominant-negative TRAF2 abrogates cytoplasmic aggregation of endogenous TRAF proteins in L-428 and L-540 cells. Transfected ΔTRAF2 was detected by the C-terminal anti-TRAF2 mouse monoclonal antibody (Santa Cruz) and the endogenous TRAF2 by N-terminal anti-TRAF2 rabbit antibody (Santa Cruz). Secondary antibodies used are Texas Red-labeled anti-mouse immunoglobulin sheep antibody and FITC-labeled anti-rabbit immunoglobulin donkey antibody (both from Amersham Pharmacia Biotech). Strong staining of the highly expressed ΔTRAF2 protein by the C-terminal antibody made it possible to discriminate it from the endogenous TRAF2 protein. In the left panel, figures show the results with anti-N-terminal TRAF2 (N-19) (α-N-TRAF2) rabbit antibody and anti-C-terminal TRAF2 (H-10) (α-C-TRAF2) mouse monoclonal antibody (both from Santa Cruz). Arrows indicate cells that express transduced ΔTRAF2. In the right panel, merged figures of immunofluorescence by anti-N-terminal and anti-C-terminal TRAF2 antibodies. Secondary antibodies used are: FITC-labeled anti-rabbit donkey antibody and Texas Red-labeled anti-mouse immunoglobulin sheep antibody (both from Amersham Pharmacia Biotech). Original magnification, ×400.

Figure 3.

TRAF2 is in the soluble fraction irrespective of CD30 signaling. A: Distribution of TRAF2 in the soluble fraction. In all cell lines examined, the majority of TRAF2 protein was found in the soluble fraction by immunoblot analysis with anti-TRAF2 (top). Successful fractionation was confirmed by blotting of the same membrane with an anti-lamin A antibody (Santa Cruz), where laminA was detected solely in the insoluble fraction (bottom). B: Subcellular distribution of TRAF2 protein in HEK293 cells transiently co-transfected with CD30. Majority of TRAF2 protein was also found in the soluble fraction by immunoblot analysis with anti-TRAF2 antibody (top). Successful fractionation was confirmed by blotting of the same membrane with an anti-lamin A antibody (bottom).

Figure 4.

Co-localization of signal transducers with TRAF5 in H-RS cell derived cell lines. A to C: Laser confocal immunofluorescence microscopy showing co-localization of TRAF5 with IKKα, NIK, and IκBα, respectively. First antibodies used are as follows: anti-TRAF5 (C-19) goat antibody, anti-IKKa (B-8) mouse monoclonal antibody, anti-NIK (A-12) mouse monoclonal antibody and anti-IκBα (H-4) mouse monoclonal antibody (all from Santa Cruz). Secondary antibodies are: FITC-labeled anti-goat donkey antibody and Texas Red-labeled anti-mouse immunoglobulin sheep antibody (both from Santa Cruz). Original magnification, ×400. D: Laser confocal immunofluorescence microscopy reveals that CD30 does not co-localize with cytoplasmic aggregates of TRAF proteins in 293CD30 cells. First antibodies are anti-TRAF2 (C-20) rabbit antibody, anti-TRAF5 (C-19) goat antibody (Santa Cruz), and anti-CD30 mouse monoclonal antibody (BerH2) (DAKO). Secondary antibodies are: FITC-labeled anti-rabbit donkey antibody, FITC-labeled anti-goat donkey antibody, and Texas Red-labeled anti-mouse immunoglobulin sheep antibody (all from Santa Cruz). Original magnification, ×400.