MALT1-dependent cleavage of CYLD promotes NF-κB signaling and growth of aggressive B-cell receptor-dependent lymphomas

Abstract

The paracaspase mucosa-associated lymphoid tissue 1 (MALT1) is a protease and scaffold protein essential in propagating B-cell receptor (BCR) signaling to NF-κB. The deubiquitinating enzyme cylindromatosis (CYLD) is a recently discovered MALT1 target that can negatively regulate NF-κB activation. Here, we show that low expression of CYLD is associated with inferior prognosis of diffuse large B-cell lymphoma (DLBCL) and mantle cell lymphoma (MCL) patients, and that chronic BCR signaling propagates MALT1-mediated cleavage and, consequently, inactivation and rapid proteasomal degradation of CYLD. Ectopic overexpression of WT CYLD or a MALT1-cleavage resistant mutant of CYLD reduced phosphorylation of IκBα, repressed transcription of canonical NF-κB target genes and impaired growth of BCR-dependent lymphoma cell lines. Furthermore, silencing of CYLD expression rendered BCR-dependent lymphoma cell lines less sensitive to inhibition of NF-κΒ signaling and cell proliferation by BCR pathway inhibitors, e.g., the BTK inhibitor ibrutinib, indicating that these effects are partially mediated by CYLD. Taken together, our findings identify an important role for MALT1-mediated CYLD cleavage in BCR signaling, NF-κB activation and cell proliferation, which provides novel insights into the underlying molecular mechanisms and clinical potential of inhibitors of MALT1 and ubiquitination enzymes as promising therapeutics for DLBCL, MCL and potentially other B-cell malignancies.

Fig. 1. CYLD is variably expressed in B-cell non-Hodgkin lymphomas (B-NHLs).

A CYLD mRNA expression analysis of publically available micro-array datasets of naïve B cells, GC (germinal center) B cells, memory B cells, plasma cells, B-cell acute lymphoblastic leukemia (B-ALL), chronic lymphocytic leukemia (CLL), follicular lymphoma (FL), mantle cell lymphoma (MCL), germinal center B-cell-like diffuse large B-cell lymphoma (GCB-DLBCL), activated B-cell-like diffuse large B-cell lymphoma (ABC-DLBCL), Waldenström’s macroglobulinemia (WM) and multiple myeloma (MM). The gray line represents the median expression value within each group. The dotted black line shows the threshold value (i.e., log-transformed probe intensity values of <2⁶). B Kaplan–Meier survival curve showing overall survival probability in CYLD high versus CYLD low expressing DLBCL and MCL patients. The cut off was based on the average CYLD expression within each cohort. The log-rank test was used to compare the survival distributions of the two groups. C RT-qPCR analysis of CYLD mRNA expression in DLBCL and MCL cell lines. RPLP0 was used as an input control and data are normalized to CYLD expression in LY1 for DLBCL cell lines and Jeko for MCL cell lines. The mean ± SD of three independent experiments performed in triplicate is shown. D Immunoblot analysis of CYLD expression in DLBCL and MCL cell lines using an antibody raised against a C‐terminal epitope which detects full-length CYLD and a C‐terminal fragment of CYLD (CYLD‐Ct). β-actin was used as a loading control. E Immunoblot analysis of CYLD expression in primary DLBCL and MCL samples. β-actin was used as a loading control.

Fig. 2. CYLD is cleaved by MALT1 protease.

A Immunoblot analysis of CYLD cleavage following treatment with phorbol myristate acetate (PMA) and ionomycin. Cells were treated for 1 h with PMA (50 ng/ml) and ionomycin (1 µg/ml) as indicated. Phosphorylated IkBα (Ser32) and total IkBα were used as positive controls for PMA/ionomycin stimulation; β-actin was used as a loading control. B Immunoblot analysis of CYLD cleavage following treatment with PMA/ionomycin and/or MALT1 inhibitor Z-VRPR-FMK. Cells were pre-treated for 1 h with 75 µM Z-VRPR-fmk before incubation for 1 h with PMA (50 ng/ml) and ionomycin (1 µg/ml) as indicated. β-actin was used as a loading control. C Immunoblot analysis of MALT1 expression in DLBCL and MCL cell lines. β-tubulin was used as a loading control.

Fig. 3. CYLD is constitutively cleaved in ABC DLBCL and BCR-dependent MCL cell lines.

A Flow cytometric analysis of the number of viable cells, as determined by 7-AAD staining, after 5 days of treatment with indicated concentrations of BTK inhibitor Ibrutinib, PKC inhibitor Sotrastaurin or MALT1 inhibitor Z-VRPR-FMK. The number of viable cells was normalized to the vehicle-treated condition. Data are presented as mean ± SD of three independent experiments. B Immunoblot analysis of CYLD cleavage in DLBCL cell lines LY3 and LY10 using an antibody raised against a C‐terminal epitope which detects full-length CYLD and a C‐terminal fragment of CYLD (CYLD‐Ct). Cells were incubated with indicated concentrations of the different BCR signalosome inhibitors for 48 h. BCL-XL protein levels were determined as a positive control for efficacy of the inhibitors; β-actin was used as loading control. C Flow cytometric analysis of the number of viable cells, as determined by 7-AAD staining, after 7 days of treatment with indicated concentrations of BTK inhibitor Ibrutinib, PKC inhibitor Sotrastaurin or MALT1 inhibitor Z-VRPR-FMK. The number of viable cells was normalized to the vehicle-treated condition. Data are presented as mean ± SD of three independent experiments. D Immunoblot analysis of CYLD cleavage in MCL cell lines Rec1 and Mino using an antibody raised against a C‐terminal epitope which detects full-length CYLD and a C‐terminal fragment of CYLD (CYLD‐Ct). Cells were incubated with indicated concentrations of the different BCR signalosome inhibitors for 72 h. BCL-XL protein levels were determined as a positive control for efficacy of the inhibitors; β-tubulin was used as loading control.

Fig. 4. Ectopic expression of non-cleavable CYLD inhibits cell growth and NF-κB pathway activity.

A Flow cytometric analysis of LY10, RIVA and Mino cells transduced with an empty vector (EV) or a CYLD (wildtype or R324A mutant) containing bicistronic vector co-expressed with YFP. The percentage of YFP positive cells was followed in time and plotted as the percentage of YFP⁺ cells, normalized to the value at day 3 following retroviral transduction. The mean ± SD of at least three independent transductions is shown. *P < 0.05; **P < 0.01 using 1-way ANOVA with Tukey’s multiple comparisons test. B Immunoblot analysis of CYLD in LY10, RIVA and Mino using an antibody raised against a C‐terminal epitope which detects full-length CYLD and a C‐terminal fragment of CYLD (CYLD‐Ct). Cells were transduced with an empty vector (EV) or an expression vector for CYLD (WT or non-cleavable R324A mutant) and sorted for YFP expression. β-actin was used as loading control. C Flow cytometric analysis of LY1 and Z138 cells transduced with an empty vector (EV) or a CYLD (wildtype or R324A mutant) containing bicistronic vector co-expressed with YFP. The percentage of YFP positive cells was followed in time and plotted as the percentage of YFP⁺ cells, normalized to the value at day 3 or day 4 following retroviral transduction. The mean ± SD of three independent transductions is shown. P > 0.05; ns (non-significant) using 1-way ANOVA with Tukey’s multiple comparisons test. D Immunoblot analysis of CYLD in LY1 and Z138 using an antibody raised against a C‐terminal epitope which detects full-length CYLD and a C‐terminal fragment of CYLD (CYLD‐Ct). Cells were transduced with an empty vector (EV) or an expression vector for CYLD (WT or non-cleavable R324A mutant) and sorted for YFP expression. β-tubulin was used as loading control. E Immunoblot analysis of (phosphorylated) IkBα in LY10 and Mino transduced with an empty vector (EV) or a CYLD (wildtype or R324A mutant) expressing vector. Three days after sorting cells were incubated with or without 5 µM proteasome inhibitor MG132 for 3 h before harvesting. β-tubulin was used as loading control. F Heatmap representing the RT-qPCR analysis of NF-κB target gene expression in LY10 and Mino transduced with an empty vector (EV) or an expression vector for CYLD (WT or non-cleavable R324A mutant). Cells were sorted for YFP expression and allowed to recover for 48 h before RNA isolation. RPLP0 was used as an input control and data are normalized to the EV control expression levels. The mean of three independent experiments performed in triplicate is shown. G Immunoblot analysis of (phosphorylated) STAT3 in LY10 transduced with an empty vector (EV) or an expression vector for CYLD (WT or non-cleavable R324A mutant). β-tubulin was used as loading control.

Fig. 5. MALT1-dependent proteolytic cleavage inhibits activity and promotes proteasomal degradation of CYLD.

A Immunoblot analysis of CYLD variants in LY10. Cells were transduced with an empty vector (EV) or an expression vector for CYLD (N-terminal fragment, C-terminal fragment, WT or non-cleavable R324A mutant) and sorted for YFP expression. CYLD was detected using an antibody raised against a C‐terminal epitope which detects full-length CYLD and a C‐terminal fragment of CYLD (CYLD‐Ct), or an antibody against an N‐terminal epitope for detection of the N-terminal fragment (CYLD-Nt). β-tubulin was used as loading control. B Flow cytometric analysis of LY10 cells transduced with an empty vector (EV) or a bicistronic expression vector for CYLD (N-terminal fragment, C-terminal fragment, WT or non-cleavable R324A mutant) co-expressing YFP. The percentage of YFP positive cells was followed in time and plotted as the percentage of YFP⁺ cells, normalized to the value at day 3 following retroviral transduction. The mean ± SD of four independent transductions is shown. P > 0.05; ns (non-significant); *P < 0.05; **P < 0.01 using 1-way ANOVA with Tukey’s multiple comparisons test. C Immunoblot analysis of (phosphorylated) IkBα in LY10 transduced with an empty vector (EV) or an expression vector for CYLD (N-terminal fragment, C-terminal fragment, WT or non-cleavable R324A mutant) expressing vector. Three days after sorting cells were incubated with or without 5 µM proteasome inhibitor MG132 for 3 h before harvesting. β-tubulin was used as loading control. D Immunoblot analysis of CYLD cleavage in LY10 and Mino. Cells were incubated with 100 nM ibrutinib for the indicated time points. β-tubulin was used as loading control. E Immunoblot analysis of endogenous CYLD cleavage in LY10 and Mino. Cells were incubated with 100 nM ibrutinib for 24 h in the presence or absence of 10 uM MG132. To prevent apoptosis, cell lines were co-incubated with 10 µM Q-VD-OPh (QVD). β-tubulin was used as loading control.

Fig. 6. CYLD knockdown promotes cell growth and NF-κB activation.

A Immunoblot analysis of CYLD in HBL1, LY10 and Mino transduced with lentiCRISPR-Cas9 (±sgCYLD) using an antibody raised against a C‐terminal epitope which detects full-length CYLD and a C‐terminal fragment of CYLD (CYLD‐Ct). Cells were treated with 50 nM BTK inhibitor Ibrutinib or 500 nM PKC inhibitor Sotrastaurin for 48 h as indicated. β-tubulin was used as loading control. B HBL1, LY10 and Mino transduced with lentiCRISPR-Cas9 without gRNA (empty vector; EV) or with sgCYLD were treated for 3 days with indicated concentrations of Ibrutinib or Sotrastaurin. The number of viable cells, as determined by 7-AAD staining, was normalized to the untreated condition. The mean ± SD of four independent experiments performed in triplicate is shown. *P < 0.05; **P < 0.01 using 2-way ANOVA with Sidak’s multiple comparisons test. C Immunoblot analysis of (phosphorylated) IkBα in HBL1, LY10 and Mino transduced with lentiCRISPR-Cas9 (±sgCYLD) treated for 48 h with 50 nM Ibrutinib or 500 nM Sotrastaurin as indicated. Cells were incubated with 5 µM proteasome inhibitor MG132 for 3 h before harvesting. β-tubulin was used as loading control. D Immunoblot analysis of (phosphorylated) STAT3 in LY10, HBL1 and Mino transduced with lentiCRISPR-Cas9 (±sgCYLD) treated 48 h with 50 nM Ibrutinib or 500 nM Sotrastaurin as indicated. β-actin was used as loading control.

Fig. 7. Model of the role of CYLD in NF-κB activation in B-cell lymphomas.

Upon B-cell receptor (BCR) ligation, tyrosine residues within the ITAM motifs of CD79 are phosphorylated by the Src-family tyrosine kinase LYN leading to activation of spleen tyrosine kinase (SYK). Subsequently, Bruton’s tyrosine kinase (BTK) is activated and can then phosphorylate phospholipase Cγ2 (PLCγ2). PLCγ2 mediates the formation of second messengers that activate protein kinase Cβ (PKCβ). PKCβ phosphorylates caspase recruitment domain-containing protein 11 (CARD11) provoking a conformational change and allowing CARD11 to interact with B-cell lymphoma 10 (BCL10), and subsequently MALT1. MALT1 is a protease that cleaves various target proteins, including CYLD. In addition, oligomerized MALT1 functions as a scaffolding protein allowing recruitment of the E3 ubiquitin ligase tumor necrosis factor receptor-associated factor 6 (TRAF6). In parallel, Toll-like receptor (TLR) engagement results in MyD88-dependent recruitment of IL-1 receptor-associated kinase-4 (IRAK4) and subsequently IRAK1. IRAK4 phosphorylates IRAK1, which then can associate with TRAF6. BCR/TLR-activated TRAF6 promotes Lys-63-linked ubiquitination of TRAF6 itself as well as transforming growth factor beta-activated kinase 1 (TAK1) and NEMO/IKK-γ. Ubiquitinated TRAF6 binds to adaptor proteins TAB1/2/3, leading to the recruitment and auto-phosphorylation of TAK1. Ubiquitination of NEMO/IKK-γ mediates the recruitment of the IKK subunits to the TAK1/TAB complex, thereby facilitating the phosphorylation of IKK-β by TAK1. IKK-β then phosphorylates IκBα resulting in Lys-48-polyubiquitination and subsequent proteasomal degradation which allows NF-κB dimers to translocate to the nucleus. The deubiquinating enzyme CYLD consists of three conserved cytoskeleton-associated protein glycine-rich (CAP-Gly) domains and a C-terminal catalytic ubiquitin-specific protease (USP) domain that is able to hydrolyze lysine 63-linked ubiquitin chains. CYLD can hydrolyze Lys-63-linked polyubiquitin chains of TRAF6, TAK1 and/or NEMO/IKK-γ, thereby suppressing NF-κB activation. Accordingly, MALT1-dependent cleavage of CYLD substantially reduces its functionality and initiates its proteasomal degradation, thereby promoting cell growth and NF-κB activation.