Phosphorylation of the tumor suppressor CYLD by the breast cancer oncogene IKKepsilon promotes cell transformation

Abstract

The noncanonical IKK family member IKKepsilon is essential for regulating antiviral signaling pathways and is a recently discovered breast cancer oncoprotein. Although several IKKepsilon targets have been described, direct IKKepsilon substrates necessary for regulating cell transformation have not been identified. Here, we performed a screen for putative IKKepsilon substrates using an unbiased proteomic and bioinformatic approach. Using a positional scanning peptide library assay, we determined the optimal phosphorylation motif for IKKepsilon and used bioinformatic approaches to predict IKKepsilon substrates. Of these potential substrates, serine 418 of the tumor suppressor CYLD was identified as a likely site of IKKepsilon phosphorylation. We confirmed that CYLD is directly phosphorylated by IKKepsilon and that IKKepsilon phosphorylates serine 418 in vivo. Phosphorylation of CYLD at serine 418 decreases its deubiquitinase activity and is necessary for IKKepsilon-driven transformation. Together, these observations define IKKepsilon and CYLD as an oncogene-tumor suppressor network that participates in tumorigenesis.

Figure 1. Identification of the optimal IKKε phosphorylation motif

(A) NIH-3T3 cells were stably transduced with Flag-IKKε, kinase-dead Flag-IKKε K38A, or GFP control via lentiviral infection. Immunoblot analysis shows similar IKKε expression. F-IKKε = Flag-IKKε. (B) Flag-IKKε or Flag-IKKε K38A-expressing NIH-3T3 cells generated in (A) were assayed for anchorage-independent growth. Colony formation was examined after 21 days. Expression of wild-type IKKε, but not IKKε K38A, induces anchorage-independent growth. Error bars depict standard deviation (SD) for 3 independent experiments. (C-D) Recombinant IKKε or K38A was used to phosphorylate 198 peptide libraries in individual kinase assays. The general sequence for these libraries is Y-A-X-X-X-Z-X-S/T-X-X-X-X-A-G-K-K-biotin (Z = fixed amino acid, X = equimolar mixture of amino acids excluding Ser, Thr, and Cys). WT IKKε (C) gave a strong, consistent motif. No motif could be identified using IKKε K38A (D). (E) Individual peptides were phosphorylated with recombinant IKKε. Phosphorylation is shown as a percentage of the rate of phosphorylation of IKKε-Tide, the optimal peptide substrate for IKKε as determined in (C). Alteration of critical amino acids to alanine decreased peptide phosphorylation. A peptide corresponding to the sequence surrounding Ser32 and Ser36 of IκBα was not efficiently phosphorylated by IKKε. Error bars depict SD. (F) Primary and secondary selections determined from IKKε phosphorylation motif

Figure 2. CYLD co-immunoprecipitates with and is phosphorylated by IKKε

(A) Myc-CYLD was cotransfected with GST-IKKε WT or K38A into HEK-293T cells. Myc-CYLD was immunoprecipitated and immune complexes were blotted with an IKKε phospho-substrate antibody. CYLD was recognized by the phospho-substrate antibody when cotransfected with WT IKKε, but not IKKε K38A. (B) Myc-CYLD was transfected into HEK-293T cells alone or with GST-IKKε. CYLD was immunoprecipitated with an anti-Myc antibody. Immune complexes were divided into two samples and treated with calf-intestinal phosphatase (CIP) or water (as a control) for 45 min at 37 °C. The IKKε phospho-substrate antibody no longer recognized CIP-treated CYLD. (C) Recombinant GST-IKKε or IKKε K38A was purified from HEK-293T cells. Myc-CYLD was transfected separately into HEK-293T cells and immunoprecipitated. CYLD immune complexes were incubated in the presence of γ-³²P-ATP with IKKε WT or K38A. Autoradiograph analysis showed phosphorylation of CYLD by WT IKKε, but not IKKε K38A. (D-E) Myc-CYLD was cotransfected into HEK-293T cells with GST-IKKε WT or GST-IKKε K38A. (D) Lysates were subjected to immunoprecipitation with an anti-Myc antibody and immune complexes were blotted with an anti-GST antibody. (E) Lysates were subjected to precipitation with glutathione-conjugated Sepharose beads and precipitates were blotted with an anti-Myc antibody. CYLD interacts with both WT and kinase-dead IKKε.

Figure 3. CYLD is phosphorylated by IKKε at Ser418

(A) Myc-CYLD was cotransfected into HEK-293T cells with GST-IKKε. Myc-CYLD was immunoprecipitated and the immune complex was subjected to SDS-PAGE followed by Coomassie staining (Figure S2). The band corresponding to CYLD was excised from the gel, and digested with trypsin and chymotrypsin. Phosphorylation sites were mapped by microcapillary LC/MS/MS, resulting in 85% coverage of the CYLD amino acid sequence. A phosphopeptide consistent with phosphorylation at Ser418 was identified. (B) Ser418 of CYLD and surrounding residues are evolutionarily conserved. (C) Site-directed mutants were created in which CYLD residues corresponding to the IKKε phosphorylation motif were changed to an alanine. These mutants were cotransfected into HEK-293T cells with GST-IKKε. Lysates were subjected to immunoprecipitation with an anti-Myc antibody and immune complexes were blotted with an anti-IKKε phospho-substrate antibody. CYLD S418A was no longer recognized by the IKKε phospho-substrate antibody. (D) Myc-CYLD WT or Myc-CYLD S418A was cotransfected into HEK-293T cells with GST-IKKε. Lysates were subjected to immunoprecipitation with an IKKε phospho-substrate antibody. Immune complexes were blotted with an anti-Myc antibody. WT CYLD was immunoprecipitated by the anti-IKKε phospho-substrate antibody, but CYLD S418A was not. (E) Endogenous CYLD was immunoprecipitated from IKKε-transformed NIH-3T3 cells generated in Figure 1A. Immune complexes were blotted with an IKKε-phospho-substrate antibody. CYLD isolated from cells expressing WT IKKε, but not IKKε K38A, was recognized by the phospho-substrate antibody.

Figure 4. Suppression of CYLD protein expression is sufficient to induce transformation

(A) Two distinct shRNA constructs targeting murine CYLD (shCYLD1 and shCYLD2) were transduced into NIH-3T3 cells via lentiviral infection. Two shRNAs targeting GFP (shGFP1 and shGFP49) were used as controls. Both shCYLD1 and shCYLD2 induce substantial knockdown of endogenous CYLD. (B-C) Anchorage-independent growth of NIH-3T3 cells was assessed following suppression of endogenous CYLD with shCYLD1 or shCYLD2 for 21 days. (B) Colony number and (C) Colony formation (at 10× magnification) were examined. Error bars depict SD for 3 independent experiments.

Figure 5. CYLD phosphorylation at serine 418 is important for IKKε-mediated transformation

(A) IKKε-transformed NIH-3T3 cells were stably transduced with WT-CYLD, CYLD S418A, CYLD S772A, or vector control. Immunoblot analysis shows similar CYLD and IKKε expression in IKKε-transformed cells. F-IKKε = Flag-IKKε. (B) Cells generated in (A) were assayed for anchorage-independent growth. Colony formation was examined after 21 days. Expression of CYLD S418A suppresses IKKε-induced anchorage-independent growth. (C) IKKε-transformed NIH-3T3 cells expressing WT CYLD, CYLD S418A, CYLD S772A, or GFP control were subcutaneously introduced into immunodeficient mice. Relative tumor volume was assessed at 21 days post injection. Expression of CYLD S418A results in a statistically-significant decrease in tumor growth. Asterisks indicate statistical significance (p≤ 0.05).

Figure 6. Phosphorylation of CYLD at Ser418 decreases CYLD activity

(A) HEK-293T cells were transfected with HA-Ub K63, Myc-TRAF2, GFP-CYLD WT or S418A, and GST-IKKε. Cells were stimulated with 10 ng/mL TNFα for 10 minutes prior to lysis. TRAF2 was immunoprecipitated using an anti-Myc antibody, and immune complexes were blotted with an anti-HA antibody. IKKε expression decreased CYLD activity (as measured by TRAF2 ubiquitination) when expressed with WT CYLD. The activity of CYLD S418A was not affected by the presence of IKKε. (B) HEK-293T cells were transfected with HA-Ubiquitin, Myc-NEMO, OMNI-RIP2, GFP-CYLD WT or S418A, and GST-IKKε. NEMO was immunoprecipitated with an anti-Myc antibody and immune complexes were blotted with an anti-HA antibody. IKKε expression decreased CYLD activity (as measured by NEMO ubiquitination) when expressed with WT CYLD. The activity of CYLD S418A was not affected by the presence of IKKε. (C) MCF-7 cells were transfected with an NF-κB-luciferase reporter gene, CMV-Renilla luciferase (to standardize transfection efficiency), and TRAF2 alone or in combination with a subsaturating amount of Myc-CYLD WT, Myc-CYLD S418A, or Myc-CYLD S772A. An NF-κB–luciferase reporter assay was performed 24 h posttransfection. TRAF2 increased NF-κB reporter activation 14.9 fold. CYLD S418A inhibited the NF-κB response induced by TRAF2 more efficiently than WT CYLD or CYLD S772A. Error bars depict SD for four separate experiments. Asterisks indicate statistical significance (p<0.05). Equivalent expression of TRAF2 and CYLD is shown in Figure S4. (D) IKKε-transformed NIH-3T3 cells stably expressing WT or mutant CYLD were transiently transfected with pTRH1-NF-κB-Luciferase reporter construct and pRL-SV40-Renilla. Additional ectopic WT and mutant CYLD was also introduced as described in the Experimental Procedures. Luciferase activity was assessed 2 days post transfection. Error bars represent SD for 4 independent experiments. (E) Anchorage-independent growth of NIH-3T3 cells was assessed following suppression of endogenous CYLD with shCYLD1 and simultaneous expression of an NF-κB “superrepressor” (IκBα S32,36A) for 21 days. Error bars depict SD for 3 independent experiments.