Fig 1.
(A) WT Constructs. Multiple sites of ubiquitination in IKKβ were identified by LC/MS-MS which were removed by site-directed mutagenesis. Initially, the IKKβ 4KR mutant was constructed with the mutations K301R, K418R, K555R, and K703R. Further analysis revealed additional ubiquitination sites, of which five more were mutated to create IKKβ 9KR with the additional mutations K310R, K428R, K509R, K614R, and K641R. (B) K171E Constructs. The kinase activating mutation K171E was introduced into the constructs shown in (A) to create IKKβ K171E, IKKβ K171E 4KR, and IKKβ K171E 9KR. (C) K147R/K171E Constructs. The mutation K147R, identified as a major site of K63-linked ubiquitination and required for IKKβ kinase activity [5], was introduced into the constructs shown in (B) to create IKKβ K147R/K171E, IKKβ K147R/K171E 4KR, and IKKβ K147R/K171E 9KR. The ubiqutin-like domain (ULD), the scaffold/dimerization domain (SDD) which contains the leucine zipper (LZ) and helix-looop-helix (HLH) regions, and NEMO binding domain (NBD) are indicated.
Fig 2.
Phosphorylation and ubiquitination of IKKβ.
HEK293T cells were transfected with the IKKβ mutants shown in Fig 1, in the same order, together with HA-Ub3. Cells were lysed in RIPA and proteins separated by SDS-PAGE. (A) Lysates were examined for activation of IKKβ kinase activity by immunoblotting for phospho-S177/S181 IKKβ. Total IKKβ expression is shown below. (B) The same lysates as in (A) were examined for STAT3 signaling by immunoblotting for phospho-Y705-STAT3. Total STAT3 is shown below. (C) HA-tagged ubiquitinated proteins from the same lysates were collected by immunoprecipitation, and HA-Ub-IKKβ was detected by immunoblotting for IKKβ.
Fig 3.
Identification of ubiquitination sites in IKKβ.
As described in Materials and Methods, ubiquitination sites in IKKβ were detected by LC-MS/MS. All Lys residues in the IKKβ primary sequence are indicated on the x-axis. For each ubiquitination site detected, the magnitude of ubiquitination is indicated as a percentage of the total IKKβ ubiquitination. Sites representing < 2.5% of the total IKKβ ubiquitination are not shown. (A) IKKβ K171E. As described previously [5], K147 (shown in red) was identified as the primary site of K63-linked ubiquitination, and the sites K301, K418, K555 and K703 were also identified. (B) IKKβ K171E 4KR. The sites K301, K418, K555 and K703 identified in (A) were removed by mutagenesis, as shown by each red X. Many new sites of ubiquitination were now detected as shown. (C) IKKβ K171E 9KR. Five additional ubiquitination sites identified in (B) were again mutated, including K310R, K428R, K509R, K614R, and K641, as shown by each red X. One minor site identified in (B), K531, now became more prominent, and new sites at K018 and K238 now appeared. However, the ubiquitination site K147 now appears as the major site observed.
Fig 4.
Examination of STAT3 activation by activated mutants of IKKβ.
(A) Requirement for K63-linked ubiquitination. HEK293T cells expressing indicated IKKβ constructs were treated with 2μM NSC697923 for 2 h to inhibit UBE2N (Ubc13). STAT3 activation (1st panel) was detected by immunoblotting for phospho-Tyr705-STAT3, with total STAT3 shown below. The 3rd panel shows IKKβ kinase activation by immunoblotting for phospho-S177/S181 IKKβ, with total IKKβ shown below. The addition of exogenous IL-6 (10 ng/ml, 10 min) results in robust STAT3 activation (Lane 10) which is only marginally affected by treatment with the inhibitor NSC697923 (Lane 9). An empty lane is included between Lanes 8 and 9 due to the intensity of the signal in Lane 9. (B) Requirement for TAK1 activation. HEK293T cells expressing various combinations of IKKβ, TAK1 and TAB1 proteins were treated with 10 μM (5Z)-7-Oxozeaenol for 2 h to inhibit TAK1 activity. 1st Panel: STAT3 activation is shown by immunoblotting for phospho-Tyr705-STAT3 and reveals that TAK1 activity is required for STAT3 activation, whether in response to IKKβ K171E (compare Lanes 3 and 8), or IKKβ WT activated by overexpression of TAK1 + TAB (compare Lanes 4 and 9). Total STAT3 is shown immediately below. 3rd Panel: MAPK activation, shown by immunoblotting for phospho-T202/Y204-MAPK, is also dependent upon TAK1 activity similar to STAT3 activation. Total MAPK is shown immediately below. 5th Panel: IKKβ kinase activation is shown by immunoblotting for phospho-S177/S181 IKKβ, with total IKKβ shown immediately below. Panels 7–9: Controls are presented for TAB and TAK1 expression and activation using phospho-T184/187-TAK1, total TAK1, and myc (9E10) to detect myc-tagged TAB. (C) Requirement for JAK activity. HEK293T cells expressing indicated IKKβ mutants were treated with the Janus kinase inhibitor JAK Inhibitor 1 (2 μM for 2 h). STAT3 activation is shown in the top panel by immunoblotting for phospho-Tyr705-STAT3 and reveals that JAK activity is required for STAT3 activation in response to activated IKKβ mutants (Lanes 3–5 compared with Lanes 9–11). The addition of exogenous IL-6 (10 ng/ml, 10 min) results in robust STAT3 activation (Lane 14) which is completely blocked by JAK Inhibitor 1 (Lane 13). (D) Requirement for GP130. A requirement for gp130 function, which serves as the β subunit of the IL-6-Receptor, was examined using the gp130 inhibitor SC144. HEK293T cells were starved and treated with 25 μM SC144 for ~20 h prior to a 2 h treatment with conditioned media from HEK293T cells expressing IKKβ mutants. STAT3 activation is shown in the top panel by immunoblotting for phospho-Tyr705-STAT3 and reveals that gp130 is required for STAT3 activation in response to activated IKKβ mutants (Lanes 3–5 compared with Lanes 9–11). The addition of exogenous IL-6 (10 ng/ml, 2 h) results in robust STAT3 activation (Lane 14) which is completely blocked by the gp130 inhibitor SC144 (Lane 13). Lysates from cells expressing the IKKβ mutants that generated the conditioned media were examined by immunoblotting to confirm IKKβ expression (data not shown). (E) Specific increase of K63-Ubiquitin-IKKβ by activated IKKβ mutants. HEK293T cells expressing indicated IKKβ mutants together with HA-tagged K63-only-Ubiquitin were immunoprecipitated with HA antiserum, and then immunblotted to detect total IKKβ. The higher MW bands of IKKβ suggest that activation by K171E, K171E 4KR, or K171E 9KR results in dramatically increased K63-conjugated Ub complexes, in comparison with IKKβ WT (compare Lanes 3–6 with Lane 2). This increase is largely, but not completely blocked, by the addition of 2μM NSC697923 for 2 h to inhibit UBE2N (Ubc13). (F) Requirement for K63-linked ubiquitination in IKKβ-deficient cells. IKKβ-deficient murine 3T3 cells expressing the indicated IKKβ mutants were treated with 5μM NSC697923 for 2 h to inhibit UBE2N (Ubc13)-catalyzed K63-linked ubiquitination. STAT3 activation (1st panel) was detected by immunoblotting for phospho-Tyr705-STAT3, with total STAT3 shown below. The 3rd panel shows IKKβ kinase activation by immunoblotting for phospho-S177/S181 IKKβ, with total IKKβ shown below.
Fig 5.
Assessing the oncogenic potential of the K171E mutation in IKKβ on the IL-6 -dependent INA-6 cell line.
The human myeloma cell line INA-6 is completely dependent on exogenous IL-6 for growth and proliferation. (A) IL-6 concentration dependence of the INA-6 cells. Triplicate cultures of cells were grown in RPMI 1640 with 10% FBS and various concentrations of IL-6 (5, 1, 0.2, 0.04 and 0 ng/ml). Duplicate samples were collected at 1, 2, 4 and 6 days and assayed by MTT metabolic assay indicating the number of viable cells. Error bars show the standard deviation (B) Proliferation of INA-6 cells treated with conditioned media. Triplicate cultures of INA-6 cells were incubated in media collected from HEK293T cells expressing IKKβ derivatives. 10% FBS was added to the starvation media. Duplicate samples were collected at 1, 2, 3, 4 and 5 days and assayed by MTT metabolic assay. Control samples were treated 5, 0.04 and 0 ng/ml of IL-6 as indicated. Error bars show the standard deviation (C) STAT3 activation induced by conditioned media from cells expressing IKKβ mutants. INA-6 cells were incubated as in B. for 48 h. Cells were lysed and immunoblotted for P-STAT3. Triplicate immunoblots were quantitated with 5 ng/ml of IL-6 set at 100%. Error bars show the standard deviation. (D). Detection of IL-6 in conditioned media from cells expressing IKKβ mutants. Using a sensitive ELISA kit for IL-6, the conditioned media was assayed in triplicate from each of two independent experiments. Concentrations were determined from a standard curve of IL-6.
Fig 6.
Network analysis in response to activated mutants of IKKβ.
(A) Proteomic Analysis Pipeline. Schematic representation of the analysis steps for proteomic mass spectrometry data that uses MaxQuant Label Free Quantification values of detected proteins as input and generates a list of significantly differentially abundant proteins (Benjamini-Hochberg adjusted p-value < 0.05) as output. (B) Top Biological Categories in Proteomic Analysis. Gene Ontology, KEGG, Panther, and Reactome pathways/categories identified as significantly enriched by WebGestalt (FDR < 0.05) for the list of differentially abundant proteins identified in (A). (C) Pathway Level Representation of Select Differentially Abundant Proteins. Representative sub-processes of the significantly enriched Ubiquitin Mediated Proteolysis KEGG category from (B) that have a high percentage overlap with the differentially abundant protein list. Red boxes indicate proteins with significantly different abundance between the IKKβ WT and IKKβ K171E 9KR sample groups. (D) Top GSEA Result for IKKβ WT. Using the list of 958 proteins identified in the two groups as input, the GSEA identified Gene set (KEGG Ubiquitin Mediated Proteolysis) with the highest Enrichment Score (ES) in IKKβ WT along with a heatmap of the top genes contributing to the ES. (E) Top GSEA Result for IKKβ K171E 9KR. As in (D), the Gene set (Reactome G1 to S Transition) with the highest ES in IKKβ K171E 9KR along with the corresponding heatmap. In the heatmaps, the range of colors (red, pink, light blue, dark blue) shows the range of expression values for each gene in each sample (high, moderate, low, lowest).
Fig 7.
Signaling pathways activated by K171E IKKβ.
A model is presented for signaling by the oncogenic mutation K171E of IKKβ, identified in hematological malignancies, which integrates the data presented here. Normally, inflammatory cytokines activate NFκB and signal the K63-linked ubiquitination of TAK1 and NEMO, leading to the activation of the IKK complex and NFκB nuclear translocation. The K171E mutation of IKKβ instead leads to the K63-linked ubiquitination of K147 by UBE2N (Ubc13), as shown by the inhibitor NSC697923, and this activation is dependent upon the activity of TAK1, as shown by the inhibitor 5Z-7-Oxozeaenol. The IKKβ K171E mutants establish an autocrine loop dependent upon the secretion of IL-6, binding to the IL-6 receptor, as shown by the inhibitor SC144 which inhibits the β-subunit gp130. The involvement of the JAK kinase family (JAK1, JAK2, JAK3, TYK2) in this system is shown by the inhibitor JAK Inhibitor 1, which also inhibits the appearance of phospho-STAT3 in response to IKKβ K171E mutants.