WSB1/2 target chromatin-bound lysine-methylated RelA for proteasomal degradation and NF-κB termination

Abstract

Proteasome-mediated degradation of chromatin-bound NF-κB is critical in terminating the transcription of pro-inflammatory genes and can be triggered by Set9-mediated lysine methylation of the RelA subunit. However, the E3 ligase targeting methylated RelA remains unknown. Here, we find that two structurally similar substrate-recognizing components of Cullin-RING E3 ligases, WSB1 and WSB2, can recognize chromatin-bound methylated RelA for polyubiquitination and proteasomal degradation. We showed that WSB1/2 negatively regulated a subset of NF-κB target genes via associating with chromatin where they targeted methylated RelA for ubiquitination, facilitating the termination of NF-κB-dependent transcription. WSB1/2 specifically interacted with methylated lysines (K) 314 and 315 of RelA via their N-terminal WD-40 repeat (WDR) domains, thereby promoting ubiquitination of RelA. Computational modeling further revealed that a conserved aspartic acid (D) at position 158 within the WDR domain of WSB2 coordinates K314/K315 of RelA, with a higher affinity when either of the lysines is methylated. Mutation of D158 abolished WSB2's ability to bind to and promote ubiquitination of methylated RelA. Together, our study identifies a novel function and the underlying mechanism for WSB1/2 in degrading chromatin-bound methylated RelA and preventing sustained NF-κB activation, providing potential new targets for therapeutic intervention of NF-κB-mediated inflammatory diseases.

Graphical Abstract

Figure 1.

Methylation of RelA by Set9 is critical for ubiquitination of chromatin-bound RelA. (A) Schematic illustration of the fractionation process for the preparation of cytoplasmic (S2), nuclear (S3), and chromatin-enriched (S4) fractions. (B) NIH3T3 cells were pulse-stimulated with TNF-α for 15 min followed by treatment with proteasome inhibitor MG132 for 5 h. The cytoplasmic, nuclear, and chromatin-enriched fractions were prepared as in (A) and heat-denatured in the presence of 1% SDS for RelA immunoprecipitation (IP) and immunoblotting (IB) for ubiquitination. (C) NIH3T3 cells were transfected with control or Set9 siRNAs, and TNF-α-induced ubiquitination of RelA in chromatin-enriched fraction was assessed as described in (B) with antibodies against ubiquitin and K48-linked or K63-linked ubiquitin chains. (D) TNF-α-induced ubiquitination of RelA in chromatin-enriched fraction from RelA or RelA-K314/315R reconstituted RelA-deficient NIH3T3 cells was assessed as described in (C). The experiments in (B-C) were repeated 3 times, and those in (D) repeated twice. A representative result for each experiment is shown.

Figure 2.

WSB1/2 promote the ubiquitination of methylated RelA. (A) HEK293T cells were transfected with plasmids encoding T7-tagged WT RelA or RelA-K314/315R and Flag-tagged Set9 together with 4 SOCS family proteins. Whole cell lysates were prepared and immunoblotted for indicated proteins. (B) HEK293T cells were transfected with the expression vectors for T7-RelA and Myc-tagged WSB1 or WSB2. At 24 h after transfection, cells were treated with MG-132 (20 μM) for 4 h, and whole cell lysates were immunoblotted with indicated antibodies. (C, D) HEK293T cells were transfected with plasmids encoding T7-tagged WT RelA or its DNA-binding defective mutant (RelA-Y36A/E39D) and Myc-tagged WSB1 or WSB2 (C), or Flag-tagged SOCS1 (D); Whole cell lysates were prepared and immunoblotted with indicated antibodies. (E, F) HEK293T cells were transfected with plasmids encoding T7-His- RelA, HA-ubiquitin, Flag-tagged Set9 or catalytic inactive Set9-H297A together with Myc-tagged WSB1 (E) or WSB2 (F). After 30 h cells were treated with MG-132 for 5 h and heat-denatured in the presence of 1% SDS for affinity-purification with Ni-NTA resin. Purified RelA was immunoblotted with anti-HA antibodies for ubiquitination. Levels of ubiquitin, RelA, Set9 and WSB1/2 are shown as input in the lower panels. (G) U2OS cells were transfected with a control siRNA or 2 siRNAs against WSB1 or WSB2 and pulse stimulated with TNF-α for 15 min followed by treatment with MG-132 for 5 h. RelA was immunoprecipitated from denatured chromatin-enriched fraction and immunoblotted for total and K48-linked ubiquitination. Fold changes of RelA ubiquitination were normalized with the levels of immunoprecipitated RelA and shown below the blots (E–G). The experiments in (B–D) were repeated 3 times, and those in (A) and (E–G) repeated at least twice. A representative result for each experiment is shown.

Figure 3.

WSB1/2 interact with methylated K314/315 of RelA through their WDR domains. (A) GST pull-down assay was performed with recombinant GST, GST-WSB1 or GST-WSB2 and whole cell lysates from HEK293T cells transfected with T7-tagged WT RelA or RelA-K314/315R and Flag-tagged Set9 as indicated. The associated RelA was assessed by IB with anti-T7 antibodies. (B) Approximately 100-fold excess molar of unmethylated or K314/315-methylated RelA peptides (a.a. 308–320) was added in GST pull-down assay performed with whole cell lysate from HEK293T cells transfected with T7-tagged WT RelA and Flag-tagged Set9. RelA associated GST-WSB1 (left) or GST-WSB2 (right) was assessed as in (A). (C) GST pull-down assay with full-length (FL), WDR or SOCS domains of WSB1 or WSB2 and the whole cell lysate from HEK293T cells transfected with T7-tagged WT RelA and Flag-tagged Set9 was performed as in (A). Asterisks to the left of each band on the two GST blots indicate the proteins of GST and GST-fusions of WSB1 or WSB2. (D) GSH sepharose beads-bound GST-WSB2 WDR protein was incubated with biotin-labeled unmethylated or K314/315-methylated RelA peptides (a.a. 308–320) and the GST-WSB2 WDR protein was eluted with GSH and the associated peptides were spotted onto a nitrocellulose membrane and assessed via dot blot analysis with anti-biotin and anti GST antibodies. Ratios of methylated peptides to unmethylated peptides bound to GST-WSB2 WDR at each dilution were quantitated and shown below the blot. The experiments in (A), (C) and (D) were repeated at least 3 times, and those in (B) repeated twice. A representative result for each experiment is shown.

Figure 4.

WSB2 negatively regulates the expression of NF-κB target genes. (A–D) U2OS cells transduced with a control shRNA or a WSB2 shRNA were stimulated with or without TNF-α for 5 h and total RNA was extracted for bulk RNA-sequencing. (A) Principal component analysis of the sequencing results. (B) Scatter plotting of up-regulated [log₂ (fold change) >1], down-regulated [log₂ (fold change) < 1], and unchanged [ 1 < log2 (fold change) <1] genes in control and WSB2-KD U2OS cells 5 h after TNF-α stimulation. (C) Gene set enrichment analysis was performed to determine the most significantly up-regulated pathways in TNF-α-stimulated WSB2-KD cells. (D) A heat map showing WSB2 and the top 24 NF-κB target genes up-regulated in TNF-α-stimulated WSB2-KD cells. (E) U2OS cells transduced with control or two different WSB2 shRNAs were stimulated with TNF-α for 5 h and the expression of indicated genes was measured by quantitative RT-PCR. (F) Levels of TNF-α-induced IL-8 and CXCL3 proteins were measured by ELISA in U2OS cells transduced with control or WSB2 shRNAs. (G) U2OS cells transduced with control or WSB2 shRNAs were stimulated with TNF-α for indicated time points and the expression of the indicated genes was measured by quantitative RT-PCR. Results in (E), (F) and (G) are shown as means ± SD from triplicate experiments. *P < 0.05; **P< 0.01; ***P < 0.001. Statistical analysis was performed using unpaired two-tailed Student's t-test for (E) and (F), and nonlinear regression for (G).

Figure 5.

WSB2 is recruited to the promoters of NF-κB target genes. (A) U2OS cells stably expressing vector or Flag-tagged WSB2 were stimulated with TNF-α for indicated time points. ChIP assays were performed using anti-Flag antibody or non-specific mouse IgG antibodies and probed for the promoters of indicated genes. (B, C) U2OS cells stably expressing control or WSB2 shRNAs were stimulated with TNF-α for indicated time points. ChIP assays were performed with antibodies against RelA (B) or methylated K314/315 of RelA (C) and probed for the promoters of indicated genes. Results are expressed as the ratio of RT-PCR signal from ChIP with specific antibodies to that from mock ChIP with IgG and shown as means ± SD from triplicate experiments. * P < 0.05; ** P < 0.01; *** P < 0.001; ns, no significant. Statistical analysis was performed using unpaired two-tailed Student's t-test.

Figure 6.

Computational modeling and experimental verification of the interaction between WSB2 and methylated RelA. (A) Alignment of WSB1 (orange) and WSB2 (blue) structures predicted by AlphaFold2. D158 in the WDR domain of WSB2 and D175 in the WDR domain of WSB2 are perfectly overlapped. (B) Structural model of human WSB2 in complex with a K314 monomethylated RelA peptide (312-IMKme1KS-316). D158 in Repeat 3 of WSB2 WDR domain coordinates methylated K314. (C) A side view of the model shown in (B). (D) Mechanistic sketch generated by Marvin Sketch showing that the Nϵ atom of monomethylated K314/315 forms a hydrogen bond with a water molecule which in turn forms a hydrogen bond with the carboxylate oxygen of E28 in WSB2. (E) GST pull-down assay with GST, GST-tagged WT WSB2, WSB2-E28A or WSB2-D158A and the whole cell lysate from HEK293T cells transfected with T7-tagged RelA and Flag-tagged Set9 was performed as in Figure 3C. (F) HEK293T cells were transfected with indicated plasmids and treated with 10 μM MG-132 for 5h before harvesting. After lysis under a denaturing condition, one half of each lysate was used for Ni-NTA purification and ubiquitination assay as in Figure 2E, and the other half was immunoprecipitated with anti-HA antibodies for immunoblotting with anti-T7 antibodies. Levels of tagged ubiquitin, RelA, Set9 and WSB2 are shown as input in the lower panels. The experiments in (E) were repeated 3 times, and those in (F) repeated twice. A representative result for each experiment is shown.