Notably, we observed a markedly higher level of ROC1 when PDLIM2 was co-expressed (Fig. 4A). Despite the significantly low expression when being expressed alone, ROC1 expression level was increased by the proteasome inhibitor MG132. On the other hand, the high level of ROC1 was not further increased by MG132 when it was co-expressed with PDLIM2. Consistently, PDLIM2 drastically increased ROC1 protein stability in the pulse-chase assays (Fig. 4B). These data suggested that PDLIM2 prevents ROC1 from its rapid proteasomal degradation.

PDLIM2 stabilizes ROC1 and synergizes with CUL1 and β-TrCP for the maximal ROC1 stabilization. A IB assays showing MG132 accumulation of ROC1 proteins in the absence but not presence of PDLIM2 in 293 cells. B In vivo protein stability assays showing PDLIM2 stabilization of ROC1 proteins in 293 cells. Cell extracts containing the same amount of ROC1 proteins at the beginning of chasing were used. C IB assays showing the synergy of PDLIM2 with CUL1 and β-TrCP for the maximal ROC1 stabilization in 293 cells. D Modeling PDLIM2 as the E5 for RelA ubiquitination and proteasomal degradation

CUL1 also stabilized ROC1, although at a much lower level compared to PDLIM2 (Fig. 4C, lane 3 vs. lanes 2 and 1). Interestingly, β-TrCP showed a similar effect in ROC1 stabilization as CUL1 (Fig. 4C, lane 4). The effect of β-TrCP would be indirectly through the endogenous CUL1 and SKP1 within the cells. Indeed, CUL1 and β-TrCP co-expression showed a synergy in stabilizing ROC1, to a level similar to that increased by PDLIM2 (Fig. 4C, lane 7 vs. lane 2). When CUL1 or β-TrCP was simultaneously expressed with PDLIM2, remarkably, ROC1 levels were further increased (Fig. 4C, lanes 5 and 6). The expression level of ROC1 reached to the highest when CUL1, β-TrCP and PDLIM2 all were expressed concomitantly (Fig. 4C, lane 8). These data suggested that in addition to recruiting RelA to the SCFβ-TrCP ubiquitin ligase, PDLIM2 plays an important role in stabilizing ROC1 and enhancing the formation of the functional SCFβ-TrCP ubiquitin ligase complex.

The PDLIM2/RelA axis has been linked to numerous pathogenic conditions and cancers in particular [2,3,4,5,6,7,8]. However, we do not know, until now, how PDLIM2 promotes nuclear RelA ubiquitination and proteasomal degradation, although PDLIM2 has been suggested to be a nuclear ubiquitin ligase [1]. The studies here show that PDLIM2 exerts the important function indirectly through the SCFβ-TrCP ubiquitin ligase. Besides this previously unidentified role, the SCFβ-TrCP ubiquitin ligase is well-known to ubiquitinate NF-κB inhibitors for degradation in the cytoplasm, freeing RelA and other NF-κB members to translocate to the nucleus and regulate gene transcription [7]. Thus, the SCFβ-TrCP ubiquitin ligase has two opposite roles in NF-κB regulation, ensuring a rapid but transient RelA activation in response to NF-κB stimuli. In the cytoplasm, it initiates NF-κB activation in response to NF-κB stimuli, and but in the nucleus, it turns off RelA activation, with the necessary help of PDLIM2.

Protein ubiquitination involves the sequential concerted action of E1, E2, and E3. This reaction starts with formation of a thiolester linkage between E1 and ubiquitin, followed by transfer of ubiquitin to an E2. Finally, E3 recruits a specific protein substrate to the E2-ubiquitin, where the ubiquitin is conjugated to a specific lysine in the protein substrate [9]. The serial actions of E1, E2, and E3 result in the poly-ubiquitination of the substrate. In certain cases, however, a ubiquitin-chain elongation factor named E4 is needed to bind to the oligo-ubiquitylated substrates for multi-ubiquitin chain assembly by E1, E2, and E3, yielding long ubiquitin chains [10].

PDLIM2 may represent a distinct and novel class of factors important for protein ubiquitination, acting as E3 enhancers, E5s (Fig. 4D). They recruit substrates that cannot be recognized by E3s, as indicated by the PDLIM2 recruitment of nuclear RelA to the SCFβ-TrCP ubiquitin ligase in the nucleus. They may also facilitate the formation and stabilization of E3s, especially the multi-subunit ones. In this regard, ROC1, the RING finger component of the SCFβ-TrCP ubiquitin ligase, is more stable within the E3 complex compared to being alone. However, PDLIM2 shows a much stronger ability in stabilizing ROC1 and synergizes with other components of the E3 for optimal ROC1 stabilization. The ROC1 stabilization function of PDLIM2 is dispensable for the ubiquitination and degradation of the targets of the SCFβ-TrCP ubiquitin ligase with β-TrCP degron sequence, such as the NF-κB inhibitor IκBα. It is highly plausible that binding to the target proteins further solidifies the SCFβ-TrCP complex, thereby preventing the dissociation from the complex and subsequent ubiquitination and degradation of ROC1.

In summary, the present studies demonstrate PDLIM2 as a novel E5 ubiquitin ligase enhancer that stabilizes ROC1 and recruits the ROC1-SCFβ-TrCP ubiquitin ligase to ubiquitinate and degrade nuclear RelA. They provide new mechanistic insights into how PDLIM2 serves as a common tumor suppressor and a critical immune regulator. They also expand our knowledge on the complex regulation and action of the ubiquitination and NF-κB pathways.