In this study, we found that LTβR knockdown induces senescence in cancer cells, which is mediated by the upregulation of p53 and p21 expression. We propose that LTβR regulates p53 activity by preventing MDMX nuclear translocation and degradation. These findings provide new insights into the potential of LTβR as a therapeutic target in cancer.
While the role of LTβR as a receptor is well-documented, recent studies have linked high LTβR expression with poor prognosis in various cancers [20], suggesting its critical role in key cellular processes. To investigate the impact of LTβR knockdown on cellular phenotype, we used specific siRNAs to knock down LTβR expression for 48 h in A375, A549, B16F10, and J774 cell lines (Fig. S1A). LTβR knockdown led to increased cell size and a reduced cell number (Fig. S1B, C). Immunofluorescence staining revealed decreased expression of Ki67, a proliferation marker, in LTβR knockdown cells (Fig. S1D). Further analysis showed an increased proportion of cells in the G1 phase and elevated SA-β-Gal activity, which was detected using the senescence green probe (Fig. S1E, F). Together, these results suggest that LTβR knockdown may induce cellular senescence, which is characterized by growth arrest and apparent morphological changes.
Among the many pathways that drive cellular senescence, we focused on exploring the well-established p53-mediated senescence pathway. To confirm whether the senescent phenotype observed in LTβR knockdown cells is related to p53-mediated pathways, we used doxorubicin (Dox), a well-known inducer of senescence, as a positive control. While Dox-treated cells exhibited reduced cell numbers and increased SA-β-Gal activity, LTβR knockdown in A375 cells similarly led to decreased cell numbers and elevated SA-β-Gal activity, which showed an additive effect when combined with Dox treatment (Fig. 1A, B). To rule out transient effects of siRNA, we prepared LTβR knockout (KO) B16F10 (B16F10) cells using the CRISPR/Cas9 system. Comparable results were also observed in B16F10 cells (Fig. 1C, D), suggesting that the LTβR depletion induces senescence. Given that p53 is a pivotal regulator of cellular senescence, p53 protein levels were examined (Fig. 1E, F). We observed increased levels of p53, along with elevated levels of MDM2, a key regulator of p53 that is also known to be upregulated during p53-mediated senescence rather than apoptosis [34, 35]. Moreover, we observed an upregulation of p21, a key mediator of p53-mediated cell cycle arrest, in LTβR-depleted cells.
These phenomena were consistent in normal human lung fibroblast IMR90 cells (Fig. S2A-C). However, in the p53 mutant human colorectal cancer cell line HT-29, which lacks p53 transcription activity, senescence was induced by Dox but not by LTβR knockdown, underscoring the role of p53 in LTβR depletion-induced senescence. (Fig. S2D-F). These findings suggest that LTβR regulates cellular senescence through a p53-dependent pathway.
Next, to investigate the effects of LTβR overexpression on cellular senescence and p53 activity, A375 and B16F10 cells were transfected with either an empty vector plasmid or an LTβR-expressing plasmid, followed by treatment with Dox. LTβR-overexpressing cells showed a less pronounced reduction in cell number and increase in SA-β-Gal activity compared to control cells when treated with Dox (Fig. 2A-D). Furthermore, western blot analysis revealed lower levels of p53, MDM2, and p21 in LTβR-overexpressing cells compared to controls (Fig. S3A), suggesting that LTβR overexpression reduces Dox-induced senescence in a p53-dependent manner (Fig. 2E, F). To further validate the effect of LTβR overexpression in cells, we restored LTβR expression in LTβR KO cells by transfecting them with an LTβR-expressing vector (Fig. S3B-D). Restored LTβR expression attenuated the senescent phenotype, as shown by increased cell numbers, decreased SA-β-Gal activity, and reduced levels of p53, p21, and MDM2 under Dox treatment. These results indicate that LTβR overexpression can suppress Dox-induced senescence.
To determine whether LTβR influences p53 transcription, p53 transcription levels in LTβR knockdown and overexpressing cells were analyzed using real-time PCR. No significant changes were observed in p53 mRNA levels (Fig. 3A). However, p21 mRNA, a downstream target of p53, was significantly upregulated, suggesting post-transcriptional regulation of p53 by LTβR (Fig. S4A). These results align with previous RNA-seq data showing increased cdkn1a (p21) levels in hematopoietic stem cells of LTβR KO mice [36] (Fig. S4B). Treatment with the proteasome inhibitor MG-132 further elevated p53 levels in LTβR knockdown cells and restored p53 and p21 levels in LTβR-overexpressing cells (Fig. 3B, C). This indicates that knockdown of LTβR prevents degradation of the p53 protein. As p53 protein degradation is regulated by MDM2 and MDMX, we assessed their expression in LTβR knockdown cells. MDMX protein level declined at 12 h after LTβR siRNA transfection, while no significant changes were observed in the level of p53, MDM2, and p21 compared to control cells before 24 h (Fig. 3D and S4C). These findings indicate that the expression of p53, MDM2, and p21 may be influenced as a consequence of changes in MDMX. Knockdown of MDMX increased p53, MDM2, and p21 levels without affecting LTβR expression (Fig. S4D), suggesting that LTβR regulates MDMX, which in turn inhibits p53 degradation. To further support the role of LTβR in p53 protein degradation, A375 cells were treated with MDM2 inhibitor nutlin-3a, which disrupts the MDM2-p53 binding and induces p53-mediated cellular senescence. MDMX overexpression has been reported to counteract the effect of nutlin-3a by preventing p53 activation [37, 38]. In LTβR knockdown cells, p53 levels increased rapidly upon nutlin-3a treatment (Fig. S4E), whereas LTβR-overexpressing cells exhibited a delayed increase (Fig. S4F).
To examine whether LTβR regulates MDMX protein degradation, A375 cells were treated with the protein synthesis inhibitor cycloheximide (CHX). MDMX protein levels decreased more rapidly in LTβR knockdown cells and more slowly in LTβR-overexpressing cells (Fig. 3E, F), indicating that LTβR stabilizes MDMX protein. Finally, changes in MDMX and p53 ubiquitination patterns were observed in both LTβR knockdown and LTβR-overexpressing cells (Fig. 3G-J). Ubiquitination of MDMX was increased in LTβR knockdown cells and decreased in LTβR-overexpressing cells, whereas p53 ubiquitination showed the opposite pattern -- decreased in LTβR knockdown cells and increased in LTβR-overexpressing cells. These results suggest that LTβR negatively regulates MDMX ubiquitination, thereby promoting p53 degradation.
We hypothesized that LTβR stabilizes the MDMX protein by binding to it, as LTβR contains an α-helix near its intracellular TRAF-binding domain and MDMX has a Zn²⁺ finger-like domain in its MDM2-binding region. To assess the potential interaction, we used the HADDOCK 2.4 web server to predict the protein-protein docking score. The computational analysis of HADDOCK score for LTβR α-helix-MDMX-MDM2 binding site was -66.1 ± 2.2 (cluster size 21, Z score -1.3), suggesting high probability of binding [39, 40]. To confirm the interaction between LTβR and MDMX, LTβR knockdown and overexpressing cells were subjected to immunoprecipitation and proximity ligation assays (PLA). LTβR-overexpressing cells showed increased LTβR-MDMX binding, whereas LTβR knockdown cells exhibited reduced interaction (Fig. 4A-D), supporting an interaction between LTβR and MDMX. To test whether this binding occurs through the intracellular domain, we generated a truncated form of LTβR (ΔECD; Δ1-227 aa), which lacks the extracellular domain but retains the transmembrane and intracellular regions. Both immunoprecipitation and PLA demonstrated that LTβR-∆ECD still interacts with MDMX, suggesting that the interaction occurs in the cytoplasm (Fig. 4E, F). We also observed increased interaction in LTβR-overexpressing B16F10 cells, as well as restored PLA signal in LTβR-overexpressing B16F10 cells (Fig. 4G, H), confirming that this interaction also occurs in mouse cells.
It has been shown that overexpression of LTβR activates non-canonical NF-κB signaling through self-oligomerization, independent of its extracellular domain [41,42,43,44]. To investigate whether the extracellular domain is required to attenuate cellular senescence, we transfected A375 cells with LTβR-ΔECD, a truncated form lacking the extracellular domain. The results were consistent with those of full-form LTβR overexpression, indicating that the effects are independent of extracellular domain (Fig. S5A-C).
Next, A375 cells were treated with LIGHT protein (a LTβR ligand) (Fig. S6A-D) to examine whether extracellular signaling contributes to p53-mediated cellular senescence. LIGHT treatment resulted in increased levels of IκBα and decreased levels of LTβR, consistent with previous research showing that ligand-induced endocytosis of LTβR limits canonical NF-κB signaling and promotes its degradation [44]. Although LIGHT treatment reduced LTβR expression, MDMX protein levels remained unaffected, which is likely due to decreased levels of MDM2, a MDMX-degrading enzyme. We observed decreased levels of p53 and MDM2 in LIGHT-treated cells, suggesting that p53 and MDM2 levels might be regulated through LTβR-dependent NF-κB signaling. LIGHT treatment led to a comparable increase in p21 levels in both control and LTβR-overexpressing cells, in contrast to the results observed in our overexpression model. The SA-β-Gal activity assay further suggests that LIGHT does not significantly affect the senescence state in LTβR knockdown cells, but induces minor changes in LTβR-overexpressing cells (Fig. S6B, D). To examine whether LIGHT treatment affects the interaction between LTβR and MDMX, we treated cells with LIGHT for 4 h (Fig. S6E). However, the interaction appeared to be primarily regulated by the expression levels of LTβR following LIGHT treatment. Taken together, p53, MDM2, and p21 may be influenced by NF-κB signaling, while MDMX-p53-mediated cellular senescence is likely associated with the expression of LTβR.
MDMX lacks a nuclear localization signal and is known to be ubiquitinated in the nucleus by MDM2 [25, 45]. To investigate whether LTβR affects nuclear localization of MDMX, we performed cytosolic and nuclear fractionation following proteasome inhibition using bortezomib (BTZ) to prevent proteasome-mediated degradation of MDMX. Western blot analysis showed that nuclear MDMX levels increased in LTβR knockdown cells but decreased in LTβR-overexpressing cells (Fig. 5A, B). Confocal microscopy results were consistent with the western blot data, showing similar patterns of nuclear MDMX localization (Fig. 5C, D). Comparable results were observed in LTβR KO cells that reconstituted with LTβR (Fig. S7A, B). PLA further confirmed that MDM2-MDMX interaction was enhanced in LTβR knockdown cells but reduced in LTβR-overexpressing cells (Fig. 5E, F). To exclude the potential involvement of the MDMX deubiquitinating enzyme USP7 [46], we performed immunoprecipitation using a USP7 antibody. USP7 has been reported to interact with TRAF6, which in turn modulates NF-κB signaling [47, 48]. Interestingly, the interaction between USP7 and TRAF6 was enhanced following LTβR overexpression, whereas the interaction between USP7 and MDMX showed no significant change upon either LTβR knockdown or overexpression (Fig. S7C, D). These findings suggest that LTβR knockdown induces the nuclear localization of MDMX, by upregulating its interaction with MDM2, which in turn facilitates MDMX degradation in the nucleus and subsequently suppresses the degradation of p53.
To investigate the senescence phenotype of LTβR KO cells in WT mice, which express potential ligands such as LIGHT and LTα1β, B16F10 cells were implanted on the right dorsal side, while B16F10 cells were implanted on the left dorsal side of 8-week-old WT C57/BL6 mice. After 9 days, the mice were treated via intraperitoneal injection with either vehicle (PBS) or Dox (4 mg/kg) to induce a robust synergistic effect on tumor senescence, and were sacrificed 7 days later (Fig. 6A). Tumor measurements indicated that B16F10 tumors were significantly smaller in weight and volume compared with B16F10 tumors (Fig. 6B, C). Western blot analysis of tumor tissues revealed increased levels of p53, MDM2, and p21 in B16F10 tumors (Fig. 6D). Moreover, fluorescent immunohistochemistry confirmed elevated levels of p21, and cryosection analysis showed higher SA-β-Gal activity in B16F10 tumors, highlighting a pronounced senescence phenotype (Fig. 6E, F).
Next, to test the additive effect of MDM2 inhibitor on enhancing p53 activation, mice implanted with B16F10 and B16F10 cells were treated with nutlin-3a. As shown in Fig.7A-D, nutlin-3a further supported the role of LTβR in regulating p53-mediated senescence through a decrease in tumor growth. Tumor tissue analysis revealed elevated p21 levels and SA-β-Gal staining in nutlin-3a-treated LTβR KO tumors (Fig. 7E, F). These findings indicate that depletion of LTβR delays tumor progression in vivo, suggesting that the combination of LTβR gene targeting and p53-activating drugs may serve as a potential therapeutic strategy for cancer treatment.