NEDD4L contributes to ferroptosis and cell growth inhibition in esophageal squamous cell carcinoma by facilitating xCT ubiquitination

Abstract

The oncogene xCT plays an indispensable role in tumor growth by protecting cancer cells from oxidative stress and ferroptosis. Emerging evidence indicated xCT function is tightly controlled by posttranslational modifications, especially ubiquitination. However, it still remains unclear what specific regulatory mechanism of xCT by ubiquitin ligases in human cancers. Here, we reported that NEDD4L, an E3 ubiquitin ligases, inhibited esophageal squamous cell carcinoma (ESCC) tumor growth and facilitated ferroptosis by ubiquitination of xCT. NEDD4L expression was declined in ESCC and was associated with tumor invasion, lymph node metastasis and distant metastasis. Silencing NEDD4L triggered ESCC tumor growth. Meanwhile, knock down of NEDD4L prevented the accumulation of ROS, elevated the level of GSH, reduced the content of MDA in ESCC cells, thereby inhibiting ferroptosis. Mechanistically, NEDD4L directly bound to the ∆CT domain of xCT through its WW and HECT domain. More importantly, NEDD4L promoted xCT degradation by facilitating its polyubiquitination in ESCC cells. Collectively, these findings suggest that NEDD4L is crucial in governing the stability of xCT and mediating ferroptosis in ESCC.

Fig. 1. NEDD4L expression is low and serves as a good prognosis marker in ESCC.

A NEDD4L protein expression was decreased in ESCC tissues (n = 25) compared with normal tissue (n = 25) as analyzed by IHC. Scale bars, 100 μm. B Representative images of NEDD4L low expression (n = 62) and high expression (n = 80) analyzed by IHC. Scale bars, 100 μm. C NEDD4L expression score in 142 cases of ESCC (n = 142). D Overall survival analysis revealed that higher expression of NEDD4L was related with poorer overall survival of ESCC patients (n = 142). p <0.01, log-rank test.

Fig. 2. NEDD4L inhibits ESCC progression in vivo and in vitro.

A, B NEDD4L silencing efficiency in ESCC cell lines. Eca109 and EC9706 cells were transfected with NEDD4L siRNAs. The silencing efficiency was measured via western blotting. C, D Silencing NEDD4L promoted the proliferation of ESCC cells. Eca109 and EC9706 were transfected with siControl or siNEDD4L. There were two different siRNAs be used. After 24 hours, the CCK8 was used to determine the cellar metabolic activity at indicated time points after infection (n = 3 per group). E, F Silencing NEDD4L increased the clone numbers of ESCC cells. Eca109 and EC9706 cells were transfected with indicated 50 nM siNEDD4L or siControl. Quantification of clone formation was shown at the indicated time points (n = 3 per group). G–I NEDD4L depletion promoted tumor growth of EC9706 cells in xenograft model. Female nude mice bearing EC9706 tumors were treated daily with control (n = 5) or shNEDD4L (n = 5) at the indicated concentrations. The growth of xenografts was monitored over 4 weeks. J, K NEDD4L overexpression efficiency in ESCC cell lines. Eca109 and EC9706 cells were transfected with NEDD4L plasmids. The overexpression efficiency was measured via western blotting. L, M Overexpression NEDD4L inhibited the proliferation of ESCC cells. Eca109 and EC9706 were transfected with Flag-NEDD4L or Flag-vector. CCK8 solution was added to determine the cellar metabolic activity at indicated time points (n = 3 per group). N, O Overexpression NEDD4L decreased the clone numbers of ESCC cells. Eca109 and EC9706 cells transfected with indicated 5 μg Flag-vector or Flag-NEDD4L plasmids. Quantification of clone formation was shown at the indicated time points (n = 3 per group). The data in I is presented as the mean ± SEMs. Statistical analysis was performed using Student’s t-test. *p < 0.05; **p < 0.01; ***p < 0.001. The other data are presented as the mean ± SD. Statistical significance was determined by one-way ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001; ns, not significant.

Fig. 3. NEDD4L promotes ferroptosis in ESCC cells.

A Volcano plot of significantly up and down-regulated genes after NEDD4L depletion. B Heat map of mRNA changes in siControl (n = 3) and siNEDD4L (n = 3) of EC9706 by bulk RNA-seq. C–G Publicly available data showed that NEDD4L expression is positively correlated with that of the ferroptosis related genes HMOX1, IREB2, KEAP1, LPCAT3 and BECN1 in esophageal cancer (http://gepia.cancer-pku.cn/). H, I ROS level in Eca109 and EC9706 cells transfected with siControl/siNEDD4L for 48 h were detected. n = 3 per group. J Silencing NEDD4L promoted GSH level in ESCC cells. Eca109 and EC9706 cells were transfected with siControl/siNEDD4L for 48 h. GSH level was determined at 412 nm. n = 3 per group. K Silencing NEDD4L reduced MDA content in ESCC cells. Eca109 and EC9706 cells were transfected with siControl/siNEDD4L 48 h. MDA content was measured at 532 nm and 600 nm. n = 3 per group. L Representative cell and mitochondrial ultrastructural images of EC9706 cells transfected with siContrlo/siNEDD4L under the treatment of erastin for 24 h. Scale bar = 5 μm. M, N Overexpressing NEDD4L promoted ROS level in ESCC cells. Eca109 and EC9706 cells were transfected with Flag-vector or Flag-NEDD4L plasmids for 48 h. n = 3 per group. O Overexpressing NEDD4L decreased GSH level in ESCC cells. Eca109 and EC9706 cells transfected with Flag-vector or Flag-NEDD4L plasmids for 48 h. n = 3 per group. P Overexpressing NEDD4L increased MDA content in ESCC cells. Eca109 and EC9706 cells transfected with Flag-vector or Flag-NEDD4L plasmids for 48 h. n = 3 per group. The data are presented as the mean ± SD. Statistical significance in I–K were determined by one-way ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001. Statistical significance in N–P Statistical analysis were performed using Student’s t-test. *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 4. NEDD4L expression reversely relates with xCT level in ESCC.

A–C Low NEDD4L expression was correlated with an increase level of xCT in ESCC samples (n = 142). D, E Silencing NEDD4L increased the xCT protein level in Eca109 and EC9706 cells. F NEDD4L overexpression reduced xCT protein level in HEK293T cells. G, H Silencing of xCT reduced NEDD4L depletion resulting in an increase in xCT protein levels. I, J Silencing xCT rescued the proliferation ability of cells with the NEDD4L depletion. n = 3 per group. K, L Silencing xCT rescued clone numbers with the NEDD4L depletion. n = 3 per group. M, N Silencing xCT rescued ROS level with the NEDD4L depletion. n = 3 per group. O Silencing xCT rescued GSH level with the NEDD4L depletion. n = 3 per group. P Silencing xCT rescued MDA content with the NEDD4L depletion. n = 3 per group. The data are presented as the mean ± SD. Statistical significance was determined by one-way ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001; ns, not significant.

Fig. 5. NEDD4L associates with ∆CT domain of xCT through its WW and HECT domain.

A Intracellular localization of NEDD4L and xCT analyzed by IF assay. EC9706 cells were cultured in normal medium before fixation. Intracellular localization of xCT (red) and NEDD4L (green) were shown. Nuclei (blue) were stained with 4’,6-diamidino-2-phenylindole (DAPI). B Cytoplasm and nuclear were separated by kit. NEDD4L is mainly localized in the cytoplasm. α-tubulin and lamin B1 were used for cytoplasm and nuclear control. NEDD4L interacted with xCT in cytoplasm. C, D Co-IP assay revealed that NEDD4L connected with xCT in EC9706 cells. EC9706 cells were transfected with indicated Flag-NEDD4L and Myc-xCT plasmids, followed by Co-IP and western blotting detection. E, F NEDD4L and xCT domain structure and deletion mutants used for Co-IP assays. G NEDD4L bound to xCT at its WW and HECT domain. HEK293T cells were transfected with the indicated NEDD4L and xCT constructs, followed by Co-IP and western blotting assays. The WW and HECT domain of NEDD4L interacted with xCT. H xCT bonded to NEDD4L at its ∆CT domain. HEK293T cells were transfected with the indicated xCT and NEDD4L constructs, followed by Co-IP and western blotting detection.

Fig. 6. NEDD4L modulates xCT stability in ubiquitination dependent manner.

Silencing NEDD4L decreased xCT half-life in EC9706 cells (A, B) and NEDD4L overexpression prolonged xCT half-life in HEK293T cells (C, D). The cells were treated with 100 μmol/L CHX for indicated time periods before being collected for western blotting assay. n = 3 per group. E, F NEDD4L upregulated xCT through proteosome. NEDD4L overexpression could inhibit xCT protein level and NEDD4L depletion could promote xCT protein level, which effect could be diminished by MG132. G Ubiquitin-based IP assays showed that overexpression NEDD4L promoted xCT overall poly-ubiquitination in HEK293T cells. H Ubiquitin-based IP assays showed that NEDD4L failed to facilitate xCT mono-ubiquitinaiton in HEK293T cells. I Silencing NEDD4L inhibited xCT overall poly-ubiquitination in EC9706 cells. J NEDD4L promoted xCT K48-linked ubiquitinaiton in HEK293T cells. K Ubiquitin-based IP assays showed that NEDD4L inhibited xCT K63-linked ubiquitinaiton in HEK293T cells. L The HECT domain of NEDD4L was necessary for NEDD4L to xCT protein suppression. M The HECT domain of NEDD4L was necessary for NEDD4L to regulate ubiquitination of xCT. N Mutations in NEDD4L that disrupted its ubiquitination activity compromised NEDD4L’s capacity to degrade the xCT protein. O, P HEK293T cells were transfected with Flag-tag or Flag NEDD4L^c942a and Myc-xCT plasmids. The cells were treated with 100 μmol/L CHX for indicated time periods before being collected for western blotting assays. n = 3 per group. Q The effects of expression of Flag NEDD4L and its mutants on ubiquitination of Myc- xCT in HEK293T cells detected by Ubiquitin-based IP assays. The data are presented as the mean ± SD. Statistical significance was determined by two-way ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 7

NEDD4L interacts with xCT, promotes xCT K48-linked ubiquitination and degradation, which activates the ferroptosis and inhibits ESCC cancer progression.