Copper-mediated SEC14L3 promotes cuproptosis to inhibit hepatocellular carcinoma growth via ERK/YY1/FDX1 axis

Abstract

Cuproptosis, a copper-triggered cell death pathway, holds therapeutic potential for cancers, but its regulatory mechanisms in hepatocellular carcinoma (HCC) remain undefined. Despite SEC14L3’s known roles in cellular signaling, its involvement in HCC progression and cuproptosis regulation is unclear. Here, we reveal that SEC14L3 expression is downregulated in HCC cells and tissues and correlates with advanced stages and poor prognosis. Copper-induced cuproptosis inhibits HCC cell viability, and SEC14L3 positively modulates cuproptosis in HCC cells by promoting DLAT lipoylation and its oligomerization. Mechanistically, SEC14L3-mediated cuproptosis suppressed HCC growth via the ERK/YY1/FDX1 axis both in vitro and in vivo. Additionally, copper enhanced the SEC14L3 expression, which in turn regulated ERK/YY1/FDX1 axis. Our findings show that copper-mediated SEC14L3 promotes cuproptosis via ERK/YY1/FDX1 axis, thereby inhibiting HCC growth. These insights provide a mechanistic foundation for targeting cuproptosis, advancing the development of SEC14L3-driven therapeutic strategies for HCC.

Copper-mediated SEC14L3 activates the ERK/YY1/FDX1 axis to promote cuproptosis and suppress hepatocellular carcinoma progression.

Introduction

Hepatocellular carcinoma (HCC), accounting for 75-85% of primary liver cancer, is the third leading cause of carcinoma-related deaths worldwide¹. Although extensive studies have focused on the pathogenesis and potential biomarkers of HCC, the treatment and prognosis of HCC remain unsatisfactory owing to their late diagnosis, early recurrence, and high levels of genetic heterogeneity². Therefore, it is imperative to further elucidate the underlying biological mechanism governing the malignant phenotype of HCC to identify novel therapeutic targets and develop efficacious strategies for prevention and treatment.

The human SEC14-like (SEC14L) proteins, mainly consisting of SEC14L1, SEC14L2, and SEC14L3, belong to the phosphatidylinositol transfer proteins (PITPs) superfamily. They commonly possess an N-terminal domain for membrane association, a small lipophilic molecules-binding domain (SEC14 domain), and a C-terminal Golgi dynamics (GOLD) domain that facilitates diverse protein-protein and protein-membrane interactions, which enables SEC14L to mediate cellular signal transduction and lipid metabolism³. Currently, the majority of studies have focused on the carcinogenesis and carcinostasis of SEC14L1 and SEC14L2, such as prostatic cancer, breast cancer, and HCC, while the role of SEC14L3 on human cancer is less well-studied^4–8. Zhu et al. have found that SEC14L3 negatively modulates the malignant behaviors of breast cancer cells via mediating the Wnt/β-catenin signaling pathway⁹. Additionally, SEC14L3 physically binds to RAB4A/5A and VEGFR2 to accelerate the latter internalization, activating the VEGF signaling pathway and thereby promoting the zebrafish vasculogenesis¹⁰. However, the function of SEC14L3 in HCC has not been reported.

The aberrant activation of the mitogen-activated protein kinase-extracellular signal-regulated kinase1/2 (MAPK-ERK1/2) signaling pathway is widely acknowledged as a pivotal factor in HCC pathogenesis and progression, affecting approximately 50% of early-stage HCC patients and nearly all patients with advanced-stage disease¹¹. Current evidence suggests that ERK with phosphorylated activation regulates the transcription factors expression by which affect the expression of respective downstream genes, thus promoting tumor cell malignant phenotype¹². Wang et al. have found that MerTK inhibits the ferroptosis regulated by SLC7A11 via ERK/SP1 axis in the resistant HCC¹³. Lei et al. have demonstrated that FGFR2 suppresses BRCA1 via ERK/YinYang 1 (YY1) axis and thereby promoting breast cancer progression¹⁴. However, few studies have explored the relationship between HCC and the ERK/YY1 axis mediated by SEC14L3.

The dynamic regulation of copper, an essential trace element in various biological processes, is crucial for maintaining cellular integrity¹⁵. Imbalances in copper homeostasis are relevant to cellular damage and various tumors, including pancreatic and colorectal cancer¹⁶. Recent study from Tsvetokv et al. have suggested that cuproptosis, a newly identified form of programmed cell death triggered by copper, exhibits distinct characteristics compared to other forms of cell death, including apoptosis, autophagy, ferroptosis, and necrosis¹⁷. The occurrence of cuproptosis is specifically associated with the direct binding of copper, transported into the cell via elesclomol (ES, an ionophore), to lipoylated dihydrolipoamide transacetylase (Lip-DLAT). This interaction leads to proteotoxic stress and ultimately results in cell death owing to the aggregation of lip-DLAT and subsequent loss of iron-sulfur cluster protein¹⁷. In particularly, ferredoxin 1 (FDX1) not only induces Cu²⁺ to its more toxic form Cu⁺, but also promotes the protein lipoylation of tricarboxylic acid (TCA) cycle¹⁷. Sun et al. have found that the lactation of METTL16 at site K299, mediated by copper, promotes the expression of FDX1 via m⁶A modification on FDX1 mRNA, thereby ultimately influencing the cuproptosis of gastric cancer cells¹⁸. Xing et al. have shown that ARID1a deletion inhibits the glycolysis process of HCC cells, thereby enhancing the sensitivity of HCC cells to cuproptosis¹⁹. The available evidence suggests that the modulation of intracellular copper concentration in tumors through ionophores represents a promising approach for cancer therapy²⁰.

In the current study, we found that copper-mediated SEC14L3 positively regulated copper-induced cuproptosis via ERK/YY1/FDX1 axis to inhibit the growth of HCC both in vitro and in vivo. Our findings may provide unfound molecular therapeutic targets for HCC by enhancing the activation of cuproptosis.

Results

SEC14L3 was downregulated in HCC and related to poor prognosis

We first explore the relationship between SEC14L3 and HCC clinicopathology in TCGA database, and the results demonstrated that compared with the adjacent non-tumor tissues (ANT), SEC14L3 was significantly downregulated in HCC tissues (Fig. 1A). In particularly, low SEC14L3 expression was significantly associated with higher pathologic stage (Fig. 1B) and histologic grade (Fig. 1C), and the survival analysis revealed that low SEC14L3 expression was significantly related to poor OS, DSS, and PFS in HCC patients (Fig. 1D–F). Subsequently, the mRNA and protein expression levels of SEC14L3 were significantly downregulated in HCC cells (HepG2, Huh7, and SK-HEP-1) and tissues compared with the normal liver cell (MIHA) or matched adjacent non-tumor tissues (MANT) (Fig. 1G–J). These results suggest that SEC14L3 is a potential prognostic biomarker for HCC.

Fig. 1. Low SEC14L3 expression was associated with poor HCC prognosis and copper promoted HCC cells death.

Comparison of SEC14L3 expression between the adjacent non-tumor tissues (ANT) and HCC tissues (A), as well as different pathological stages (B) and histological grades (C) in TCGA database. Data in each graph indicate median with interquartile range. The survival analysis of SEC14L3 for overall survival (D), disease specific survival (E), and progress free survival (F) in TCGA database. The mRNA and protein expression levels of SEC14L3 were detected by qRT-PCR (G) and immunoblot (H) in normal liver cells and HCC cells. I, J The immunohistochemistry (IHC) staining was used to determine the SEC14L3 protein expression between the matched adjacent non-tumor tissues (MANT) and HCC tissues; the H score was used to semi-quantitative analysis for IHC. Data in each graph indicate the median with interquartile range. Scale bars: 100 μm (10X). Vitality of HepG2 (K) and SK-HEP-1 (L) cells was assessed following treatment of different concentrations of elesclomol (ES) and 1 μM CuCl₂ or 10 μM tetrathiomolybdate (TTM). n  =  3 biologically independent experiments for each group. Data in each graph indicate mean ± SEM.

Copper ion inhibited the growth on HCC cells

To explore the impact of copper on HCC cells viability, we employed the copper chelator tetrathiomolybdate (TTM, 10 μM) either alone or in combination with various dose of ES and 1 μM CuCl₂ to treat HCC cells (HepG2 and SK-HEP-1). The results demonstrated that ES–CuCl₂ exhibited a dose-dependent inhibition on the viability of HCC cells (HepG2 and SK-HEP-1), whereas the removal of copper by TTM significantly counteracted this impact, indicating that the viability inhibition was caused by copper rather than ES or ES–CuCl₂ complex (Fig. 1K-L).

SEC14L3 positively modulated cuproptosis on HCC cells

To explore the biological process involved in SEC14L3, GO enrichment was performed using gene-set enrichment analysis (GSEA) in TCGA database. Interesting, we found that SEC14L3 was significantly associated with protein complex oligomerization, pyruvate metabolic process, and mitochondrial gene expression, which played crucial role in cuproptosis (Fig. 2A). Therefore, we hypothesized that SEC14L3 exerted an influence on the growth of HCC via cuproptosis.

Fig. 2. SEC14L3 promoted the cuproptosis in HepG2 cells.

A The GO term gene-set enrichment analysis (GSEA) were conducted in TCGA database. B Vitality of HepG2 cells with negative control, SEC14L3 overexpression or silencing was assessed following treatment of different concentrations of elesclomol (ES) and 1 μM CuCl₂. ** P <0.01 vs. OE-CON; ## P <0.01 vs. si-CON. The protein (C) expression levels of lip-DLAT and DLAT were detected by immunoblot (SDS-PAGE) in HepG2 cells with SEC14L3 overexpression or silencing at a specified concentration of ES-CuCl₂ (30 nM: 1 μM) treatment; semi-quantitative analysis for protein expression was shown (D). NS, not significant vs. negative control. Lip-DLAT oligomerization was determined after 48 h treatment of ES-CuCl₂ (30 nM: 1 μM) by immunoblot (Native-PAGE) (E) and immunofluorescence imaging (F) in HepG2 cells with SEC14L3 overexpression or silencing (DAPI - blue, DLAT – red, mitochondria - green). Scale bars: 100 μm (10X). G Foci counts from immunofluorescence imaging were segmented and quantified in each condition. NES normalized enrichment scores, Lip lipoylation, OE-CON negative control for SEC14L3-OE, SEC14L3-OE SEC14L3 overexpression, si-CON negative control for siSEC14L3, siSEC14L3 siRNA against SEC14L3. n  =  3 biologically independent experiments for each group. Data in each graph indicate mean ± SEM.

Subsequently, to explore whether SEC14L3 regulated the copper-mediated effect on HCC cells (HepG2 and SK-HEP-1) viability, CCK-8 assays were performed initially to determine the impact of SEC14L3 on copper-mediated HCC cells viability. The results revealed that compared with the negative control, transient transfection of SEC14L3 overexpression (SEC14L3-OE) significantly enhanced the copper-mediated inhibitory effect on HCC cells viability, whereas transient transfection of silencing SEC14L3 (siSEC14L3) by siRNA significantly attenuated the copper-mediated inhibitory effect on HCC cells viability (Fig. 2B and Supplementary Fig. 1A). These results indicate that SEC14L3 amplifies the cytotoxic effects of copper ions on HCC cells. Then, to determine whether the copper-mediated inhibitory effect on the growth of HCC cells (HepG2 and SK-HEP-1) was mediated by SEC14L3 via cuproptosis, immunoblot and immunofluorescence assays were performed to detect lip-DLAT and lip-DLAT oligomerization, which are distinctive characteristics associated with cuproptosis. We found that compared with the negative control, SEC14L3-OE significantly promoted the lipoylation of DLAT as measured by SDS-PAGE following treatment of fixed concentration of ES- CuCl₂ (30 nM: 1 μM), and resulted in lip-DLAT oligomerization detectable by native-PAGE, which could be further confirmed by immunofluorescence assays, observing obviously DLAT foci by ES-CuCl₂ (30 nM: 1 μM) treatment. On the other hand, compared with the negative control, siSEC14L3 significantly reduced the lipoylation of DLAT and lip-DLAT oligomerization following treatment of fixed concentration of ES-CuCl₂ (30 nM: 1 μM) (Fig. 2C–G and Supplementary Fig. 1B–E). These results suggest that SEC14L3 positively promotes the cuproptosis on HCC cells.

SEC14L3 regulated the expression of cuproptosis key gene FDX1 via ERK/YY1 axis in HCC cells

To investigate the mechanism by which SEC14L3 regulated the cuproptosis on HCC cells, the subsequent correlation analysis in TCGA database was conducted to determine the expression relationship between SEC14L3 and FDX1, a pivotal mediator of cuproptosis. The results showed that SEC14L3 mRNA expression positively correlated with FDX1 mRNA expression, indicating that SEC14L3 potentially regulated the expression of FDX1 (Fig. 3A). Next, we initially found that the mRNA and protein expression of FDX1 was downregulated in HCC cells (HepG2, Huh7, and SK-HEP-1) compared to MIHA cells as measured by qRT-PCR and immunoblot (Fig. 3B-C). Subsequently, compared with the negative control, SEC14L3-OE significantly promoted the mRNA and protein expression of FDX1 on HCC cells (HepG2 and SK-HEP-1), whereas siSEC14L3 remarkably reduced them, which validated the findings in TCGA database (Fig. 3D–G and Supplementary Fig. 2A–C). Meanwhile, overexpressed or silencing FDX1 did not affect the expression of SEC14L3 (Supplementary Fig. 2D–F).

Fig. 3. SEC14L3 positively regulated the expression of FDX1 in HepG2 cells.

A The correlation analysis was performed between SEC14L3 and FDX1 in TCGA database. The mRNA (B) and protein (C) expression levels of FDX1 in normal liver cells and HCC cells were detected by qRT-PCR and immunoblot. D The mRNA expression levels of SEC14L3 were detected by qRT-PCR in HepG2 cells with SEC14L3 overexpression or silencing. The mRNA (E) and protein (F) expression levels of FDX1 were detected by qRT-PCR and immunoblot in HepG2 cells transiently transfected with SEC14L3 expression plasmid or siRNA specific for SEC14L3; semi-quantitative analysis for protein expression was shown (G). OE-CON, negative control for SEC14L3-OE; SEC14L3-OE, SEC14L3 overexpression; si-CON, negative control for siSEC14L3; siSEC14L3, siRNA against SEC14L3. n  =  3 biologically independent experiments for each group. Data in each graph indicate mean ± SEM.

Further investigation was performed in TCGA database to determine the underlying mechanism by which SEC14L3 modulated FDX1 expression. The GO and KEGG results by GSEA revealed that low SEC14L3 expression was significantly associated with the activation of ERK1/2 cascade and MAPK signaling pathway (Fig. 4A). Hence, we hypothesized that SEC14L3 might regulate MAPK/ERK signaling pathway and transcriptional suppressors to affect FDX1 expression. Previous studies have shown that phosphorylated ERK (p-ERK) exerts regulatory control over YY1, thereby suppressing the expression of downstream genes¹⁴. Furthermore, the JASPAR database predicted two binding sites for YY1 in the FDX1 promoter (Fig. 4B). Thus, we hypothesized that SEC14L3 regulated the expression of FDX1 via ERK/YY1 axis on HCC cells. As expected, SEC14L3-OE significantly decreased the p-ERK levels on HCC cells (HepG2 and SK-HEP-1), whereas siSEC14L3 significantly increased the p-ERK levels as measured by immunoblot, indicating that SEC14L3 negatively modulated the activation of MAPK/ERK signaling pathway (Fig. 4C-Dand Supplementary Fig. 3A-B). Given that the ERK signaling pathway has the potential to influence apoptosis, we employed flow cytometry to assess the apoptotic status of the HCC cells. The results showed that overexpressed SEC14L3 promoted apoptosis of HepG2 and SK-HEP-1 cells, indicating that overexpressed SEC14L3 efficiently initiates apoptosis pathways and accumulated Cu⁺ to promote cuproptosis (Supplementary Fig. 3C).

Fig. 4. SEC14L3 positively regulated FDX1 via ERK/ YY1 axis in HepG2 cells.

A The GO and KEGG term gene-set enrichment analysis (GSEA) were conducted in TCGA database. B The JASPAR database predicted the YY1 binding motif and binding site at FDX1 promoter. The protein (C) expression levels of ERK and p-ERK were detected by immunoblot in HepG2 cells transiently transfected with SEC14L3 expression plasmid or siRNA specific for SEC14L3; semi-quantitative analysis for protein expression was shown (D). The mRNA and protein expression levels of YY1 (E and F) and FDX1 (G and H) were detected by qRT-PCR and immunoblot and the bindings of YY1 to the FDX1 promoter were detected by ChIP assays (I) in HepG2 cells transiently transfected with negative control, SEC14L3 expression plasmid, SEC14L3 expression plasmid coupled with Ro 67-7476. The mRNA and protein expression levels of YY1 (J and K) and FDX1 (L and M) were detected by qRT-PCR and immunoblot and the bindings of YY1 to the FDX1 promoter were detected by ChIP assays (N) in HepG2 cells transiently transfected with negative control, siRNA specific for SEC14L3 or siRNA specific for SEC14L3 coupled with PD98059. NES, normalized enrichment scores; OE-CON, negative control for SEC14L3-OE; SEC14L3-OE, SEC14L3 overexpression; si-CON, negative control for siSEC14L3; siSEC14L3, siRNA against SEC14L3. n  =  3 biologically independent experiments for each group. Data in each graph indicate mean ± SEM.

The subsequent qRT-PCR, immunoblot, and ChIP assays revealed that compared with the negative control, SEC14L3-OE significantly suppressed YY1 mRNA and protein expression, while enhancing FDX1 mRNA and protein expression by a significant reduction of YY1 enrichment on the FDX1 promoter in HCC cells (HepG2 and SK-HEP-1). After the additional Ro 67-7476 treatment (p-ERK agonist), YY1 mRNA and protein expression were significantly higher than the SEC14L3-OE group, while FDX1 mRNA and protein expression were significantly lower owing to a significant elevation of YY1 occupied in FDX1 promoter on HCC cells (HepG2 and SK-HEP-1) (Fig. 4E–Iand Supplementary Fig. 3D–H). Conversely, compared with the negative control, siSEC14L3 significantly enhanced YY1 mRNA and protein expression, while suppressing FDX1 mRNA and protein expression by a significant elevation of YY1 occupied in FDX1 promoter on HCC cells (HepG2 and SK-HEP-1). After the additional PD98059 treatment (p-ERK inhibitor), YY1 mRNA and protein expression were significantly lower than the siSEC14L3 group, while FDX1 mRNA and protein expression were significantly higher owing to a significant reduction of YY1 enrichment on the FDX1 promoter on HCC cells (HepG2 and SK-HEP-1) (Fig. 4J–Nand Supplementary Fig. 3I–M). These results indicate that SEC14L3 positively regulates FDX1 expression via suppressing the ERK/YY1 axis on HCC cells.

SEC14L3 regulated cuproptosis via ERK/YY1/FDX1 axis to inhibit HCC cells viability

Subsequently, CCK-8, immunoblot, and immunofluorescence assays were performed to determine whether SEC14L3 modulated cuproptosis via ERK/YY1/FDX1 axis to inhibit HCC cells (HepG2 and SK-HEP-1) viability. Compared with the negative control, SEC14L3-OE following treatment of fixed concentration of ES-CuCl₂ (30 nM: 1 μM) significantly decreased HCC cells viability (Fig. 5Aand Supplementary Fig. 4A), while lip-DLAT and its oligomerization (Supplementary Fig. 5B–E, Fig. 5B–D, and Supplementary Fig. 4B–D) were significantly increased. After the additional Ro 67-7476 treatment, there was a significant increase in cell viability, accompanied by a notable decrease in lip-DLAT levels and its oligomerization compared to the SEC14L3-OE + ES-CuCl₂ group. Furthermore, the combination of transient transfection of SEC14L3 and FDX1 significantly inhibited the viability, and significantly increased lip-DLAT and its oligomerization compared to SEC14L3-OE + ES-CuCl₂ + Ro 67-7476 group. On the other hand, compared with the negative control, siSEC14L3 following treatment of fixed concentration of ES-CuCl₂ (30 nM: 1 μM) significantly increased HCC cells viability (Fig. 5E and Supplementary Fig. 4E), while lip-DLAT and its oligomerization (Supplementary Fig. 5F–I, Fig. 5F-H, and Supplementary Fig. 4F–H) were significantly decreased. After the additional PD98059 treatment, the viability was significantly decreased, and lip-DLAT and its oligomerization were significantly increased than siSEC14L3+ES-CuCl₂ group following treatment of fixed concentration of ES-CuCl₂ (30 nM: 1 μM). Furthermore, the combination of transient transfection of siSEC14L3 and siFDX1 significantly enhanced the viability, and significantly decreased lip-DLAT and its oligomerization compared to siSEC14L3+ES-CuCl₂ + PD98059 group. These results suggest that SEC14L3 positively modulates cuproptosis via ERK/YY1/FDX1 axis to inhibit HCC cells viability.

Fig. 5. SEC14L3 increased copper-induced lip-DLAT oligomerization in HepG2 cells via ERK/ YY1/ FDX1 signaling axis.

A Vitality of HepG2 cells was assessed following treatment with 30 nM elesclomol (ES) and 1 μM CuCl₂ in each condition. The Lip-DLAT oligomerization was detected by immunoblot (Native-PAGE) (B) and immunofluorescence imaging (C) in HepG2 cells with SEC14L3 overexpression, SEC14L3 overexpression + Ro 67-7476, or SEC14L3 overexpression + Ro 67-7476 + FDX1 overexpression following treatment with 30 nM ES and 1 μM CuCl₂ (DAPI - blue, DLAT – red, mitochondria - green); Scale bars: 100 μm (10X). D Foci counts from immunofluorescence imaging were quantified in each condition. E Vitality of HepG2 cells was assessed following treatment with 30 nM elesclomol (ES) and 1 μM CuCl₂ in each condition. The Lip-DLAT oligomerization was detected by immunoblot (Native-PAGE) (F) and immunofluorescence imaging (G) in HepG2 cells with SEC14L3 silencing, SEC14L3 silencing + PD98059, or SEC14L3 silencing + PD98059 + FDX1 silencing following treatment with 30 nM ES and 1 μM CuCl₂ (DAPI - blue, DLAT – red, mitochondria - green); Scale bars: 100 μm (10X). H Foci counts from immunofluorescence imaging were quantified in each condition. OE-CON, negative control for SEC14L3-OE; SEC14L3-OE, SEC14L3 overexpression; si-CON, negative control for siSEC14L3; siSEC14L3, siRNA against SEC14L3;FDX1-OE, FDX1 overexpression; siFDX1, siRNA against FDX1. n  =  3 biologically independent experiments for each group. Data in each graph indicate mean ± SEM.

SEC14L3-mediated cuproptosis inhibits HCC growth via ERK/YY1/FDX1 axis in vivo

Encouraged by the promising in vitro results, empty vector, SEC14L3 stable expression HepG2, si-control HepG2, or siSEC14L3 HepG2 cells were subcutaneously injected into nude mice above their right front legs to construct xenograft nude mice. Then, the xenograft nude mice were intraperitoneally injected with normal saline or ES-CuCl₂ five times a week, respectively (Fig. 6A). The results demonstrated that SEC14L3 sensitized the in vivo response to ES-CuCl2 treatment, whereas SEC14L3 knockdown produced opposite effects, which was evidenced by a significant reduction in tumor size, tumor weight, and growth rate following the ectopic expression of SEC14L3 (Fig. 6B–D). What is more, the immunofluorescence assays in tumor xenografts showed that combination of SEC14L3 and ES-CuCl₂ treatment was more significantly increased DLAT aggregation compared to treatment with ES-CuCl₂ alone (Fig. 6E-F). Additionally, combination of SEC14L3 and ES-CuCl₂ treatment demonstrated significantly reduced levels of Ki-67, p-ERK, and YY1 staining, while exhibiting elevating levels of SEC14L3 and FDX1 staining in tumor sections compared to treatment with ES-CuCl₂ alone. The treatment with ES-CuCl₂ or the combination of SEC14L3 did not result in any significant changes in ERK staining levels (Fig. 6E and G-H, and Supplementary Fig. 6A, B). Conversely, SEC14L3 knockdown resulted in effects that were opposite to those observed with the ectopic expression of SEC14L3 (Supplementary Fig. 6C–F). These results indicates that SEC14L3 positively promotes cuproptosis via ERK/YY1/FDX1 axis, which subsequently suppresses the growth of HCC in vivo.

Fig. 6. Copper-mediated SEC14L3 promoted cuproptosis of HCC via ERK/YY1/FDX1 signaling axis in vivo.

A Schematic illustration of the establishment of xenografted nude mice model and treatment procedures. Tumor size (B), weight (C), and volume (D) were monitored and measured starting from the 7th day after tumor cell inoculation in each group. (E) The lipoylated DLAT oligomerization, Ki-67, SEC14L3, p-ERK, YY1, and FDX1 were detected by immunofluorescence imaging (DAPI - blue, DLAT – red) and immunohistochemical staining (IHC) in HCC tissues from each group. Scale bars: 100 μm (10X). F The fluorescence intensity of DLAT from immunofluorescence imaging was quantified in each condition. G, H The proportion of Ki-67-expressing cells and the H score from immunohistochemical staining was used to semi-quantitative analysis for IHC in each condition. n  =  5 biologically independent animals for each group. Data in each graph indicate mean ± SEM. I, J The mRNA and protein expression levels of SEC14L3 were detected in HepG2 cells by qRT-PCR and immunoblot following treatment of different concentration of elesclomol (ES) and 1 μM CuCl₂. n  =  3 biologically independent experiments for each group. Data in each graph indicate mean ± SEM.

Interestingly, we found that the application of ES-CuCl₂ was observed to enhance the protein expression of SEC14L3, which in turn, appeared to stimulate the expression of FDX1 via ERK/YY1 axis (Fig. 6Eand Supplementary Fig. 6C). Hence, qRT-PCR and immunoblot assays were performed to assess the effects of copper on the mRNA and protein expression of SEC14L3 following exposure to various dose of ES and 1 μM CuCl₂ in HCC cells (HepG2 and SK-HEP-1). Consistent with our hypothesis, ES–CuCl₂ treatment dose-dependently increased SEC14L3 mRNA and protein levels in HCC cells (HepG2 and SK-HEP-1) (Fig. 6I,Jand Supplementary Fig. 6G,H).

Discussion

In this study, to the best of our knowledge, we uncover a mechanism by which copper-mediated SEC14L3 regulation of cuproptosis, a newly recognized programmed cell death pathway, constrains the growth of HCC. Mechanistically, elesclomol-facilitated copper intracellular transport upregulates SEC14L3 expression, which in turn enhances the expression of FDX1, a key gene in cuproptosis, via the ERK/YY1 axis. Notably, FDX1 reduces Cu²⁺ to Cu⁺ within cells, inducing lipoylation of DLAT. Subsequently, Cu⁺ directly binds to lipoylated DLAT, triggering its oligomerization and ultimately leading to cell death and inhibition of HCC growth (Fig. 7).

Fig. 7.

Illustration of the hypothesized mechanism by which copper-mediated SEC14L3 activation promoted cuproptosis of HCC via ERK/YY1/FDX1 signaling axis.

The SEC14L3 protein, predominantly localized in the cytoplasm of hepatocytes, plays a crucial role in cellular signal transduction and cell membrane transportation^3,21. Previous studies have found that SEC14L3 is downregulated in breast cancer cells, leading to the activation of the Wnt/β-catenin pathway and promoting tumor cell proliferation, migration, and invasion⁹. However, its role in HCC remains to be elucidated. Similarly, our results demonstrated that SEC14L3 expression was downregulated in HCC cells and tissues, which was associated with advanced clinicopathological stages and poorer prognosis of HCC patients. Additionally, we discovered that SEC14L3 positively suppressed the viability of HCC cells in a dose-dependent manner, with increasing copper ion concentrations. Concomitantly, SEC14L3 positively promoted a significant elevation of lip-DLAT levels and oligomerized lip-DLAT levels. These findings collectively suggest that SEC14L3 plays a crucial role in the development and progression of HCC, and that it exerts its anti-tumor effect by enhancing cuproptosis on HCC cells. However, given the potent cytotoxicity of ES-CuCl₂, future research should focus on determining the safe concentration threshold for normal cells.

Next, we delved deeper into the potential mechanism by which SEC14L3 regulates HCC cells growth through cuproptosis. Based on the results of bioinformatics analysis, we hypothesized that SEC14L3 might regulate the FDX1 via ERK/YY1 axis. Previous study has shown that the MAPK/ERK signaling pathway can upregulate CDK6 mRNA expression, which in turn regulates GSK3β activity and activates the Wnt/β-catenin signaling pathway, contributing to HCC development and lenvatinib resistance²². Previous evidence has demonstrated that FGFR2 inhibits BRCA1 expression via the ERK-YY1 signaling axis, promoting breast cancer development¹⁴. Furthermore, Wang et al. showed that activating the MAPK-p38 signaling pathway in renal clear cell carcinoma promotes phosphorylation of FOXO3, which negatively regulates FDX1 expression in the nucleus²³. Consistent to the previous findings, our results showed that FDX1 expression was significantly lower in HCC cells compared to normal liver cells. Notably, SEC14L3 was found to negatively regulate the p-ERK, which in turn led to a significant downregulation of YY1 expression. Consequently, this decrease in YY1 expression resulted in a significant upregulation of FDX1 expression, due to the reduced YY1 enrichment on the FDX1 promoter. Our results support the notion that MAPK/ERK activates YY1 expression in HCC, which in turn negatively regulates FDX1 expression by binding to its promoter. This study reveals that SEC14L3 promotes FDX1 expression, a key gene of cuproptosis, by mediating ERK/YY1 axis. Although we found that SEC14L3 negatively modulated the p-ERK, further clarification was still required regarding its specific regulatory mechanism in terms of ERK phosphorylation modification.

The regulation of cuproptosis by SEC14L3 was further explored via ERK/YY1/FDX1 axis to inhibit the growth on HCC. Study by Li et al. has revealed that MELK overexpression promotes the oligomerization of lip-DLAT and improved mitochondrial respiratory function via PI3K/mTOR pathway activation, which enhances the cuproptosis on HCC cells²⁴. Francisco et al. have shown the evidence that copper overload, mediated by copper ionophore disulfides, enhances the activity of trametinib to inhibit MAPK-ERK and synergistically inhibits melanoma growth by inducing oxidative stress and endoplasmic reticulum stress²⁵. Xie et al. have demonstrated that copper ionophoretic disulfiram/copper ions inhibit MAPK/ERK and PI3K/AKT signaling by increasing endogenous ROS levels, and selectively kill BRAFV600E-mutated thyroid cancer cells²⁶. The previous studies support our view that the MAPK/ERK signaling pathway is intimately linked to copper toxicity. Notably, we found that SEC14L3 significantly increased the levels of lip-DLAT and its oligomerized forms through the ERK/YY1/FDX1 axis, leading to enhanced cuproptosis and inhibition on HCC growth both in vitro and in vivo. Our current study provides evidence that SEC14L3-mediated ERK/YY1/FDX1 axis activity modulates cuproptosis in HCC.

The mechanism by which copper regulates gene transcription is not fully understood. Kersey et al. performed leverage whole transcriptome sequencing to reveal the transcriptomic changes induced by copper in hMSC cells, indicating a significant impact of copper on gene transcription²⁷. In fact, copper-induces transcription is tightly modulated by the chromatin accessibility, RNA polymerase II, transcription factors and even RNA-binding proteins²⁸. For instance, the copper-sensing transcription factor MAC1 plays a crucial role in regulating the downstream expression of CTR1 in Saccharomyces cerevisiae²⁹. Previous studies have found that DsbRS, a newly discovered protein in Pseudomonas aeruginosa, exhibits direct binding to copper and triggers the transcription of genes involved in the protein disulfide bonds formation³⁰. Interestingly, we unexpectedly found that in vivo treatment of ES-CuCl₂ promoted SEC14L3 expression, resulting in cuproptosis via ERK/YY1/FDX1 axis, thereby inhibiting HCC growth. Furthermore, our in vitro experiment consistently showed that a dose dose-dependent increase in SEC14L3 expression was observed following ES–CuCl₂ treatment, indicating that copper upregulated SEC14L3 expression in HCC.

In summary, the results presented in this study provide an unfound mechanism by which copper-mediated SEC14L3 enhances cuproptosis in HCC. Concretely, the upregulation of SEC14L3 expression by copper, which subsequently promotes FDX1 expression through the suppression of ERK/YY1 axis activity, leads to enhanced oligomerization of lip-DLAT and ultimately induces cuproptosis, thereby inhibiting the growth of HCC. Our study suggests that SEC14L3 represents a promising prognostic and therapeutic target for HCC, while also elucidating the underlying mechanisms involved in cuproptosis regulation. These findings lay a solid foundation for the future development of targeted drugs.

Materials and methods

HCC tissues, cell lines, and cell culture

46 pairs of HCC tissue and matched adjacent non-tumor tissues (MANT) specimens were gained from patients who underwent curative HCC resection at the First Affiliated Hospital of Jinan University (Guangzhou, China) from January 2004 to December 2018. The diagnosis of HCC in the pathological examination was made by experienced pathologists, and none of the patients had received preoperative treatment. The gender data were not collected in this study. The study was approved by the medical ethics committee of the First Affiliated Hospital of Jinan University, and written informed consent was obtained from each participant. All ethical regulations relevant to human research participants were followed.

The human HCC cell lines, HepG2, Huh7, and SK-HEP1, as well as the normal human liver cells (MIHA), were procured from the Abiowell (Changsha, China). These cell lines were cultured under standard conditions in DMEM (Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco) at 37 °C with a 5% CO₂ atmosphere. Elesclomol (ES) and tetrathiomolybdate (TTM) were dissolved in DMSO, while CuCl₂ in ddH₂O. Cells were treated with different concentrations of elesclomol (0 nM, 10 nM, 30 nM, 50 nM), fixed concentration of CuCl₂ (1 μM), and tetrathiomolybdate (TTM, 10 μM) for in vitro research.

Cell transfection

Cells were seeded in 24-well plates, incubated for 24 h, and then transfected with small interfering RNA (siRNA) for SEC14L3, siRNA for FDX1, SEC14L3 expression plasmids, and FDX1 expression plasmids using Lipofectamine 2000 (Invitrogen, USA) based on the manufacturer’s instructions. Cells were harvested 48 h after transfection for analysis. The SEC14L3 and FDX1 expression plasmids were synthesized and constructed by Synbio Technologies (Suzhou, China). The siRNA sequences were listed in Supplementary Table 1.

Quantitative real-time PCR (qRT-PCR)

The total RNA was isolated using TRIzol reagent (Invitrogen, USA) according to the protocols, and cDNA synthesis was performed using EasyScript First-Strand cDNA Synthesis SuperMix (TransGen Biotech, China). Subsequently, qPCR was carried out using Power SYBR Green qPCR SuperMix (DBI, China). The qRT-PCR primers were listed in Supplementary Table 1.

Immunoblot analysis

Total cellular protein were extracted using a total protein extraction kit (Solarbio, China) according to the manufacturer’s instructions at 4 °C for 30 min. The protein concentration was quantified using the bicinchoninic acid (BCA) assay kit (Beyotime Biotech, China) and samples were diluted to the same protein concentration. For the insoluble fraction, equal amounts of denatured protein were added to 4%-20% precast protein plus gels (Yeasen, China) and separated with by SDS-PAGE electrophoresis (Bio-Rad Biotech, China). For the soluble fraction, equal amounts of non-denatured protein were added to 4%-20% precast protein plus gels (Yeasen, China) and separated with by Native-PAGE electrophoresis (Bio-Rad Biotech, Beijing, China). Then, those protein were transferred on ice onto PVDF membranes (Millipore, USA) using eBlot® L1 wet transfer system (GenScript, NJ). The membranes were blocked using 5% nonfat milk prior to incubation with appropriate primary antibodies overnight, including anti-SEC14L3 (Abcam, Cat.Ab235110, 1:3000), anti-FDX1 (Proteintech, Cat.12592-1-AP, 1:2000), anti-ERK1/2 (Proteintech, Cat.11257-1-AP, 1:10000), anti-phospho-ERK1/2 (Thr202/Tyr204) (Proteintech, Cat.80031-1-RR, 1:5000), anti-YY1 (Proteintech, Cat.66281-1-Ig, 1:25000), anti-DLAT (Proteintech, Cat.13426-1-AP, 1:50000), anti-Lipoic Acid (Abcam, Cat. Ab58724, 1:5000), or GAPDH (Abcam, Cat. Ab8245, 1:10000). Subsequently, the membranes were incubated with secondary antibodies (HRP Goat anti-Rabbit IgG (BOSTER, Cat. BA1054, 1:20000)) for 40 min at room temperature and visualized using ECL (Biosharp, China) and analyzed using ImageJ software (version 1.48).

Determination of apoptosis

Apoptosis was analyzed using Elabscience’s Annexin V-FITC/PI Kit (E-CK-A211) with BD FACSCanto II. Cells were harvested, washed with PBS, and stained with Annexin V-FITC/PI following manufacturer’s protocol. Flow cytometry analysis (BD Bioscience) quantified apoptotic populations using the CellQuest Program.

Cell Counting Kit-8 (CCK-8) assay

The CCK-8 assay was utilized to assess cell viability, with cells being plated into 96-well plates. Subsequently, 10 µL of CCK-8 reagent (Sigma-Aldrich, USA) was aliquoted into each well for a 4-hour incubation period. Thereafter, the optical density (OD) was determined using a microplate spectrophotometer (Bio-DL K3 Plus, China) at a wavelength of 450 nm.

Immunohistochemistry (IHC) staining

IHC staining was performed as described previously³¹. In brief, paraffin tissue sections were deparaffinized using xylene and then rehydrated with a gradient of ethanol. Subsequently, 3% hydrogen peroxide was added for 15 min, followed by blocking with 10% goat serum for 30 min. The slides were incubated overnight at 4 °C with the corresponding primary antibody, including anti-SEC14L3 (Abcam, Cat.Ab235110, 1:200), anti-FDX1 (Proteintech, Cat.12592-1-AP, 1:200), anti-ERK1/2 (Proteintech, Cat.11257-1-AP, 1:200), anti-phospho-ERK1/2 (Thr202/Tyr204) (Proteintech, Cat.80031-1-RR, 1:200), anti-YY1 (Proteintech, Cat.66281-1-Ig, 1:200), and anti-Ki-67 (Proteintech, Cat.27309-1-AP, 1:200). On the next day, the primary antibody was removed and the slides were incubated for 30 min with the secondary antibody (HRP Goat anti-Rabbit IgG (BOSTER, Cat. BA1054, 1:200)). Tissue slides were immersed in horseradish lyase-labeled horseradish workup for 20 min. Then, the appropriate volume of the DAB staining solution (Sigma-Aldrich, USA) was added dropwise at room temperature. The protein expression intensity and extent were microscopically determined using the histochemical scoring system (H-score), which is calculated as the sum of the products obtained by multiplying the staining intensity (score ranging from 0 to 3) with the percentage of cells stained at a given intensity (ranging from 0 to 100)³¹.

Immunofluorescence (IF) staining

Cells were rinsed with PBS, fixed with 4% paraformaldehyde at room temperature for 15 min, permeabilized with 0.5% Triton X-100 for 20 min, and incubated with 3% BSA for 30 min. Next, the samples were incubated at 4 °C overnight with the primary antibody DLAT (Proteintech, Cat.13426-1-AP, 1:200) diluted in PBS. Subsequently, the cells were rinsed three times with PBS and incubated with fluorescein-labeled secondary antibodies for 50 min at room temperature. The cells were stained with MitoTracker Green FM (Invitrogen, USA) for 30 min at room temperature and counterstained with DAPI (Yeasen, China) for 10 min. IF was observed with fluorescence microscope.

Chromatin immunoprecipitation (ChIP)

ChIP assays were performed using a kit and following the manufacturer’s instructions (Pierce Agarose ChIP Kit, ThermoFisher Scientific). Anti-YY1 (Proteintech, Cat.66281-1-Ig, 5 μg/IP) was used to precipitate protein-bound DNA and normal rabbit IgG was used as a control. Isolated DNAs were quantitated using SYBR Green PCR SuperMix (DBI, China) on an ABI Real-Time PCR System. qRT-PCR results were normalized to input values. Primer sequences for ChIP qPCR were presented in Supplementary Table 1.

Construction of lentivirus vectors, and lentivirus production and infection

The sequences of SEC14L3 (NM_174975.5) and empty control were constructed into pLV-puro lentivirus vectors (pLV-puro-control and pLV-puro-SEC14L3). Lentiviruses were produced in HEK293T cells by cotransfection of lentiviral and packaging plasmids using Lipofectamine 2000 transfection reagent (Invitrogen, USA). Viruses were harvested 48 hours after transfection, concentrated using the Lenti-X concentrator, aliquoted, and stored at −80 °C for future use. HepG2 cells were transfected with LV-SEC14L3 or LV-control, respectively to generate HepG2 cells stably expressing SEC14L3 and control (HepG2-SEC14L3 and HepG2-control). Puromycin (6 μg/mL) was used for the stable cell lines selection.

Xenografts tumor experiment in nude mice

A total of 30 male BALB/c nude mice, aged 4 weeks and weighing 12–14 g, were sourced from Guangdong Medical Laboratory Animal Center, and housed under specific pathogen-free (SPF) conditions at the Department of Jinan University Animal Center and fed with enough water and food. ES were dissolved in DMSO and diluted by saline (5 mg/mL), while CuCl₂ were dissolved in ddH2O (6 μg/mL).

Then, they were randomly divided into 6 groups The experimental groups, which were subcutaneously injected with either HepG2-SEC14L3 cells or HepG2 cells transiently transfected with SEC14L3-siRNA (2 × 10⁶), pre-suspended in 100 µL PBS, received intraperitoneal administration of ES (50 mg/kg) + CuCl₂ (0.06 mg/kg), administered five times per week. The negative and positive control groups, which were subcutaneously injected with HepG2-control cells or HepG2 cells transiently transfected with siRNA-control (2 × 10⁶), pre-suspended in 100 µL PBS, received intraperitoneal administration of an equivalent volume of saline or ES (50 mg/kg) + CuCl₂ (0.06 mg/kg), administered five times per week. Additionally, for siSEC14L3 and si-control groups, 1 nmol cholesterol-modified SEC14L3 siRNA and corresponding siRNA control (RiboBio) dissolved in 10 μl saline were injected intratumorally (0.1 nmol per injection) at multiple points once every three days.

The tumor size was measured with calipers, and the tumor volume was calculated every week with the following equation: volume=width²×length/2. The tumor was excised prior to reaching a maximum volume of 2000 mm³. Additionally, measures were taken to ensure that the maximum permissible tumor volume was not exceeded throughout the experimental procedure. After 4 weeks, the mice were sacrificed by cervical dislocation under deep anesthesia with CO₂, and the tumors were collected, measured, and photographed. Tumor tissues were either fixed in 4% paraformaldehyde or snap-frozen at −80 °C for IHC staining. The animal experiments acquired ethical authorization from the Animal Research Ethics Committee of Jinan University. We have complied with all relevant ethical regulations for animal use.

Bioinformatics analysis

A total of 424 human HCC tissues and 50 adjacent non-tumor tissues (ANT) were obtained from The Cancer Genome Atlas (TCGA, http://gdc-portal.nci.nih.gov) datasets that were available in public repositories. RNA-seq count data was normalized using Transcripts Per Million (TPM). Differential expressions analysis was performed using the R package ‘DESeq2’. Survival analysis for overall survival (OS), disease specific survival (DSS), and progression-free survival (PFS) was calculated using the Kaplan–Meier and cox-proportional hazard model. The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) term gene-set enrichment analysis (GSEA) were conducted using the R package ‘ClusterProfiler’³². The Spearman correlation analysis was done to explore the gene expression relationship. The JASPAR database (https://jaspar.genereg.net/) was used to predict the YY1 binding site at FDX1 promoter.

Statistics and reproducibility

The experiments were conducted in triplicate. The Kolmogorov-Smirnov Test was employed to assess the normality of the continuous data distribution. If the data followed a normal distribution, either an independent T-test or paired T-test was utilized, and the results were presented as mean values with standard error of the mean (SEM). In case of non-normal distribution, either non-parametric K-independent Wilcoxon signed-ranks test or non-parametric two-tailed Wilcoxon matched-pairs signed-ranks test was applied, and the results were expressed as medians along with interquartile ranges. Detailed sample size (n) are shown in the figure legends. R software (version 4.3.3) and GraphPad Prism (version 9.0) were employed, and statistical significance was set at P < 0.05.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

Chutian Wu, Linjing Long, Min Wang, Lianli Shen, and Jianjun Hu contributed equally to this paper. The authors confirm contribution to the paper as follows: study conception and design: Chutian Wu, Min Wang, Shaohui Tang, and Yanfang Chen; data collection: Linjing Long, Lianli Shen, and Jianjun Hu; analysis and interpretation of results: Linjing Long, Lianli Shen, Jianjun Hu, Huijun Tang, Yuting Li, Shufen Feng, Xiongxiu Liu, and Ying Shi; draft manuscript preparation: Chutian Wu, Linjing Long, Shaohui Tang, and Yanfang Chen. All authors contributed to the manuscript for important intellectual content and approved the final submission of the manuscript.

Peer review

Peer review information

Communications Biology thanks Lih Wen Deng and the other, anonymous, reviewer for their contribution to the peer review of this work. Primary Handling Editors: Ruby Huang & Rosie Bunton-Stasyshyn. A peer review file is available.

Data availability

The data supporting the conclusions of this article is included within the article and its additional files. The source data can be found in the Supplementary Data.

Code availability

The R code is provided in github (https://github.com/biomedt/commsbio.git).

Competing interests

The authors declare no competing interests.

Ethics approval

The study was approved by the ethics committee of Jinan university first affiliated hospital (Approval number: KY-2024-201).

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-025-08101-z.