SQLE drives bladder cancer progression by boosting mitochondrial oxidative phosphorylation

Abstract

Bladder cancer (BCa) remains a prevalent malignancy with limited therapeutic options. Although cholesterol elevation links to BCa progression, the specific role of cholesterol metabolism remains unclear. Here, we demonstrate that squalene epoxidase (SQLE), a key cholesterol biosynthesis enzyme, drives BCa oncogenesis. SQLE is upregulated in BCa patients and correlates with poor survival. Functionally, bladder-specific Sqle transgenic (tg) mice showed accelerated tumorigenesis, while Sqle knockout (ko) demonstrated opposite effects in vivo. Mechanistically, SQLE localizes to mitochondria and directly interacts with Lon peptidase 1 (LONP1) to stabilize mitochondrial transcription factor A (TFAM) by preventing its proteolysis, leading to elevated oxidative phosphorylation (OXPHOS) and mitochondrial reactive oxygen species (mtROS). Pharmacological clearance of mtROS via Mito-TEMPO suppressed tumor growth in Sqle-overexpressing models. Importantly, the FDA-approved SQLE inhibitor terbinafine significantly suppressed BCa progression in preclinical models. Our findings establish SQLE as a critical regulator of mitochondrial metabolism in BCa, supporting SQLE inhibitors as potential therapeutics.

In bladder cancer, overexpression of SQLE impairs LONP1-mediated TFAM degradation through direct interaction with LONP1, thereby leading to increased mitochondrial OXPHOS and the accumulation of mtROS, which ultimately contributes to tumor growth. Treatment with the SQLE inhibitor terbinafine effectively blocks this process, providing a potential therapeutic strategy to inhibit tumor progression. The Graphical Abstract was created using Smart.Servie (https://smart.servier.com/citation-sharing/).

Introduction

Bladder cancer represents the second most prevalent urological malignancy globally [1, 2]. Elevated serum cholesterol has been recognized as a significant risk factor for the development of BCa [3, 4], and statin therapy, which effectively reduces serum cholesterol levels, has demonstrated beneficial effects in its treatment [3–5]. Nonetheless, the relationship between elevated tissue cholesterol levels and cancer pathogenesis remains poorly understood, primarily due to the intricate feedback mechanisms governing cholesterol metabolism. SQLE, a rate-limiting enzyme in cholesterol biosynthesis, is predominantly localized within the endoplasmic reticulum (ER) [6, 7]. Emerging evidence has indicated that SQLE may function as an oncogene across various cancer types, including colorectal cancer [8], hepatocellular carcinoma [9, 10] and lymphomas [11]. While recent investigations have associated SQLE with poor survival outcomes in BCa [12, 13], its specific role in the initiation and progression of this malignancy is not yet elucidated.

Mitochondria serve as crucial hubs for cholesterol metabolism [14]. Current understanding of mitochondrial cholesterol homeostasis has primarily focused on its exogenous sources, with well-characterized mechanisms involving StAR and TSPO-mediated transport of ER-derived cholesterol for steroid, oxysterol and bile acid synthesis [15]. While these studies have established that excessive imported cholesterol disrupts mitochondrial function and redox balance contributing to tumor progression and chemoresistance [16–18], the potential role of endogenous mitochondrial cholesterol synthesis remains surprisingly unexplored. This knowledge gap is particularly notable given that pharmacological agents like lovastatin that modulate mitochondrial cholesterol show anti-tumor efficacy [19]. Importantly, whether mitochondria possess autonomous cholesterol synthesis capability, and how such potential endogenous production might differentially impact cancer biology compared to imported cholesterol, represents a fundamental unanswered question in the field. Investigation of cholesterol synthesis enzyme localization and activity within mitochondria thus holds significant promise for revealing novel mechanisms of metabolic regulation in cancer progression, particularly in malignancies like bladder cancer where cholesterol metabolism is increasingly recognized as a key pathogenic factor.

Here, we provide evidence that SQLE expression is upregulated in BCa as compared to adjacent normal tissues, with its elevated expression correlating with poor survival outcomes in clinical contexts. We conducted a comprehensive evaluation of the oncogenic role of SQLE by using bladder cancer cell lines in vitro and generating bladder-specific Sqle transgenic and knockout mice to assess its impact on BCa development in vivo. Furthermore, we demonstrated that SQLE facilitates BCa growth by enhancing mitochondrial function and the excess production of mtROS, while also showing that the clearance of mtROS effectively inhibits BCa both in vitro and in vivo. We further demonstrated that SQLE localizes to mitochondria, where it protects TFAM from proteolytic degradation through direct binding to LONP1. Via this mechanism, upregulation of SQLE promotes OXPHOS and the progression of BCa. Lastly, we evaluated the therapeutic efficacy of targeting SQLE in the treatment of BCa, both in vitro and in vivo.

Materials and methods

Patients and samples

All human BCa tumor tissues and adjacent normal tissues utilized in this study were procured from Harbin Medical University Cancer Hospital. No specific inclusion or exclusion criteria were established for this study. The participants’ sex was randomly selected, encompassing both male and female study participants, to ensure the representativeness and balance of the samples. Following surgical procedures, all samples were promptly frozen in liquid nitrogen and subsequently stored in a -80°C freezer or prepared as paraffin-embedded wax blocks for further analysis. This research project has received approval from the Ethics Committee of Harbin Medical University Cancer Hospital (Approval No. KY-2024-70), and informed consent forms have been obtained from each participating patient. All methods were performed in accordance with the relevant guidelines and regulations.

The human BCa tissue microarray (HBlaU079Su01) utilized in this study was procured from OUTDO Biotech Company (Shanghai, China).

Multi-omics data analysis

The single-cell RNA sequencing raw data of public dataset CNP0000460 [20] were processed through a standardized analytical pipeline. Sequencing reads were aligned to the human reference genome GRCh38 using Cell Ranger (version 8.0.0), followed by quality control and normalization in Seurat [21] (version 5.1.0) under R environment (version 4.3.3). In brief, low-quality cells exhibiting >25% mitochondrial gene content or <500 detected genes were excluded to ensure data robustness. Normalization was performed using the Normalize Data function with a scale factor of 10,000, preserving biological variance while minimizing technical noise. Cell clusters were annotated using canonical markers, from which epithelial cells were rigorously identified through co-expression of EPCAM, KRT13, and KRT7 [20]. Expression of SQLE was visualized via FeaturePlot, complemented by quantitative comparisons between normal and tumor tissues using non-parametric Wilcoxon tests.

Transcriptomic and proteomic datasets were quantile-normalized followed by differential analysis with limma [22]. Enriched pathways were identified through clusterProfiler [23] (v4.10.1), and spearman correlations were computed using corrplot (v0.92). SQLE mRNA levels between tumor and normal tissues from TCGA (https://portal.gdc.cancer.gov/) were analyzed using paired Student’s t-test, from GSE3167 and GSE188715 were analyzed using unpaired Student’s t-test. For survival analysis, 407 patients from TCGA were categorized into high and low SQLE expression groups. Kaplan-Meier curves were generated to assess overall survival (OS), and differences between groups were evaluated by log-rank test.

For targeted metabolomics data, differential metabolites were screened by integrating multivariate and univariate statistical approaches. First, orthogonal partial least squares-discriminant analysis was performed using the ropls [24] package to evaluate metabolic profile separation between groups. Metabolites with variable importance in projection (VIP) > 1, |log2(FC)| > 0.263 and p-value < 0.05 were selected as potential candidates for further analysis. Finally, a heatmap was generated to visualize the expression patterns of these differential metabolites across samples.

Cell lines and cell culture

The J82, RT112, tccsup were cells obtained from ATCC (Manassas, VA). The 647 V and UMUC3 cells were provided by Professor Wanghai Xu from Harbin Medical University Cancer Hospital, China. All cells were cultured in DMEM supplemented with 10% FBS and 1% penicillin–streptomycin at 37 °C in a humidified atmosphere containing 5% CO₂. Additionally, all cell cultures were confirmed to be free of mycoplasma contamination through appropriate testing.

Animal models

This research project has received approval from the Experimental Ethical Inspection of Fourth Clinical Hospital of Harbin Medical University (Approval No. 2024-DWSYLLCZ-50). All methods were performed in accordance with the relevant guidelines and regulations.

A xenograft human BCa mouse model was established using the SQLE mutation-stable cell lines (J82 and UMUC3). Each cell (3×10⁶ cells in 0.05 mL PBS and 0.05 mL Matrigel Matrix) was injected subcutaneously into the right dorsal flank of 6-week-old male Balb/c nude mice. At the indicated time point, the mice were killed and examined.

For bladder-specific Sqle overexpression mice, Sqle transgenic mice (pCAG-loxp-stop-loxp-Rosa26-Sqle, Rosa26-Sqle) and Tcf21-iCreETR2 mice on a C57BL/6 J background were generated by Biocytogen Company (Beijing, China). To drive the bladder-specific overexpression of Sqle (Sqle bladTg), Rosa26-Sqle mice were crossed with Tcf21-iCreETR2 mice. At the age of 6 weeks, Sqle bladTg male mice and Rosa26-Sqle control male mice were used to establish a BCa model.

For bladder-specific Sqle knockout mice, Sqle knockout mice on a C57BL/6 J background were also generated using CRISPR/Cas9 technology by Biocytogen Company. LoxP sites were inserted into non-conserved regions flanking exon 2 (Sqle^fl/fl). Then, Sqlefl/fl mice were crossed with Tcf21-iCreETR2 mice to generate bladder-specific knockout Sqle mice (Sqle bladKO). At the age of 6 weeks, Sqle bladKO and Sqle^fl/fl male mice were used to induce BCa.

For functional studies, to activate Sqle overexpression, tamoxifen (100 mg/kg, intraperitoneally) was first injected into Sqle bladTg mice once a day for five days to activate CreERT2, and their Rosa26-Sqle control mice were also treated with tamoxifen at the age of 6 weeks. Seven days after the last injection of tamoxifen, the mice were fed with 0.05% BBN in drink water for 12 weeks and then with carcinogen-free water for 8 weeks to establish bladder cancer. Mice were then sacrificed at week 20 from the start of the carcinogen (BBN) treatment. Histology was scored after H&E staining by a pathologist blinded to the nature of the samples. Similarly, the Sqle bladKO mice were established and also employed with similar methods to induce BCa.

For the ROS inhibitor (Mito-TEMPO) experiments, Sqle bladTg mice and their Rosa26-Sqle control mice were first induced with tamoxifen once a day for five days at the age of 6 weeks. Seven days after last injection of tamoxifen, the mice were fed with 0.05% BBN in drink water for 20 weeks. Nine weeks after carcinogen (BBN) induction, these mice were randomized and received: (1) Rosa26-Sqle + normal saline (intraperitoneally, once per week); (2) Sqle bladTg + normal saline (intraperitoneally, once per week); (3) Sqle bladTg + Mito-TEMPO (0.1 mg/kg, intraperitoneally, once per week) till the termination of experiment. The mice were sacrificed and examined at 20 weeks after carcinogen (BBN)induction.

For therapeutic study, Sqle bladKO and littermate Sqle^fl/fl male mice were fed with 0.05% BBN in drink water for 20 weeks at the age of 6 weeks. Also, to study the treatment of blocking Sqle expression in the early stage of tumor induction, mice were injected with tamoxifen (once a day for five times) to activate CreERT2 to genetically delete Sqle at day 49 from the start of the carcinogen (BBN) treatment. The mice were sacrificed and examined at 20 weeks after carcinogen (BBN) induction.

For the SQLE inhibitor (terbinafine) experiments, A xenograft human BCa mouse model was established using the human BCa cell lines (RT112). At the indicated time point, 6-week-old male Balb/c nude mice were randomly divided into two groups. Mice were treated with terbinafine (80 mg/kg/day) and phosphate-buffered saline (PBS) (vehicle) through oral gavage daily. At the indicated time point, the mice were killed and examined.

The therapeutic effect of terbinafine was also evaluated in the orthotopic mice model. At the age of 6 weeks, wild-type (C57BL/6 J background) mice were fed with 0.05% BBN in drink water for 20 weeks to induce BCa. Ten weeks after carcinogen (BBN) induction, mice were randomly divided into two groups. Mice were treated with terbinafine (80 mg/kg/day) and PBS (vehicle) through oral gavage daily until the end of the experiment.

All experimental procedures mentioned above were approved by the Animal Ethics Committee of the 4th Affiliated Hospital of Harbin Medical University (Approval no. 2024-DWSYLLCZ-50). All mice had free access to food and water, with a strict 12-hour light and 12-hour dark cycle, in a temperature-controlled environment (22 ± 2°C). Male mice of the same age (6 weeks old) were chosen for the study, as described in the figure legends, and were randomly assigned to different experimental groups. Prior to treatment, there were no significant differences in body weights among the groups. Euthanasia of the mice was performed by CO₂ asphyxiation, and no mice were excluded from the subsequent analysis.

MTT assay

Cell proliferation activity was assessed using the MTT assay. Cells were cultured at a density of 1,000 cells per well in 96-well plates. At specified time points, 10 µl of MTT solution (5 mg/ml) was added to each well. Subsequently, optical density at 490 nm (OD490) was measured using a spectrophotometer to construct a cell growth curve. All experiments were conducted in triplicate.

Colony formation assay

In the colony formation assay, cells were plated at a density of 1,000 cells per well in 6-well plates. Following a culture period of 8 to 14 days, cells were fixed with 70% ethanol and subsequently stained with a 0.5% crystal violet solution for 15 minutes. Colonies containing more than 50 cells were counted. All experiments were performed in triplicate.

Monolayer wound healing assay

Cells were seeded in 6-well plates and incubated overnight at 37 °C. Wounds were created in the cell monolayer using a pipette tip, after which the cells were cultured in serum-free medium. Images were captured immediately and at various intervals during cell migration to assess wound closure. Migration distances were subsequently compared. All experiments were conducted in triplicate.

Transwell migration assay

Transwell migration assays were conducted using a transwell chamber according to established protocols. Cells were resuspended in 200 µl of serum-free starvation medium and placed in the upper chamber, while 600 µl of complete medium was added to the lower chamber. The cells were incubated at 37°C for 24 to 48 h, after which they were fixed with 4% formaldehyde and stained with 0.5% crystal violet. All experiments were executed in triplicate.

Cell cycle analyses

For cell cycle analysis, cells were serum-starved overnight and then stimulated with complete medium for 4 to 8 hours. Following stimulation, the cells were fixed in 70% ethanol, stained with propidium iodide, and analyzed via flow cytometry. All experiments were conducted in triplicate.

Cholesterol concentrations

Cells (10⁶ cells) were used for quantifying cholesterol concentrations, measured with a Cholesterol Quantification Kit (ab65359, Abcam). All experiments were carried out in triplicate.

Western blot analysis

Western blotting was performed as previously described [25, 26]. Approximately 25 µg of protein was loaded and separated using SDS-PAGE, followed by transfer onto nitrocellulose membranes. After blocking with 3% BSA in TBST solution at room temperature for 2 h, the membrane was incubated with the primary antibody in 1.5% BSA overnight at 4 °C, followed by incubation with the secondary antibody at room temperature for 2 h. The proteins of interest were visualized using an enhanced chemiluminescence reagent (Omin-ECL, Yamei, China).

Immunohistochemistry

Fresh tumors were fixed in 4% paraformaldehyde overnight and then transferred to 70% ethanol for 12 hours. The specimens were embedded in paraffin and sectioned into 4 µm slices. The sections were dewaxed by baking at 60 °C for 30 minutes. Paraffin-embedded tissue slides were dewaxed using xylene and treated sequentially with 100%, 90%, and 70% ethanol, followed by a wash with water. The slides underwent antigen unmasking. For immunohistochemical staining, endogenous peroxidase activity was inhibited with 3% hydrogen peroxide at room temperature for 15 minutes. Tissues were blocked with 5% goat serum for one hour and incubated with primary antibodies at 4 °C overnight. After washing with PBS-T, secondary antibodies were applied for one hour at room temperature.

Immunofluorescence

Cells were collected and washed three times. Following fixation with 4% paraformaldehyde, permeabilization with 0.1% Triton X-100, and blocking with 5% BSA, the slides were treated with primary antibodies overnight at 4 °C. After washing three times with PBST, cells were incubated with secondary antibodies conjugated to fluorescent dyes for one hour. After three additional washes with PBST, cell nuclei were stained with DAPI. Imaging was performed using a laser scanning confocal microscope (Zeiss LSM 800) with a 63× oil immersion objective. For tissue analysis, sections were dewaxed by baking at 60 °C for 30 minutes. After dewaxing and rehydration, antigen retrieval was conducted. The sections were incubated with primary antibodies overnight at 4 °C, followed by three washes with PBST and incubation with fluorescent dye-labeled secondary antibodies for one hour. After three washes with PBST, cell nuclei were stained with DAPI, and the tissues were imaged using a laser scanning confocal microscope (Zeiss LSM 800) with a 20× immersion objective.

FACS analysis of mitochondrial mass

Cells were collected and washed three times with PBS. The cells were stained with 100 nM MitoTracker Red (Thermo) at 37 °C for 30 minutes. After centrifugation, the cells were resuspended in PBS for subsequent assays. Samples were analyzed using an LSRFortessa flow cytometer (BD FACSCalibur), and data analysis was conducted with FlowJo software. All experiments were performed in triplicate.

Seahorse XF96 respirometry

Cells were seeded at a density of 10,000 cells per well in XFe96 cell culture microplates (Agilent) and allowed to adhere overnight. After complete cell adhesion, the cells were incubated with Seahorse XF DMEM (DMEM-100) supplemented with 10 mM glucose, 2 mM glutamine, and 1 mM pyruvate at 37 °C for 1 h in a CO₂-free incubator. After preincubation, oligomycin (1.5 mM), FCCP (2 mM), and rotenone/antimycin A (0.5 mM) were sequentially injected to assess the OCR according to the manufacturer’s instructions. Data were analyzed using Seahorse XF96 Wave software. All experiments were performed in triplicate.

Co-Immunoprecipitation (Co-IP)

Samples were collected and incubated with specific antibodies overnight at 4°C. Magnetic beads (MCE) were equilibrated with lysis buffer and incubated with the protein-antibody mixture for four hours at 4°C with gentle rotation. The protein mixture was denatured by heating at 100°C for five minutes in sodium dodecyl sulfate (SDS) loading buffer. All experiments were performed in triplicate.

Statistical analysis

All statistical analyses were conducted using GraphPad Prism (version 8.0.1) or SPSS (version 21.0) software. The outlier tests were performed per ROUT method [27]. The sample size was determined based on prior experience, sample availability and previously published studies [9, 28]. Data are presented as mean ± standard deviation (SD). One-way analysis of variance (ANOVA) was used for comparisons among multiple groups. A two-sided Student’s t-test was utilized for comparisons between two variables. Kaplan-Meier survival curves and log-rank tests were used to evaluate the association between overall survival and expression levels. P values < 0.05 were considered to be significantly different.

Results

SQLE is highly expressed in cases of BCa and correlates with poor prognosis

To investigate the relationship between BCa and cholesterol synthesis, we analyzed the mRNA expression of cholesterol biosynthesis genes (Fig. S1A) using data from The Cancer Genome Atlas (TCGA), GSE3167 [29] and GSE188715 [30] databases. We found significant upregulation of HMGCR, MVK, MVD, FDPS, FDFT1, SQLE, CYP51A1, DHCR24 and DHCR7 in the TCGA database; HMGCR, FDPS, FDFT1, SQLE, CYP51A1, DHCR24 and DHCR7 in the GSE3167 database; and MVK, MVD, FDPS, FDFT1, SQLE and DHCR7 in the GSE188715 database (Fig. S1B, C, D, and Fig. 1A). Conversely, MVK mRNA expression showed marked downregulation in the GSE3167 database compared to adjacent normal tissues (Fig. S1C). Given the overexpression of SQLE in BCa, we evaluated its clinical significance. By Kaplan-Meier survival analysis of the TCGA database we found that elevated SQLE mRNA levels correlated with poor survival outcomes in patients with BCa (p < 0.01) (Fig. 1B). Additionally, by multivariate Cox proportional hazards regression analysis we found that high SQLE mRNA expression was an independent prognostic factor for reduced disease-specific survival in patients with BCa (p < 0.05; HR 1.969; 95% CI 1.023 to 3.792) (Fig. 1B). Furthermore, SQLE protein levels were elevated in BCa tissues compared to adjacent normal tissues, as confirmed by Western blotting and immunofluorescence analyses (Fig. 1C, D). Also, by tissue microarray (TMA) analysis of 63 cases of BCa using immunohistochemistry (IHC) we found higher SQLE protein expression in BCa tissues compared to adjacent normal tissue (Fig. 1E). Elevated SQLE protein levels were also associated with poor survival in patients with BCa, as demonstrated by Kaplan-Meier survival analysis (Fig. 1F). Single-cell RNA sequencing from Dataset CNP0000460 further indicated increased mRNA levels of SQLE and six additional genes involved in cholesterol biosynthesis pathway in epithelial cells of BCa tumors (Fig. S1E and Fig. 1G). These findings suggest that SQLE is overexpressed in human BCa and correlates with a poorer prognosis.

Fig. 1. SQLE is aberrantly elevated in human BCa.

A The mRNA expression of SQLE in tumor and adjacent normal tissues from the TCGA cohort, the GSE3167 cohort, and the GSE188715 cohort. B The relationship between SQLE mRNA expression and poor survival outcomes as analyzed by Kaplan-Meier curves (left) and Cox regression analysis (right) within the TCGA cohort. C SQLE protein expression in BCa (T) vs. paired adjacent normal tissue (N) in an independent cohort (n = 8 from the Harbin Medical University Cancer Hospital). D Representative immunofluorescent staining of SQLE protein in BCa tissues and adjacent normal tissues. Staining for SQLE (Orange), DAPI (Blue) (n = 5 from the Harbin Medical University Cancer Hospital). E TMA analysis of SQLE protein expression in BCa (T) vs. paired adjacent normal tissue (N) (n = 16). F Representative images of BCa samples with high, medium, and low SQLE protein expression levels from TMA (left), and Kaplan–Meier survival analysis demonstrating the association between SQLE expression and survival (right). G UMAP plots showing the SQLE expression among cells in adjacent normal and tumor tissues from Dataset CNP0000460 (left). SQLE expression was determined in the epithelial cells among adjacent normal and tumor tissues (right). Data are represented as means ± SD. *P < 0.05, **P < 0.01, **P < 0.001. Paired, two-tailed Student’s t-test (A(left), C, D, E, G). Unpaired, two-tailed Student’s t-test (A (middle, right)). Kaplan–Meier survival analysis (B(left), F) and Cox regression analysis (B (right)). Scale bars, 20 µm (D). Scale bars, 500 µm (low magnification) and 200 µm (higher magnification) (E, F).

SQLE promotes cell growth and migration and regulates cell cycle progression in BCa cell lines

To investigate the potential biological role of SQLE in BCa cells, we first quantified its baseline expression across three cell lines. SQLE was highest in 647 V cells, lowest in J82 cells, and intermediate in UMUC3 cells (Fig. 2A). Based on these patterns, we established UMUC3 and J82 cell lines that stably overexpress SQLE, as well as 647 V and UMUC3 cells with SQLE knockout, confirmed through Western blot analysis (Fig. 2B). SQLE overexpression significantly enhanced cell proliferation, as shown by cell growth curves and colony formation assays (Fig. 2C, D). Conversely, silencing SQLE in 647 V and UMUC3 cells using sgRNA markedly inhibited cell proliferation compared to control groups (Fig. 2C, D). Additionally, SQLE overexpression increased the migration capability of UMUC3 and J82 cells, while SQLE knockout diminished the migration capacity of 647 V and UMUC3 cells (Fig. 2E, F). We then found that SQLE regulates cell cycle progression, as its overexpression resulted in an increased population of cells in the S-phase and a decreased population in the G0/G1 phase in UMUC3 and J82 cells (Fig. S2A and Fig. 2G). In contrast, SQLE knockout produced the opposite effects in 647 V and UMUC3 cells (Fig. S2A and Fig. 2G), indicating that SQLE accelerates the G1-S phase transition. Consistent with these findings, SQLE overexpression induced the protein expression of cyclin D1 and proliferating cell nuclear antigen (PCNA) in UMUC3 and J82 cells, whereas SQLE knockout resulted in reduced expression of these proteins in 647 V and UMUC3 cells (Fig. 2H and Fig. S2B). Notably, the re-expression of SQLE in 647V-sgSQLE and UMUC3-sgSQLE cells (Fig. S2C) restored the viability of BCa cells, as evidenced by cell growth curves, colony formation assays and wound healing assays (Fig. S2D–F). These results demonstrate that SQLE promotes malignant behaviors in BCa cell lines.

Fig. 2. SQLE promotes BCa cell proliferation, migration and cell cycle progression.

A Western blots analysis of SQLE expression in 647 V, J82 and UMUC3 cells. B Western blots analysis of SQLE expression in UMUC3 and J82 cells overexpressing SQLE (upper) and in 647 V and UMUC3 cells with SQLE knockout (bottom). The effects of SQLE overexpression (upper) or knockout (bottom) on the viability of BCa cells, as evidenced by cell growth curves (C) (n = 3, performed in triplicate), colony formation assays (D) (n = 3, performed in triplicate), wound healing assays (E) (n = 3, performed in triplicate), and transwell assays (F) (n = 3, performed in triplicate). G The effects of SQLE overexpression (upper) or knockout (bottom) on the cell cycle, as evidenced by flow cytometry analysis (n = 3, performed in triplicate). H Representative Western blots analysis of PCNA and Cyclin D1 expression in BCa cells with SQLE overexpression (upper) or knockout (bottom) (n = 3, performed in triplicate). Data are represented as means ± SD.*P < 0.05, **P < 0.01, ***P < 0.001. Unpaired, two-tailed Student’s t-test (C, D, E, F, G).

Bladder-specific Sqle manipulation modulates tumorigenesis in mouse models

To elucidate the pro-tumoral role of SQLE in BCa development in vivo, we inoculated nude mice with cells stably expressing either the control vector or SQLE, establishing a subcutaneous tumor formation model. After 22 days, tumor volume and weight were significantly greater in the SQLE overexpression group compared to controls (Fig. 3A–C). Subcutaneous tumors derived from J82 cells overexpressing SQLE demonstrated elevated levels of cyclin D1 and PCNA (Fig. 3D and Fig. S3A). Conversely, SQLE knockout resulted in a marked reduction in tumor growth, accompanied by decreased levels of cyclin D1 and PCNA (Fig. 3E–H and Fig. S3B).

Fig. 3. Bladder-specific Sqle manipulation modulates tumorigenesis in mouse models.

Subcutaneous tumors derived from J82 cells overexpressing SQLE analyzed for tumor volume (A, B) and weight (C). D Western blots analysis of PCNA and cyclin D1 expression in subcutaneous tumors derived from J82 cells with or without SQLE overexpression. Subcutaneous tumors derived from UMUC3 cells with SQLE knockout analyzed for tumor volume (E, F) and weight (G). H Western blots analysis of PCNA and cyclin D1 expression in subcutaneous tumors derived from UMUC3 cells with or without SQLE knockout. I A schematic diagram depicting the experimental design for the induction of bladder cancer in Rosa26-Sqle mice and Sqle bladTg mice utilizing 0.05% BBN is presented. J Representative MRI imaging, gross histology, and H&E staining (left), along with tumor incidence and severity (right) of the bladder in Sqle bladTg mice (n = 5) compared to Rosa26-Sqle mice (n = 6) mice. K Representative Ki-67 staining in the bladders of BBN-treated Rosa26-Sqle mice (n = 6) versus Sqle bladTg mice (n = 5). L Western blots analysis of PCNA and cyclin D1 expression in bladder tissues from BBN-treated Rosa26-Sqle mice versus Sqle bladTg mice. M A schematic diagram illustrating the experimental design for the induction of bladder cancer in Sqle^fl/fl mice and Sqle bladKO mice using 0.05% BBN is shown. N Representative MRI imaging, gross histology, and H&E staining (left), along with tumor incidence and severity (right) of the bladder in Sqle bladKO mice (n = 6) compared to Sqle^fl/fl mice (n = 6). O Representative Ki-67 staining in the bladders of BBN-treated Sqle^fl/fl mice (n = 6) versus Sqle bladKO mice (n = 6). P Western blots analysis of PCNA and cyclin D1 expression in bladder tissues from BBN-treated Sqle^fl/fl mice versus Sqle bladKO mice. Data are expressed as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Paired, two-tailed Student’s t-test (B, F). Unpaired, two-tailed Student’s t-test (C, G, K, O). Scale bars, 500 µm (low magnification) and 200 µm (high magnification) (J, N). Scale bars, 200 µm (K, O).

Next, we generated bladder-specific Sqle transgenic mice (Sqle bladTg) by crossing Rosa26-Sqle mice with Tcf21-iCreERT2 mice (Fig. S3C). The bladder-specific overexpression of Sqle was induced through tamoxifen injection, which activates Cre in the bladder at the age of 6 weeks (Fig. 3I). To initiate bladder tumorigenesis, the mice were subjected to a 12-week exposure to 0.05% N-butyl-N-(4-hydroxybutyl)-nitrosamine (BBN), followed by an 8-week period of carcinogen-free water to promote bladder cancer in both Sqle bladTg and Rosa26-Sqle mice (Fig. 3I). Following 20 weeks of treatment, magnetic resonance imaging (MRI) conducted prior to sacrifice revealed the presence of histological morphology consistent with the development of BCa (Fig. 3J). By H&E analysis we further confirmed that Sqle bladTg mice exhibited a higher tumor incidence (3/5 in the Sqle bladTg mice vs. 1/6 in the control mice) and greater disease severity compared to Rosa26-Sqle mice (Fig. 3J). By Ki-67 staining we further found that there was greater cell proliferation in non-tumor bladder tissues from Sqle bladTg mice compared to Rosa26-Sqle mice, with tumors demonstrating the highest Ki-67 scores (Fig. 3K). Furthermore, non-tumor regions of bladder tissues from Sqle bladTg mice exhibited elevated protein expression levels of cyclin D1 and PCNA compared to control mice (Fig. 3L and Fig. S3D).

Moreover, we further investigated the role of Sqle in BCa by utilizing Sqle bladKO mice (Fig. S3E) in the BBN-induced BCa model (Fig. 3M). Sqle bladKO mice displayed reduced bladder abnormalities, with a lower degree of tumor incidence (1/6 in the Sqle bladKO mice vs. 3/6 in the control mice) and disease severity in comparison to Sqle^fl/fl mice (Fig. 3N). Ki-67 staining revealed diminished cell proliferation in non-tumorous bladder tissues of Sqle knockout mice compared to Sqle^fl/fl mice, with tumors from Sqle^fl/fl mice exhibiting the highest Ki-67 scores (Fig. 3O). Additionally, non-tumor regions of bladder tissues from Sqle knockout mice demonstrated reduced levels of cyclin D1 and PCNA compared to control mice (Fig. 3P and Fig. S3F). These findings substantiate the notion that SQLE plays an oncogenic role in bladder tumorigenesis in vivo.

SQLE enhances mitochondrial oxidative phosphorylation

To elucidate the mechanisms by which SQLE influences the progression of BCa, we conducted a comparative proteomics analysis on BCa cells with SQLE knockout relative to a control vector. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed significant alterations in OXPHOS in response to SQLE deficiency in 647 V cells (Fig. 4A). Gene Set Enrichment Analysis (GSEA) indicated the enrichment of OXPHOS signatures in J82-SQLE cells (Fig. 4B). RNA sequencing data demonstrated an increased percentage of upregulated mitochondrial mRNA content in UMUC3-SQLE and J82-SQLE cells (Fig. S4A–D). Comparative proteomics data indicated a greater proportion of upregulated mitochondrial protein content in UMUC3-SQLE and J82-SQLE cells (Fig. S4E–H), whereas SQLE knockout resulted in the opposite effect in 647V-sgSQLE and UMUC3-sgSQLE cells (Fig. S4I–L). Subcutaneous tumors derived from UMUC3-sgSQLE cells exhibited a higher percentage of downregulated mitochondrial proteins and SQLE are positively correlated with many mitochondrial-encoded proteins (Fig. 4C, D). These results suggest that SQLE expression causes mitochondrial dysfunction, specifically activating OXPHOS.

Fig. 4. SQLE knockout suppresses oxidative phosphorylation in vitro.

A KEGG pathway analysis of the proteomic data revealed that SQLE deficiency suppressed the OXPHOS pathway in 647V cells. B GSEA revealed a positive correlation between SQLE protein expression and OXPHOS pathway enrichment in J82 cells. A volcano plot and heatmap (C) along with a correlation analysis (D) to evaluate the proteomic data from the subcutaneous tumors of UMUC3 cells with or without SQLE knockout. E The effects of SQLE knockout on respiratory capacity in 647 V and UMUC3 cells (n = 3, performed in triplicate). The effects of SQLE knockout on mtMPs as assessed by flow cytometry (F) (n = 3, performed in triplicate) and immunofluorescence (G) (mitochondria: red, DAPI: blue) (n = 3, performed in triplicate). H Representative Western blot analysis of MTCO2 and CIV-MTCO1 in 647 V and UMUC3 cells with or without SQLE knockout (n = 3, performed in triplicate). I Immunofluorescence analysis of MTCO2 in BCa (T) versus normal tissues (N) (n = 5 from the Harbin Medical University Cancer Hospital). Data are represented as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Spearman correlation analysis (D). Unpaired, two-tailed Student’s t-test (E, F, G). Paired, two-tailed Student’s t-test (I). Scale bars, 10 µm (G), Scale bars, 20 µm (I).

To confirm the association between SQLE expression and mitochondrial OXPHOS, we assessed the cell-autonomous effects of SQLE on mitochondrial respiration in bladder cancer cells. SQLE knockout resulted in decreased basal respiratory capacity, maximum respiratory capacity and spare respiratory capacity in 647 V and UMUC3 cells (Fig. 4E). Mitochondrial membrane potentials (mtMPs) were also diminished following SQLE knockout (Fig. 4F, G). The reduced SQLE levels decreased the expression of mitochondrial-encoded proteins MTCO2 and CIV-MTCO1 in 647 V and UMUC3 cells (Fig. 4H and Fig. S5A). Conversely, SQLE overexpression in J82 cells resulted in increased oxygen consumption rate (OCR) and mtMPs (Fig. S5B, C) and upregulated the proteins MTCO2 and CIV-MTCO1 (Fig. S5D). Immunofluorescence analysis of BCa samples revealed elevated levels of MTCO2 protein in BCa tissues compared to adjacent normal tissues (Fig. 4I). These findings confirm that SQLE expression activates mitochondrial OXPHOS in bladder cancer.

Mitochondrial ROS clearance attenuates the oncogenic function of SQLE

Next, we characterize the impact of mitochondrial oxidative phosphorylation (mtOXPHOS) on SQLE-induced BCa progression by modulating OXPHOS in cell lines with stable SQLE expression. Metformin, a known inhibitor of mitochondrial respiratory chain complex I [31, 32] (Fig. S6A), did not inhibit SQLE-promoted cell proliferation, although it exhibited a slight reduction in SQLE-induced migration (Fig. S6B–D). Furthermore, succinate, a metabolite associated with mitochondrial respiratory chain complex II and the tricarboxylic acid (TCA) cycle [33, 34] (Fig. S6A), was significantly elevated in J82-SQLE cells (Fig. S6E) and in human BCa tissues compared to control adjacent normal tissue (Fig. S6F). Reintroducing succinate to SQLE knockout cells resulted in only a marginal restoration of cell growth and migration, with no observable impact on colony formation (Fig. S6G–I). These findings imply that the mitochondrial respiratory chain, which underpins the function of OXPHOS, does not fully mediate the oncogenic effects of SQLE in bladder cancer.

Subsequently, we found that overexpression of SQLE in BCa cells significantly increased the levels of intracellular mtROS and ROS. (Fig. S6J and Fig. 5A). Given that ROS, particularly mtROS, are primarily generated by mtOXPHOS [35], we aimed to investigate the role of increased mtROS in SQLE-driven BCa development. We treated UMUC3-SQLE and J82-SQLE cells with Mito-TEMPO, a mitochondria-targeted chemical with mtROS [36] scavenging properties, at a concentration of 40 µM. Mito-TEMPO treatment in UMUC3-SQLE and J82-SQLE cells significantly reduced the viability of bladder cancer cells, as demonstrated by cell growth curves (Fig. 5B), colony formation assays (Fig. 5C), wound healing assays (Fig. 5D), and transwell assays (Fig. 5E). Furthermore, Mito-TEMPO treatment of Sqle bladTg mice (Fig. 5F) significantly decreased BCa formation, as evidenced by histological examination (Fig. 5G). In conjunction with the reduced BCa development, Mito-TEMPO treatment of Sqle bladTg mice also resulted in significantly less tumor incidence and less disease severity compared to normal saline-treated Sqle bladTg mice (Fig. 5G). Additionally, Ki-67 scores of non-tumor bladder tissues from Mito-TEMPO-treated Sqle bladTg mice were significantly lower than those from normal saline-treated Sqle bladTg mice (Fig. 5H). We also observed lower protein expression of PCNA (Fig. 5I) in Mito-TEMPO-treated Sqle bladTg mice compared to controls. These results suggest that SQLE accelerates BCa development under the influence of mitochondrial OXPHOS-increased mtROS.

Fig. 5. mtROS clearance attenuates the oncogenic function of SQLE.

A Intracellular ROS levels as assessed by flow cytometry in UMUC3 and J82 cells with or without SQLE overexpression (n = 3, performed in triplicate). The impact of Mito-TEMPO (TEMPO) (40 µM) treatment on the viability of UMUC3 and J82 cells with SQLE overexpression, as determined by growth curves (B) (n = 3, performed in triplicate), colony formation assays (C) (n = 3, performed in triplicate), wound healing assays (D) (n = 3, performed in triplicate), and transwell assays (E) (n = 3, performed in triplicate). F A schematic diagram for in vivo Mito-TEMPO treatment. G Representative gross morphology and H&E staining of the bladder following in vivo Mito-TEMPO treatment (left). The effects of Mito-TEMPO on BCa formation are illustrated by tumor incidence and disease severity (right). (Rosa26-Sqle mice treated with NS controls (n = 5), Sqle bladTg micetreated with NS controls (n = 4), Sqle bladTg mice treated with Mito-TEMPO (n = 4)). H Ki-67 staining of the bladder from Rosa26-Sqle mice treated with NS controls (n = 5), Sqle bladTg micetreated with NS controls (n = 4), and Sqle bladTg mice treated with Mito-TEMPO (n = 4). I Western blots analysis of PCNA expression in the bladder from Rosa26-Sqle mice treated with NS controls, Sqle bladTg mice treated with NS controls, and Sqle bladTg mice treated with Mito-TEMPO. Data are represented as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Unpaired, two-tailed Student’s t-test (A, C, D, E, H, I). Difference in cell viability between the two groups was evaluated through ANOVA with repeated-measures analysis of variances (B). Scale bars, 500μm (low magnification) and 200 μm (higher magnification) (G). Scale bars, 200 μm (H).

SQLE protects TFAM protein from degradation by direct binding to LONP1 in mitochondria

To investigate the mechanisms underlying the effects of SQLE on mtOXPHOS, we performed a coimmunoprecipitation (co-IP) assay from extracts of UMUC3 and J82 cells and identified LONP1 as a potential interacting partner of SQLE (Fig. 6A). Proximity ligation assays (PLA) demonstrated a specific interaction between SQLE and LONP1, with increased binding capacity observed in SQLE-overexpressing cells compared to control cells (Fig. 6B). Immunofluorescence and MitoTracker staining indicated substantial colocalization of SQLE and LONP1 within the mitochondria (Fig. 6C). Fractionation assays and confocal imaging confirmed the localization of SQLE within mitochondria in BCa cells (Fig. S7A, B). Notably, SQLE does not affect LONP1 protein expression in BCa cell lines, subcutaneous tumor tissue and Sqle bladTg or Sqle bladKO mice (Fig. 6D, E and Fig. S7C–J), while SQLE overexpression led to increased TFAM protein expression, and SQLE knockout resulted in decreased TFAM levels (Fig. 6D, E and Fig. S7C–J). Furthermore, LONP1 protein levels showed no significant difference between human BCa tissues and adjacent normal tissues (Fig. S7K). LONP1, a specialized AAA + ATP-dependent protease, mediates the degradation of TFAM. Co-IP analysis further revealed that SQLE knockout enhanced the binding of LONP1 to TFAM (Fig. 6F). Therefore, we hypothesize that SQLE may bind with LONP1, thereby attenuating its ability to degrade TFAM. Next, chloroquine (CQ) treatment increased TFAM protein levels in UMUC3 and J82 control cells, but not in UMUC3-SQLE and J82-SQLE cells (Fig. S7L). Knockout of LONP1 significantly restored TFAM protein expression in SQLE-silenced cell lines (Fig. 6G), while knockout of LONP1 failed to restore TFAM protein expression in SQLE-overexpressing cell lines (Fig. S7M). Furthermore, treatment with the protein synthesis inhibitor cycloheximide revealed that SQLE-deficient cells exhibited a shorter TFAM half-life, while knockout of LONP1 significantly restored TFAM half-life in SQLE-silenced cell lines, indicating that SQLE may prevent the degradation of TFAM by binding with LONP1 (Fig. 6H). Collectively, these findings suggest that upregulated SQLE may compete with TFAM for binding to LONP1, thereby attenuating the LONP1-TFAM interaction and reducing TFAM proteolytic degradation.

Fig. 6. SQLE directly binds to LONP1 in mitochondria.

A Immunoprecipitation assays demonstrated the interaction between endogenous SQLE and LONP1 in UMUC3 and J82 BCa cells (n = 3, performed in triplicate). B The in situ interaction between SQLE and LONP1 as assessed through PLA (n = 3, performed in triplicate). C Immunofluorescence assays to determine the co-localization of SQLE and LONP1 in UMUC3 and J82 BCa cells (n = 3, performed in triplicate). D Representative Western blots analysis of LONP1 and TFAM expression in UMUC3 and J82 cells, both with and without SQLE overexpression, as well as in 647 V and UMUC3 cells both with and without SQLE knockout (n = 3, performed in triplicate). E Western blot analysis of LONP1 and TFAM expression in: (i) subcutaneous tumor tissue derived from J82 cells with or without SQLE overexpression; (ii) subcutaneous tumor tissue derived from UMUC3 cells with or without SQLE knockout; (iii) non-tumorous bladder tissues from Rosa26-Sqle mice versus Sqle bladTg mice; and (iiii) non-tumorous bladder tissues from Sqle^fl/fl mice versus Sqle bladKO mice. F Immunoprecipitation assays showed the impact of SQLE knockout for the interaction of LONP1 and TFAM in 647 V cell (n = 3, performed in triplicate). G The effects of LONP1 knockout in 647V-sgSQLE and UMUC3-sgSQLE cells on TFAM protein levels (n = 3, performed in triplicate). H Western blot analysis of TFAM in sgControl + shControl, sgSQLE + shControl and sgSQLE + shLONP1-transfected 647 V and UMUC3 cells treated with cycloheximide (CHX; 50ug/mL) for 0, 4, 6, and 8 h. Right: fold change in TFAM expression relative to 0 h (n = 3, performed in triplicate). I The effects of LONP1 knockout in 647V-sgSQLE and UMUC3-sgSQLE cells on respiratory capacity (n = 3, performed in triplicate). The effects of LONP1 knockout in 647V-sgSQLE and UMUC3-sgSQLE cells on viability, as determined by growth curves (J) (n = 3, performed in triplicate) and colony formation assays (K) (n = 3, performed in triplicate) (n = 3, performed in triplicate). Data are represented as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Unpaired, two-tailed Student’s t-test (B, G, I, K). Paired, two-tailed Student’s t-test (H). Difference in cell viability between the two groups was evaluated through ANOVA with repeated-measures analysis of variances (J). Scale bars, 10 μm (B, C).

Given that the knockout of LONP1 in SQLE-silenced cells resulted in the restoration of TFAM protein abundance, we examined whether the knockout of LONP1 in these cells could reinstate oxidative phosphorylation and bladder cancer cell activity. We found that the knockout of LONP1 effectively reversed the impact of SQLE knockout on respiratory capacity in both 647 V and UMUC3 cell lines (Fig. 6I). Furthermore, the knockout of LONP1 in 647V-sgSQLE and UMUC3-sgSQLE cells also restored cell viability and migration ability (Fig. 6J, K and Fig. S7N, O). These data suggest that LONP1-dependent degradation of TFAM may play a crucial role in mediating the effects of SQLE on oxidative phosphorylation and cell proliferation.

TFAM is required for SQLE-mediated OXPHOS and cell proliferation

As TFAM may play a pivotal role in mediating the effects of SQLE on oxidative phosphorylation and cellular proliferation, we analyzed TFAM protein expression in tumor tissues compared to adjacent normal tissues. We found that TFAM was significantly higher in expression in BCa compared to adjacent normal tissues, as determined by Western blotting and immunofluorescence (Fig. 7A, B). Also, by tissue microarray (TMA) analysis of 63 cases of BCa using immunohistochemistry (IHC) we found higher TFAM protein expression in BCa tissues compared to adjacent normal tissues (Fig. S8A). Consistent with this finding, Kaplan-Meier survival analysis indicated that elevated TFAM expression was associated with poor survival in BCa patients (Fig. S8B). Additionally, we evaluated the clinical relevance of SQLE and TFAM in human BCa through TMA analysis. We found that tumor samples with elevated SQLE expression also demonstrated a concomitant increase in TFAM expression, while samples with low SQLE expression showed a corresponding reduction in TFAM levels (Fig. 7C). Furthermore, subcutaneous tumors derived from UMUC3-sgSQLE cells exhibited decreased TFAM levels, as evidenced by immunofluorescence (Fig. 7D). This observation suggests a positive correlation between the protein expressions of SQLE and TFAM.

Fig. 7. TFAM mediates SQLE function in activating mitochondrial OXPHOS and cell proliferation.

A Western blot analysis of TFAM in paired tumor and adjacent normal tissues (n = 8, obtained from the Harbin Medical University Cancer Hospital). B Representative immunofluorescent staining illustrating the expression of TFAM protein in BCa tissues as compared to adjacent normal tissues. Staining for TFAM (Green), DAPI (Blue) (n = 5 from the Harbin Medical University Cancer Hospital). C Correlation analysis between SQLE and TFAM protein expression from TMA. D Representative immunofluorescent staining of TFAM protein in subcutaneous tumors derived from UMUC3 cells with or without SQLE knockout (n = 4). E Western blots analysis of TFAM expression in 647V-sgSQLE and UMUC3-sgSQLE cells following the overexpression of TFAM. F The effects of TFAM overexpression in 647V-sgSQLE and UMUC3-sgSQLE cells on respiratory capacity (n = 3, performed in triplicate). The impact of TFAM overexpression in 647V-sgSQLE and UMUC3-sgSQLE cells on cellular viability as assessed by growth curves (G) (n = 3, performed in triplicate), colony formation assays (H) (n = 3, performed in triplicate), wound healing assays (I) (n = 3, performed in triplicate), and transwell migration assays (J) (n = 3, performed in triplicate). Data are presented as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Paired, two-tailed Student’s t-test (A, B). Unpaired, two-tailed Student’s t-test (D, F, H, I, J). Difference in cell viability between the two groups was evaluated through ANOVA with repeated-measures analysis of variances (G). Scale bars, 20 μm (B, D). Scale bars, 500 μm (C).

To further investigate the functional implications of TFAM, we overexpressed TFAM in SQLE knockout cells (Fig. 7E). We found that TFAM overexpression effectively reversed maximum respiratory capacity, spare respiratory capacity and basal respiratory capacity in both 647V-sgSQLE and UMUC3-sgSQLE cells, thereby recapitulating the cell-autonomous effects of SQLE on mitochondrial respiration in bladder cancer cells (Fig. 7F). Additionally, TFAM overexpression restored cell viability in 647V-sgSQLE and UMUC3-sgSQLE cells, as demonstrated by cell growth curves and colony formation assays (Fig. 7G, H). Moreover, the overexpression of TFAM also restored the migratory ability of 647V-sgSQLE and UMUC3-sgSQLE cells (Fig. 7I, J). These findings suggest that SQLE promotes mitochondrial oxidative phosphorylation by modulating TFAM, thereby facilitating the proliferation of bladder cancer cells.

Building on prior reports that SQLE promotes tumor progression by increasing ROS through depletion of squalene (SQA), an antioxidant that lowers ROS and lipid peroxidation and thereby protects cells from oxidative injury [37, 38], we further investigated this process by supplementing with exogenous SQA. Our results showed that SQA significantly suppressed ROS and inhibited cell proliferation in both control and SQLE-overexpressing bladder cancer cells (Fig. S9A-C). Notably, ROS levels in SQA-treated SQLE-overexpressing cells failed to return to the baseline observed in control cells, indicating that these cells retain SQA-independent ROS-generating mechanisms (Fig. S9A). Importantly, SQA treatment did not alter LONP1 or TFAM protein expression (Fig. S9D), indicating that its effects are primarily attributable to direct antioxidant activity rather than modulation of the LONP1–TFAM axis. Therefore, our findings not only confirm that SQLE drives tumorigenesis through squalene depletion-induced ROS production, but also uncover a novel parallel oncogenic pathway mediated by the SQLE–LONP1–TFAM–mtROS axis.

Pharmacological inhibition of SQLE suppresses BCa growth

Given the oncogenic role of SQLE in BCa, we investigated the efficacy of SQLE-targeted therapeutic strategies in bladder tissue. In the initial stages of chemically-induced BCa formation in mice, tamoxifen was used to selectively activate Cre in the bladder to induce Sqle knockout in the bladder tissue of Sqle bladKO mice, simulating SQLE inhibition during the pre-carcinogenic phase (Fig. 8A). This genetically-targeted disruption notably reduced the incidence of BCa and decreased disease severity compared to littermate Sqle^fl/fl control mice (Fig. 8B). Early inhibition of SQLE during the initiation of BCa also resulted in lower Ki-67 expression levels (Fig. 8C).

Fig. 8. The SQLE inhibitor terbinafine suppresses BCa growth in vitro and in vivo.

A A schematic diagram for the knockout of Sqle in bladder tissue during the early stages of chemically induced bladder cancer formation. B Representative gross morphology and H&E staining (upper) of bladder. The impact of SQLE blockade in bladder tissue on bladder cancer formation during the pre-carcinogenic phase, as assessed by tumor incidence and disease severity (below). (Sqle^fl/fl mice (n = 11), Sqle bladKO mice (n = 9)). C Ki-67 staining of bladder tissue from 0.05% BBN-treated Sqle^fl/fl mice (n = 11) versus Sqle bladKO mice (n = 9). D The effects of terbinafine (60 µM) treatment on intracellular cholesterol levels in J82, UMUC3, RT112, and TCCSUP cell lines (n = 3, performed in triplicate). The effects of terbinafine treatment on the viability of J82, UMUC3, RT112, and TCCSUP cell lines, as determined by growth curves (E) (n = 3, performed in triplicate) and colony formation assays (F) (n = 3, performed in triplicate). The effects of oral terbinafine treatment (80 mg/kg per day) on subcutaneous tumor derived from RT112 cells, as measured by tumor volume (G, H) and weight (I). (J) Western blot analysis of LONP1 and TFAM expression in subcutaneous tumor tissues from nude mice treated with PBS or terbinafine. K A schematic diagram for terbinafine treatment in the in situ model. L Representative gross morphology and H&E staining of bladder (uper). The effects of terbinafine on bladder cancer formation as assessed by tumor incidence and disease severity (bottom). (WT mice treated with PBS (n = 13), WT mice treated with terbinafine mice (n = 14)). M Ki-67 staining of bladder tissue from WT mice treated with PBS (n = 13) and WT treated with terbinafine mice (n = 14). N Western blot analysis of PCNA, LONP1 and TFAM protein expression in bladder tissue from WT mice treated with PBS versus WT mice treated with terbinafine. Data are represented as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Paired, two-tailed Student’s t-test H. Unpaired, two-tailed Student’s t-test (C, D, E, F, I, M). Scale bars, 500 μm (low magnification) and 200 μm (higher magnification) (B, L). Scale bars, 200 μm (C, M).

Overexpression of SQLE led to increased cellular cholesterol levels, while SQLE knockout resulted in decreased cholesterol content (Fig. S10A, B). Administration of NB598 or terbinafine, both SQLE inhibitors, caused a reduction in total cholesterol levels in BCa cells (Fig. S10C and Fig. 8D). NB598 significantly decreased the viability of BCa cell lines and inhibited cell migration (Fig. S10D–F). Terbinafine treatment effectively suppressed respiratory capacity in 647 V and J82 cells (Fig. S10G), and reduced respiratory capacity and ROS levels in J82-SQLE cells (Fig. S10H, I). It also significantly impaired proliferation and migration in UMUC3, J82, RT112, and tccsup cell lines (Fig. 8E, F and Fig. S10J).

Subsequently, we examined the effectiveness of terbinafine in vivo. Terbinafine significantly inhibited the growth of subcutaneous RT112 xenografts (Fig. 8G–I) and decreased TFAM protein expression in subcutaneous tumor tissue (Fig. 8J and Fig. S10K). Additionally, terbinafine suppressed the growth of orthotopic BCa models (Fig. 8K, L) and reduced both Ki-67 scores, as well as PCNA and TFAM expression in bladder tissue (Fig. 8M, N and Fig. S10L). Collectively, these findings indicate that pharmacological inhibition of SQLE represents a promising approach for the prevention and treatment of BCa.

Discussion

In this study, we identified SQLE as an oncogenic factor with high expression in cases of BCa. Specifically, SQLE promoted BCa growth by modulating mitochondrial OXPHOS, and the elimination of SQLE-induced mtROS effectively suppressed BCa progression both in vitro and in vivo. We found that SQLE localized to mitochondria and protected TFAM from LONP1-mediated degradation by directly binding LONP1, thereby influencing the degree of OXPHOS that contributed to BCa progression. Treatment with terbinafine conferred therapeutic benefits in BCa in cell culture and animal models, further validating SQLE as a promising therapeutic target for the treatment of BCa.

Previous studies reported that SQLE mRNA levels are elevated in BCa relative to adjacent normal tissues, suggesting an association with poor prognosis, as evidenced by public databases [12, 13]. Our study reinforced this conclusion by incorporating an analysis of the GSE3167 database [29]. Moreover, we demonstrated frequent overexpression of the SQLE protein in BCa tissues via Western blotting and tissue immunofluorescence analyses. Tissue microarray analysis further validated the overexpression of SQLE in BCa and its correlation with poor prognostic outcomes. Additionally, single-cell RNA-sequencing of Dataset CNP0000460 [20] revealed heightened mRNA levels of SQLE in epithelial cells of BCa tumors, thereby supporting its role as both an oncogenic driver and a prognostic factor in BCa. Our work provides definitive validation of its independent prognostic value through multi-institutional cohorts and single-cell transcriptomics, establishing its clinical relevance with unprecedented robustness.

Our research unequivocally demonstrated that SQLE functions as a tumor cell-intrinsic oncogenic factor in BCa. Prior studies have suggested that the silencing of SQLE expression in BCa cells can suppress cell proliferation, migration and the G1-S phase transition [13, 39]. In this study, we investigated the oncogenic function of SQLE through ectopic expression, knockout, and rescue experiments in various BCa cell lines. Furthermore, we substantiated the oncogenic potential of SQLE in BCa using xenograft models and Sqle transgenic mice with bladder-specific Sqle overexpression or knockout. Sqle overexpression in vivo fully recapitulated the signaling cascades induced by ectopic SQLE expression in vitro. Moreover, Sqle transgenic expression in mice accelerated BCa development and disease severity in an experimental model of BBN-induced BCa, concomitant with increased cell growth, while Sqle knockout exhibited the opposite effect. Our findings provide compelling evidence supporting the notion that SQLE functions as an oncogene during the tumorigenesis of BCa.

Our findings support the concept that mitochondrial OXPHOS is critical for tumor cell proliferation [40, 41]. In this study, we conducted a comparative proteomics analysis and identified that silencing SQLE significantly inhibited mitochondrial OXPHOS, as evidenced by a reduced OCR and down-regulated protein expression of mitochondrial-related genes. Restoration of OCR through the knockout of LONP1 or the re-expression of TFAM in SQLE knockout cell lines markedly restored cell proliferation. This suggests that activated mitochondrial OXPHOS is a key mediator of the oncogenic function of SQLE in BCa.

ROS, particularly mtROS generated by OXPHOS, exhibit oncogenic properties [35, 42]. In support of this notion, we observed that the inhibition of mtROS using Mito-TEMPO [43] negated the growth-promoting effects of SQLE in BCa cells, as well as the tumorigenic role of SQLE in bladder-specific Sqle transgenic mice in vivo. Our findings established SQLE as a critical mediator of mtROS production, and targeting the SQLE-mtROS axis represents a precision therapeutic strategy particularly suitable for patients with SQLE-high tumor subtypes, highlighting its potential for personalized medicine.

Previous studies have primarily focused on SQLE located in the endoplasmic reticulum membrane and cell membrane [6, 28]. Our study revealed a novel mitochondrial localization and function of SQLE in BCa cells, as demonstrated by fractionation and confocal imaging. We identified a previously unreported mechanism whereby mitochondrial SQLE competitively binds to the inner mitochondrial protein LONP1 [44], disrupting LONP1-TFAM interaction and leading to TFAM stabilization. This SQLE-LONP1-TFAM axis activates mtOXPHOS and promotes tumorigenesis. These findings redefine the role of SQLE beyond its canonical function in cholesterol synthesis, highlighting its critical involvement in mitochondrial metabolic reprogramming. In parallel, our squalene add-back experiments refined prior models that SQLE drives tumor progression by depleting the antioxidant SQA to elevate ROS [37, 38]. Exogenous SQA add-back suppressed ROS and proliferation in both control and SQLE-overexpressing BCa cells. However, ROS in SQLE-overexpressing cells did not return to control baseline, indicating SQA-independent ROS drivers. Moreover, SQA did not alter LONP1 or TFAM levels, supporting a direct antioxidant mechanism rather than involvement in the LONP1–TFAM axis. Collectively, these data support parallel oncogenic pathways: substrate depletion–driven ROS and SQLE-mediated disruption of the LONP1–TFAM axis that elevates mtROS.

Patients with BCa experience poor survival rates due to a lack of effective treatment alternatives [45, 46]. Consequently, the identification of supplementary treatment strategies for BCa is both urgent and of paramount importance. Our studies employing bladder-specific Sqle transgenic mouse models demonstrated that targeted SQLE inhibition during early carcinogenesis effectively suppressed bladder tumor progression. Notably, the clinically approved SQLE inhibitor terbinafine also exhibited potent anti-tumor activity. We found that terbinafine effectively suppressed the viability of BCa cells in vitro. Utilizing xenograft and orthotopic tumor models, we demonstrated that terbinafine was both efficacious and safe in inhibiting BCa growth in vivo. Terbinafine is a US FDA-approved drug that possesses an excellent safety profile [47]. Previous studies have indicated that terbinafine suppresses tumor development by inhibiting cholesterol biosynthesis and decreasing intracellular ROS in animal and cell models [9]. Our findings indicated that pharmacologic inhibition of SQLE with terbinafine reduced TFAM expression, suppressed mitochondrial oxidative phosphorylation, and lowered ROS levels, thereby restraining cell proliferation and exerting antitumor activity in bladder cancer. While prior studies have examined anti-cancer effects of terbinafine in other contexts, our research uniquely validated its therapeutic potential in BCa, offering a novel therapeutic strategy for clinical translation. However, clinical testing in humans will be necessary to validate the therapeutic benefits of terbinafine for the treatment of BCa. Furthermore, the detailed mechanisms and implications of terbinafine-induced mitochondrial dysfunction will need to be elucidated in future studies.

In conclusion, our research elucidates the oncogenic role of SQLE in BCa. The findings suggest that SQLE may localize to mitochondria and stabilize TFAM by regulating its degradation through interaction with LONP1, thereby influencing OXPHOS and promoting BCa progression. Furthermore, the application of Mito-TEMPO for mtROS clearance significantly mitigates SQLE-induced cell proliferation and tumor formation both in vitro and in vivo. Moreover, the targeted inhibition of SQLE using terbinafine considerably suppresses tumor progression, highlighting its potential for future clinical application in the treatment of BCa.

Limitations of the study

While our study provides compelling evidence for the role of SQLE in BCa, several limitations merit consideration. Firstly, the reliance on mouse models may not fully replicate human tumor heterogeneity or microenvironmental interactions. Secondly, the precise contribution of cholesterol accumulation versus OXPHOS dysregulation to SQLE-driven tumorigenesis remains to be elucidated. Although SQLE knockout reduced cellular cholesterol levels, the incomplete rescue of OXPHOS defects by succinate supplementation suggests that additional mechanisms may underlie its oncogenic effects. Future investigations should explore the crosstalk between SQLE, lipid metabolism and other mitochondrial pathways, including fatty acid oxidation and apoptosis regulation. Thirdly, clinical trials assessing the efficacy and safety of terbinafine in patients with BCa, particularly those with SQLE-high tumors, are essential for validation. Furthermore, while we obtained reliable results through multiple experiments confirming that SQLE can localize to mitochondria, the specific role of mitochondrial SQLE in BCa requires further investigation in future studies.

Author contributions

Conceptualization: XY, XH, HM, DL, SL; Methodology: YD, XJ, XY, JZ, QF, YZ, XY; Investigation: YD, YJ, XY, JZ, QF, YZ; Visualization: YD, YJ, XY; Funding acquisition: HM, DL, SL; Project administration: YF, YH, JY, JX, YY, YW, CG, QZ, AW, XL, XZ, SC, TA, LJ, XH, JH; Supervision: XH, HM, DL, SL; Writing – original draft: YD, YJ, JZ; Writing – review & editing: DL, SL.

Funding

This study was supported by the National Natural Science Foundation of China (82273225, 82273244); Scientific Research Foundation for Returned Scholars, Ministry of Human Resources and Social Security of China (6101020101); HMU Marshal Initiative Funding (HMUMIF-21002); Science Fund Program for Young Talents of the Fourth Hospital of Harbin Medical University (HYDSYRCYJ01); Nn10 program of Harbin Medical University Cancer Hospital (Nn10 2024-05); Heilongjiang Province Innovation Base Award Project (JD2023SJ03); Heilongjiang Chunyan Innovation Team Program (2022CYCX0179); Priority Areas of Heilongjiang Provincial Key Research and Development Program, (2024ZX12C05).

Data availability

Bulk RNA-seq datasets analyzed in this study were obtained from TCGA, GSE316, and GSE188715. The raw single-cell RNA-seq dataset was retrieved from the China National GeneBank (CNGB, https://db.cngb.org/cnsa/) under the accession code CNP0000460. All other data supporting the findings of this study are available from the corresponding author upon reasonable request.

Code availability

All bioinformatic tools and software utilized for data analysis in this study were executed in accordance with their respective manuals and protocols. No custom code was utilized.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41388-025-03626-3.