Tumor-derived apoptotic extracellular vesicles impede hepatocarcinoma progression by disrupting mitochondrial metabolism

Abstract

Background

Apoptosis undergoes dynamic changes during tumor progression and treatment and has complex effects on tumor development. Apoptotic cell-derived extracellular vesicles (Apo-EVs) have recently been recognized as key mediators of intercellular communication, but their role in hepatocellular carcinoma (HCC) remains unclear.

Methods

We performed a series of in vitro and in vivo experiments to investigate the effects of tumor-derived Apo-EVs on HCC cell proliferation, apoptosis, and metabolic remodeling. Additional mechanistic studies were conducted to identify the functional cargo molecules carried by Apo-EVs and to determine their roles in ATP synthesis and metabolic regulation.

Results

Tumor-derived Apo-EVs suppressed HCC cell proliferation and increased apoptosis. Mechanistically, Apo-EVs promoted apoptotic cell death by shifting cellular metabolism from aerobic glycolysis to oxidative phosphorylation. Notably, transmembrane protein 70 (TMEM70), which is transported by Apo-EVs, was identified as a key mediator of this process. TMEM70 participated in ATP synthesis and contributed to metabolic reprogramming in HCC cells.

Conclusions

Our study reveals a potential mechanism by which tumor-derived Apo-EVs promote apoptosis and metabolic remodeling in HCC. Apo-EV-associated TMEM70 may contribute to this process through its involvement in mitochondrial metabolism. These findings suggest that Apo-EVs and selected cargo molecules may represent promising therapeutic targets for HCC.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12967-026-08581-5.

Highlights

Engineered apoptotic vesicles inhibit hepatocarcinoma proliferation by disrupting mitochondrial function.

Tumor growth is suppressed while energy metabolism is altered.

Incresed Tmem70 expression promotes oxidative phosphorylation over glycolysis.

A metabolic reprogramming strategy enables potent suppression of hepatocarcinoma.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12967-026-08581-5.

Introduction

Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related death worldwide and ranks among the most common cancers in many countries [1]. HCC development is a multistep process driven by the accumulation of genetic and epigenetic alterations, including he inactivation of tumor suppressor gene and the activation of oncogenes, which disrupt fundamental cellular processes. Currently, curative therapies are suitable only for a small subset of patients diagnosed at an early stage, and fewer than 30% benefit from surgical resection, whereas local ablation and chemoembolization offer limited efficacy in delaying disease progression. Given the frequent delays in diagnosis and the scarcity of effective treatments, elucidating the signaling networks that govern the proliferation and death of liver cancer cells is essential for identifying the intrinsic mechanisms driving HCC progression [2].

Apoptosis is a closely regulated and evolutionarily conserved cell death program with essential functions in normal physiological processes, including embryogenesis and adult tissue homeostasis [3]. Furthermore, the implications of apoptosis on tumor growth are significant. Recent studies have indicated that cancer cells are characterized by a short lifespan and high rate of proliferation rates, in addition to high rate of apoptosis [4]. Notably, the effects of apoptosis on cancer cells are complex. On the one hand, apoptosis slows the rate of cancer cell proliferation [5]. On the other hand, various signaling molecules are released during apoptosis that modify the tumor microenvironment (TME) and influence tumor growth [6].

During apoptosis, cells disassemble and release vesicles of varying sizes. Among these, apoptotic cell-derived extracellular vesicles (Apo-EVs), which are regulated by pannexin channel activity and driven by phosphatidylserine, have diameters ranging from 0.05 to 1 μm [7]. Apo-EVs function as critical intercellular messengers by carrying multiple potentially biologically active components and participate in differentiation, damage repair, and regeneration [8–10]. Additionally, these vesicles are involved in pathophysiological processes such as inflammatory responses [11]. Apo-EVs not only directly promote tumor cell proliferation or increase tumor resistance but also indirectly orchestrate various systemic pathological processes, including coagulation, drug resistance and immune tolerance [12–14]. Apo‑EVs produced by oroxylin A‑treated HCC cells are enriched in glycolytic proteins and act on macrophages to regulate their metabolic state, thereby inhibiting tumor development [15]. Similarly, Apo‑EVs derived from chemically induced hepatocyte‑like cells (ciHeps) can induce an anti‑inflammatory phenotype in macrophages, inhibit hepatic stellate cell activation, and promote hepatocyte survival, thus slowing the progression from liver fibrosis to HCC [16]. However, although these studies have highlighted the important role of Apo‑EVs in regulating immune cells and the liver microenvironment, direct experimental evidence demonstrating their regulatory effects on HCC cells themselves remains limited.

As energy-generating organelles, mitochondria are the primary source of energy for various metabolic activities, including cell proliferation and apoptosis; therefore, mitochondria are closely linked to tumor cell growth and development [17]. Genetic disruption and the stalling of mitochondrial dynamics exacerbate liver damage and promote tumorigenesis [18], whereas loss of cyclic guanylate-adenylate synthetase (cGAS) on the outer mitochondrial membrane leads to increased oxygen accumulation and cell death, thereby inhibiting tumor growth [19]. Oxidative phosphorylation (OXPHOS) is among the main pathways by which mitochondria produce energy [20]. As energy-generating organelles, mitochondria serve as the primary source of energy for various metabolic activities, including cell proliferation and apoptosis. When energy is generated, the electron transport chain (ETC) complexes (complexes I-IV) located on the inner mitochondrial membrane facilitate the conversion of chemical energy stored in nutrients into an electrochemical gradient. Subsequently, ATP synthase, designated complex Ⅴ, subsequently catalyzes the generation of ATP from ADP [21]. Metabolism in tumor cells differs from that in normal cells. Even when there is sufficient oxygen to support OXPHOS, tumor cells use glycolysis, a phenomenon referred to as the Warburg effect [22]. While this effect generates less metabolic energy overall, it provides proliferative advantages to tumor cells [23]. In cells with compromised ETC supercomplexes, mitochondrial dysfunction drives a metabolic shift from OXPHOS to glycolysis and exacerbates tumor growth in xenograft models [24, 25]. Consequently, impairments in the ETC can result in the dysregulation of cellular proliferation and metabolism through various pathways.

TMEM70 is an inner mitochondrial membrane protein that plays a crucial role in shaping the biogenesis of mitochondrial ATP synthase and is essential for OXPHOS [26, 27]. In TMEM70 loss-of-function mutants, mitochondrial ATP synthase activity is impaired, triggering a metabolic shift from OXPHOS to glycolysis [28]. These changes can result in severe mitochondrial-related diseases and perturbations of apoptosis and other metabolic pathways [29]. In addition, several studies have reported an association between TMEM70 and cancer progression [30, 31].

In this study, we observed that the apoptosis of cancer cells in liver cancer tissues was accompanied by a significant increase in Apo-EV secretion. These Apo-EVs inhibited the proliferation of peripheral HCC cells and demonstrated pro-apoptotic properties. This phenomenon is attributed to the presence of the mitochondrial-associated protein TMEM70 within Apo-EVs, which plays a crucial role in regulating mitochondrial OXPHOS and glycolysis. These findings offer novel insights into the metabolic and proliferative effects of apoptotic vesicles in the context of cancer, highlighting their potential implications for the development of new therapeutic strategies for liver cancer patients.

Materials and methods

Cell culture

Hepa1-6 liver cancer cells were obtained from FuHeng BioLogy (Shanghai, China; Cat. No. FH0862). The cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Meilunbio, Cat. No. MA0212-2) supplemented with 10% fetal bovine serum (FBS, Diagnovum, Cat. No. D016) and 1% penicillin–streptomycin (Biolab-Tech, Cat. No. 10,378,016). The cultures were maintained at 37 °C in a humidified atmosphere of 5% CO₂ and passaged at approximately 80% confluence.

To fluorescently label the cells, they were seeded in 6-well Transwell plates at a density of 2 × 10⁵ cells/well. Upon reaching confluence, the cells were transfected using adenovirus pcSLenti-ET1-EGFP-P2A-Puro-CMV-MCS-3×FLAG-WPRE (GenePharma, China) and Lipofectamine™ 3000 transfection reagent (Thermo Fisher Scientific, Cat. No. L3000015) in accordance with the manufacturer’s instructions. Transfection efficiency was subsequently assessed through reverse transcription quantitative polymerase chain reaction (RT-qPCR).

To induce apoptosis, Hepa1-6 cells were washed with phosphate-buffered saline (PBS) and then cultured in basal medium supplemented with 500 nM staurosporine (STS; TargetMol; Cat. No. T2532) for 16 h.

Mice and tumor models

Male C57BL/6 mice aged 6–8 weeks were obtained from the Hangzhou Institute of Medical Sciences, Chinese Academy of Sciences (Zhejiang, China). All the mice were housed under specific pathogen-free (SPF) conditions. This study was approved by the Animal Ethics Committee of the Hangzhou Institute of Medical Sciences, Chinese Academy of Sciences, Zhejiang, China. All the experimental procedures complied with the relevant institutional and national guidelines and regulations.

After 7‑week‑old C57BL/6 mice were anesthetized, their abdomens were shaved. The abdominal skin and muscles were then incised layer by layer to expose a portion of the right liver lobe. A suspension of 2 × 10⁶ EGFP‑labeled cells in 20 μL of PBS mixed with matrix glue (1:1 ratio) was injected onto the surface of the liver lobe. Seven days later, tumor growth was monitored using an in vivo imaging system (IVIS Lumina III; PerkinElmer, USA) with a 480 nm laser. During the second week after tumor implantation, we randomly selected mice according to the experimental requirements and injected them via the tail vein with 20 μL of Apo‑EVs (10⁹ particles), 20 μg/mL TMEM70, or control treatments (empty vesicles or PBS) via the tail vein. During the third week, the mice were euthanized, and the livers were harvested for histological analysis.

Analysis of hepatocarcinoma-specific Apo-EVs

In accordance with the methodology described by Brandel and colleagues, Apo-EVs were analyzed indirectly via flow cytometry [8]. Nonhepatic extracellular vesicles (EVs) were identified through the expression of specific surface markers, including CD45 (leukocytes), CD61 (platelets), CD31 (endothelial cells and platelets), CD4 (T lymphocytes), and CD144 (endothelial cells). Apo-EVs were defined as large EVs that were positively stained for both Calcein Violet™ 450 and annexin V.

The specific procedures were as follows. First, blood was collected from the orbital sinus into CTAD tubes and centrifuged at 450×g for 10 min at 4 °C. Aliquots (20 µL) of the supernatant were stained with 80 µL of antibody cocktail for 90 min at 4 °C in the dark. The antibody panel included the following conjugates: anti-CD45 (PE-CY7; Cat. No. 157,203), anti-CD61 (PE-CY; Cat. No. 104,317), anti-CD31 (PE; Cat. No. 157,203), anti-CD44 (Per CP/Cy5.5; Cat. No. 103,031), anti-CD144 (APC; Cat. No. 138,011), Calcein Violet 450 (Thermo Fisher Cat. No. 65–0854-39), and anti-annexin V (FITC; Cat. No. 640,905). Flow cytometry was performed using a BD LSR Fortessa™ flow cytometer analyzer (BD Biosciences, USA).

Isolation and characterization of Apo-EVs

Following previously reported experimental methods [32, 33], apoptosis was induced in Hepa 1‑6 cells with STS, and Apo‑EVs were subsequently isolated and purified through sequential centrifugation and filtration. Briefly, confluent Hepa 1‑6 cells were washed with PBS and incubated in serum‑free medium supplemented with 500 nM STS. Sixteen hours later, the supernatant was centrifuged at 800 × g for 10 min and then at 2,000 × g for 15 min to remove cell debris. Finally, the Apo‑EVs were pelleted by centrifugation at 16,000 × g for 45 min.

The morphology of the Apo-EVs was examined by transmission electron microscopy (TEM). Apo-EVs were fixed in 1% glutaraldehyde for 30 min and applied to formvar-coated copper grids, allowing 20 min for adsorption. After the samples were washed with distilled water, the grids were negatively stained with 2.5% uranyl acetate for 2 min. Imaging was performed using a JEM-1200EX transmission electron microscope (JEOL, Japan).

To analyze the size and concentration of Apo-EVs, we prepared experimental samples according to the requirements of nanoparticle tracking analysis (NTA). Apo-EVs were diluted in filtered PBS and analyzed with a ZetaView PMX120™ (Particle Metrix, Germany).

Apo-EVs were diluted in filtered PBS and stained with Annexin V–FITC in accordance with the manufacturer’s protocol for nanoflow cytometry. Sample analysis was performed using a Flow NanoAnalyzer™ U30 instrument (NanoFCM, China). The particle size, concentration, and purity were determined directly using the instrument.

To quantify the relative fluorescence intensity, we stained Apo-EVs with primary antibodies against Lamin B1, cleaved caspase 3, caspase 3, TSG101, CD63, CD81, and GAPDH. This was followed by incubation with appropriate secondary antibodies.

Label-free quantification and analysis

The Label-free Quantification (LFQ) experimental workflow begins with protein extraction using RIPA buffer and sonication, followed by centrifugation to isolate the supernatant. The extracted proteins are then quantified using a BCA assay. Next, 100 µg of protein per sample is precipitated overnight with cold acetone, centrifuged, and washed to obtain a protein pellet. This pellet is redissolved, reduced using DTT at 55 °C, and subsequently alkylated with IAA in the dark. The proteins are then subjected to enzymatic digestion overnight at 37 °C using Trypsin at a 1:50 enzyme-to-protein ratio. Following digestion, sodium deoxycholate (SDC) is removed from the mixture by precipitation with TFA and high-speed centrifugation. The resulting peptides undergo desalting through C18 spin columns and are vacuum-dried. Finally, the prepared peptide samples are separated by a nano-UPLC system and analyzed via a Q-Exactive HFX mass spectrometer utilizing a data-dependent acquisition (DDA) mode.

Data analysis was performed by Shanghai NewCore Biotechnology Co., Ltd. Briefly, raw data were searched against the Mus musculus UniProt database (uniprot-Mus+musculus-10090–2022-8.fasta) using the Sequest HT search engine integrated into Proteome Discoverer software (v2.4.0.305, Thermo Fisher Scientific). Proteins were strictly filtered to retain those identified by at least one unique peptide. Subsequently, differential protein expression analysis and a comprehensive bioinformatics workflow were performed, including volcano plot generation, hierarchical clustering, and GO and KEGG enrichment analyses.

Statistical analysis

All the statistical analyses were performed in GraphPad Prism version 9.0 (GraphPad Software, USA). The data are expressed as the mean ± standard deviation (SD). Comparisons between groups were made using an unpaired Student’s t test or one-way analysis of variance (ANOVA), as appropriate. A p value of less than 0.05 was considered to indicate statistical significance.

Additional methods are available in the Supplementary Materials.

Results

Increased apoptosis of HCC cells is accompanied by the release of Apo-EVs from the liver

Cancer cell metabolism is known to be altered in tumor tissues. To verify the increased frequency of apoptosis observed in HCC cells compared with normal cells, we established a mouse model of HCC by injecting Hepa1-6 hepatoma cells. At the three-week endpoint, in vivo imaging revealed that the fluorescence signal from the HCC cells was localized exclusively to the liver (Fig. S1). Subsequent necropsy revealed negligible ascites accumulation and peritoneal adhesion across all groups. Macroscopic liver tumors were observed in approximately two-thirds of the tumor-bearing mice; however, no detectable metastases were found in the bone or lung upon examination of other organs. Furthermore, the results of the TUNEL assay indicated increased apoptosis in the HCC tissue samples compared with the adjacent normal tissue samples (Fig. 1A). TEM revealed that the apoptotic cells subsequently exhibited vesicle separation (Fig. 1B). To determine the characteristics and quantity of Apo-EVs, we used flow cytometry to evaluate circulating blood, and the results demonstrated that the number of apoptotic nonblood cell-derived EVs was increased in the circulating blood of HCC mice (Fig. 1C).

Fig. 1.

Increased HCC cell apoptosis is accompanied by the release of Apo-EVs from the liver. A HCC model was established by orthotopic injection of Hepa 1–6 cells. Liver samples were collected 3 weeks later. Representative images of DAPI-positive (blue) and TUNEL-positive (green) apoptotic hepatocytes among normal or HCC cells (bar = 100 μm). B Apo-EVs (white arrows) were released from apoptotic HCC cells and characterized via TEM (bar = 1 μm). C. Apo-EVs from peripheral blood were identified through flow cytometry as CD45^-/CD44^-/CD31^-/CD144^-/annexin V⁺/calcein⁺. **, p < 0.01 (n = 6–8)

Previous experiments have demonstrated that apoptotic epithelial cell-derived EVs [32], when administered via the tail vein, accumulate mainly in the liver, skin, spleen and lung tissues. To determine the biodistribution and binding site of Apo-EVs, radioisotope-labeled Apo-EVs were intravenously injected via the tail vein. Immunofluorescence staining revealed that Apo-EVs were primarily engrafted in the liver (Fig. S2).

Identification of apoptotic HCC cell-derived EVs

To verify the characteristics of Apo-EVs released by apoptotic HCC cells, we used STS to induce apoptosis in culture-expanded HCC cells and then isolated the Apo-EVs using a sequential centrifugation system. After 16 h of STS induction, the Hepa 1–6 cells exhibited significant morphological changes and apoptotic responses, as assessed by flow cytometry analysis with annexin V and 7-AAD staining (Fig. 2A–B). The morphology of the Apo-EVs was determined by transmission electron microscopy (TEM) (Fig. 2C). Nanoparticle tracking analysis (NTA) revealed that the particle size of Apo-EVs was predominantly distributed between 100 and 1000 nm, with a particle count of approximately 10¹¹ (Fig. S3). Nanoflow cytometry revealed that Apo-EVs expressed Annexin V (Fig. 2D), and Western blotting (WB) was conducted to confirm their purity Analysis of EVs generated from apoptotic HCC cells revealed that Apo-EVs did not express the exosome-specific markers Lamin B1 and CD63 but did expressed the common extracellular vesicle marker TSG101 and the apoptosis-associated marker cleaved caspase-3 (Fig. 2E).

Fig. 2.

Identification of apoptotic HCC cell-derived EVs, A Morphology of normal and STS-treated Hepa1-6 cells. B Validation of apoptosis through flow cytometry. C The morphology of Apo-EVs was examined through TEM. D. Nanoflow cytometry staining revealed that Apo-EVs expressed the apoptotic body-specific surface markers ptdser (indicated by Annexin V staining) and Lamin B2. E. The protein expression levels of Lamin B1, cleaved caspase 3, TSG101, caspase 3, CD63, CD81 and GAPDH were detected by Western blotting. (n = 3)

Apo-EVs specifically inhibit the proliferation of hepatoma cells and promote their apoptosis

On the basis of the observations that Apo-EVs are secreted during HCC progression and preferentially accumulate in the liver, we hypothesized that they may regulate HCC growth and metabolism. To confirm this hypothesis, we cocultured Hepa1-6 cells with different concentrations of Apo-EVs (0, 20, and 40 μg/mL) and monitored changes in cell proliferation and apoptosis. Notably, coculture of Hepa1-6 cells with Apo-EVs for 12 or 24 hours resulted in the significant inhibition of cell proliferation and survival, as demonstrated by the results of the CCK-8, EdU and CFSE assays (Fig. 3A–B&S4). The results of the scratch tests and colony formation assays demonstrated that HCC cell colony formation and migration were inhibited in response to Apo-EVs (Fig. 3C–D).

Fig. 3.

Apo-EVs specifically inhibit the proliferation of hepatoma cells and promote their apoptosis, A Cell viability was determined by a cell counting kit-8 assay, B EdU and C. Colony formation assays were conducted to assess cell proliferation, D. Wound healing assays were used to evaluate cell migration ability. E. The mRNA expression levels of cyclin A2, cyclin B1, cyclin D, cyclin E, p16 INK4a, and p27 Kip1 were measured using RT‒qPCR. F. Western blotting analysis and quantification of cyclin A2, cyclin B1, cyclin D, cyclin E, and p27 Kip1 expression. *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001 (n = 3)

The expression of cyclin proteins is regulated by Apo-EVs. The effects of Apo-EVs on the expression level of cyclins at different times and concentrations were evaluated by RT‒qPCR and WB. The mRNA and protein expression of Cyc A, Cyc B, Cyc D, and Cyc E decreased in a concentration-dependent manner, while that of p27 increased in comparison with that in the control group (Fig. 3E–F). In summary, these data indicate that Apo-EVs impair cell proliferation, which is related to the regulation of cyclin proteins and the promotion of specific cyclin inhibitors.

To determine whether the inhibitory effects of Apo-EVs are conserved in a human HCC model, we treated Huh7 cells with Apo-EVs under the same experimental conditions. Consistent with the results observed in Hepa1-6 cells, Apo-EV treatment significantly reduced Huh7 cell viability and proliferation and increased apoptosis (Fig. S5A-C). To further evaluate the involvement of TMEM70 in this response, TMEM70 was knocked down in Huh7 cells prior to Apo-EV treatment. TMEM70 knockdown partially reversed the Apo-EVs-induced decrease in cell viability and attenuated Apo-EVs-induced apoptosis. Together, these findings indicate that the anti-proliferative and pro-apoptotic effects of Apo-EVs are conserved in a human HCC cell model and further support the contribution of TMEM70 to Apo-EV-mediated responses in HCC cells (Fig. S5D-F).

To exclude the potential influence of Apo-EVs on the proliferation and apoptosis of normal liver cells, we cocultured Apo-EVs with primary mouse hepatocytes (PMHs) under comparable conditions. As demonstrated by CCK-8 and Annexin V/FITC double staining assays (Fig. S6A-B), we found that the activity, apoptosis and invasive capacity of the PMHs remained unchanged. These results demonstrate that Apo-EVs did not significantly affect the proliferation or apoptosis of normal liver cells. Conversely, Apo-EVs specifically inhibited the proliferation and promoted the apoptosis of liver cancer cells.

TMEM70 is enriched in apo-EVs and inhibits HCC proliferation via a mitochondrial metabolism-dependent mechanism

Treatment with Apo-EVs inhibited the proliferation of HCC cells and promoted their apoptosis. These findings prompted us to investigate whether Apo-EVs are involved in the physiological regulation of liver function through the transport of apoptosis-related molecules to modify the HCC cell TME. To elucidate the mechanism underlying the effects of Hepa1-6-derived Apo-EVs, we conducted a proteomic analysis comparing Apo-EVs and EVs from Hepa1-6 cells were compared. We identified 419 differentially expressed proteins, including 41 upregulated and 378 downregulated proteins (Fig. 4A–B). These differentially expressed proteins were significantly enriched in mitochondrial-related pathways, particularly cellular respiration, generation of precursor metabolites and energy, and energy derivation through the oxidation of organic compounds (Fig. 4C). Further enrichment analysis revealed prominentTMEM70 expression among the proteins associated with precursor metabolites and energy production (Fig. 4D). Subsequent Western blot analysis confirmed the high TMEM70 expression in Hepa1-6-derived Apo-EVs (Fig. 4E). TMEM70 participates in the biogenesis of mitochondrial ATP synthase and plays a role in regulating tumor progression.

Fig. 4.

Comparative proteomic analysis of Apo-EVs. A The differentially expressed proteins are displayed in a volcano map. B Heatmap of the enrichment of differentially expressed proteins in different types of EVs. C. Bubble plot constructed on the basis of the GO biological process data. D. Circle plot illustrating the relationships between the genes and GO terms. E. TMEM70 protein expression was detected by Western blotting (n = 3)

TMEM70 encodes a nuclear protein essential for the assembly of respiratory chain ATP synthase. Although mutations in multiple genes can cause ATP synthase deficiency, we found that TMEM70 knockdown in Hepa1-6 cells markedly impaired the activity of this enzyme (Fig. 5E). The XF Cell Mito Stress assay assesses mitochondrial function by directly measuring the cellular oxygen consumption rate. Among the parameters measured, basal respiration reflects OXPHOS capacity under baseline conditions; proton (H⁺) leakage represents H⁺ influx independent of the proton pump; and maximal respiration indicates the function of the ETC in addition to ATP synthase. Our study revealed that TMEM70-deficient cells exhibit increased proton leakage, suggesting that loss of TMEM70 leads to ATP synthase dysfunction and damage to the inner mitochondrial membrane. However, the observed increases in both basal and maximal respiration indicate that TMEM70 deficiency can increase the oxygen consumption rate (Fig. 5A), which contradicts our initial expectation that reduced TMEM70 expression would impair OXPHOS capacity. To further explore the underlying mechanism, we performed ATP bioluminescence assays, which revealed substantial reductions in mitochondrial ATP production following TMEM70 knockout (Fig. 5B). In conclusion, we propose that ATP synthase dysfunction and impaired the inner mitochondrial membrane integrity in TMEM70-knockdown cells result in oxygen consumption that is uncoupled from ATP synthesis. Notably, the relative contribution of glycolysis to cellular energy metabolism increases, possibly because TMEM70 deficiency primarily impairs OXPHOS, prompting a compensatory metabolic shift toward glycolysis. In addition, TMEM70 knockdown resulted in an increase in the mitochondrial membrane potential (ΔΨm), a parameter that when diminished is often associated with early apoptosis (Fig. 5C). To determine the effect of TMEM70 on the glycolytic ability of liver cancer cells, glycolytic capacity was evaluated by assessing lactate production under certain conditions. While TMEM70 knockout did not affect basal lactate levels, inhibition of oligomycin-induced OXPHOS induced a marked increase in lactate output in TMEM70-deficient cells compared with controls. In contrast, no difference in the response to 2-DG was detected between the groups (Fig. 5D). These findings indicate that the loss of TMEM70 shifts the reliance of HCC cells toward glycolysis from OXPHOS for energy production. Collectively, the results of this study reveal that TMEM70 deficiency disrupts oxidative phosphorylation and glycolytic equilibrium through a complex V-dependent mechanism, leading to respiratory chain dysfunction and an altered mitochondrial membrane potential.

Fig. 5.

TMEM70 inhibits hepatocarcinoma proliferation via a mitochondrial metabolism-dependent mechanism. A The OCR was measured using a seahorse XFe96 extracellular flux analyzer. Basal respiration, maximal respiration and proton leak-related OCRs were quantified. B Relative ATP levels were determined using an ATP assay kit C. The mitochondrial membrane potential was determined through JC-1 staining D. Lactate production in Hepa1-6 cells treated with or without oligomycin or 2-DG E. Complex V activity was detected by an enzyme-coupled spectrophotometric assay F. Cell viability was determined by a cell counting kit-8 assay G. Detection of apoptosis through flow cytometry H. Detection of the cell cycle distribution through flow cytometry I. mRNA expression levels of cyclin A2, cyclin B1, cyclin D, cyclin E, p16 INK4a, and p27 Kip1 were measured using RT‒qPCR. *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001 (n = 3)

On the basis of the aforementioned evidence, we propose that TMEM70 modulates HCC cell proliferation and apoptosis through metabolic reprogramming. To investigate the functional mechanism of TMEM70 in liver cancer cells, we subjected control and STS-treated cells to TMEM70 knockout with or without oligomycin treatment. TMEM70 knockdown substantially attenuated the STS-induced loss of cell viability, as determined by CCK-8 assays, and this effect was maintained with oligomycin cotreatment (Fig. 5F). Conversely, flow cytometry revealed that TMEM70 deficiency reduced apoptosis, and this effect was reversed by oligomycin (Fig. 5G). Cell cycle analysis demonstrated that TMEM70 knockout alleviated STS-induced G1 arrest and increased S/G2-M phase progression, promoting proliferation. This proproliferative effect was negated by cotreatment with oligomycin (Fig. 5H). Molecular analysis via RT‒qPCR confirmed that TMEM70 depletion upregulated cyclins A and D while downregulating p27. Combined TMEM70 knockdown and oligomycin treatment further increased the expression of cyclins A, B, and E, with a concomitant reduction in p27 expression (Fig. 5I).

To further determine whether Apo-EV-associated TMEM70 functionally contributes to the effects of Apo-EVs on HCC cells, we generated shNC-Apo-EVs, shTMEM70-Apo-EVs, and shTMEM70+TMEM70 OE-Apo-EVs. Compared with shNC-Apo-EVs, shTMEM70-Apo-EVs exhibited a significantly reduced ability to suppress HCC cell viability and proliferation (Fig. S7A). TMEM70 depletion in Apo-EVs also weakened Apo-EV-induced apoptosis and mitochondrial functional (Fig. S7B-C). Notably, restoration of TMEM70 expression in Apo-EV-producing cells partially rescued these effects. Together, these findings indicate that Apo-EV-associated TMEM70 is required for the full in vitro activity of Apo-EVs in regulating HCC cell proliferation, apoptosis, and mitochondrial metabolism (Fig. S7). Taken together, these findings establish TMEM70 as a key regulator of HCC cell proliferation and apoptosis via the OXPHOS metabolic axis.

TMEM70 in Apo-EVs effectively suppressed tumor development in a murine model of HCC

To validate the increase in apoptotic frequency among HCC cells compared with that in normal hepatocytes, we established a murine HCC model through Hepa1-6 cell implantation. Three weeks post-inoculation, compared with those receiving Apo-EVs or TMEM70-overexpressing Hepa1-6 cells, control animals had significantly larger tumor volumes and more tumor nodules (Fig. 6A–C). Notably, the tumor-suppressive efficacy was comparable between the two intervention groups, with no statistically significant differences in these parameters. Consistent with the gross pathological observations, both Apo-EV administration and TMEM70 overexpression significantly reduced the absolute liver weight and liver-to-body weight ratios, indicating the suppression of tumor-associated hepatomegaly (Fig. 6D, E). Histological evaluation demonstrated a decreased tumor burden and improved parenchymal architecture in both treatment groups. We also performed a Ki67 immunohistochemical analysis, which revealed a marked reduction in proliferative indices in Apo-EVs- and TMEM70-treated tumors compared with those in control tumors (Fig. 6F). The similar suppression of proliferation between the two intervention groups suggests functional convergence of their antitumor mechanisms.

Fig. 6.

TMEM70 from Apo-EVs effectively suppressed tumor development in a murine model of HCC A. Representative images of HCC tumors from the orthotopic HCC model B. liver weight, C. tumor number, D. liver/body weight ratio and E. tumor volume of mice subjected to different treatments Representative IHC staining images of paraffin-embedded tumor sections for the visual analysis of ki-67 expressio (bar = 200 μm). **, p < 0.01; ***, p < 0.001; ****, p < 0.0001 (n = 6–8)

These findings collectively demonstrate that Apo-EVs exert notable antitumor effects against HCC through the efficient delivery of TMEM70 to neoplastic cells, resulting in coordinated cell cycle arrest and the activation of apoptosis. The functional equivalence between Apo-EV administration and direct TMEM70 overexpression establishes a causal role for TMEM70 in mediating the tumor-suppressive activity of apoptotic vesicles.

Discussion

The growth of cancerous cells is the result of the disruption of the balance between cell proliferation and cell death [4]. Apoptosis, which is an extraordinary form of programmed cell death, serves as a key physiological mechanism to eliminate potentially harmful cells and limit the expansion of cell populations [3]. Consequently, apoptosis has a major effect on cancer progression. The suppression of apoptosis, as one of the acquired characteristics of cancer cells, is advantageous for tumor development. Previous studies have indicated that cell autonomous apoptosis is a common tumor suppressor mechanism in the context of cancer [34]. However, subsequent studies have revealed the complexity of the relationship between cancer cells and apoptosis [4]. Apoptosis may reduce the number of abnormal tumor cells while also regulating the TME through the release of molecules and vesicles [35, 36]. Our study confirmed that apoptosis of HCC cells is significantly increased, indicating that apoptosis plays a role in the development of liver cancer.

Apoptosis is accompanied by the release of Apo-EVs that carry donor cell-derived components and surface markers [7]. Apo-EVs act as carriers of information, regulating the proliferation or apoptosis of tumor cells by influencing the TME in tumor tissue [8, 10]. Apo-EVs derived from normal mesenchymal stem cells (MSCs) activate Fas ligands in multiple myeloma (MM) cells, thereby inducing their apoptosis [37]. In contrast, Apo-EVs derived from cervical cancer (CC) cells carry MTA1, which promotes CC dormancy and confers resistance to radiotherapy [38]. Our study demonstrated that apoptotic HCC cells release increased quantities of Apo-EVs into tissue fluid. The substantial increase in apoptosis observed in these cells may therefore mediate the systemic dissemination of Apo-EVs. In line with methodologies described by Brand et al., our experiments further revealed a significant increase in circulating Apo-EVs of nonhematopoietic origin in HCC-bearing mice without distant metastasis. These findings argue against the direct role of tumor-induced systemic immunoinflammation in increasing vesicle counts and instead support the view that apoptotic HCC cells contribute to the release of Apo-EVs into the circulation. On the basis of these findings, we hypothesized that Apo-EV-mediated apoptosis regulates tumor development.

We investigated the underlying mechanism and observed that coculture with Apo-EVs reduced the viability and increased the apoptosis of HCC cells, suggesting that Apo-EVs may directly modulate apoptotic pathways to influence cancer progression. To assess the specificity of this effect, we performed parallel experiments under identical conditions and found that revealed that Apo-EVs did not significantly affect the growth of normal hepatocytes. These results collectively indicate that Apo-EVs selectively impede HCC progression without compromising normal hepatocyte function. Furthermore, using PKH26-based tracing, we observed specific enrichment of Apo-EVs in HCC tissues, supporting their targeted localization within hepatic tumor regions.

Metabolic reprogramming is a recognized hallmark of cancer [39]. In pancreatic cancer cells, energy metabolism remodeling is closely associated with tumor proliferation, invasion, metastasis, drug resistance, and immune evasion [40]. During OXPHOS, the chemical energy stored in nutrients is converted into ATP thus providing the energy essential for tumor proliferation and the regulation of apoptosis [41]. Although glycolysis yields ATP less efficiently, it promotes biosynthetic processes, suppresses apoptosis, and generates signaling metabolites that increase cancer cell survival under stress conditions [31]. The Warburg effect is widely regarded as the foundational paradigm of glucose metabolic reprogramming in cancer cells and is characterized by increased glycolysis coupled with decreased OXPHOS efficiency [23, 42]. Following the disruption of PDHX, a key enzyme in the aerobic respiratory chain, tumor cells exhibit increased expression of genes related to aerobic glycolysis and lactylation [25]. Consequently, most glucose is converted into lactate, thereby promoting tumor progression.

Mitochondria serve as cellular energy factories and metabolic hubs, coordinating the balance between glycolysis and oxidative phosphorylation (OXPHOS) [43, 44]. Mitochondria can be translocated between cells via large EVs, which target and enter recipient cells to increase the efficiency of aerobic respiration or participate in the regulation of other mitochondrial functions [45]. This transfer mechanism can effectively interfere with apoptosis in recipient cells and promote the recovery of their metabolic function [46]. Studies have confirmed that Apo-EVs are enriched in mitochondria-associated proteins and mitochondrial DNA (mtDNA). Such vesicles can increase the mitochondrial membrane potential in recipient cells by modulating mitochondrial activity, thus altering cellular proliferation and metabolic status [47]. The results of the present study revealed that Apo-EVs carry large amounts of mitochondria-associated proteins, suggesting that their biological effects may be closely linked to mitochondria-mediated metabolic regulation. In cancer cells, the coordinated interplay between OXPHOS and glycolysis represents a fundamental characteristic of glucose metabolism [48]. Notably, OXPHOS activity suppresses glycolytic flux [42]. As the core enzymatic component of OXPHOS, ATP synthase increases mitochondrial oxidative respiration while attenuating glycolysis, thus inhibiting cancer cell proliferation [23, 49]. Furthermore, ATP synthase modulates apoptosis by regulating cristae architecture, which influences the pH gradient and membrane curvature and facilitates the release of cytochrome C into the cytoplasm. As a central component of OXPHOS, ATP synthase has emerged as a significant predictor of the response to chemotherapy [50]. We demonstrated that Apo-EVs contained large amounts of mitochondria-related proteins. These results suggest that the ability of Apo-EVs to inhibit liver cancer growth may be mediated by ATP synthase. Through coproteomic technology, we demonstrated that Apo-EVs express high levels of mitochondria-related proteins. These findings suggest that the inhibitory effect of Apo-EVs on the growth of liver cancer may be achieved through interference with mitochondrial-mediated cellular metabolism.

TMEM70 is an accessory protein localized to the inner mitochondrial membrane that facilitates ATP synthase biogenesis [27]. The absence of TMEM70 disrupts ATP synthase assembly, leading to a metabolic shift from OXPHOS n toward glycolysis in cancer cells [30], a transition closely linked to the regulation of apoptosis. Recent studies have reported that the TMEM70 promoter is hypermethylated in thioacetamide-induced rat HCC models [31], whereas the downregulation of TMEM70 expression has been shown to suppress mitochondrial respiration and increase proliferative activity [51]. In a rat model of early-stage liver cancer, the downregulation of TMEM70 expression was found to suppress mitochondrial ATP synthase activity and upregulate the expression of glycolysis-related proteins [30]. This shift reflects metabolic reprogramming from OXPHOS to glycolysis, which is consistent with the Warburg effect. In our work, we detected elevated levels of TMEM70 in Apo-EVs. Functional studies confirmed that TMEM70 regulates the metabolic phenotype of liver cancer cells, promoting a shift between OXPHOS and glycolysis and altering mitochondrial membrane potential—effects potentially mediated through impaired ATP synthase assembly. These findings suggest that Apo-EV-delivered TMEM70 further attenuates OXPHOS in HCC, which is consistent with its limited effect on normal hepatocytes.

To our knowledge, this study is the first to establish a link between Apo-EVs and mitochondrial metabolism in cancer cells, demonstrating that HCC cells utilize Apo-EVs to regulate oxidative respiration and influence tumor progression. However, several limitations should be acknowledged. First, because no specific marker protein for apoptotic HCC cell-derived EVs has been identified to date, the specificity of these vesicles cannot be accurately determined. Moreover, flow cytometry only indirectly suggested an increase in the number of apoptotic HCC cell-derived EVs in peripheral blood through an exclusion‑based approach. Additionally, since vesicles from various tissues and organs may enter the circulation, the quantification of apoptotic HCC cell-derived EVs in our experiments may be associated with errors. Second, and most importantly, HCC cell lines exhibit high heterogeneity and different subtypes; using a single cell line may fail to capture key features of proliferative or metastatic HCC, potentially leading to misinterpretation of the proliferation and invasion mechanisms of HCC cells. Third, to validate the mechanism underlying the effects of Apo-EVs on HCC cells, we examined metabolism and activity in these cells following TMEM70 knockdown, as well as the effect of intravenous TMEM70 injection on HCC progression. However, we did not perform in vitro TMEM70 overexpression experiments. As a result, our study did not address whether TMEM70 expression is sufficient to induce apoptosis in HCC cells. Furthermore, the dose-dependent effect of TMEM70 on HCC cells requires additional experimental validation. Finally, while our study demonstrates that apoptotic HCC cell-derived EVs could enter the bloodstream, their effects on the physiological functions of other vital organs (such as the heart, liver, spleen, lungs, and kidneys) remain unknown. These aspects will be investigated in future studies to fully assess the biosafety profile of this therapeutic approach and provide a robust foundation for its clinical translation.

Conclusion

This study reveals the crucial regulatory role of Apo-EVs in HCC progression: Apo-EVs transport the mitochondrial protein TMEM70, shifting the energy metabolism of tumor cells from OXPHOS to glycolysis and thereby effectively inhibiting liver cancer growth. These findings not only deepen our understanding of the pathogenesis of liver cancer and expand our knowledge of the functions of Apo-EVs in tumor biology but also, more importantly, provide a solid experimental basis for the clinical translation of this theoretical achievement to human liver cancer. Additionally, this study provides a theoretical foundation for engineering Apo-EVs as targeted delivery vectors for precise cancer treatment and suggests that strategies for reprogramming mitochondrial glucose metabolism have broad application prospects in broad-spectrum targeted anticancer therapies.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Yi Chen: Writing-Original Draft, Writing-review & editing, Methodology, Validation, Project administration, Formal analysis, Data curation, Conceptualization. Xihui Wang: Writing-Original Draft, Methodology, Visualization, Software, Investigation, Formal analysis. Tianjie Yuan: Methodology, Resources, Software, Investigation. Chenchen Ma: Validation, Project administration, Investigation. KangJie Xie: Writing-review & editing, Methodology, Software, Project administration, Investigation. Feixiang Wu: Resources, Project administration, Investigation. Weifeng Yu: Writing-review & editing, Supervision, Funding acquisition, Conceptualization.

Funding

This work was supported by the National Natural Science Foundation of China (32030043), the Fundamental Research Funds for the Central Universities (No.25X010202804) and the Shanghai Engineering Research Center of Perioperative Organ Support and Function Preservation (20DZ2254200).

Data availability

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD079444.

Declarations

Ethics approval and consent to participate

All animal experiments were carefully performed in accordance with the International Guide for the Care and Use of Laboratory Animals and approved by the Laboratory Animal Research Center of Shanghai Jiaotong University.

Consent to publish

Not applicable.

Competing interests

The authors declare that they have no conflicts of interest.