TCEB2/HIF1A signaling axis promotes chemoresistance in ovarian cancer cells by enhancing glycolysis and angiogenesis

Abstract

Ovarian cancer is an extremely malignant gynaecological tumour with a poor patient prognosis and is often associated with chemoresistance. Thus, exploring new therapeutic approaches to improving tumour chemosensitivity is important. The expression of transcription elongation factor B polypeptide 2 (TCEB2) gene is reportedly upregulated in ovarian cancer tumour tissues with acquired resistance, but the specific mechanism involved in tumour resistance remains unclear. In this study, we found that TCEB2 was abnormally highly expressed in cisplatin-resistant tumour tissues and cells. TCEB2 silencing also inhibited the growth and glycolysis of SKOV-3/cisplatin (DDP) and A2780/DDP cells. We further incubated human umbilical vein endothelial cells (HUVECs) with culture supernatants from cisplatin-resistant cells having TCEB2 knockdown. Results revealed that the migration, invasion, and angiogenesis of HUVECs were significantly inhibited. Online bioinformatics analysis revealed that the hypoxia-inducible factor-1A (HIF-1A) protein may bind to TCEB2, and TCEB2 silencing inhibited SKOV-3/DDP cell growth and glycolysis by downregulating HIF1A expression. Similarly, TCEB2 promoted HUVEC migration, invasion, and angiogenesis by upregulating HIF1A expression. In vivo experiments showed that TCEB2 silencing enhanced the sensitivity of ovarian cancer nude mice to cisplatin and that TCEB2 knockdown inhibited the glycolysis and angiogenesis of tumour cells. Our findings can serve as a reference for treating chemoresistant ovarian cancer.

Introduction

Ovarian cancer is one of the three major common malignant cancers of the gynaecologic reproductive system. At the time of diagnosis, about 70–80% of patients are already in advanced stages of the disease. Thus, the mortality rate of patients with ovarian cancer ranks high amongst gynaecological malignancies [1]. In China, about 52,971 new cases of ovarian cancer and 30,886 related deaths are reported annually, and this figure increases yearly [2]. More pathological types of ovarian cancer exist, amongst which epithelial ovarian cancer is the most common, accounting for approximately 70% of ovarian cancer cases [3].

The treatment of choice in ovarian cancer involves effective tumour cytoreductive surgery combined with platinum-based combination chemotherapy. In chemotherapy, platinum-based agents such as cisplatin, carboplatin, and oxaliplatin can promote tumour cell death and apoptosis by inducing DNA damage [4, 5]. However, chemoresistance has become an important cause of poor prognosis in ovarian cancer during the course of clinical treatment. About 80% or more ovarian cancer patients experience the invasion, metastasis, and recurrence of tumour lesions after treatment, and the 5 year survival rate is only ~ 40% [6, 7]. Therefore, identifying the molecular targets that regulate chemoresistance in ovarian cancer has great significance. In recent years, increasing evidence has revealed that tumour angiogenesis plays an important role in the progression of ovarian cancer.

Tumour angiogenesis is a complex process involving endothelial proliferation, migration of pericapillary extracellular matrix and endothelial cell migration, and differentiation into blood vessels of organic energy [8]. New blood vessels form through the extension of primary blood vessels and then provide nutrition to the continuously infiltrating primary tumour. Therefore, tumour angiogenesis is the basis of tumour initiation, progression, invasion, and metastasis [9]. A study in a nude mouse model has shown that ovarian cancer cell-derived exosomal miR-205 significantly promotes endothelial cell angiogenesis and accelerates angiogenesis and tumour growth [10]. Chen et al. also found that vascular endothelial growth factor (VEGF) silencing results in decreased expression of CD31, matrix metallopeptidase 2 (MMP2), and V-cadherin, markers of angiogenesis and epithelial mesenchymal transition [11]. A previous study has demonstrated that transcription elongation factor B polypeptide 2 (TCEB2) gene expression increases in ovarian cancer tumours with acquired resistance compared with sensitive tumours, and TCEB2 plays a critical role in these tumour’s acquired resistance to the anti-VEGF drug bevacizumab [12]. TCEB2 is also reportedly related closely to tumour angiogenesis. However, the specific mechanism for these phenomena remain unclear. In the present study, we deeply explored the possible molecular mechanism of TCEB2’s involvement in ovarian cancer cell resistance to provide a new reference for ovarian cancer treatment.

Materials and methods

Cell culture and clinical samples

The human ovarian adenocarcinoma cell line SKOV-3, human ovarian cancer cell line A2780, cisplatin-resistant SKOV-3 cells (SKOV-3/DDP), cisplatin-resistant A2780 cells (A2780/DDP), and human umbilical vein endothelial cells (HUVECs) were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). These cells were cultured in Roswell Park Memorial Inatitute-1640 medium (Sigma–Aldrich, St. Louis, MO, USA) with 10% foetal bovine serum and in a humidified atmosphere containing 5% carbon dioxide (CO₂). A total of 40 cases of ovarian cancer diagnosed pathologically and treated in Shaanxi Provincial People’s Hospital from February 2020 to November 2022 were selected. They included 20 cases in the cisplatin chemotherapy-sensitive group and 20 cases in the chemotherapy-resistant group. We also collected 20 cases of patients with other diseases (normal ovarian tissue according to the pathological diagnosis) or benign ovarian tumours in the same period. Tumour tissue (about 1 cm × 1 cm) without fat and blood clot was collected and divided into two parts. One portion was immediately placed in a 2 mL cryopreservation tube, added with RNA protection solution, and stored in liquid nitrogen for later RNA extraction. One part was placed in an embedding box, fixed in 4% formaldehyde solution for 24 h, and embedded in paraffin for immunohistochemical experiments. The protocol in this study was approved by the Ethics Committee of Shaanxi Provincial People’s Hospital, and each subject signed an informed consent.

Cell transfection

pcDNA-HIF1A and vector were purchased from RiboBio Co., Ltd. (Guangzhou, China). TCEB2 expression was knocked down using shRNA (shTCEB2) designed at GenePharma (Shanghai, China), and blank shRNA served as a control (shNC). Cells were seeded onto six-well plates and cultured for 24 h. When the cells reached 60%–70% confluence, shTCEB2 was transiently transfected into cells according to the instructions of Lipofectamine 2000 kits. After incubating the transfected cells in the incubator for 6 h, the medium was replaced with complete medium. Forty-eight hours later, the inhibitory efficiency of TCEB2 shRNA was verified by immunofluorescence (Supplementary Fig. 1).

Animals

Fourteen female BALB/c nude mice (4–6 weeks old; 18–22 g) were raised in specific pathogen-free environment. SKOV-3/DDP cells were transfected with shNC or shTCEB2, and these cells (2 × 10⁶ cells) were then implanted subcutaneously into the right flank of mice. Seven days after inoculation, mice received 10 mg/kg cisplatin treatment every 3 days. These mice were euthanized on day 28 after injection, and tumour tissues were collected. All procedures in this work were performed in accordance with the approval of the animal ethics committee.

Real time-quantitative polymerase chain reaction (RT-qPCR)

Total RNA from cells was isolated using TRIzol. Single-stranded cDNA was synthesized with a PrimeScript Reagent Kit. Real-time qPCR was conducted by using a SYBR Premix Ex TaqTM Kit. TCEB2 forward, 5′—GAG GCC CAT TTC CCC CAA TA—3′, reverse, 5′-ACA GGA CAG CAC AGG AAC TG—3′; HIF1A forward, 5′—TAC TCA GCA CTT TTA GAT GCT GTT—3′, reverse, 5′—ACG TTC AGA ACT TAT CCT ACC AT—3′; GAPDH forward, 5′—TAC CAC AGG CAT TGT GAT GG—3′, reverse, 5′—TTT GAT GTC ACG CAC GAT TT—3′. The relative expression levels were normalized with that of the GAPDH gene and calculated using the 2^−ΔΔCt method.

Western blotting

Proteins in cell and tissues were extracted using RIPA lysis buffer. Then, proteins were separated by polyacrylamide-gel electrophoresis and transferred onto polyvinylidene fluoride membranes. After blocking for 2 h at 37 °C, the following primary antibodies were added and incubated at 4 °C. The primary antibodies (Abcam) are as follows: GAPDH (1:3000), TCEB2 (1:1000), hexokinase 2 (HK2; 1:500), lactate dehydrogenase A (LDHA; 1:5000), hypoxia-inducible factor-1A (HIF-1α; 1:1000), B-cell lymphoma 2 (Bcl-2; 1:1000), Bcl2-associated X (Bax; 1:2000), and β-catenin (1:5000). Then, the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG for 1 h at room temperature. The protein bands were visualized using enhanced chemiluminescence detection reagents (National Institutes of Health, Bethesda, MD, USA).

Immunohistochemical assay

Cartilage tissue sections were digested in pancreatin for 10 min and treated with 3% H₂O₂. After blocking the cartilage tissue sections in 10% goat serum for 1 h, the sections were rinsed twice with PBS. The rinsed clean sections were incubated with anti-HIF1A antibody and anti-TCEB2 antibody (1:50) overnight at 4 °C in a refrigerator. Sections were rinsed thrice with PBS and incubated with biotinylated secondary antibodies. Next, sections were rinsed and incubated with 3,30-diaminobenzidine substrate for 2 min. Next, the sections were rinsed twice with PBS and counterstained with hematoxylin for 1 min. After washing with tap water and returning to blue, the sections were dehydrated in graded alcohol for 3 min each and made transparent with xylene. Finally, the slices were observed under a microscope. The nuclei were blue under the microscope, and the positive results were brown in shades. Under a light microscope, positively stained cells were counted in 10 random fields, and the positivity percentages were determined as follows: (number of positive cells) / (total number of cells) × 100.

MTT assay

Cells were first cultured on 96-well plates (2 × 10⁴ cells/well) and transfected according to experimental grouping. After transfection, the cell-culture supernatant was discarded and MTT solution was added to each well, followed by incubation at 37 °C for 4 h. The absorbance of each well was measured and recorded at 490 nm.

Cell colony-formation assay

The cells were first cultured on six-well plates, and fresh medium was added every 3 days. After 14 days of cell culture, the cell supernatant was aspirated with a pipette and discarded. Next, paraformaldehyde and crystal violet (sigma) were added and co-incubated in culture plates, and cell colonies were counted.

Glucose uptake, lactate production, and VEGF content

Cells were first seeded onto 12-well plates (1 × 10⁵ cells per well) and cultured overnight. Then, the medium was removed and cells were rinsed with PBS. After culturing the cells in glucose-free DMEM containing 18F-FDG, cell lysates were obtained by washing the cells and supplementing each well with NaOH (0.5 mol/L) for 1 h in culture. Finally, the radioactivity of the lysates was detected to quantify 18F-FDG uptake. Cell supernatant was collected, and lactate concentration and VEGF content were determined using corresponding kits.

Flow cytometry

FITC-labelled annexin V is a phospholipid binding protein with high affinity for phosphatidylserine. It binds to the cell membrane of early apoptotic cells through phosphatidylserine exposed outside the cell and is a sensitive indicator for detecting early apoptosis. The cell-culture medium was collected and rinsed twice with PBS. Next, the cell density (1 × 10⁶ cells/mL) was adjusted and 100 μL cell suspension was aspirated to inoculate into a cell-culture flask. After adding annexin V-FITC to the cells, they were incubated in darkness. Finally, binding buffer was added into the culture tube, and flow cytometry was used to determine apoptosis rate.

Wound-healing assay

HUVECs were seeded onto conditioned plates (1 × 10⁵ per well) and cultured for 12 h. After rinsing the cells twice with PBS, a straight line was drawn on the medium with a pipette tip and culture was continued for 24 h before observing the cell migration under an inverted microscope.

In vitro lumen formation

Matrigel was added onto 96-well plates and allowed to solidify by standing for 1 h. HUVECs were then cultured on 96-well plates (2 × 10⁴), and after adherence, the original medium was discarded and replaced with the supernatant of the transfected cisplatin-resistant cells. After incubation in a 37 °C incubator for 48 h, cells were observed microscopically for lumen formation and imaged by ImageJ software.

Transwell assay

HUVECs (1 × 10⁴) were seeded onto the upper chamber of the Transwell chamber coated with Matrigel. Complete medium was added into the lower chamber. After 48 h of culture, cells in the upper chamber were removed, and cells in the lower chamber were fixed with ethanol and stained with crystal violet. The invaded cells were counted under a light microscope (Olympus, Tokyo, Japan).

Co-immunoprecipitation (Co-IP) assay

Cells were rinsed twice with PBS and then lysed extensively by adding RIPA lysis buffer. The supernatant was obtained after centrifuging the lysed cells and incubated overnight with the corresponding antibodies overnight at 4 °C. Then, protein A agarose beads were added to obtain antigen–antibody complexes, followed by incubation overnight at 4 °C on a shaker to obtain antigen–antibody complexes. After rinsing the agarose-bead antigen–antibody complexes with PBS, we obtained free antigen, antibody, and beads by adding protein loading buffer to the complexes and boiling this mixture. Finally, the expression levels of the obtained proteins were detected by Western blotting.

Statistical analysis

SPSS software was used to analyse all data (ver. 21.0). Data obtained from three independent experiments are expressed as the mean ± standard deviation. Comparisons between two groups were performed by t-tests. For comparison amongst multiple groups, one-way ANOVA was applied to assess homogeneity of variance, and nonparametric tests were used to assess uneven variance. For non-normally distributed data, the Mann–Whitney U-test was used for comparison between two groups, whereas the Kruskal–Wallis test was used for comparison between multiple groups. Values with P < 0.05 were statistically significant.

Results

TCEB2 expression increased in ovarian cancer DDP-resistant cells and tissues

Through online bioinformatics database analysis, we found abnormally increased TCEB2 expression in ovarian cancer tissues (Fig. 1A, B; http://gepia.cancer-pku.cn/). TCEB2 expression increased in tumour tissues (Fig. 1C) and DDP-resistant tumour tissues (Fig. 1D). TCEB2 protein expression significantly increased in cisplatin resistance cells compared with parental cells (Fig. 1E). Tumour cells and their parental cells were further treated with cisplatin. Results showed that with increased cisplatin concentration, the viability of cisplatin-sensitive cells was inhibited compared with that of cisplatin-resistant cells (Fig. 1F). The IC₅₀ of cisplatin in resistant cells also increased (Fig. 1G). Moreover, silencing TCEB2 expression significantly enhanced cisplatin sensitivity in resistant cells (Fig. 1H).

Fig. 1

Expression of TCEB2 in cisplatin-resistant cell lines and tissues of ovarian cancer. A and B TCEB2 expression in ovarian cancer tissues were analysed using an online bioinformatics database (http://gepia.cancer-pku.cn). C TCEB2 mRNA expression in tissues (20 pairs). D TCEB2 mRNA expression in tumour tissues resistant and sensitive to DDP (20 pairs). E TCEB2 protein expression. F Viability of cisplatin-resistant cells treated with cisplatin. G IC₅₀of cisplatin in cisplatin-resistant cells. H Viability of TCEB2 silenced cisplatin-resistant cells treated with cisplatin. *P < 0.01

TCEB2 knockdown inhibited growth and glycolysis in SKOV-3/DDP and A2780/DDP cells

We found that shTCEB2 inhibited TCEB2 protein expression in cisplatin-sensitive cells (Fig. 2A). TCEB2 knockdown also attenuated the proliferation ability (Fig. 2B) and clonogenic ability (Fig. 2C) of cisplatin-sensitive cells, as well as promoted apoptosis (Fig. 2D) and reduced glucose uptake (Fig. 2E) and lactate production (Fig. 2F) in SKOV-3/DDP and A2780/DDP cells. Western blotting results showed that TCEB2 silencing significantly inhibited HK2 and LDHA protein expression in cells (Fig. 2G).

Fig. 2

Effect of TCEB2 silencing on the growth and glycolysis of SKOV-3/DDP and A2780/DDP cells. Cisplatin-resistant cells were transfected with shNC or shTCEB2. A TCEB2 protein expression. B Cell viability. C Cell proliferation. D Cell apoptosis. E Glucose-uptake levels. F Lactate secretion levels. G HK2 and LDHA protein expression. *P < 0.01

TBEC2 knockdown inhibited the angiogenesis of HUVECs

SKOV-3/DDP cells and A2780/DDP cells were transfected with shNC or shTCEB2, and the cell culture supernatants were collected and co-incubated with HUVECs. We observed a slower migration rate of HUVECs incubated with the culture supernatant of shTCEB2 transfected cisplatin-sensitive cells compared with the control (Fig. 3A). Additionally, HUVECs incubated with culture medium from cisplatin-resistant cells transfected with shTCEB2 had fewer branch points (Fig. 3B). Similarly, the invasion of HUVECs was inhibited in SKOV-3/DDP-shTCEB2 and A2780/DDP-shTCEB2 medium groups (Fig. 3C). We also found that TCEB2 knockdown inhibited VEGF secretion in HUVECs (Fig. 3D), and silencing TCEB2 decreased the protein expression of angiogenic regulatory molecules, including HIF-1α, MMP2, and β-catenin (Fig. 3E).

Fig. 3

Effect of TCEB2 silencing on the angiogenesis of HUVECs. SKOV-3/DDP cells and A2780/DDP cells were transfected with shNC or shTCEB2, and cell-culture supernatants were collected and co-incubated with HUVECs. A Cell migration. B In vitro tube-formation assay. C Cell invasion. D VEGF secretion levels. E HIF-1α, MMP2, and β-catenin protein expression. *P < 0.01

TCEB2 upregulated HIF1A protein expression

Through online bioinformatics analysis, we found that the HIF1A protein may bind to and be regulated by TCEB2 (Fig. 4A; https://thebiogrid.org; https://cn.string-db.org/). pcDNA-TCEB2 increased HIF1A expression and shTCEB2 decreased HIF1A expression (Fig. 4B). Co-IP assay verified the binding between TCEB2 and HIF1A (Fig. 4C). Next, we found that HIF1A mRNA expression was significantly higher in tumour tissues from 20 ovarian cancer patients (Fig. 4D). By immunohistochemical staining, we observed more HIF1A positive cells in tumour tissue sections (Fig. 4E).

Fig. 4

HIF1A is a potential binding protein for TCEB2. A Online bioinformatics databases were used to predict the potential binding proteins of TCEB2 (https://thebiogrid.org/; https://cn.string-db.org/). B HIF1A protein expression. C The binding of TCEB2 and HIF1A was validated by co-immunoprecipitation assay. D HIF1A mRNA expression in tissues (20 pairs). E Representative images of immunohistochemical staining of HIF1A in tissues. *P < 0.01

TCEB2 promoted SKOV-3/DDP cell growth and glycolysis by upregulating HIF1A expression

shTCEB2 inhibited HIF1A protein expression in SKOV-3/DDP cells, and pcDNA-HIF1A removed the effect of shTCEB2 (Fig. 5A). TCEB2 silencing inhibited cell proliferation (Fig. 5B) and clonogenicity (Fig. 5C), promoted cell apoptosis (Fig. 5D), inhibited the cell uptake of glucose (Fig. 5E), and reduced lactate production (Fig. 5F). However, this effect was abolished by pcDNA-HIF1A. Western blotting results showed that pcDNA-HIF1A removed the inhibitory effect of TCEB2 silencing on glycolysis-related molecule expression in SKOV-3/DDP cells (Fig. 5G).

Fig. 5

TCEB2 affects SKOV-3/DDP cell growth and glycolysis via HIF1A. SKOV-3/DDP cells were transfected with shTCEB2 alone or together with pcDNA-HIF1A. A HIF1A protein expression. B Cell viability. C Cell proliferation. D Cell apoptosis. E Glucose-uptake levels. F Lactate secretion levels. G HK2 and LDHA protein expression. *P < 0.01

TCEB2 promoted HUVEC angiogenesis by upregulating HIF1A expression

SKOV-3/DDP cells were transfected with shTCEB2 alone or together with pcDNA-HIF1A, and cell culture supernatants were collected and co-incubated with HUVECs. We found that silencing TCEB2 inhibited the migration of HUVECs, and HIF1A overexpression accelerated cell migration again (Fig. 6A). Compared with HUVECs incubated with medium from SKOV-3/DDP cells transfected with shTCEB2 alone, we observed increased branch points in HUVECs incubated with medium from SKOV-3/DDP cells overexpressing HIF1A (Fig. 6B). Moreover, pcDNA-HIF1A removed the inhibitory effect of TCEB2 silencing on HUVEC invasion (Fig. 6C). Furthermore, we found that TCEB2 knockdown decreased VEGF secretion (Fig. 6D) and angiogenesis regulatory molecules expression in HUVECs, but this effect was abolished after HIF1A overexpression (Fig. 6E).

Fig. 6

TCEB2 affects HUVEC angiogenesis through HIF1A. SKOV-3/DDP cells were transfected with shTCEB2 alone or together with pcDNA-HIF1A, and the cell-culture supernatants were collected and co-incubated with HUVECs. A Cell migration. B In vitro tube-formation assay. C Cell invasion. D VEGF secretion levels. E HIF-1α, MMP2 and β-catenin protein expression. *P < 0.01

TCEB2 knockdown inhibited SKOV-3/DDP cell growth in vivo

We observed that TCEB2 silenced SKOV-3/DDP cells induced the generated tumour tissues to be more sensitive to cisplatin, and representative images are shown in Fig. 7A. The weight (Fig. 7B) and volume (Fig. 7C) in TCEB2 knockdown tumour tissues were smaller. Next, we collected tumour tissues from nude mice for subsequent studies. Immunohistochemical staining results suggested a significant reduction in the number of TCEB2 and HIF1A protein positive cells in tumour tissues of nude mice with TCEB2 silencing (Fig. 7D). Western blotting results indicated that TCEB2 silencing decreased Bcl-2 and increased Bax protein expression in tumour tissues (Fig. 7E and F). Similarly, we observed less protein expression of glycolytic molecules HK2 and LDHA in tumour tissues of nude mice with TCEB2 silencing (Fig. 7E and G). CD31 is a glycoprotein on the cell surface, also known as the attachment molecule of vascular endothelial cells. It is highly expressed in vascular endothelial cells, monocytes/macrophages, lymphocytes, and other cell types, and it serves as one of the markers of vascular endothelial cells. We found that after DDP treatment, the positive signal of CD31 in the tumour tissue induced by TCEB2 silenced SKOV-3/DDP cells was less than that induced by SKOV-3/DDP cells (Fig. 7H). Additionally, VEGF secretion (Fig. 7I) and MMP2 protein expression (Fig. 7J) also significantly decreased in tumour tissues of TCEB2 silenced nude mice.

Fig. 7

Effect of TCEB2 silencing on tumour growth in vivo. A Representative tumour images at day 28 post-injection. B Tumour weight. C Tumour volume. D Representative images of immunohistochemical staining of TCEB2 and HIF1A. E and F Bcl-2 and Bax protein expression. E and G HK2 and LDHA protein expression. H Representative images of immunofluorescence of CD31. I VEGF secretion levels in tissues. J MMP2 protein expression. *P < 0.01

Discussion

Ovarian cancer is one of the three most common malignancies of the gynaecologic reproductive system. Most patients are already at an advanced stage at the time of diagnosis, so the mortality rate of patients with ovarian cancer ranks first amongst gynaecologic malignancies. In China, new cases and deaths from ovarian cancer increases yearly.

TCEB2 gene encodes the protein elongin B, a subunit of the transcription factor B (SIII) complex. The SIII complex, comprising elongins A/A2, B, and C, activates elongation through RNA polymerase II by suppressing the transient pausing of the polymerase at many sites within transcription units. Elongin A functions as the transcriptionally active component of the SIII complex, whereas elongins B and C are regulatory subunits. Few studies have been conducted on TCEB2, but we focused on its family member TCEB1. TCEB1 can reportedly bind to HIF protein and mediate its degradation through ubiquitination proteasome pathway [13]. Mutations in TCEB1 gene lead to abnormal binding of its encoded protein Elongin C with protein VHL, affecting the ubiquitination and degradation of HIF protein and causing HIF accumulation, which may be an important pathogenesis of TCEB1 gene mutant renal cell carcinoma [14]. Accordingly, we speculated that TCEB2 also played an important role in malignant tumour progression.

Mammalian cells have three major energy sources, including glucose, protein, and fat metabolism, amongst which glucose metabolism predominates. The transformation of normal cells into malignant cells is accompanied by the reprogramming of energy metabolism, the most typical of which is glycolysis, also known as the Warburg effect [15]. Unlike normal cells, glycolysis of malignant tumour cells is active. As such, even under aerobic conditions, the process is manifested by high rates of glucose uptake and high levels of the metabolite lactate. Ovarian cancer cells reportedly overexpress the GLUT1 glucose transporter to adequately take up glucose, thereby meeting the nutritional requirements of rapid tumour cell growth compared with normal ovarian epithelial cells [16]. Gao et al. showed that salt-inducible kinase 2 promotes glycolysis-related molecule transcription and accelerates ovarian cancer cell growth and metastasis by upregulating HIF-1α through PI3K/AKT signalling pathway activation [17]. Another study has shown that FBP1 overexpression inhibits ovarian cancer cell proliferation, decreases glycolysis, and promotes apoptosis [18]. Additionally, online bioinformatics data analysis shows that an increasing number of glycolysis-related genes are closely associated with ovarian cancer progression [19, 20]. Therefore, inhibiting the glycolysis of tumour cells holds promise as a new target for ovarian cancer therapy.

HIF-1 is a heterodimer with two subunits, α and β. Subunit α influences HIF-1 activity, whereas β is the structural subunit. HIF-1 is a master regulator of a ubiquitous transcriptional activator in the cell that regulates the expression of numerous genes at the DNA and epigenetic levels. HIF-1 affects mitochondrial oxidative metabolism, glucose uptake, energy production, and angiogenesis by regulating gene expression. It can regulate multiple target-gene transcription by binding to DNA binding sites on target genes and subsequently activate glycolysis rate-limiting enzymes to increase the aerobic glycolysis supply and improve the glucose-uptake rate. Thus, tumour cells can proliferate, invade, and metastasize under hypoxic or hypoxic conditions [21, 22]. Chen et al. showed that TRPM7 silencing enhances AMPK activation to shift glycolysis into oxidative phosphorylation by promoting the ubiquitinated degradation of HIF-1, suggesting that TRPM7 may be involved in the reprogramming of glucose metabolism to inhibit the growth of ovarian cancer cells [23]. Another study has shown that ABT737, an inhibitor of Bcl2, decreases glucose uptake and lactate secretion in SKOV3 cells by downregulating glucose uptake-related protein expression through the Sirt3–HIF1α axis, thereby inhibiting ovarian cancer cell growth [24]. An independent study has pointed out that inhibin induced by HIF-1 promoting tumour growth and vascular permeability in ovarian cancer [25]. Moreover, Zhang et al. observed a significant inhibition of VEGF secretion, vascularization, and tumour growth in nude mice of ovarian cancer treated with cinnamon extract, which are mediated by the downregulation of HIF1α expression [26].

In this study, we found that TCEB2 expression increased in ovarian cancer cisplatin-resistant cell lines and tissues. TCEB2 silencing was found to inhibit the growth and glycolysis of SKOV-3/DDP and A2780/DDP cells and also HUVEC angiogenesis, which was mediated by downregulation of HIF1A protein expression. Additionally, in vivo findings further validated that TCEB2 expression knockdown enhanced the sensitivity of tumour cells on cisplatin. Taken together, our findings may have important implications for clarifying the molecular targets of chemoresistance in ovarian cancer.