Sodium citrate targeting Ca²⁺/CAMKK2 pathway exhibits anti-tumor activity through inducing apoptosis and ferroptosis in ovarian cancer

Graphical abstract

Highlights

• •Sodium citrate induces apoptosis and ferroptosis of ovarian cancer cells.
• •Sodium citrate chelates intracellular Ca²⁺ and inhibits the CAMKK2/AKT/mTOR/HIF1α-dependent glycolysis pathway, thereby inducing cell apoptosis.
• •Sodium citrate inhibits activation of CAMKK2/AMPK pathway and increases NCOA4-mediated ferritinophagy.
• •Inactivation of CAMKK2/AMPK pathway reduces Ca²⁺ level in the mitochondria by inhibiting the activity of the MCU, resulting in excessive ROS production.
• •Sodium citrate increases the sensitivity of ovarian cancer cells to chemo-drugs.

Abstract

Introduction

Ovarian cancer (OC) is known for its high mortality rate. Although sodium citrate has anti-tumor effects in various cancers, its effect and mechanism in OC remain unclear.

Objectives

To analyze the inhibitory effect of sodium citrate on ovarian cancer cells and the underlying mechanism.

Methods

Cell apoptosis was examined by TUNEL staining, flow cytometry, and ferroptosis was examined intracellular Fe²⁺, MDA, LPO assays, respectively. Cell metabolism was examined by OCR and ECAR measurements. Immunoblotting and immunoprecipitation were used to elucidate the mechanism.

Results

This study suggested that sodium citrate not only promoted ovarian cancer cell apoptosis but also triggered ferroptosis, manifested as elevated levels of Fe²⁺, LPO, MDA and lipid ROS production. On one hand, sodium citrate treatment led to a decrease of Ca²⁺ content in the cytosol by chelating Ca²⁺, which further inhibited the Ca²⁺/CAMKK2/AKT/mTOR signaling, thereby suppressing HIF1α-dependent glycolysis pathway and inducing cell apoptosis. On the other hand, the chelation of Ca²⁺ by sodium citrate resulted in inactivation of CAMKK2 and AMPK, leading to increase of NCOA4-mediated ferritinophagy, causing increased intracellular Fe²⁺ levels. More importantly, the inhibition of Ca²⁺/CAMKK2/AMPK signaling pathway reduced the activity of the MCU and Ca²⁺ concentration within the mitochondria, resulting in an increase in mitochondrial ROS. Additionally, metabolomic analysis indicated that sodium citrate treatment significantly increased de novo lipid synthesis. Altogether, these factors contributed to ferroptosis. As expected, Ca²⁺ supplementation successfully reversed the cell death and decreased tumor growth induced by sodium citrate. Inspiringly, it was found that coadministration of sodium citrate increased the sensitivity of OC cells to chemo-drugs.

Conclusion

These results revealed that the sodium citrate exerted its anti-cancer activity by inhibiting Ca²⁺/CAMKK2-dependent cell apoptosis and ferroptosis. Sodium citrate will hopefully serve as a prospective compound for OC treatment and for improving the efficacy of chemo-drugs.

Introduction

Ovarian cancer, a multifaceted ailment predominantly affecting postmenopausal women, is linked to unfavorable survival outcomes. It ranks as the sixth most prevalent cancer and the fifth leading cause of cancer-related mortality in women residing in developed nations [1]. In contrast to other types of female cancers that present with identifiable early warning signs, ovarian cancer is characterized by non-specific symptoms, leading to delayed detection typically occurring at advanced stages (III or IV) [2]. Despite the initial positive response of most patients to chemotherapy involving paclitaxel and platinum-based drugs, nearly 80 % of women experience relapse due to the development of chemoresistance, greatly diminishing the effectiveness of chemotherapy [3]. Therefore, it is an urgent need to explore new anti-tumor therapeutic drugs for OC therapy. Increasing evidence suggests that abnormal energy metabolism is one of the main characteristics of cancer, and targeting the metabolism of cancer cells could be a valuable option to improve cancer therapeutics [4], [5], [6].

Aerobic glycolysis is an important characteristic of cancer cells, and blocking glycolysis can trigger cell apoptosis [7], [8]. It has been reported that the key enzymes involved in glycolysis were directly regulated by hypoxia inducible factor 1 subunit alpha (HIF1α) [9]. Activation of the AKT/mTOR pathway has been shown to promote glycolysis and suppress apoptosis by upregulating HIF1α [10], [11], [12], [13].

In addition to apoptosis, various other forms of cell death have been discovered that help prevent tumor growth. Ferroptosis is a newly recognized form of non-apoptotic cell death distinguished by the Fe²⁺-dependent buildup of lipid peroxides and disruption of redox balance [14], [15], [16], [17]. This alternative mechanism of cell death bypasses the resistance commonly observed with therapies targeting apoptosis by regulating the cellular redox balance. This strategy shows significant potential for the management of tumors, especially in instances where genetic mutations hinder the apoptosis pathway [14].

It has been reported that ferroportin (FPN) and ferritin heavy chain 1 (FTH1) could affect intracellular iron homeostasis and ferroptosis [18], [19]. Autophagy has been shown to be associated not only with apoptosis, but also with ferroptosis [20], [21]. Research has shown that autophagy-related (ATG) proteins (ATG5 and ATG7) are critical for the formation of the autophagosome and degradation of ferritin. Moreover, nuclear receptor coactivator 4 (NCOA4) has the capability to interact with FTH1 within autophagosomes, resulting in the degradation of ferritin and the subsequent release of Fe²⁺ [22]. The Fe²⁺ released from FTH1 degradation could accumulate and induce fenton reactions, which produces excessive ROS to promote ferroptosis [23], [24]. Additionally, fatty acid supply is critical for the induction of ferroptosis [14] Recent studies have shown that AMPK-mediated acetyl-CoA carboxylase (ACC) phosphorylation inhibits ferroptosis [25], and AMPK inhibits fatty acid synthesis through its phosphorylation and inhibition of ACC [26].

Numerous cancer types display the Warburg effect while still maintaining mitochondrial respiration. The adaptability provided by mitochondria to cancer cells, including changes in fuel usage and response to oxidative stress, enables their survival under challenging conditions, such as exposure to chemotherapy and targeted cancer therapies. Therefore, it is crucial to take into consideration the escape pathways facilitated by mitochondria when devising effective therapeutic interventions for cancer treatment [27].

The anti-tumor activity of mitochondrial inhibitors has been recognized for a considerable period of time [28]. As a regulator of cell fitness, mitochondrial Ca²⁺ affects oxidative metabolism, mitochondrial respiration, and ATP synthesis through its ability to impact cell energetics [29], [30], [31]. Ca²⁺ within the mitochondrial matrix modulates the activity of pyruvate dehydrogenase (PDH), isocitrate dehydrogenase (IDH), α-ketoglutarate dehydrogenase (OGDH), complex III and complex V, thereby affecting the NADH and ATP levels [29], [32]. Previous studies have demonstrated that CAMKK2 is a serine/threonine (Ser/Thr) kinase and CAMKK2/AMPK pathway is sensitive to intracellular Ca²⁺ concentration [33], [34]. Mitochondrial calcium uniporter (MCU) is an essential component of the mitochondrial Ca²⁺ uniporter [35], and phosphorylation of MCU by AMPK is crucial for mitochondrial Ca²⁺ uptake [36], [37]. Enhancing the cytotoxic effect of anti-cancer drugs by modulation of mitochondrial Ca²⁺ homeostasis would be a new therapeutic approach [29].

Sodium citrate is a sodium salt of citric acid and is commonly used in food as an acidity regulator, emulsifier, and flavor enhancer. It has been reported to have an inhibitory effect on a variety of malignant tumor cells [38], [39], [40], [41]. Previous studies have suggested that sodium citrate is an inhibitor of PFK1 and PFK2, the main regulators of glycolysis [42], [43], [44]. Several studies have shown that administration of sodium citrate inhibits PFK1 and PFK2, decreases ATP production, counteracts PI3K/AKT signaling, and induces apoptosis [42], [45], [46], [47]. These findings suggest that sodium citrate may be applicable to ovarian cancer treatment, and the detailed mechanism needs further elucidation.

The findings in this study indicated that the sodium citrate not only promoted ovarian cancer cells apoptosis by inhibiting glycolysis, but also induced ferritinophagy-dependent ferroptosis. Mechanistic study revealed that sodium citrate inhibited glycolysis and induced apoptosis through Ca²⁺ chelation-mediated inhibition of CAMKK2/AKT/mTOR/ HIF1α signaling pathway. Moreover, Ca²⁺ chelation by sodium citrate promoted NCOA4-mediated ferritinophagy and weakened mitochondrial Ca²⁺ uptake by inhibiting the activation of MCU activity, leading to high ROS level in the mitochondria. Ultimately, these changes induced the occurrence of ferroptosis. In summary, the results of the study clarified the mechanisms of cell death induced by sodium citrate in ovarian cancer cells, offering novel insights for potential therapeutic strategies in the management of ovarian cancer and overcoming drug resistance.

Materials and methods

Cell culture

Human ovarian cancer cells (SKOV3 and A2780), human ovarian epithelial cells (HOSEpiC and IOSE-80) and human embryonic kidney cells (HEK-293 T) were procured from the Chinese Academy of Sciences. SKOV3, A2780, HOSEpiC and IOSE-80 cells were cultured in RPMI1640 medium (Gibco) supplemented with 10 % FBS (Gibco), while HEK-293 T cells were cultured in high glucose DMEM (Gibco) supplemented with 10 % FBS (Gibco). The culture media were further supplemented with 100 IU/mL penicillin and 100 μg/mL streptomycin, and the cells were maintained at 37 °C in a 5 % CO₂ environment.

MTT assay

Cells were plated in 96-well cell plates at a density of 1 × 10⁴ cells per well, with three replicate wells per experimental group. After cell attachment, the medium in the sodium citrate treatment group was exchanged with sodium citrate-containing medium and incubated for varying periods. Subsequently, 20 μL of MTT solution (5 mg/ml in PBS) was introduced to each well, and the cells were further incubated for 4 h. The medium was then removed, and DMSO was added to dissolve the formazan product. Absorbance measurements at a wavelength of 570 nm were taken using a microplate reader. The cell viability was determined by analyzing the absorbance values.

BrdU cell proliferation assay

The Cell Proliferation ELISA BrdU kit (Roche, Cat. No. 116472290011) was utilized for the quantification of BrdU uptake in this study. Cells were seeded in 96-well cell plates at a density of 1 × 10⁴ cells per well, with three replicate wells for each experimental group. Upon cell adhesion, the sodium citrate treatment group was exposed to a medium supplemented with sodium citrate and 10 μM BrdU labeling solution. The cells underwent a 12-hour incubation period, followed by measurement of BrdU uptake in accordance with the manufacturer's guidelines. Subsequently, specific anti-BrdU antibody was introduced to facilitate the formation of BrdU-anti-BrdU antibody complexes. Chromogenic substrate was then applied, and color variances were analyzed using a microplate reader set to a wavelength of 450 nm for the evaluation of cell proliferation.

RNA isolation and real-time PCR

Cells were seeded in 6-well cell plates and allowed to adhere. Subsequently, the medium in the sodium citrate treatment group was substituted with sodium citrate-containing medium, and the cells were cultured for a duration of 12 h. Total RNA extraction was performed using TRIzol reagent (Invitrogen, Cat. No. 15596018CN), followed by cDNA synthesis with TransScript First-Strand cDNA Synthesis SuperMix (TransGen Biotech, Cat. AT301-02). The mRNA expression levels were evaluated through Real-Time PCR utilizing Fast Green qPCR SuperMix (TransGen Biotech, Cat. AQ611-01). The Real-Time PCR reaction mixture was composed of 1 μg of cDNA, 10 μL of Fast Green qPCR SuperMix, 0.2 μM of Forward Primer, 0.2 μM of Reverse Primer, and Nuclease-free H₂O in a total volume of 20 μL. The Real-Time PCR protocol included an initial denaturation step at 95 °C for 1 min, followed by 40 cycles of denaturation at 95 °C for 5 s and annealing at 60 °C for 15 s. Data analysis was performed using the 2^−ΔΔCT method with β-actin as the reference gene. Primer design for Real-Time PCR was conducted using Primer 5.0 software, and all primers were synthesized by GENEWIZ. The sequences of the primers used were provided in Supplementary Table 1.

TUNEL staining

The TUNEL staining procedure utilized the DeadEnd Fluorometric TUNEL System (Promega, Cat. G3250) for measurement. Cells were plated on slides in 24-well cell plates and allowed to attach before being subjected to different treatments. In the sodium citrate treatment group, the medium was replaced with sodium citrate-containing medium and cells were cultured for 24 h. Following this, the medium was removed and cells were fixed in 4 % methanol-free formaldehyde in PBS. Subsequently, cells were permeabilized with Triton® X-100 and incubated in a buffer containing Equlibration Buffer, Nucleotide Mix, and rTdT Enzyme at 37 °C for 1 h. DAPI was utilized as a nuclear stain for cell visualization. Apoptotic cells were identified through green fluorescence (fluorescein-12-dUTP) using fluorescence microscopy.

Annexin V-FITC apoptosis detection assay

The quantification of Annexin V-FITC apoptosis detection was conducted utilizing the FITC Annexin V Apoptosis Detection Kit I (BD Biosciences Cat. 556547). Cells were initially seeded in 6-well cell culture plates and permitted to adhere. Following this, the medium in the sodium citrate treatment group was replaced with medium containing sodium citrate and incubated for a duration of 24 h. The cells were subsequently washed twice with cold PBS and reconstituted in Binding Buffer at a concentration of 1 × 10⁶ cells/mL. Subsequently, 1 × 10⁵ cells were transferred to a new tube for further analysis. 5 µL of FITC Annexin V and 5 µL of propidium iodide (PI) were added and incubated for 15 min at room temperature in the absence of light. Subsequently, 400 µL of Binding Buffer was added to each tube. The stained cells were then analyzed using flow cytometry within one hour. Cells that exhibited positive staining for FITC Annexin V and negative staining for PI were determined to be undergoing apoptosis. Cells that displayed positive staining for both FITC Annexin V and PI were classified as either in the late stages of apoptosis or undergoing necrosis. Cells that showed negative staining for both FITC Annexin V and PI were identified as viable and not undergoing measurable apoptosis.

Immunoblotting

The cells were initially seeded into 6-well cell plates and allowed to adhere. Subsequently, the medium in the sodium citrate treatment group was replaced with sodium citrate-containing medium, and the cells were cultured for a period of 24 h. Following the removal of the culture medium and PBS washing, the cells were lysed to extract proteins utilizing RIPA Lysis Buffer (Beyotime Biotechnology, Cat. P0013B). The protein concentration was then quantified using the BCA Enhanced Protein Kit (Beyotime Biotechnology, Cat. P0012). Protein samples underwent separation based on molecular weight through polyacrylamide gel electrophoresis (SDS-PAGE), followed by transfer onto a polyvinylidene fluoride (PVDF) membrane. Subsequently, the PVDF membranes were blocked with 5 % skim milk, and incubated at 4 °C with diluted primary antibodies, and then at room temperature with Horseradish peroxidase (HRP)-conjugated secondary antibodies. The protein-antibody complexes were visualized in the final immunoblot using ECL luminescent solution (Tanon, Cat. 180-5001) and a chemiluminescence imaging system, with imaging data subsequently acquired. Film exposure ranged from 10 s to 1 min. The quantification of imaging data was utilized to determine the relative expression levels of the target protein, with β-actin serving as the reference protein. Antibodies for western blotting were as follows: Caspase 3 antibody (Proteintech, Cat. 19677-1-AP, 1:1000), Caspase 9 antibody (Proteintech, Cat.10380-1-AP, 1:1000), BAX antibody (Proteintech, Cat.50599-2-Ig, 1:1000), BCL2 antibody (Proteintech, Cat. 68103-1-Ig, 1:1000), β-actin antibody (Proteintech, Cat. 81115-1-RR, 1:2000), HIF-1 alpha antibody (Proteintech, Cat. 20960-1-AP, 1:1000), Glut1 antibody (Proteintech, Cat. 21829-1-AP, 1:1000) Hexokinase 2 antibody (Proteintech, Cat. 66974-1-Ig, 1:1000), PKM2 antibody (Proteintech, Cat. 15822-1-AP, 1:1000), GPI antibody (Proteintech, Cat. 15171-1-AP, 1:1000), PFKP antibody (Proteintech, Cat. 13389-1-AP, 1:1000), PGK1 antibody (Proteintech, Cat. 17811-1-AP, 1:1000), PGAM1 antibody (Proteintech, Cat. 16126-1-AP, 1:1000), GAPDH antibody (Proteintech, Cat. 60004-1-Ig, 1:1000), GPX4 antibody (Proteintech, Cat.67763-1-Ig, 1:1000), ACSL4 antibody (Proteintech, Cat. 22401-1-AP, 1:1000), PTGS2 antibody (Proteintech, Cat. 80455-1-RR, 1:1000), CAMKK2 antibody (Proteintech, Cat. 11549-1-AP, 1:1000), Phospho-CaMKK2 (Ser511) Antibody (Cell Signaling, Cat.12818, 1:1000), HIF-1 alpha antibody (Proteintech, Cat.20960-1-AP, 1:1000), AMPK Alpha antibody (Proteintech, Cat.10929-2-AP, 1:1000), Phospho-AMPKα (Thr172) antibody (Cell Signaling, Cat.50081, 1:1000), MCU antibody (Proteintech, Cat.26312-1-AP, 1:1000), Pan Phospho-Serine/Threonine antibody (Abclonal, Cat.AP0893, 1:1000), Ferritin heavy chain antibody (SANTA CNUZ, Cat.sc-376594, 1:1000), NCOA4 (E8H8Z) antibody (Cell Signaling, Cat.66849, 1:1000), APAF1 antibody (Proteintech, Cat.29022-1-AP, 1:1000), Cytochrome c antibody (Cell Signaling, Cat.12963, 1:1000), ACC1 antibody (Proteintech, Cat.21923-1-AP, 1:1000), Phospho-ACC1 (Ser79) antibody (Proteintech, Cat.29119-1-AP, 1:1000), HRP-conjugated Affinipure Goat Anti-Mouse IgG (H + L) (Proteintech, Cat.SA00001-1, 1:3000), HRP-conjugated Affinipure Goat Anti-Rabbit IgG (H + L) (Proteintech, Cat.SA00001-2, 1:3000).

Measurement of lactate production, glucose consumption and pyruvate content

Cells were seeded into 96-well plates with three replicate wells per group. Following cell attachment, the medium in the sodium citrate treatment group was substituted with medium containing sodium citrate and incubated for an additional 12 h. Glucose consumption, lactate production, and pyruvate content were quantified utilizing the Glucose Analysis Kit (Nanjing Jiancheng Bioengineering Institute, Cat. F0006-1-1), Lactate Assay Kit (Nanjing Jiancheng Bioengineering, Cat. A019-2-1), and Pyruvate Assay Kit (Solarbio, Cat. BC2205), respectively, in accordance with the manufacturer's guidelines. All values were adjusted based on protein concentrations.

Intracellular reactive oxygen species (ROS) measurement

Cells were initially seeded into 6-well plates and allowed to adhere prior to the replacement of the medium in the sodium citrate treatment group with a medium containing sodium citrate. The cells were then cultured for a duration of 12 h. Following this, for the detection of ROS, the cells underwent a rinse with phosphate-buffered saline (PBS) and were exposed to either 10 μM dihydroethidium (DHE) from Pulley Biologicals (Cat. C1300-2) or the fluorescent probe 2′,7′-dichlorofluorescein diacetate (DCFH-DA) from Beyotime Biotechnology (Cat. S0033S) for a period of 30 min at 37 °C. Subsequent to this treatment, the cells were washed with PBS and observed using laser scanning confocal microscopy from Olympus. The cells were incubated with the fluorescent probes DCFH-DA (excitation wavelength = 488 nm; emission wavelength = 525 nm) or DHE (excitation wavelength = 510 nm; emission wavelength = 610 nm) and subsequently analyzed using flow cytometry.

Lipid peroxidation (LPO) and malondialdehyde (MDA) and iron measurement

The cells were seeded into 6-well cell culture plates and allowed to adhere before the medium in the sodium citrate treatment group was replaced with sodium citrate-containing medium. The cells were then cultured for 12 h. LPO, MDA, and iron content were quantified using specific assay kits (Nanjing Jiancheng Bioengineering Institute, Cat. A106-1-1 for LPO, Cat. A003-4-1 for MDA, and Cat. A039-2-1 for iron) following the manufacturer's instructions. All values were adjusted based on protein concentrations.

Measurement of NADP⁺/NADPH ratio and GSH

The cells were seeded into 6-well cell plates and allowed to adhere before the medium in the sodium citrate treatment group was exchanged with a sodium citrate-containing medium. The cells were then cultured for 12 h. The NADP⁺/NADPH ratio and levels of GSH were determined using the NADP⁺/NADPH Assay Kit with WST-8 (Beyotime, Cat. S0179) and the GSH and GSSG Assay Kit (Beyotime, Cat. S0053), respectively, following the manufacturer's instructions. All values were adjusted based on protein concentrations.

Oxygen consumption rate (OCR) detection

The Seahorse XFp Extracellular Flow Analyzer (Seahorse Bioscience) was employed for conducting OCR assays. A total of 1 × 10⁴ cells were seeded in hippocampal XFp cell culture microplates and incubated in a 5 % CO₂ environment at 37 °C until cell attachment was achieved. Subsequently, the medium in the sodium citrate treatment group was replaced with a medium containing sodium citrate, and cells were cultured for an additional 12 h. Oligomycin, p-trifluoromethoxycarbonyl cyanide phenylhydrazone (FCCP), and rotenone were administered at specified time intervals. The OCR was measured using the Seahorse XF Cell Mito Stress Test kit (Seahorse Bioscience, Cat.103015-100) in accordance with the manufacturer's instructions. Subsequently, the data obtained were analyzed utilizing the Seahorse XFp Wave software.

Measurements of Ca²⁺

To visualize mitochondrial Ca²⁺ levels, 293 T cells were co-transfected with the plasmid 4mt-GCaMP6, provided by Xin Pan of Fudan University, in conjunction with packaging vectors psPAX2 and envelop plasmids pMD2.G. The viral supernatant was collected 48 h after transfection. Subsequently, SKOV3 cells were exposed to the viral supernatant and incubated at 37˚C for 24 h. Following this, the viral supernatant was replaced with fresh RPMI-1640 medium containing puromycin. Cells were subjected to 488 nm light for excitation, and the fluorescence emitted through 527 nm filters for GFP was recorded. The resulting image was then captured and analyzed using Harmony High Content Analysis (HCA) Software (Perkin Elmer).

To detect intracellular Ca²⁺, cells were treated with 1 μM of Fluo-3 AM (Beyotime, Cat. S1056) or Rhod-2 AM (MCE, Cat. HY-D0989) fluorescent probes and subsequently washed. The intracellular Ca²⁺ levels were quantified by measuring the fluorescence intensity of Fluo-3/AM and Rhod-2 AM using HCA Software.

Extracellular acidification rate (ECAR) detection

The Seahorse XFp Extracellular Flow Analyzer from Seahorse Bioscience was employed for conducting ECAR assays. A quantity of 1 × 10⁴ cells were seeded into hippocampal XFp cell culture microplates and incubated in a 5 % CO₂ environment at 37 °C until cell attachment. Subsequent to attachment, the medium in the sodium citrate treatment group was replaced with medium containing sodium citrate, and cells were cultured for an additional 12 h. Glucose, 2-deoxyglucose, and Antimycin A/Rotenone were administered at designated time points. The ECAR was measured using the Seahorse XF Glycolytic Rate Assay Kit (Seahorse Bioscience, Cat. 103344-100) following the manufacturer's instructions. Subsequently, the collected data was analyzed utilizing the Seahorse XFp Wave software.

Measurement of TG (Triglyceride) content and NEFA (Nonesterified free fatty acids) content

The cells were seeded into 6-well plates and allowed to adhere before the medium in the sodium citrate treatment group was exchanged with sodium citrate-containing medium and cultured for an additional 12 h. The levels of TG and NEFA were quantified using the Triglyceride assay kit (Nanjing Jiancheng Bioengineering Institute, Cat. A110-1-1) and Nonesterified Free fatty acids assay kit (Nanjing Jiancheng Bioengineering Institute, Cat. A042-1-1), respectively, following the manufacturer's instructions. All values were adjusted based on protein concentrations.

Immunoprecipitation

The cells were seeded into 6-well cell plates and allowed to adhere before replacing the medium in the sodium citrate treatment group with a sodium citrate-containing medium. Following removal of the culture medium and PBS washing, cell lysis was performed using IP lysis buffer (Beyotime, Cat.P0013) to extract proteins. Subsequently, 1 μg of MCU antibody (Proteintech, Cat.26312-1-AP) or NCOA4 antibody (Cell Signaling, Cat.66849) and 40 μL of protein A/G-beads (Beyotime, Cat.P2195M) were added to the lysis solution, and the mixture was incubated overnight at 4 °C with gentle agitation. Following the immunoprecipitation reaction, the protein A/G-beads underwent washing with lysis buffer. Subsequently, a mixture of 4 × SDS sample buffer and 15 μL of lysis buffer was added, and the resulting solution was subjected to boiling for 10 min. The samples were then subjected to immunoblotting.

Immunofluorescence

Cells were seeded onto slides within 24-well cell plates and subsequently fixed with 4 % paraformaldehyde for 20 min. Following fixation, the cells were permeabilized with 0.1 % Triton X-100 for 10 min and subsequently blocked with 5 % BSA for 30 min to minimize non-specific antibody interactions. The primary antibodies FTH (SANTA CRUZ, Cat. sc-376594, 1:200) and NCOA4 (Cell Signaling, Cat. 66849, 1:200) were incubated at 4 °C overnight. Subsequent to incubation, cells were rinsed with PBS to eliminate any unbound antibodies. Following this, fluorescently labeled FITC-Goat anti-Mouse IgG (Proteintech, Cat. SA00003-1, 1:100), Cy3-Goat anti-Rabbit IgG (Proteintech, Cat. SA00009-2, 1:100), and DAPI were introduced for additional incubation. The cells were then observed utilizing laser scanning confocal microscopy (Olympus) for image capture.

Non-targeted metabolomic analysis

The SKOV3 cells were cultured in 10 cm cell culture dishes, with the sodium citrate treatment group receiving a medium containing sodium citrate for 12 h post cell attachment. Following this, the cells were collected, centrifuged, and the resultant samples were rapidly frozen in liquid nitrogen prior to storage at −80 °C. Subsequently, the cell samples underwent analysis by Allwegene (Beijing, China) utilizing an ultra-high performance liquid chromatography Q Exactive mass spectrometer (UHPLC-QE-MS).

Tumor xenograft studies

The animal experiments were conducted in accordance with the guidelines set forth by the Institutional Animal Care and Utilization Committee of Northeast Normal University (NENU/IACUC, AP20210105). Female BALB/c nude mice, aged 4 to 6 weeks, were obtained from Beijing Vitong Lever Experimental Animal Technology and were inoculated subcutaneously with 5 × 10⁶ SKOV3 cells. Subsequently, the SKOV3 tumor-bearing mice were randomly divided into three groups (control group, sodium citrate treatment group, and Ca²⁺ plus sodium citrate treatment group), each consisting of 6 mice. Mice in the sodium citrate treatment group were administered a total of 4 g/kg/day of sodium citrate once daily for a duration of 4 consecutive weeks. Mice in the Ca²⁺ plus sodium citrate treatment groups were provided with 15 mg/mL of CaCl₂ in their drinking water simultaneously. The control group of mice received an equivalent volume of normal saline.

In the tumor xenograft studies evaluating the efficacy of carboplatin and sodium citrate combination treatment, SKOV3 tumor-bearing mice were randomly assigned to 4 groups (control, carboplatin treatment, sodium citrate treatment, and carboplatin and sodium citrate combination treatment), with 6 mice in each group. Mice in the carboplatin group received daily intraperitoneal injections of 50 mg/kg for a duration of 4 weeks, while mice in the sodium citrate treatment group were administered a dose of 4 g/kg/day via intragastric injection. Mice in the combination treatment group received both treatments, and mice in the control group were given an equivalent volume of normal saline. Following a 4-week treatment period, xenografts were excised from the mice and their weight and volume were evaluated. The xenograft volume was assessed using vernier calipers and determined by the formula: V = L × W² × 0.5, where L represents the length and W denotes the width of the xenograft.

Statistical analysis

All results were replicated a minimum of three times and subjected to analysis using IBM SPSS, GraphPad, and Excel software. The t-test was employed to assess variance between two data sets, while the ANOVA test was utilized for comparisons involving three or more groups. Statistical significance was denoted by *p < 0.05, **p < 0.01, and ***p < 0.001.

Ethics statement

All animal studies were conducted with approval from the Animal Research Ethics Committee of Northeast Normal University (NENU/IACUC, AP20210105) of China and performed in accordance with established guidelines.

Results

Sodium citrate induces apoptosis in ovarian cancer cells

To investigate the effect of sodium citrate on ovarian cancer, two human ovarian cancer cell lines, SKOV3 and A2780, were treated with different concentrations of sodium citrate for different time points (12 h, 24 h and 48 h). The results from MTT cell viability assay and BrdU labeling assay showed that sodium citrate potently inhibited cell viability and proliferation in a dose- and time-dependent manner (Fig. 1A, B, Supplementary Fig. S1A). Next, the effect of sodium citrate on human ovarian epithelial cells was investigated. The result suggested that, compared to HOSEpiC cells (IC50: 33.63 mM) and IOSE-80 cells (IC50: 42.44 mM), sodium citrate significantly inhibited the growth of SKOV3 and A2780 cells with an IC50 of 15.93 mM and 16.8 mM, respectively (Fig. 1C). Inhibition of cell viability and proliferation may be attributed to apoptosis induction, thus cell apoptosis was assessed by flow cytometry. The results showed that cell death was significantly increased in SKOV3 and A2780 cells after the treatment with sodium citrate (Fig. 1D). Meanwhile, TUNEL staining images revealed the similar results (Fig. 1E).

Fig. 1.

Sodium citrate induces apoptosis in ovarian cancer cells. A. The cell viability was examined by MTT assay after treatment with various concentrations of sodium citrate for 24 h in SKOV3 and A2780 cells. B. The cell viability was examined by MTT assay after treatment with 15 mM sodium citrate for various time points in SKOV3 and A2780 cells. C. The cell viability was examined by MTT assay after treatment with various concentrations of sodium citrate for 24 h in SKOV3, A2780, HOSEpiC and IOSE-80 cells. D. Measurement of cell death by Annexin V-FITC/PI staining after the treatment with various concentrations of sodium citrate for 24 h. E. Apoptosis of SKOV3 and A2780 was determined by TUNEL staining after the treatment with 15 mM sodium citrate for 24 h. Quantitative data were shown as mean ± SEM. Scale bar, 50 μm. F. Immunoblot analysis of apoptosis-related proteins was performed after treatment with 15 mM sodium citrate for 24 h in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. G-I. Glucose consumption, lactate production and pyruvate production were analyzed after treatment with 15 mM sodium citrate for 12 h in SKOV3 and A2780 cells. J-L. Expression of glycolysis-related proteins and activation of AKT/mTOR signal was analyzed by immunoblot analysis after treatment with 15 mM sodium citrate for 24 h in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. M. The cell viability was examined by MTT assay after treatment with 15 mM sodium citrate plus 20 μM Z-VAD-FMK (Z-VAD) or 100 μM deferoxamine (DFO) for 24 h in SKOV3 and A2780 cells. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used.

To elucidate the underlying mechanisms by which sodium citrate induced cell death in ovarian cancer cells, the expression of several apoptotic-related proteins was examined. The results showed that the levels of activated Caspase3, activated Caspase9 and BAX were markedly elevated while the level of anti-apoptotic protein BCL2 was decreased (Fig. 1F). Furthermore, immunofluorescent staining confirmed that the co-localization of cytochrome c and Apaf-1 was increased significantly after treatment with sodium citrate (Supplementary Fig. S1B). Based on these results, the preliminary judgment was that sodium citrate induced apoptosis of ovarian cancer cells via mitochondrial pathway.

Tumorigenesis is characterized by aerobic glycolysis, and glycolysis inhibition often induces cell apoptosis. The subsequent investigation aimed to investigate whether glycolysis inhibition was involved in sodium citrate-induced apoptosis. The functional colorimetric validation showed that glucose consumption, lactate production and pyruvate content were significantly reduced (Fig. 1G-I). Moreover, the effect of sodium citrate on the levels of the rate-limiting enzymes in the glycolytic pathway was examined. The results indicated that sodium citrate decreased both mRNA and protein expression levels of glycolysis-related proteins such as Glut1, HK2 and PFKP (Fig. 1J, Fig. Supplementary Fig. S1C, D). These results proved that sodium citrate inhibited glycolysis of SKOV3 and A2780 cells through regulating expression of metabolic enzymes. To determine whether HIF1α was responsible for reduced expression of glycolysis-related proteins after sodium citrate, its protein expression was measured. The result suggested that HIF1α expression was decreased significantly after sodium citrate treatment (Fig. 1K). In line with this, 83 common HIF1α target genes that covered all 10 enzymes of the glycolytic pathway were predicted from Database of Human Tranion Factor Targets（hTFtarget）and CHEA Transcription Factor Targets (Supplementary Fig. S1E). Furthermore, data from Fig. 1L demonstrated that phosphorylation of AKT and mTOR was notably suppressed after sodium citrate treatment. Consequently, AKT/mTOR/HIF1α signaling pathway was involved in the sodium citrate-induced glycolysis inhibition and cell apoptosis.

To confirm that sodium citrate induced apoptosis by reducing HIF1α expression, SKOV3 and A2780 cells were co-treated with sodium citrate and caspase inhibitor Z-VAD-FMK (Z-VAD) or the HIF1α activator deferoxamine (DFO). The results revealed that both Z-VAD and DFO could partially reduce sodium citrate toxicity (Fig. 1M). Surprisingly, treatment with DFO had a better rescue effect on the cytotoxicity induced by sodium citrate. Since DFO is also a Fe²⁺ chelator which can inhibit ferroptosis by chelating Fe²⁺, it was inferred that ferroptosis may be also involved in sodium citrate-induced cell death of ovarian cancer cells.

Sodium citrate induces ferroptosis in ovarian cancer cells

To determine whether ferroptosis contributes to sodium citrate cytotoxicity in ovarian cancer cells, Fe²⁺ content, LPO levels, MDA levels, ROS production and mitochondrial H₂O₂ generation were measured. The results showed that ovarian cancer cells treated with sodium citrate exhibited higher Fe²⁺ levels, LPO levels, MDA levels, ROS and mitochondrial H₂O₂ levels (Fig. 2A-E). Further results showed that shrunken mitochondria, an increase in mitochondrial membrane density and disruption of mitochondrial cristae could be observed by transmission electron microscopy (TEM) after sodium citrate treatment (Fig. 2F). As expected, the reduced glutathione (GSH) levels, GPX activity and expression levels of GPX4 were significantly reduced in SKOV3 and A2780 cells with sodium citrate treatment (Fig. 2G-I). Moreover, a significant elevation in the NADP⁺/NADPH ratio was observed with sodium citrate treatment (Fig. 2J).

Fig. 2.

Sodium citrate induces ferroptosis in ovarian cancer cells. A-C. Fe²⁺ content, LPO and MDA content were analyzed after treatment with 15 mM sodium citrate for 12 h in SKOV3 and A2780 cells. D. ROS levels were measured using DCF-fluorescence by flow cytometry. E. Representative fluorescence images of Mito-LX staining (H₂O₂ indicator). Scale bar, 50 μm. Quantitative data were shown as mean ± SEM. F. Ultrastructure images of SKOV3 cells were examined by TEM. Red arrow, mitochondria. G, H. Intracellular GSH content and GSH-Px activity were measured. I. Immunoblot analysis of GPX4 was performed after treatment with 15 mM sodium citrate. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. J. Determination of the ratios of NADP⁺/NADPH. K. The cell viability was examined by MTT assay after treatment with 15 mM sodium citrate plus 10 μM ferrostatin-1(Fer-1), 20 μM acetylcysteine (NAC) or 5 mM NADPH for 24 h in SKOV3 and A2780 cells. L. Immunoblot analysis of apoptosis-related proteins and ferroptosis-related proteins was performed after treatment with 15 mM sodium citrate for various time points in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. N = 3 biologically independent replicates. Data were presented as means mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

To further confirm that ferroptosis was involved in sodium citrate-induced cell death, cells were co-treated with sodium citrate and ferrostatin-1 (Fer-1, ferroptosis inhibitor), N-Acetylcysteine (NAC, a ROS scavenger) or NADPH. The data showed that Fer-1, NAC and NADPH significantly restored the cell viability inhibited by sodium citrate (Fig. 2K). Collectively, these results suggested that sodium citrate significantly induced ferroptosis of ovarian cancer cells. Furthermore, the expression levels of the ferroptosis-related protein GPX4 and the apoptosis-related proteins Cleaved Caspase3 and Cleaved Caspase9 were assessed following a short-term exposure of cells to sodium citrate in order to determine which pathway started first. The results suggested that both of them started almost simultaneously (Fig. 2L).

Ferroptosis induced by sodium citrate initiates via ferritinophagy

Next, more detailed mechanisms of ferroptosis induced by sodium citrate were explored. Firstly, the expression of ACSL4, PTGS2, NOX1, FSP1 and FTH1 was examined. The results showed that the expression of ACSL4 and FSP1 did not change significantly while the expression of PTGS2 and NOX1 increased slightly in the two cell lines after sodium citrate treatment (Fig. 3A). Among these, the decreased expression of FTH1 caught the attention given that the mRNA levels of FTH1 were increased (Fig. 3B, C). Since FTH1 is a substrate of ferritinophagy, it is thus reasonable to conjecture that sodium citrate may promote ferritinophagy in ovarian cancer cells.

Fig. 3.

Ferroptosis induced by sodium citrate initiates via ferritinophagy. A. Immunoblot analysis of ferroptosis-related proteins was performed after treatment with 15 mM sodium citrate for 24 h in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. B, C. mRNA expression of FTH1 gene in SKOV3 and A2780 cells was examined by qPCR after treatment with 15 mM sodium citrate. D. Immunoblot analysis of FTH1 proteins was performed after treatment with 15 mM sodium citrate plus 100 μM chloroquine (CQ) in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. E. Fe²⁺content was analyzed after the treatment with 15 mM sodium citrate and CQ in SKOV3 and A2780 cells. F. Immunoblot analysis of LC3 proteins was performed after treatment with 15 mM sodium citrate in SKOV3 and A2780 cells. G. Ultrastructure images of SKOV3 cells were examined by TEM. Arrow, autophagosome. H. Representative images of GFP–RFP-LC3 labeled SKOV3 cells to monitor autophagic flux. Scale bar, 10 μm. I. Cell lysates were immunoprecipitated using an anti-NCOA4 antibody and then subjected to immunoblot analysis using an anti-FTH1 antibody. J. SKOV3 cells were fixed and subjected to immunofluorescence analysis with indicated antibodies. Red: NCOA4; Green: FTH1. Scale bar, 10 μm. K. Immunoblot analysis of ferritinophagy-related proteins was performed after treatment with sodium citrate in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. L. The cell viability was examined by MTT assay after treatment with 15 mM sodium citrate plus100 μM CQ in SKOV3 and A2780 cells. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

To test this assumption, activation of autophagy was examined. The results indicated that the decreased FTH1 levels and increased Fe²⁺ content induced by sodium citrate were almost completely reverted upon treatment with the autophagy inhibitor Chloroquine (CQ) (Fig. 3D, E). Autophagy-related indicators were then further detected. First, immunoblot results showed that sodium citrate increased the conversion of cytosolic LC3 (LC3-I) to the lipidated form of LC3 (LC3-II) (Fig. 3F). Next, the double membraned vesicles (presumed to be autophagosomes) were observed by TEM with sodium citrate treatment (Fig. 3G). To monitor autophagic flux, stable SKOV3 cell lines that express tandem fluorescent-tagged LC3 (RFP-GFP-LC3) were constructed. As expected, an increased autophagic flux by sodium citrate was visualized by fluorescence analysis. Furthermore, the co-administration of sodium citrate and CQ resulted in the accumulation of punctate LC3 fluorescence (Fig. 3H). In a word, sodium citrate promoted ferritinophagy in ovarian cancer cells.

Further co-immunoprecipitation results showed that sodium citrate treatment resulted in higher levels of NCOA4 and increased co-localization of FTH1 with NCOA4 (Fig. 3I-K). Importantly, CQ partially restored the cell viability inhibited by sodium citrate (Fig. 3L). In conclusion, these results indicated that sodium citrate induced ferroptosis via ferritinophagy in ovarian cancer cells.

Ca²⁺ chelation by sodium citrate leads to cell death

Next, the intracellular action targets of sodium citrate were identified. Considering that sodium citrate is Ca²⁺ chelator, the cytoplasmic free Ca²⁺ was examined through fluorescent calcium indicators. The result suggested that sodium citrate treatment led to a decrease in cytoplasmic Ca²⁺ by the cell-permeable Ca²⁺ fluorescent probe (Fluo-3 AM and Rhod-2 AM) (Fig. 4A, B). This reduction of Ca²⁺ was more pronounced after histamine stimulation, and the decreased Ca²⁺ almost recovered to baseline with Ca²⁺ replenished (Fig. 4C, D).

Fig. 4.

Ca²⁺ chelation by sodium citrate leads to cell death. A, B. Representative images of Fluo-3 AM and Rhod-2 AM fluorescence. Quantitative data were shown as mean ± SEM. Scale bar, 50 μm. C, D. Changes in the Fluo-3 AM fluorescence intensity were recorded after treatment with 200 μM histamine. E. Measurement of cell death by Annexin V-FITC/PI staining after treatment with 15 mM sodium citrate plus 2 mM CaCl₂ for 24 h. F, G. Glucose, oligomycin, and 2-deoxyglucose (2-DG) were sequentially added to measure ECAR in the XFp analyzer from Seahorse Bioscience. H. Immunoblot analysis of AKT activation was performed after treatment with sodium citrate plus CaCl₂ in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. I, J. MDA and LPO content in SKOV3 and A2780 cells were analyzed following different treatments. K. Representative fluorescence images of DHE staining represented ROS levels. Scale bar, 10 μm. Quantitative data were shown as mean ± SEM. L. SKOV3 cells were fixed and subjected to immunofluorescence analysis with indicated antibodies. Red: NCOA4; Green: FTH1. Scale bar, 10 μm. M. Immunoblot analysis of ferritinophagy-related proteins was performed following different treatments in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. N. Cell lysates were immunoprecipitated using an anti-NCOA4 antibody and then subjected to immunoblot analysis using an anti-FTH1 antibody. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Based on the results above, the effect of Ca²⁺ supplementation on cell death induced by sodium citrate was detected. The results from flow cytometry, MTT assay and TUNEL staining showed that supplementation of Ca²⁺ greatly reversed the sodium citrate-induced cell death (Fig. 4E, Supplementary Fig. S2A, B). Similarly, the glucose consumption, ECAR levels and expression levels of glycolysis-related enzymes (such as Glut1, PFK and HK2) were significantly recovered by Ca²⁺ supplementation (Fig. 4F, G, Supplementary Fig. S2C, D). Additionally, Ca²⁺ supplementation remarkably reversed the decreased activation of AKT/mTOR signal (Fig. 4H).

Subsequent studies were performed to test whether Ca²⁺ supplementation could rescue sodium citrate-induced ferroptosis. The results showed that Ca²⁺ dramatically reversed the enhanced levels of MDA, LPO and ROS triggered by sodium citrate (Fig. 4I-K). Next, autophagy-related proteins were detected. The results showed that cells treated with sodium citrate plus Ca²⁺ exhibited lower expression levels of ATG5, ATG7 and NCOA4 compared with those treated with sodium citrate alone. Importantly, the FTH1 level and the co-localization of FTH1 with NCOA4 almost returned to the normal level after Ca²⁺ supplementation (Fig. 4L-N). Consistent with these findings, the abnormal mitochondrial morphology was recovered after Ca²⁺ supplementation by TEM (Fig. Supplementary Fig. S2E). Taken together, these results suggested that Ca²⁺ supplementation could prevent ovarian cancer cell apoptosis and ferroptosis induced by sodium citrate.

Ca²⁺ chelation by sodium citrate leads to mitochondrial dysfunction

To further determine the effect of sodium citrate on mitochondrial Ca²⁺ levels, a SKOV3 cell line expressing a mitochondria-localized Ca²⁺ indicator (4mt-GCaMP6) was generated. As expected, the results showed that SKOV3 cells also exhibited decreased levels of mitochondrial Ca²⁺ with sodium citrate treatment (Fig. 5A). The data from Fig. 5B suggested that expression of PDH, IDH and OGDH, which were modulated by Ca²⁺ within the mitochondrial matrix, did not exhibit significant change after treatment with sodium citrate. However, the ATP level and the activity of complex III and complex V were decreased significantly after sodium citrate treatment, and Ca²⁺ supplementation largely recovered these changes (Supplementary Fig. S3A, Fig. 5C, D). Since mitochondria are the main sources of ROS generation, mitochondrial H₂O₂ levels and mitochondrial OCR were assessed. The data indicated that cells treated with sodium citrate had higher H₂O₂ levels and lower OCR levels, while the alteration were largely recovered after Ca²⁺ supplementation (Fig. 5E-G). Together, these results demonstrated that Ca²⁺ chelation by sodium citrate induced a decrease of intracellular and mitochondrial Ca²⁺ level and excessive ROS production.

Fig. 5.

Ca²⁺ chelation by sodium citrate leads to mitochondrial dysfunction. A. Representative traces of mitochondrial Ca²⁺ dynamics (indicated by fluorescence of 4mt-GCaMP6) following histamine (200 μM) treatment in SKOV3 cells. B. Immunoblot analysis of TCA cycle enzymes was performed after treatment with sodium citrate plus CaCl₂ in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. C, D. Complex Ⅲ and complex V activity in mitochondrial lysates were detected. E. Representative fluorescence images of Mito-LX staining represented H₂O₂ levels. Scale bar, 10 μm. Quantitative data were shown as mean ± SEM. F, G. Oligomycin, FCCP, antimycin and rotenone were sequentially added to measure OCR in the XFp analyzer from Seahorse Bioscience. H. Immunoblot analysis of CAMKK2 and AMPK activation was performed following different treatments in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. I, J. Cell lysates were immunoprecipitated using an anti-MCU antibody and then subjected to immunoblot analysis using an anti-phospho-serine/threonine antibody and anti-AMPK antibody. K, L. The cell viability was examined by MTT assay after treatment with 15 mM sodium citrate plus 8 μM Ionomycin (Iono) or 150 μM STO-609 for 24 h in SKOV3 and A2780 cells. M, N. MDA and LPO content in SKOV3 and A2780 cells were analyzed following Ionomycin treatment. O. Immunoblot analysis of GPX4 and FTH1 level was performed following Ionomycin treatment in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used.

Next, the mechanism by which the decreased Ca²⁺ level regulates cell death was investigated. Firstly, the effect of sodium citrate on the activation of Ca²⁺/CAMKK2/AMPK signaling pathway was analyzed. As expected, the level of phosphorylation of CAMKK2 and AMPK was significantly reduced after sodium citrate treatment (Fig. 5H). While the activity of LKB1, another upstream kinase of AMPK, was not affected by sodium citrate (Supplementary Fig. S3B). Secondly, AMPK activator (AICAR) significantly reduced sodium citrate-induced cell death, indicating that Ca²⁺/CAMKK2/AMPK pathway played a critical role in this process (Supplementary Fig. S3C, D). Moreover, the way by which sodium citrate decreased mitochondrial Ca²⁺ level was tested. It was found that the phosphorylated MCU, a mitochondrial Ca²⁺ uniporter, was dramatically decreased after sodium citrate treatment (Fig. 5I, J). Co-immunoprecipitation assays further revealed that the association between AMPK and MCU was markedly reduced and Ca²⁺ supplementation successfully rescued this effect (Fig. 5I, J). These data suggested that sodium citrate treatment decreased the Ca²⁺ uptake by mitochondria through regulating the activity of MCU, causing mitochondrial dysfunction and excessive ROS.

To further confirm that CAMKK2 played a crucial role in the sodium citrate-mediated ferroptosis, cells were co-treated with sodium citrate and Ionomycin (a CAMKK2 activator) or STO-609 (a CAMKK2 inhibitor). The data indicated that Ionomycin effectively restored cell viability that had been inhibited by sodium citrate, whereas STO-609 exacerbated the decrease in cell viability. (Fig. 5K, L). Moreover, Ionomycin successfully reversed the enhanced levels of MDA and LPO as well as the decreased levels of FTH1 and GPX4 triggered by sodium citrate. These results suggested that CAMKK2 inhibition was required for sodium citrate-induced cell death. (Fig. 5M-O). In a word, sodium citrate remarkably inhibited the activation of Ca²⁺/CAMKK2 pathway, thereby inhibiting AKT and AMPK signal, causing apoptosis and ferroptosis respectively.

Ca²⁺ chelation by sodium citrate drives metabolic reprogramming

Metabolomics data analysis based on UHPLC-QE-MS was performed to explore the influence of sodium citrate-induced Ca²⁺ chelation on the metabolism of ovarian cancer cells. The relative abundance of metabolites was analyzed and the results were listed in Supplementary Table 2. The 2-dimensional PCA plot using all available metabolomic data revealed three discrete clusters (Fig. 6A). In detail, it showed a distinct separation between the control and the sodium citrate groups, while the Ca²⁺ plus sodium citrate group was close to the control group. Likewise, heat map cluster showed the same results (Fig. 6A, B, Supplementary Fig. S4A, B). The Venn plot of differential metabolites visually showed the number of differential metabolites unique or common between comparison groups. Most of the significantly different metabolites in sodium citrate vs control groups overlapped with those in Ca²⁺ plus sodium citrate vs sodium citrate groups (Fig. 6C, D).

Fig. 6.

Ca²⁺ chelation by sodium citrate drives metabolic reprogramming. A. Principal component analysis (PCA) was performed. B. Heatmap of hierarchical clustering of all differential metabolites in all samples. C, D. Venn diagrams showing differential metabolites shared or unique among the different comparisons. E. Differential metabolite screening results presented using volcano map. Each point in the volcano map represents a metabolite, in which the metabolites with significantly up-regulated expression are represented by red dots, and those with significantly down-regulated expression are represented by green dots. F, G. Structure-based clustering analysis of up-regulated and down-regulated metabolites. H, I. Pathway-based clustering analysis of up-regulated and down-regulated metabolites. J. Ultrastructure images of SKOV3 cells were examined by TEM. Arrow, lipid droplet. K, L. NEFA and TG content was analyzed after treatment with 10 mM sodium citrate for 12h in SKOV3 and A2780 cells. M. The cell viability was examined by MTT assay after treatment with 15 mM sodium citrate plus 200 mM HCA for 24h in SKOV3 and A2780 cells. N, O. SKOV3 and A2780 cells were transfected with negative si-NC, FASN siRNA (si-FASN-1) and FASN siRNA (si-FASN-2) and then treated with sodium citrate. The cell viability was examined by MTT assay. P. Immunoblot analysis of phosphorylated ACC (p-ACC), total ACC, phosphorylated AMPK (p-AMPK) and total AMPK proteins was performed following different treatments in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. Q-U. The levels of α-ketoglutarate, malate, oxaloacetate, succinate and fumarate were compared in control and different treated groups. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used.

218 significantly different metabolites were obtained between the sodium citrate group and the control group by volcano plot analysis (Fig. 6E, Supplementary Fig. S4C, Supplementary Table 3). Chemical structure and pathway-based clustering analysis of up-regulated and down-regulated metabolites were performed by online MetaboAnalyst 5.0 software (Fig. 6F-I). Specifically, the Small Molecule Pathway Database (SMPDB) analysis revealed that the up-regulated metabolites were mainly enriched in pathways associated with biosynthesis of phospholipid. Among them, phosphatidylethanolamine (PE) was the main substrate for lipid peroxidation in ferroptosis. Data from TEM exhibited an increased lipid droplets in the SKOV3 cells treated with sodium citrate (Fig. 6J). Consistent with this finding, NEFA and TG were elevated after sodium citrate treatment (Fig. 6K, L). It was found that both ACLY inhibitor hydroxycitric acid (HCA) and siRNA-mediated FASN knockdown partially reversed the increased levels of NEFA and TG induced by sodium citrate (Fig. 6M-O, Supplementary Fig. S4D-K). Moreover, the phosphorylation of ACC regulated by AMPK was significantly reduced after sodium citrate treatment, and Ca²⁺ supplementation remarkably rescued this effect (Fig. 6P). In contrast, AMPK activator AICAR greatly reduced the production of NEFA and TG induced by sodium citrate (Supplementary Fig. S4L-O). Thus, these results collectively suggested that the lipotoxicity induced by sodium citrate participated in aggravating ferroptosis.

Additionally, the down-regulated metabolites were mainly organic acids. Of note, TCA intermediates, such as α-ketoglutarate, malate, oxaloacetate, succinate and fumarate were markedly decreased in sodium citrate treatment groups relative to control groups (Fig. 6Q-U). These results indicated that mitochondria were damaged after sodium citrate treatment, which was consistent with results from Fig. 5E-G. Altogether, these results demonstrated that sodium citrate treatment resulted in increased phospholipid synthesis and disordered cell metabolism, thereby inducing ferroptosis.

Sodium citrate suppresses tumor growth in xenograft mouse models and improves anti-cancer efficacy of carboplatin

In order to study the effect of sodium citrate on tumor growth in vivo, SKOV3 cells were subcutaneously injected into nude mice to observe tumor progression. It was observed that sodium citrate treatment significantly suppressed tumor growth at week 4 compared to the control group (Fig. 7A-D). As expected, sodium citrate plus Ca²⁺ treatment restored tumor growth attenuated by sodium citrate (Fig. 7A-D), suggesting that sodium citrate inhibited tumor growth by chelation of Ca²⁺ in vivo. Furthermore, the expression of CD31 in tumor was analyzed by immunofluorescence, and the result suggested that there were no significant differences in CD31 staining in tumors between the control and those treated with sodium citrate. Therefore, sodium citrate directly targeted ovarian cancer cells instead of endothelial cells in this study (Fig. 7E).

Fig. 7.

Sodium citrate suppresses tumor growth in xenograft mouse models and improves anti-cancer efficacy of carboplatin. A, B. Photography of the xenograft tumor-bearing mice injected with SKOV3 cells following different treatments. C, D. Tumor weight and volume for each group. E. Representative images of CD31 immunofluorescence analysis in tumors. Red: CD31; Blue: DAPI. Scale bar, 50 μm. Quantitative data were shown as mean ± SEM. F. Plate cloning formation assay showing cell proliferation of SKOV3 and A2780 cells following different treatments. G, H. Photography of the xenograft tumor-bearing mice injected with SKOV3 cells following different treatments. Ctrl, treated with 0.9 % NaCl; CBP, treated with carboplatin; SCT, treated with sodium citrate; CBP + SCT, treated with carboplatin and sodium citrate. I, J. Tumor weight and volume for each group. N = 6 in each group. Data were presented as mean ± SEM.*P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Next, sodium citrate was tested for its ability to potentiate the sensitivity of chemotherapy drugs used in clinical treatment of OC. OC cells were treated with sodium citrate in combination with different kinds of chemotherapy drugs. The results showed that sodium citrate increased the sensitivity of OC cells to chemotherapy drugs (carboplatin, cisplatin, doxorubicin hydrochloride and vincristine) (Fig. 7F, Supplementary Fig. S5A-H).

Ultimately, an investigation was conducted to determine if the effectiveness of carboplatin could be enhanced through the co-administration of sodium citrate in an in vivo setting. As shown in Fig. 7G-J, both sodium citrate and carboplatin inhibited xenograft tumor growth individually, and sodium citrate significantly enhanced the sensitivity of ovarian cancer to carboplatin. Altogether, these data provided a new perspective on the ovarian cancer treatment.

Conclusion

This study demonstrated that sodium citrate decreased free Ca²⁺ level in the cytosol and mitochondria by chelating Ca²⁺, and the decreased Ca²⁺ level resulted in inhibition of CAMKK2 activity. On one hand, the reduced CAMKK2 activity by sodium citrate greatly inhibited the AKT/mTOR/HIF1α-mediated glycolysis pathway, thereby inducing cell apoptosis. On the other hand, the reduced CAMKK2 activity significantly suppressed AMPK signal. Inhibition of AMPK not only increased NCOA4-mediated ferritinophagy and intracellular Fe²⁺ levels but also decreased the activity of MCU and Ca²⁺ level in the mitochondria, leading to increased mitochondrial ROS. In addition, AMPK inhibition also increased de novo lipid synthesis. Together, these factors induced by sodium citrate contributed to the ferroptosis. Furthermore, sodium citrate significantly inhibited growth of subcutaneously xenografted ovarian tumors and enhanced the sensitivity of ovarian cancer cells to chemotherapy drugs in vivo. Collectively, this study not only has practical value in clinical applications but also has far-reaching implications for understanding the mechanism of ovarian cancer and exploring more effective treatment strategies.

Discussion

Understanding the mechanisms underlying tumor cell death is essential for the progression of cancer research and the improvement of treatment outcomes. Historically, the study of caspase-mediated apoptosis has been key in understanding cancer cell death and developing cancer treatments. The efficacy of current anti-cancer drugs is significantly hindered by the tumor cells' intrinsic or developed ability to resist apoptosis. This resistance can be attributed to various factors, including the presence of multi-drug resistance and heightened activity of anti-apoptotic proteins. Therefore, there is an urgent need to find new ways to kill tumor cells that are different from apoptosis to improve current treatment options.

Apoptosis and ferroptosis are two distinct types of cell death mechanisms that play important roles in maintaining cellular homeostasis and regulating tissue development and function. In present study, it was found that sodium citrate not only promoted ovarian cancer cells apoptosis but also triggered ferroptosis. To determine whether one pathway initiated prior to the other, or if both pathways were activated simultaneously, ovarian cancer cells were subjected to brief exposure to sodium citrate, followed by the detection of the onset of cell apoptosis and ferroptosis. The data obtained demonstrated that apoptosis and ferroptosis stated almost simultaneously (Fig. 2L). In summary, these data suggested that both apoptosis and ferroptosis were crucial in sodium citrate-induced cell death. In this study, apoptosis and ferroptosis may occur independently of each other, with different triggers and regulatory pathways driving each process after treatment in ovarian cancer cells.

After analyzing the mechanism of sodium citrate-induced cell death in ovarian cancer cells, it was concluded that the observed effects were due to the Ca²⁺ chelating properties of sodium citrate. Sodium citrate treatment led to a decrease in cytosolic Ca²⁺ levels in ovarian cancer cells, which on one hand inhibited the CAMKK2/AKT/mTOR/HIF1α-dependent glycolysis pathway, thereby inducing cell apoptosis. On the other hand, the inhibition of CAMKK2 by sodium citrate inactivated AMPK, leading to reduced mitochondrial Ca²⁺ levels and an abnormal surge in mitochondrial ROS. Moreover, inhibition of CAMKK2/AMPK pathway increased intracellular Fe²⁺ levels through promoting ferritinophagy and lipid synthesis by inhibiting ACC activity. These factors ultimately promoted ferroptosis of ovarian cancer cells (Fig. 8). This study presented a novel contribution to the existing literature by identifying sodium citrate as a dual inducer of apoptosis and ferroptosis, and offered a comprehensive analysis of the underlying mechanisms. The study validated the significance of the Ca²⁺/CAMKK2/AMPK pathway, which was associated with mitochondrial dysfunction, increased Fe²⁺ levels and enhanced lipid synthesis. These findings indicated that targeting metabolic pathways in ovarian cancer cells with sodium citrate held great promise as a therapeutic strategy.

Fig. 8.

A schematic of the network in which sodium citrate induces apoptosis and ferroptosis in ovarian cancer cells. A schematic of the network in which sodium citrate induces apoptosis and ferroptosis in ovarian cancer cells.

In this study, it was determined that activation of the CAMKK2/AMPK pathway by sodium citrate led to an increase in intracellular Fe²⁺ levels by promoting ferritinophagy. However, there have been no reports about the direct link between CAMKK/AMPK signaling and NCOA4-mediated ferritinophagy. CAMKK/AMPK signaling is crucial for controlling cellular energy and metabolism, which can impact iron homeostasis through the regulation of iron-related protein expression. Given that NCOA4-mediated ferritinophagy was involved in iron homeostasis, it's plausible that CAMKK/AMPK signaling could influence NCOA4 levels or the process of ferritinophagy through its effects on iron metabolism. For example, CAMKK/AMPK signaling could modulate the expression of HERC2, the E3 ubiquitin ligase that mediates NCOA4 degradation in iron-replete conditions, thereby indirectly influencing NCOA4-mediated ferritinophagy. Further research is needed to elucidate the specific pathways and mechanisms involved in this regulation.

CAMKK2 was found to play a central role in sodium citrate-induced ovarian cancer cell death in this study. Previous research has indicated that CAMKK2 is prominently expressed in high-grade serous ovarian cancer (HGSOC), suggesting the potential efficacy of inhibiting ovarian cancer cells by targeting the CAMKK2/AMPK signaling pathway [48]. Although the important role of CAMKK2 in sodium citrate-induced ovarian cancer cell death has been confirmed by using the CAMKK2 inhibitor in this study, knockdown and overexpression of CAMKK2 protein are needed to further confirm this result in future study.

MCU is a key component of the inner mitochondrial membrane (IMM) that facilitates the uptake of mitochondrial Ca²⁺. This study demonstrated that sodium citrate reduced mitochondrial Ca²⁺ by suppressing the activity of MCU. The activation of MCU by AMPK is achieved through increasing its phosphorylation, and the phosphorylated MCU is then translocated into the mitochondria. The data in Fig. 5 indicated that reduced mitochondrial Ca²⁺ levels inhibited Ca²⁺-dependent respiratory chain complexes III and V in mitochondria, resulting in increased ROS production and enhanced ferroptosis. However, there are studies that have shown that inhibition of mitochondrial TCA cycle or electron transfer chain (ETC) mitigated ferroptosis [49]. The oxidative respiratory chain plays a significant role in the generation of ROS, and a reduction in mitochondrial respiration may result in a corresponding decrease in ROS production. However, severe damage to the mitochondrial respiratory chain following sodium citrate treatment can lead to a substantial increase in ROS levels, ultimately causing mitochondrial dysfunction. It is important to recognize that the inhibitory impact of sodium citrate on mitochondria offers novel insights into potential anti-tumor strategies.

It was found that biosynthesis of phospholipid was significantly elevated after sodium citrate treatment, and AMPK-mediated inhibition of ACC was involved in this process (Fig. 6K, L). It is noteworthy that sodium citrate is a substrate for ATP citrate lyase (ACLY), which plays a crucial role in facilitating the conversion of citrate to acetyl-CoA and biosynthesis of phospholipid. Therefore, elevated levels of sodium citrate could potentially increase lipid metabolism in ovarian cancer cells. The studies have uncovered a partial mechanism behind sodium citrate-induced cell death in ovarian cancer cells, it is likely that sodium citrate targets multiple pathways given its role as an intermediary in cellular metabolism.

In conclusion, these findings suggest potential applications of sodium citrate in ovarian cancer treatment and this work will encourage further translational studies in the clinical setting.

CRediT authorship contribution statement

Yulun Wu: Investigation, Methodology, Writing – original draft. Chaoran Jia: Investigation, Methodology, Validation. Wei Liu: Data curation, Software, Formal analysis. Wei Zhan: Formal analysis, Data curation. Yao Chen: Formal analysis, Validation. Junlin Lu: Software, Data curation. Yongli Bao: Project administration, Resources. Shuyue Wang: Resources, Methodology. Chunlei Yu: Supervision, Methodology. Lihua Zheng: Supervision, Funding acquisition. Luguo Sun: Project administration, Conceptualization. Zhenbo Song: Writing – review & editing, Funding acquisition, Project administration.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

The following are the Supplementary data to this article:

Supplementary Fig. 1.

Sodium citrate induces apoptosis in ovarian cancer cells. A. The BrdU absorbance of SKOV3 and A2780 was detected after treatment with various concentrations of sodium citrate for 12 h. B. SKOV3 and A2780 cells were fixed and subjected to immunofluorescence analysis with indicated antibodies. Red: Cytochrome c; Green: APAF-1. Scale bar, 10 μm. C, D. mRNA expression of glycolysis-related genes in SKOV3 and A2780 cells was measured by qPCR after treatment with 15 mM sodium citrate for 12 h. E. Venn diagrams showed shared target genes between 699 and 314 HIF1A target genes predicted from Database of Human Tranion Factor Targets (hTFtarget) and CHEA Transcription Factor Targets respectively. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P< 0.01, ***P< 0.001. The t-test and one-way ANOVA statistical analysis were used.

Supplementary Fig. 2.

Ca²⁺chelation by sodium citrate leads to apoptosis.A. The cell viability was examined by MTT assay after supplement of 2 mM CaCl₂ in SKOV3 and A2780 cells. B. Apoptosis of SKOV3 and A2780 cells was determined by TUNEL staining after supplement of CaCl₂ for 24 h. Scale bar, 50 μm. C. Glucose consumption was analyzed following different treatments in SKOV3 and A2780 cells. D. Immunoblot analysis of glycolysis-related proteins was performed following different treatments. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. E. Ultrastructure images of SKOV3 cells were examined by TEM. Red arrow, mitochondria. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used.

Supplementary Fig. 3.

Sodium citrate inhibits the activation of Ca²⁺/CAMKK2/AMPK pathway. A. ATP levels were determined following different treatments in SKOV3 and A2780 cells. B. Immunoblot analysis of phosphorylated LKB1 (p-LKB1) and total LKB1 proteins was performed following different treatments in SKOV3 and A2780 cells. Expression of proteins was normalized to β-actin. Quantitative data were shown as mean ± SEM. C, D. The effect of AICAR on decreased cell viability induced by sodium citrate was examined by MTT assay in SKOV3 and A2780 cells. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used.

Supplementary Fig. 4.

Ca²⁺chelation by sodium citrate drives metabolic reprogramming.A, B. Heatmap of hierarchical clustering of all differential metabolites. C. Differential metabolite screening results presented using volcano map. Each point in the volcano map represents a metabolite, in which the metabolites with significantly up-regulated expression are represented by red dots, and those with significantly down-regulated expression are represented by green dots.D-G. NEFA and TG content was analyzed in SKOV3 and A2780 cells transfected with FASN siRNA (si-FASN-1) and FASN siRNA (si-FASN-2) after being treated with sodium citrate. H-O. NEFA and TG content were analyzed in SKOV3 and A2780 cells after treatment with 15 mM sodium citrate plus 200 mM HCA or 2 mM AICAR for 12 h. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P< 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used.

Supplementary Fig. 5.

Sodium citrate increases the sensitivity of chemotherapy drugs. A-H. The SKOV3 and A2780 cells were incubated with sodium citrate and sodium citrate plus different chemotherapy drugs and MTT assay was performed 24 h after treatment. CBP, carboplatin; DDP, cisplatin; Vin, vincristine; DOX, doxorubicin hydrochloride. N = 3 biologically independent replicates. Data were presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. The t-test and one-way ANOVA statistical analysis were used.