Cadmium induces ferroptosis in mouse spermatocytes by activating the ROS–TCA pathway

Abstract

Cadmium (Cd) can cause testis toxicity, and we have demonstrated Cd induced ferroptosis in testis. However, the underlying toxic mechanism has not yet been fully elucidated. Here, we performed in vitro experiments on a mouse spermatocyte cell line. GC–2spd cells were divided into control, Cd, and Cd+ferroptosis inhibitor groups and cultured in high-glucose DMEM for 36 h. We conducted metabolome analysis, RNA sequencing, western blot, and immunofluorescence on GC-2spd cells to determine whether Cd exposure induced ferroptosis in spermatocyte and explore the potential mechanism. The results showed Cd exposure significantly decreased cell viability. Cd exposure significantly decreased GPX4 expression but increased malondialdehyde, mitochondrial ROS, succinate, and α-ketoglutarate contents, as well as FTH1, SLC40A1, Nrf2, Ho–1, and pyruvate carboxylase expression. Ferroptosis inhibitors (deferoxamine and liproxstatin-1) partly attenuated these effects. These findings indicate that Cd exposure directly damages mitochondria and promotes excessive ROS production, causing paradoxical activation of the mitochondrial TCA cycle, which enhances ROS production and triggers ferroptosis. This study elucidates the mechanisms of Cd-induced ferroptosis in spermatocytes and provides support for future research into the impacts of Cd on the mitochondrial TCA cycle.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-38827-7.

Introduction

Cadmium (Cd) is an environmental toxicant that exerts detrimental impacts on male reproductive health, particularly by inducing spermatogenic abnormalities^1,2. Epidemiological and experimental studies have consistently emphasized the role of Cd in disrupting testicular function and sperm quality^3,4. Cross-sectional studies on human populations have revealed a strong correlation between Cd exposure and impaired spermatogenesis^3,5. Occupational cohorts with prolonged Cd exposure exhibit elevated urinary Cd levels, which are negatively correlated with sperm motility, sperm count, and sperm morphology⁶. Notably, non-occupational populations in Cd-polluted regions demonstrate increased sperm DNA fragmentation and reduced semen quality, even at environmental exposure levels^7,8. Thus, Cd exhibits dose-dependent toxicity, with potential mechanisms involving oxidative stress and endocrine disruption.

Animal experiments have helped elucidate the pathological pathways of Cd toxicity. In rodent models, chronic Cd exposure induces testicular atrophy, disrupts blood–testis barrier integrity, and triggers germ cell apoptosis via mitochondrial dysfunction and caspase-3 activation^4,9. Cd accumulation in Leydig cells suppresses testosterone synthesis by downregulating steroidogenic enzymes (e.g., 3β-HSD), which exacerbates spermatogenic failure^10,11. Additionally, epigenetic modifications—such as altered DNA methylation in spermatogonial stem cells and ovarian granulosa cells—have been implicated in transgenerational reproductive toxicity and female reproductive toxicity^12,13. Cd can induce testicular inflammation by mediating genomic DNA release into the cytoplasm, which activates the AIM2 inflammasome and leads to pyroptosis¹⁴. Furthermore, paternal Cd exposure has been shown to affect metabolic reprogramming in offspring¹⁵. Under external stress conditions, the survival ability of cells can be restored by regulating metabolic reprogramming¹⁶. These insights highlight the multifaceted toxicity of Cd. Cd cause damage to male reproduction by activating oxidative stress, triggering inflammation, affecting the secretion of steroid hormones, competing for targets with trace elements, influencing immune regulation and epigenetics. However, further research is needed to identify biomarkers and therapeutic interventions.

Previously, we reported abnormal iron metabolism in Cd-induced ferroptosis in testicular cells¹. Ferroptosis is characterized by excess reactive iron, excessive reactive oxygen species (ROS) production, and abnormal lipid metabolism^17–19. Unlike apoptosis or necrosis, ferroptosis is mechanistically linked to iron metabolism and redox imbalance, in which excessive ROS production overwhelms the capacity of antioxidant systems, leading to cell demise²⁰. Central to this process is the iron-mediated fenton reaction, where Fe²⁺ reacts chemically with hydrogen peroxide to produce excess hydroxyl radicals (•OH), exacerbating cell damage²¹. Ferroptosis inhibitors such as deferoxamine, liprostatin-1, and natural compounds (e.g. quercetin) alleviate ferroptosis by inhibiting ROS production²².

ROS has a dual function in ferroptosis: physiological ROS levels are critical for signaling, but pathological ROS accumulation triggers oxidative damage²³. For example, impaired mitochondrial metabolism, through disruption of the tricarboxylic acid (TCA) cycle or electron transport chain (ETC), exacerbates mitochondrial ROS (mtROS) generation, further depleting GPX4 activity and promoting ferroptosis²⁴. Mitochondrial dysfunction, particularly in the TCA cycle, is increasingly recognized as a critical driver of ROS-mediated cellular damage culminating in ferroptosis²⁵. The TCA cycle, a central hub for energy production and redox homeostasis, generates NADH and FADH₂ to fuel the ETC. Disruption of this cycle, whether through enzymatic deficiencies, substrate limitations, or metabolic rewiring in pathological states, leads to electron leakage from the ETC, resulting in excessive superoxide (O₂⁻) production^26,27. Subsequent dismutation of O₂⁻ into hydrogen peroxide and its interaction with labile iron via the fenton reaction generates •OH and further initiates lipid peroxidation, which is a hallmark of ferroptosis²⁸.

Emerging evidence indicates that impairing the TCA cycle exacerbates mtROS generation; excessive mtROS overwhelms cellular antioxidant defenses and then damage cells²⁹. For example, glutaminolysis—a process that feeds the TCA cycle using α-ketoglutarate derived from glutamine—can amplify mtROS under nutrient-stressed conditions, sensitizing cells to ferroptosis³⁰. Similarly, defects in enzymes such as aconitase or succinate dehydrogenase disrupt the TCA cycle flux, promoting ROS accumulation and subsequent ferroptosis³¹. Notably, the metabolic byproducts of the TCA cycle modulate ferroptosis sensitivity by altering the redox balance and mitochondrial membrane integrity^32,33. An in vivo study reported that low-dose Cd exposure caused pulmonary disorders in mice by disrupting the mitochondrial TCA cycle and inducing lipid accumulation³⁴. Moreover, Cd induced renal toxicity in rats by reducing enzyme activity in the TCA cycle and mitochondrial ETC and triggering oxidative stress³⁵.

However, the relationship between Cd-induced ferroptosis and the TCA cycle has not been fully explored. In this study, we hypothesize that Cd induces ferroptosis in the GC-2spd mouse cell line by affecting the mitochondrial TCA cycle and triggering excessive mtROS production. We will conduct in vitro experiments on spermatocytes to detect ferroptosis-related markers and TCA cycle intermediates following Cd exposure, in order to clarify the relationship between Cd-induced ferroptosis and the TCA cycle. This study provides theoretical support for further research on Cd-induced ferroptosis in spermatocytes.

Results

Effects of Cd on cell viability and cell proliferation

Cell viability was performed using a CCK8 assay kit. Treatment with 10 µM of Cd for 36 h led to the greatest decrease in cell viability to 45% (Supplementary Fig. 1a, 1b). Compared to cells treated with Cd alone, those in the Cd+ deferoxamine and Cd+liproxstatin-1 groups exhibited a marked concentration-dependent increase in viability, with 2 µM identified as the optimal dose (Supplementary Fig. 1c). Cd treatment significantly reduced proliferating cell nuclear antigen (PCNA) expression compared to the control group (p < 0.05), whereas Cd+liproxstatin-1 treatment partly reversed this trend, although the difference from the control group was not significant (Fig. 1c, p > 0.05). PCNA expression was also higher in the Cd+deferoxamine group than that in the Cd group; however, this difference was not significant (Fig.1c, p > 0.05).

Fig. 1.

Cd inhibited cell proliferation and induced ferroptosis in GC-2spd cells. (a) Fe²⁺ content; (b) malondialdehyde (MDA) content; (c) Proliferating nuclear antigen (PCNA) expression; (d) GPX4 protein expression; (e) FTH1 protein expression; and (f) SLC40A1 expression among the experimental groups. *p < 0.05, ** p < 0.005, *** p < 0.0005.

Cd induced ferroptosis in GC-2spd cells

Cd treatment significantly increased cellular Fe²⁺ concentration (p < 0.05, Fig. 1a), significantly decreased GPX4 expression (p < 0.05, Fig. 1d), and significantly increased the protein expressions of ferritin heavy-chain 1 (FTH1) and SLC40A1 (p < 0.05, Fig. 1e and f) and contents of malondialdehyde (MDA) (p < 0.05, Fig. 1b). Moreover, Cd treatment significantly increased the levels of several antioxidant enzymes, including Nrf2 and HO-1 (p < 0.05, Fig. 2b–d), but did not alter the expression of cytoplasmic Nrf2 (p > 0.05, Fig. 2a). According to the mitochondrial morphology observed using 3D Cell Explorer, healthy GC-2spd cells exhibited long filamentous mitochondria, whereas Cd exposure led to increased mitochondrial fragmentation (p < 0.05; Fig. 2e).

Fig. 2.

Cd causes mitochondrial damage and over activates the oxidative enzyme system. (a) Cytosolic Nrf2 protein expression; (b) cytosolic phosphorylated Nrf2 expression; (c) nuclear Nrf2 protein expression; and (d) HO–1 expression among the experimental groups. (e) Mitochondrial morphology observed using 3D Cell Explorer. *p < 0.05, ** p < 0.005, *** p < 0.0005, **** p < 0.0001.

Deferoxamine and liproxstatin-1 partially rescued Cd-induced ferroptosis

Compared to Cd treatment alone, Cd plus deferoxamine or liproxstatin-1 treatment induced a significant decrease in the cellular Fe²⁺ concentration (p < 0.05; Fig. 1a) and reduced the MDA content (p < 0.05; Fig. 1b), although with no significant difference from the control group (p > 0.05). FTH1 expression was significantly decreased after treatment with Cd plus deferoxamine or liproxstatin-1 (p < 0.05; Fig. 1e), reverting to close to that in the control group (Fig. 1e). The addition of liproxstatin-1 significantly increased GPX4 expression from that in the Cd group (p < 0.05; Fig. 1d); deferoxamine treatment also increased GPX4 expression from that in the Cd group, although the difference was not significant (p > 0.05; Fig. 1d). Conversely, deferoxamine significantly decreased SLC40A1 expression from that in the Cd group, whereas liproxstatin-1 did not have a significant effect (p < 0.05; Fig. 1f). Finally, compared with Cd group, deferoxamine and liproxstatin-1 significantly decreased the expression of nuclei Nrf2, phosphor-Nrf2, and HO-1 expression in GC-2spd cells (p < 0.05; Fig. 2a–d).

RNA sequencing (RNA–seq) results

RNA- seq analysis identified 2,677 differentially expressed genes between the control and Cd groups, of which 1,954 genes were upregulated and 723 were downregulated (Fig. 3). Table 1 reveals the main terms and signaling pathways of the differentially expressed genes obtained by Gene Ontology and KEGG enrichment analyses.

Fig. 3.

RNA sequencing (RNA-Seq) analysis results. (a) Clusters of differentially expressed genes. b) Volcano map of differentially expressed genes. (c) Top 10 Gene Ontology terms of differentially expressed genes. (d) Top 10 KEGG signaling pathways of differentially expressed genes (Kanehisa et al., 2000; Kanehisa et al., 2016).

Table 1.

Top 10 significantly upregulated/downregulated genes and top 10 terms according to gene ontology and KEGG enrichment analyses of RNA sequencing data.

Upregulated genes	Downregulated genes	Gene Ontology enrichment analysis	KEGG enrichment analysis
Hspa1a	Sgce	Cellular response to chemical stimulus	Drug metabolism-cytochrome P450
Mt2	Proz	System development	GSH metabolism
Hmox1	Ahsg	Multicellular organism development	Fluid shear stress and atherosclerosis
Prl2c2	Gnat1	Regulation of multicellular organismal process	Metabolism of xenobiotics by cytochrome P450
Calml4	Ppp1r3d	Developmental process	Hepatocellular carcinoma
Acod1	Fyb2	Anatomical structure development	Cell adhesion molecules
Mgp	Sirpa	Anatomical structure morphogenesis	Longevity regulating pathway - worm
Bmpr1b	Wnt2b	Animal organ development	Hypertrophic cardiomyopathy
Tssk6	Cep85l	Protein binding	Platinum drug resistance
Fut1	Alkal1	Response to external stimulus	Pathways in cancer

Metabolomics analysis

Metabolomic analysis revealed 352 differential metabolites between the control and Cd groups. In these differential metabolites, 296 metabolites were upregulated, and 56 metabolites were downregulated (Fig. 4). Table 2 reveals the main terms and signaling pathways of the differentially expressed genes analyzed by Gene Ontology and KEGG enrichment.

Fig. 4.

Metabolomics analysis results. (a) Clusters of differentially expressed genes. (b) Volcano map of differentially expressed genes. (c) Top 30 KEGG signaling pathways of differentially expressed genes (Kanehisa et al., 2000; Kanehisa et al., 2016).

Table 2.

Top 10 terms according to gene ontology and KEGG enrichment analyses of metabolomics data.

Gene Ontology enrichment analysis	KEGG enrichment analysis
Cimifugin	Biosynthesis of plant secondary metabolites
N-myristoylsphinganine	Biosynthesis of alkaloids derived from histidine and purine
Coniferin	Fatty acid biosynthesis
Heptanoic acid	Metabolic pathways
Siduron	Purine metabolism
Curvulin5-(3-(2-(7-chloroquinolin-2-yl) ethenyl) phenyl) -8-dimethylcarbamyl-4,6dithiaoctanoicacidinosine	Beta-alanine metabolism
Oxadiazon	Morphine addiction
C17-sphinganine	Monobactam biosynthesis
2-phosphoglycerate	AMPK signaling pathway
L-homocystine	Insulin resistance

Combined analysis of RNA–seqand metabolomics analysis

Metabolome and RNA–seq association analysis identified 194 KEGG pathways that are common to both genes and metabolites, and a total of 1181 candidate genes and 206 metabolites with pathway annotations were obtained. The top 10 terms of KEGG pathway includes glutathione metabolism, CGMP-PKG signaling pathway, ferroptosis, CAMP signaling pathway, pyruvate metabolism, fatty acid biosynthesis, PI3K-Akt signaling pathway, AMPK signaling pathway, citrate cycle (TCA cycle), and Oxidative phosphorylation (Fig. 5).

Fig. 5.

Association analysis of RNA-Seq and metabolomics data. (a) Load diagram of differentially expressed genes common to RNA-Seq and metabolomics results. (b) Top 10 KEGG signaling pathways of differentially expressed genes.

Cd induced MtROS production and enhanced the TCA cycle

Cd exposure induced excess mtROS production in GC-2spd cells (p < 0.05; Fig. 6) and significantly decreased the mitochondrial membrane potential (MMP) in GC-2spd cells (p < 0.05; Fig. 7). Conversely, deferoxamine and liproxstatin-1 treatment significantly reduced mtROS levels (p < 0.05), although the difference from the control group was not significant (Fig. 6). Compared with Cd-only treatment, deferoxamine and liproxstatin-1 treatment significantly increased the MMP (p < 0.05), although to a level significantly lower than that in the control group (p < 0.05; Fig. 7).

Fig. 6.

Cd triggered excessive ROS production. (a) Images of mitochondrial ROS staining among groups. (b) Semi-quantitative analysis of mitochondrial ROS staining. *** p < 0.0005, **** p < 0.0001.

Fig. 7.

Cd reduced the mitochondrial membrane potential (MMP). (a) Images of MMP staining. (b) Semi-quantitative analysis of MMP staining. *p < 0.05, **** p < 0.0001.

Furthermore, Cd exposure significantly increased the contents of α-ketoglutaric acid and succinic acid (p < 0.05; Fig. 8a and b), which are intermediate products of the TCA cycle in mitochondria. Compared with Cd treatment, deferoxamine and liproxstatin-1 treatment significantly decreased the contents of α-ketoglutaric acid and succinic acid (p < 0.05; Fig. 8a and b). Cd exposure also significantly increased both the activity and expression of pyruvate carboxylase (p < 0.05), whereas deferoxamine and liproxstatin-1 significantly reversed these effects (p < 0.05; Fig. 8c and d).

Fig. 8.

Cd exposure affects the mitochondrial TCA cycle. (a) α-ketoglutaric acid content; (b) succinic acid content; (c) pyruvate carboxylase (PC) activity; and (d) PC expression among the experimental groups. *p < 0.05, ** p < 0.005, *** p < 0.0005, **** p < 0.0001.

Discussion

Previously, we reported that Cd caused injury and ferroptosis in testicular cells, mainly in spermatocytes¹. In this study, we used the spermatocyte GC–2spd cell line to verify our results and explore the potential underlying mechanisms. Cd exposure damaged GC–2spd cells, which manifested as decreased cell viability, inhibited cell proliferation, and increased cell death. However, the specific mechanism of Cd-induced damage on spermatocyte remained unclear. Therefore, we performed RNA–seq and metabolomic analysis on cells from control and Cd groups. Based on previous research³⁶, we conducted an integrative analysis of the transcriptomics and metabolomics results, which indicated that Cd-induced cell damage was predominantly related to ferroptosis, the mitochondrial TCA cycle, and energy metabolism.

Regarding ferroptosis markers, Cd exposure significantly increased the contents of Fe²⁺ and MDA, as well as the expression of FTH and SLC40A1 in spermatocytes, but reduced the expression of GPX4, suggesting that Cd exposure may cause spermatocyte damage by inducing ferroptosis. The expression changes of FTH and SLC40A1 indicate that Cd led to Fe²⁺ accumulation by dysregulating iron metabolism, inducing ferroptosis. This agrees with our previous findings¹ and those of other researchers³⁷. Fe²⁺ accumulation is a hallmark of ferroptosis. The iron-mediated fenton reaction causes lipid peroxidation and cell damage; therefore, iron-removing reagents are beneficial for cell survival. In our study, cells treated with Cd and deferoxamine—a type of iron chelating agent, increased cell proliferation, decreased Fe²⁺ and MDA, and reduced expression of FTH and SLC40A1. These results indicate that deferoxamine reduced intracellular Fe²⁺ and alleviated Cd-induced ferroptosis in GC-2spd cells.

Another characteristic of ferroptosis is excessive ROS production. In this study, Cd exposure significantly enhanced mtROS levels and Nrf2 and HO-1 expressions. These results contrast with those of some studies, which reported that Cd increases ROS and inhibits the activity and expression of reductase^38,39. However, our results agree with those of Li et al.⁴⁰, who attributed this phenomenon to the strong stress response of spermatocytes to Cd toxicants. Although reductase expression increased, the increase was not sufficient to remove the excessive mtROS caused by Cd exposure.

The origin of mtROS is unclear. Cd toxicity is related to immune responses and epigenetic regulation^14,41. During testis exposure to Cd, Cd recognizes AIM2 as the receptor and activates it, resulting in DNA release into the cytoplasm and ROS production and inflammation in the testes¹⁴. Moreover, Cd disturbs histone acetylation by increasing HDAC1 levels, thus enhancing ROS levels and damaging DNA⁴¹. While epigenetic regulation can alleviate ferroptosis⁴². However, in our study, we propose two origins for mtROS. As mentioned above, some mtROS originate from the iron-mediated fenton reaction, and the rest from the mitochondria itself.

In this study, Cd exposure caused mitochondrial fragmentation, which is consistent with a reduction of the mitochondrial area and mitochondrial ridges observed by Zhang et al.⁴³, and indicates that Cd damage mitochondria directly. Reduced MMP in the Cd group also indicates mitochondrial damage. As the main energy production center and one of the primary sources of ROS, mitochondria generate ATP through oxidative phosphorylation, as well as some ROS, during normal metabolic processes. When mitochondrial function is dysregulated or iron metabolism is imbalanced, ROS production increases significantly, leading to intensified oxidative stress responses that affect cell survival⁴⁴. Fortunately, liproxstatin-1 reversed the redundant mtROS produced by Cd exposure and reduced the excessive stress response of the cells. Liproxstatin-1 can effectively inhibit lipid peroxidation reactions, thereby reducing the production of ROS and reactive lipid species⁴⁵.

Subsequently, we examined intermediate metabolites of the mitochondrial TCA cycle and related enzymes. Our results revealed that Cd exposure caused a significant increase in the contents of succinate and α-ketoglutarate in GC-2spd cells, whereas deferoxamine and liproxstatin-1 mitigated the impact of Cd on the mitochondrial TCA cycle. Our findings differ from those of previous studies, which suggested that Cd exposure reduces the activity of α-ketoglutarate dehydrogenase and succinate dehydrogenase, and that α-ketoglutaric acid supplementation restores the mitochondrial TCA cycle and thus inhibits ferroptosis and alleviates damage^35,46. This difference may be attributed to the enhanced activity and expression of pyruvate carboxylase— a critical mitochondrial enzyme for gluconeogenesis and anaplerosis catalyzes pyruvate to oxaloacetate — thereby replenishing TCA cycle intermediates⁴⁷. Despite the metabolic role of pyruvate carboxylase, pyruvate carboxylase-mediated TCA cycle activation influences cellular redox homeostasis^48,49. Enhanced pyruvate carboxylase activity increases oxaloacetate production, replenishing TCA cycle intermediates and promoting flux through the cycle⁵⁰. This anaplerotic effect stimulates mitochondrial NADH generation via accelerated TCA cycling, subsequently driving ETC activity⁵⁰. The resulting increased electron flow through ETC complexes I and III elevates electron leakage, particularly during high membrane potential states, leading to greater superoxide formation⁵¹. Furthermore, pyruvate carboxylase-mediated TCA cycle activation enhances succinate accumulation, which can reverse electron transport in complex I via succinate dehydrogenase, amplifying ROS production⁵². The potential mechanism described above is supported by studies showing that mtROS generation is correlated with TCA cycle flux rates and NADH/NAD+ ratios⁵³. In addition, elevated ROS levels may activate the Nrf2–Keap1 pathway and further trigger the expression of genes encoding detoxifying enzymes to mitigate ROS damage. This is consistent with our results demonstrating the upregulation of Nrf2 and its downstream gene HO-1.

Consistent with the above theory, our results showed that Cd exposure caused excessive ROS production in the mitochondria as well as MMP polarization. The overproduction of mtROS may be related to direct mitochondrial damage caused by Cd exposure. Oxidative stress drives paradoxical activation of the TCA cycle, likely as a compensatory mechanism to sustain ATP production despite mitochondrial impairment⁵⁴. Increased TCA cycle activity further amplifies mtROS production through an enhanced ETC flux and reverse electron transport drive by succinate in complex I, creating a self-reinforcing cycle of oxidative damage⁵⁵. This vicious cycle promotes iron overload via the ROS-mediated degradation of iron-sulfur cluster proteins, which facilitates fenton reaction-driven lipid peroxidation^56,57.

Conclusion

The results of this study support our hypothesis that Cd exposure directly damages the mitochondria in GC-2spd cells, resulting in paradoxical activation of the TCA cycle, further driving excessive ROS production and finally triggering ferroptosis, which provides new insights for future research on Cd-induced damage in the male reproductive system. As we only tested our hypothesis in the GC-2spd cell line and did not further explore the relevant signaling pathways, subsequent research will include animal experiments to verify our results, as well as a more in-depth analysis of how Cd affects the mitochondrial TCA cycle.

Materials and methods

Cell culture

We obtained GC-2spd cells (CL-0593) from the Wuhan Pricella Biotechnology Company (Wuhan, China). GC-2spd cells are spermatocytes isolated from a six-week-old male mouse and arrested at a premeiotic stage, which have lost their differentiation potential. In a previous in vivo study, we found that Cd caused cell death in testes, concentrated in the spermatocytes¹. Hence, in this study, we selected the GC-2spd cell line for further research.

The cells were cultured in high-glucose DMEM with 10% fetal bovine serum, then incubated with 5% CO₂ at 37 °C. The cells were divided into four groups: control (cultured in high-glucose DMEM), Cd, Cd+deferoxamine, and Cd+liproxstatin-1. Cells in the Cd group were cultured in DMEM with Cd. The final concentration of Cd was 10 µM; preliminary experiments confirmed that this concentration inhibited cell proliferation after 36 h of culture. For the Cd+deferoxamine/Cd+liproxstatin-1 groups, 10 µM Cd and 2 µM deferoxamine/liproxstatin-1 were added to the cell culture. Cells were cultured for 36 h then collected for follow-up experiments.

Non‑invasive optical nanoscopy

We performed Live-cell imaging using 3D Cell Explorer (Nanolive, Switzerland) to observe changes in mitochondrial morphology. Refractive index images (holographic reconstructions) were produced every 15 s using a 60× objective, which captured the real-time mitochondrial morphology and cell health.

Measurement of iron content

The Fe²⁺ content in GC-2spd cells was measured according to the test kit instructions. Iron was released into the supernatant to react with the chromogen, resulting in a colorimetric product. Absorbance was measured at 593 nm using a microplate reader. Finally, the Fe²⁺ content was calculated as per the instructions and expressed as nmol/10⁶.

Measurement of malondialdehyde content

The MDA content was measured according to the test kit instructions. The cell culture supernatant was discarded, and the cells were collected and assayed according to the instructions. Finally, the absorbance was measured at 530 nm. The MDA content in the sample was calculated according to the standard formula and expressed as nmol/mg protein.

Measurement of mitochondrial superoxide content

Mitochondrial superoxide contents were measured to determine mtROS levels. After culture, the supernatant was discarded. MitoSOX™ Green reagent (1 µM) working solution was added to cover cells adhering to coverslip(s) in the well of a 35-mm dish and incubated for 30 min. We then washed the cells gently three times with warm HBSS buffer. Finally, we observed the spectral properties of the cells within 2 h of staining using a fluorescence microscope. Image fluorescence intensity was analyzed using ImageJ software.

TCA cycle assay

We also measured key metabolites of the TCA cycle, including succinate and α-ketoglutarate. Specifically, α-ketoglutarate was detected with a α-ketoglutarate Assay Kit (MAK054, Sigma, USA), and succinate was measured using a succinate colorimetric assay kit (MAK184, Sigma, USA) according to the manufacturer’s instructions.

Western blotting

Experimental methods were performed as described in our previous studies^1,2. An enhanced chemiluminescence kit (cat# C190601, UElandy, Suzhou, China) was used to visualize the bands. β-actin was used as an internal control to quantify the density of the protein bands.

Measurement of mitochondrial membrane potential

Cells (3 × 10⁵ cells/ml) were seeded in a six-well plate and cultured overnight. After removing the supernatant, cells were treated according to the manufacturer’s protocols. Finally, the cells were treated with imaging buffer solution then observed under a fluorescence microscope. Image fluorescence intensity was analyzed using ImageJ software.

RNA sequencing (RNA-seq)

The total RNA from GC-2spd cells in each group (n = 3) were extracted using TRIzol reagent then enriched using mRNA Capture Beads. After purification with the beads, the mRNA was fragmented at high temperatures and used as a template to synthesize the first strand of cDNA in a reverse-transcription enzyme mixture. Briefly, an Illumina sequencing platform produced by GeneDenovo Biotechnology Co., Ltd. (Guangzhou, China) was used to sequence the cDNA libraries. Fastp was used to filter the raw RNA sequencing data to remove adapters and low-quality bases. All clean reads were annotated using the GenBank database. We then calculated the fragment per kilobase of transcript per million mapped reads to quantify gene expression abundance and variations and compare differentially expressed genes among samples. Genes were considered to be differentially expressed when genes/transcripts exhibited a false discovery rate of < 0.05 and an absolute fold change of ≥ 2. Finally, Reactome and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases were used to analyze pathway enrichment^58–60.

Metabolomics analysis

GC-2spd cells were promptly washed with precooled PBS and collected. The cells were then digested with trypsin, resuspended in PBS, and promptly counted. Finally, cells were flash-frozen in liquid nitrogen and temporarily stored at − 80 °C. Following pretreatment, the samples were subjected to non-targeted mass spectrometry-based metabolomic profiling using an AB SCIEX Triple TOF 6600 mass spectrometers followed by bioinformatic analysis.

Combined transcriptome and metabolome analysis

To screen and obtain sets of genes and metabolites associated with sample grouping and analyze their association characteristics, three types of model analyses were performed based on gene expression and metabolite abundance data. (1) Pathway functional model: we queried the KEGG metabolic pathways shared by genes and metabolites and analyzed the association characteristics of genes and metabolites within the shared pathways. (2) O2PLS (Bidirectional Orthogonal Projections to Latent Structures) model: we constructed an O2PLS model using gene expression and metabolite abundance data, then performed predictions to obtain and analyze sets of associated genes and metabolites. (3) Correlation coefficient model (applied for sample groups ≥ 3): we calculated the Pearson correlation coefficients between gene expression and metabolite abundance, then visualized the results using heatmaps and network diagrams.

Statistical analysis

Analysis of variance was used for multiple comparisons. Multiple comparisons were performed using one-way analysis of variance (ANOVA) corrected by Tukey’s test. All data are presented as the mean ± the standard error of the mean. The level of statistical significance was set to p < 0.05 and was shown using asterisks as follows: *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001. The p-values that did not reach a significant level greater than 0.05 but less than 0.1 were reported as indicating a trend that is close to being statistically significant. As for the statistics of cell proliferation rate, data was normalized first and then fitted using nonlinear regression. 95% confidence intervals were used to represent the errors and determine the significance of the differences. GraphPad Prism 8 software (GraphPad Software Inc., San Diego, CA, USA) was used to analyze the data and prepare the graphs.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Data curation: Lijuan Xiong: data acquisition, analysis, work, and draft of the manuscript. Lijun Yi: data acquisition. Xingying Zeng and Jiyi Huang: Data analysis. Hong Li ang Hong Liu: manuscript reviewing and editing. All authors read and approved of the final manuscript.

Funding

This study was supported by the Natural Science Foundation of Jiangxi Province (No. 20224BAB216028).

Data availability

The datasets generated during the current study are available in the Sequence Read Archive (SRA) repository (https://www.ncbi.nlm.nih.gov/sra, accession number: SRR35103192). The statistical results supporting our findings can be found in the article and in the supplementary information.

Declarations

Competing interests

The authors declare no competing interests.