FBXO39 knockdown promotes spermatogenesis impairment by inducing mitochondrial dysfunction and ferroptosis via inhibiting KDM5A ubiquitination and regulating H3K4me3 demethylation

Abstract

The global decline in male fertility highlights the need to understand the mechanisms of spermatogenesis. Mitochondrial dysfunction and ferroptosis have emerged as key contributors to spermatogenic impairment, although the molecular basis of this process is still poorly defined. F-Box Protein 39 (FBXO39), a testis-enriched F-box protein, has been preliminarily associated with cell survival. However, whether FBXO39 participates in mitochondrial functional regulation or ferroptosis signaling during spermatogenesis remains largely unexplored. In our study, FBXO39 knockdown resulted in abnormal testicular development, impaired spermatogenesis, abnormal sperm morphology, and reduced testicular cell viability. Further analysis revealed that FBXO39 deficiency caused mitochondrial dysfunction and ferroptosis, as reflected by decreased ATP production, reduced mitochondrial DNA content, elevated eactive oxygen species (ROS) levels, diminished expression of key mitochondrial proteins, and elevated lipid peroxidation. Mechanistically, FBXO39 maintains mitochondrial homeostasis by targeting lysine-specific demethylase 5A (KDM5A) for ubiquitination-dependent degradation. Conversely, the accumulation of KDM5A upon FBXO39 loss suppressed single-stranded DNA-binding protein 1 (SSBP1) levels through demethylation of Histone H3 lysine 4 trimethylation (H3K4me3) at the SSBP1 promoter. Importantly, restoration of SSBP1 expression functionally ameliorated mitochondrial dysfunction induced by FBXO39 knockdown. Overall, FBXO39 regulates mitochondrial function and ferroptosis in testicular cells through ubiquitinating KDM5A, which affects SSBP1 expression by modulating H3K4me3 demethylation at the SSBP1 promoter. This study elucidates the role of FBXO39 in spermatogenesis and suggested that targeting this regulatory axis may offer novel therapeutic strategies for male infertility.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10565-026-10167-2.

Introduction

Infertility has become a major global health issue, affecting 12% of couples worldwide (Agarwal and Baskaran et al. 2021). A marked decline in semen quality and quantity has led to one in every 20 men being diagnosed with fertility disorders (Bhattacharya and Sharma et al. 2024). The etiology of male infertility is highly diverse. It commonly involves congenital, acquired, or idiopathic factors that lead to impaired spermatogenesis (Agarwal and Baskaran et al. 2021; Rambhatla and Shah et al. 2024; Feng and He et al. 2022). This process is characterized by the differentiation of spermatogonial stem cells into mature spermatozoa within the testicular seminiferous tubules (Cannarella and Condorelli et al. 2020). This process involves three key phases: mitotic proliferation to produce spermatocytes, meiosis to generate haploid spermatids, and spermiogenesis, which transforms them into motile sperm through nuclear condensation, acrosome formation, and flagellum development. Spermatogenesis is supported by Sertoli cells and protected by the blood-testis barrier. Dysfunctional spermatogenesis has been established as the most common cause of male infertility (Meng and Liu et al. 2025). It has been suggested by recent research that the underlying mechanisms of spermatogenic impairment may be closely linked to mitochondrial dysfunction and subsequent ferroptosis, although the precise contributions of this dysfunction remain unclear (Aitken and Drevet et al. 2022).

Mitochondria, as the primary source of reactive oxygen species (ROS) within cells and as key sites for iron metabolism and storage, play a crucial role in ferroptosis (Ahola and Langer 2024; Huang and Lane et al. 2011). Maintaining mitochondrial homeostasis is essential for cellular structure and function. Imbalances in fission and fusion, along with impaired autophagy, can aberrantly activate ferroptosis, leading to cellular damage (Li and Jia et al. 2023). Ferroptosis is an iron-dependent, regulated form of cell death characterized by morphological features such as membrane shrinkage and reduced mitochondrial cristae. It typically involves the accumulation of iron, peroxidation of lipids, and dysregulation of redox homeostasis (Dixon and Lemberg et al. 2012; Friedmann Angeli and Schneider et al. 2014; Li and Cao et al. 2020). Moreover, iron is a critical element for male reproductive function, and adequate iron levels are necessary for spermatogenesis (Tsao and Liao et al. 2022). Therefore, the dynamic balance between mitochondrial homeostasis and ferroptosis is vital during spermatogenesis (Meng and Liu et al. 2025; Cao and Jin et al. 2024; Yang and Chen et al. 2022). Disruption of this balance may represent a key pathological mechanism underlying spermatogenic disorders. Investigating this mechanism provides new insights into the molecular pathology of impaired spermatogenesis.

Regulation of mitochondrial homeostasis relies on the coordinated action of multiple systems, including maintaining antioxidant defenses, controlling protein quality, repairing mitochondrial DNA, regulating dynamics, mediating mitophagy, and supporting biogenesis (Tang and Cai et al. 2021). By serving as the core substrate-recognition modules of the SKP1-CUL1-F-box (SCF) complex, F-box proteins function as a central regulatory node in coordinating mitochondrial homeostasis and ferroptosis through targeted degradation of specific substrate proteins (Ho and Tsai et al. 2006; Naseem and Zhang et al. 2023; Baek and Scott et al. 2023; Skaar and Pagan et al. 2013). There are 69 F-box proteins encoded in the human genome – each one confers unique substrate specificity to the SCF complex, thereby enabling fine-tuned biological regulation (Skaar and Pagan et al. 2013). For example, deficiency of F-box protein 7 (FBXO7) decreases mitochondrial membrane potential and ATP content, increases ROS production, and disruption of mitochondrial homeostasis, ultimately resulting in mitochondrial dysfunction (Delgado-Camprubi and Esteras et al. 2017). F-box proteins also play a critical regulatory role in ferroptosis in neuronal and cancer cells (Bao and Liu et al. 2021; Zhang and Wang et al. 2024). Emerging evidence indicates that F-box proteins have been implicated in the pathogenesis of diverse spermatogenic disorders. For instance, mutations in the FBXO43 gene have been identified as genetic causes of human teratozoospermia and non-obstructive azoospermia, while FBXO24 deficiency disrupts mitochondrial positioning and mRNA alternative splicing, leading to aberrant sperm tail axoneme assembly and mitochondrial coiling defects along the flagellum (Wu and Zhang et al. 2022; Li and Liu et al. 2024). Furthermore, FBXO7 governs spermatocyte differentiation by modulating meiotic substrates, and FBXL10 is essential for maintaining spermatogonial stem cell homeostasis (Zhuang and Ruan et al. 2024). Notably, F-box protein 39 (FBXO39), a testis-enriched F-box protein, has been preliminarily related to cell survival (Yang and Zhao et al. 2022; Zheng and You et al. 2018). However, the specific contributions of FBXO39 to nuclear-cytoplasmic remodeling, mitochondrial functional regulation, or ferroptosis signaling during spermatogenesis remain largely undefined, highlighting a critical gap in our current understanding.

Through alterations in chromatin structure and accessibility, histone modifications—such as methylation, acetylation, and phosphorylation—serve as key epigenetic mechanisms for regulating gene expression (Allis and Jenuwein 2016). Histone modifications play a crucial regulatory role in the mitochondrial unfolded protein response and mitochondrial stress-dependent lifespan regulation (Matilainen and Quirós et al. 2017). Research has shown that trimethylated histone 3 lysine 4 (H3K4me3) is associated with transcriptional activation, promoting cell proliferation and migration, inhibiting ferroptosis, and playing a key role in transmitting mitochondrial stress adaptation information to progeny cells (Wang and Zu et al. 2024; Ma and Niu et al. 2019). KDM5A and its related family members (KDM5B, KDM5C, KDM5D) are the major demethylases for H3K4me3, effectively regulating gene expression and transcriptional activity (Liu and Wang et al. 2022). Notably, a bioinformatic analysis using the Ubibrowser platform identified KDM5A as a putative substrate of the ubiquitin ligase adaptor FBXO39. This finding is significant in light of previous reports that KDM5A promotes mitochondrial dysfunction (Váraljai and Islam et al. 2015) and potentiates ferroptosis (You and Lee et al. 2021). This study sought to elucidate the molecular mechanisms underlying FBXO39's role in spermatogenesis, focusing on how FBXO39 promotes the ubiquitination of KDM5A, thereby regulating the demethylation of H3K4me3, mitigating mitochondrial dysfunction, and inhibiting ferroptosis. Unraveling this mechanism provides new targets for understanding the pathogenesis of spermatogenic disorders and potential therapeutic interventions, with significant clinical implications.

Materials and methods

Animal studies

This study was conducted under an approved animal protocol (The Second Affiliated Hospital of Hainan Medical University IACUC #JENNIO-IACUC-2024-A065) and followed ARRIVE guidelines (Percie du Sert and Hurst et al. 2020). Male Kunming mice (4–6 weeks old, body weight 15–25 g) from the Laboratory Animal Center of Southern Medical University (Guangzhou, China) were acclimated for one week in a specific pathogen-free (SPF) environment with controlled temperature (24 ± 2 °C), humidity (30–70%), and a 12-h light/dark cycle. Sterilized rodent chow and water were provided ad libitum. Twelve mice were then weighed and randomly divided into either the negative control (NC) or the FBXO39 knockdown (shFBXO39) group (n = 6 mice/group). All lentiviral particles (NC or shFBXO39) expressing the red florescence protein (RFP) fluorescent protein (1.0 × 10⁹ PFU/mL, GIEE0280213, GeneChem, Shanghai, China) were delivered via a single bilateral intratesticular injection at the superior pole (50 μL saline-diluted suspension) under aseptic conditions. Four weeks post-injection, animals were euthanized by CO₂ asphyxiation, and testicular tissues were systematically harvested for subsequent analysis.

Detection of testicular weight and volume

Following euthanasia, body weights were recorded. Testes were then collected and immediately weighed. Testicular volume quantification (n = 6 per group) was calculated by dividing fresh tissue weight (g) by the standardized density coefficient of 0.93 g/cm³. This density value was empirically validated and uniformly applied to all specimens during volumetric analysis.

Histopathological assessment and spermatogenesis scoring

Testicular specimens were fixed in 4% paraformaldehyde (PFA) solution, processed through standardized paraffin embedding protocols, and sectioned coronally at 5 μm (RM2016, Leica, Wetzlar, Hesse, Germany). After histopathological analysis (H&E staining), whole-slide images were digitally scanned using a Nikon DS-U3 slide scanner (Tokyo, Japan). Seminiferous tubule staging was performed according to the Johnsen Score by two blinded, independent observers. A 10-point grading system was applied, with 10-point grading criteria based on the sequential presence of spermatogonia, spermatocytes, and mature spermatids within 20 randomly selected tubule cross-sections per specimen (Johnsen 1970).

Assessment of sperm morphology

Sperm were collected by dispersing the cauda epididymis in physiological saline (37 °C, 5% CO₂). Sperm morphological abnormalities and count were assessed at 400 × magnification (ML31, Mshot, Guangzhou, Guangdong, China), with emphasis on defects in the head, neck/midpiece, and tail regions for each sample.

Immunohistochemical staining

After xylene-mediated deparaffinization and sequential rehydration through a graded ethanol series, testicular tissue sections underwent antigen retrieval via microwave heating in citrate buffer (pH 6.0) for 15 min. The sections were then permeabilized with 0.3% Triton X-100 in PBS for 10 min and blocked with 5% BSA in PBS for 1 h, both at room temperature (RT). Sections were subsequently probed overnight at 4 °C with the primary antibodies diluted in PBS containing 1% BSA: FBXO39 (GTX04746, GeneTex, Irvine, ​CA, USA), Ki67 (ab16667, Abcam, Cambridge, ​England, ​UK), Bcl-2 (ab182858, Abcam, Cambridge, ​England, UK), peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α, 66,369–1-Ig, proteintech, Wuhan, Hubei, China), mitochondrial transcription factor A (Tfam, HPA040648, Atlas Antibodies, Stockholm, ​Stockholm County, ​SWE), translocase of outer mitochondrial membrane 20 (TOMM20, MCE, HY-P86575, Monmouth Junction, NJ, USA), 4-HNE (MCE, HY-P81208, Monmouth Junction, NJ, USA), lysine-specific demethylase 5 A (KDM5A, MCE, HY-P81208, Monmouth Junction, NJ, USA). Following PBS washes, sections were probed with the appropriate secondary antibody for 1 h at RT. Diaminobenzidine was used for chromogenic detection, followed by hematoxylin counterstaining. Images were acquired using an Olympus BX53M microscope (Tokyo, Japan) equipped with a digital camera. Quantitative analysis was then conducted on the acquired images using Image-Pro Plus (version 6.0), wherein positive staining was defined based on a standardized threshold for color (brown DAB signal) and optical density.

Western blotting

Total protein was isolated from testicular tissues or cells using a standard RIPA buffer (P0013B, Beyotime Biotechnology, Shanghai, China) per the manufacturer's instructions, and concentrations were subsequently quantified with a BCA assay kit (ab102536, Abcam, Cambridge, ​England, UK). Twenty micrograms of protein aliquots derived from each experimental sample were adjusted to a concentration within the linear detection range prior to sodium dodecyl sulfate–polyacrylamide gel electrophoresis analysis. Following denaturation at 100 °C for 5 min, proteins were resolved on 10–12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and subsequently transferred onto polyvinylidene difluoride membranes. After blocking with 5% non-fat milk for 1 h at RT, the membranes were probed with the following primary antibodies diluted in the appropriate blocking buffe: FBXO39 (1:1000, GTX04746, GeneTex, Irvine, ​CA, USA), PGC-1α (1:1000, HY-P80783, MCE, Monmouth Junction, NJ, USA), Tfam (1:1000, HPA040648, Atlas Antibodies, Stockholm, ​Stockholm County, SWE), TOMM20 (1:5000, HY-P86575, MCE, Monmouth Junction, NJ, USA), KDM5A (1:1000, ab78322, Abcam), GPX4 (1:1000, HY-P80692, MCE), FTH1 (1:1000, HY-P80670, MCE), and β-actin (1:2000, AF7018, Affinity Biosciences, Cincinnati, ​OH, ​USA) at 4 °C overnight. The membranes were then incubated with a horseradish peroxidase-conjugated secondary antibody (1:10,000, SA00001-2, ProteinTech Group, Chicago, IL, USA) for 2 h. For the ubiquitination assay, the protein complexes were detected by western blotting using an anti-ubiquitin primary antibody (1:1000, 43,124, Cell Signaling Technology, Danvers, ​MA, USA). The signals were ultimately visualized with an enhanced chemiluminescence detection system, followed by quantitative analysis of band intensities with ImageJ software, normalized to β-actin.

Pull-down assay using glutathiones-transferase (GST)

For the GST pull-down assay, Glutathione Sepharose 4B beads (17,075,601, Cytiva, Marlborough, MA, USA) were immobilized with 1 µg of either Glutathione S-transferase (GST, control) or GST-FBXO39 protein. The bead-bound proteins were then probed with bacterially purified 6 × His-tagged KDM5A protein at 4 °C for 2 h in a binding buffer containing protease inhibitors (#P1008, Beyotime Biotechnology). After extensive washing, the bound proteins were eluted and subjected to western blotting analysis.

Assessment of the intracellular mitochondria ultrastructure

Testicular tissue was fixed in glutaraldehyde (2.5%, G1102, Servicebio, Wuhan, ​Hubei Province, ​China) and osmium tetroxide (1%, 18,463, Ted Pella Inc, Redding, CA, ​USA), followed by graded ethanol dehydration and epoxy resin embedding. Subsequently, 70-nm ultrathin sections were cut, double-stained with uranyl acetate and lead citrate, and visualized under a Hitachi transmission electron microscope (TEM, HT7800/HT7700, ​Tokyo, ​Japan).

Assessment of mitochondrial superoxide content

For superoxide measurement by flow cytometry, testicular tissue was enzymatically dissociated into a single-cell suspension using a medium of DMEM/F12 containing collagenase IV (1 mg/mL), trypsin (0.5 mg/mL), and DNase I (10 µg/mL), followed by filtration. The cells were then stained with MitoSOX™ (M36008, Thermo Fisher Scientific, Waltham, MA, USA) for 20 min at 37 °C in the dark and washed with PBS before analysis. The samples were analyzed using a flow cytometer (CytoFLEX, Beckman Coulter, Brea, CA, USA).

The cells were seeded onto confocal laser microscopy plates (NEST, Wuxi, Jiangsu, China), treated under identical conditions, and imaged using laser confocal microscopy (FV3000, CLSM, Olympus) for fluorescence analysis.

Measurement of ATP levels

An ATP assay kit (S0026, Beyotime Biotechnology) was used to measure ATP levels. In brief, testicular tissues were homogenized mechanically and transfected cells were lysed in the provided ATP lysis buffer. Then, the samples were centrifuged at 12,000 × g for 5 min at 4 °C. ATP concentration in the supernatant was determined by a Centro XS3 LB 960 luminometer (Berthold Technologies, Bad Wildbad, Baden-Württemberg, Germany), quantified via a standard curve, normalized to protein concentration, and expressed as nmol per mg protein.

Mitochondrial DNA (mtDNA) copy number quantification

To quantify mtDNA, a FastPure® Cell/Tissue DNA Isolation Mini Kit (DC102, Vazyme, ​Nanjing, Jiangsu, China) was used to isolated total DNA from testicular tissue. Briefly, the procedure involved Proteinase K digestion for 15–30 min at 65 °C, ethanol precipitation, and column purification. DNA was eluted in 70 °C-preheated Elution Buffer. Quantitative real-time reverse transcription polymerase chain reaction (RT-qPCR) was performed in triplicate using 96-well reaction plates (436,110, Applied Biosystems, ​Foster City, CA, USA). Each 20 µL reaction mixture comprised 100 ng of DNA, 10 µL of GoTaq® RT-qPCR Master Mix (A6001, Promega, Madison, Wisconsin, USA), and 10 µM of each primer (see Supplementary Table 1). The ΔΔCt method was used for relative quantification.

Assessment of Malondialdehyde (MDA) content

A Lipid Peroxidation MDA Assay Kit (A003-4–1, Jiancheng Bioengineering Institute, Nanjing, China) was used to measure MDA levels. Briefly, Testicular homogenates/cell lysates were mixed with anhydrous ethanol and the assay working solution alongside standards, vortexed, and heated at 95 °C for 40 min before cooling.

After centrifugation (4,000 rpm, 10 min), 0.25 mL of supernatant was analyzed for absorbance at 530 nm on a plate reader (1,410,101, Thermo Fisher). The concentration of MDA was determined from a standard curve, and normalized to protein (nmol/mg protein).

Glutathione (GSH) assay

A total glutathione assay kit (S0052, Beyotime Biotechnology) was used to determine total glutathione concentration in cell and tissue lysates. The principle involves the enzymatic reduction of GSSG to GSH by glutathione reductase, followed by the reaction of GSH with DTNB to generate the chromogen TNB. The cumulative production of TNB, proportional to the total glutathione (GSH + GSSG) present, was monitored by measuring absorbance at 412 nm, enabling quantification of total glutathione content.

Cell Culture and transfection

GC-1 and GC-2 (spermatogonia-derived) and TM4 (mouse Sertoli) cell lines were maintained in Dulbecco’s Modified Eagle Medium (DMEM, 12,800–017, Gibco, Grand Island, NY, USA) supplemented with 10% Fetal Bovine Serum (FBS, 16,000–044, Gibco) and 1% penicillin–streptomycin (C0222, Beyotime Biotechnology) at 37 °C with 5% CO₂.

Lentiviral transduction was performed on GC-1, GC-2 and TM4 cells using NC, shFBXO39, shKDM5A or SSBP1 constructs at MOIs of 10, 20, and 50 (16–20 h). After aspirating and discarding the supernatant, the culture was continued for 48 h to identify the optimal MOI. For the co-transfected experiments, cells were simultaneously transduced with lentiviruses encoding shFBXO39 and shKDM5A, FBXO39 or SSBP1, using optimized MOIs. Puromycin selection (initial 5 μg/mL) was dynamically adjusted (3 → 1 → 0.5 μg/mL if mortality > 40%) over 3–5 cycles. Stable polyclonal populations were validated by sustained proliferation (> 90% viability) under antibiotic pressure with parallel confirmation of untransduced control mortality (> 90%). The transduction efficiency of shFBXO39, shKDM5A, or SSBP1 was evaluated by Western blotting or RT-qPCR. During the 24-h transfection at 37 °C, cells were cultured in the presence of 2 μΜ Ferrostatin-1 (Fer-1, MCE HY-100579) to inhibit ferroptosis and thereby assess its contribution to FBXO39 knockdown-mediated cytotoxicity.

RT-qPCR

Total RNA, extracted with TRIzol (Invitrogen, AM9738), was reverse-transcribed by a Promega kit (M1705). RT-qPCR amplification was carried out on an ABI 7500 system via GoTaq Master Mix (A6002, Promega). Expression levels were calculated relative to β-actin using the 2 − ΔΔCt method. Primer sequences are provided in Supplementary Table 1.

Cell viability assay

The transfected cells (1.0 × 10⁴/well) were plated in 96-well plates at 1.0 × 10⁴ cells/well. Following respective incubation periods (24, 48, 72, and 96 h), add the MTT reagent (11,465,007,001, Sigma, St. Louis, MO, USA) to each well and incubate for 4 h. The crystals were then solubilized via DMSO, and the absorbance was read at 490 nm on a plate reader (1,410,101, Thermo Fisher Scientific).

Propidium iodide (PI) staining

The apoptosis of cells was assessed via PI staining. Cells were plated in 6-well plates at a density of 1 × 10⁶ cells per well. 24 h post-seeding, the cells were fixed using 4% PFA, and then co-stained with PI (SL31141204, Coolaber, Beijing, China), and 4',6-diamidino-2-phenylindole (DAPI, 10,236,276,001, Sigma) for nuclear visualization. Fluorescent imaging was performed using an Olympus BX53M inverted fluorescence microscope.

ROS and lipid peroxidation assay

The transfected cells were seeded into 6-well plates for 24 h. The C11 BoDiPY 581/591 assay kit (Dojindo, Kumamoto, Japan), and the ROS detection kit (L248, Dojindo, Kumamoto, Japan) were performed to assess lipid peroxidation and ROS levels, respectively. Fluorescence was first observed via inverted fluorescence microscopy (BX53M, Olympus) to visualize subcellular localization and morphological patterns of oxidation. For quantitative analysis, the fluorescence signals were then subjected to the BD FACSCalibur flow cytometer (​Franklin Lakes, NJ, USA) to obtain statistically robust measurements at the single-cell level.

Assessment of mitochondrial membrane potential

MitoTracker fluorescent probes (C1035, Beyotime Biotechnology), and JC-1 (C2003S, Beyotime Biotechnology) were performed to assess Mitochondrial morphology and membrane potential, respectively. For the JC-1 staining, cells were treated with 5 µM JC-1 working solution and incubated at 37 °C for 20 min in the dark. In this assay, JC-1 forms red-fluorescent polymers under high membrane potential and remains green-fluorescent monomers under low membrane potential. Images were captured using an Olympus BX53M inverted fluorescence microscope.

For the Mito Tracker staining assay, the transfected cells were seeded onto confocal laser microscopy plates for 24 h. Cells were incubated with 100 nM Mito Tracker working solution at 37 °C in the dark for 30 min. Following cell fixation and nuclear counterstaining using 1 µg/mL DAPI in the dark at 37 °C for 5 min, CLSM (Olympus) was systematically employed for high-resolution fluorescence imaging and subsequent quantitative analysis of subcellular structures. The intensity of MitoTracker staining correlates with mitochondrial membrane potential, with decreased fluorescence indicating potential depolarization or dysfunction.

Co-immunoprecipitation (Co-IP) assay

The transfected cells were lysed using a low-salt buffer, and the lysates were incubated overnight at 4 °C with antibodies specific to KDM5A (ab194286, Abcam, Cambridge, ​England, ​UK), FBXO39 (GTX04746, GeneTex, Irvine, ​CA, USA), or IgG (ab6708, Abcam, Cambridge, ​England, UK) as a control. Immune complexes were isolated by immunoprecipitation with protein A/G agarose beads (16–266, Millipore, Boston, MA, USA). After elution of the precipitated complexes, the presence of KDM5A and FBXO39 in the precipitates was examined by immunoblotting (western blotting).

Detection of protein ubiquitination

Following lysis of transfected cells, immunoprecipitation was carried out on the lysates. Briefly, agarose beads coupled to recombinant protein A and G (Cat# 20,422, Sigma) were pre-conjugated with either an anti-KDM5A antibody (1:30, ab194286, Abcam) or a control anti-IgG antibody (1:30, ab6708, Abcam). Immunoblotting with an anti-ubiquitin antibody (1:1000, #43,124, Cell Signaling Technology) was performed to analyze the precipitated complexes.

Immunofluorescence staining

Cells were grown overnight on coverslips, fixed in 4% paraformaldehyde, treated with 0.2% Triton X-100 for permeabilization, and then blocked with 5% goat serum at RT for 40 min. Primary antibodies, including mouse anti-KDM5A (1:100, ab78322, Abcam) and rabbit anti-FBXO39 (1:100, GTX04746, GeneTex, Irvine, CA, USA) were applied to the cells and incubated overnight at 4 °C. Following this, a 30-min incubation at RT with fluorophore-conjugated secondary antibodies was conducted: goat anti-mouse IgG-Cy3 (1:200, ab97035, Abcam) and anti-rabbit IgG-FITC (1:200, BL033A, biosharp, Hefei, China). Nuclei were counterstained with DAPI (BL105A, Biosharp, Hefei, China), and the samples were subsequently visualized and imaged using a fluorescence microscope (CKX53, Mshot, Guangzhou, Guangdong, China).

Protein half-life detection

Cells were treated with 10 μM cycloheximide (CHX, C7698, Sigma) for varying durations (0, 2, 4, and 8 h) to inhibit protein synthesis. Proteins were extracted from each group, and KDM5A levels were analyzed by Western blotting.

Chromatin immunoprecipitation (ChIP) assay

ChIP was conducted via the EZ-ChIP kit (17–409, Sigma) in strict accordance with the provided protocol. In brief, cells were fixed with 1% PFA (RT, 10 min), washed, and resuspended in lysis buffer. Chromatin was fragmented by sonication, generating DNA fragments ranging from 400 to 800 bp. ChIP was performed overnight with purified anti-H3K4me3 antibodies (1:30, PA5-27,029, Thermo Fisher Scientific), anti-KDM5A antibodies (1:30, ab194286, Abcam), or with mouse IgG isotype control. Antibody-bound complexes were isolated with agarose beads coupled to recombinant Protein A and G, followed by sequential washing with washing buffer. Reversal of crosslinks was achieved by incubation in NaCl for 6 h at 65 °C. The purified DNA was subsequently analyzed by RT-qPCR.

Statistical analysis

Continuous variables are reported as mean ± standard deviation (SD). Statistical analyses were carried out with SPSS 21.0. Inter-group differences were evaluated by Student’s t-test for two-group comparisons. When comparing multiple groups, one-way analysis of variance (ANOVA, for a single independent variable) or two-way ANOVA (for two independent variables) was utilized, each followed by Tukey’s multiple comparison test. A threshold of p < 0.05 was adopted to determine statistical significance.

Results

FBXO39 knockdown led to testicular damage and spermatogenesis impairment in mice

A single bilateral intratesticular injection of lentiviruses (NC or shFBXO39) into the superior pole resulted in successful transduction of both germ and Sertoli cells (Supplementary Fig. 3a, 3b). Representative images of testes from each experimental group are presented in Fig. 1a, the testes from the shFBXO39 group were visibly smaller than those from the non-targeting control (NC) group. No significant difference in body weight was observed between the shFBXO39 and NC groups (Supplementary Fig. 1b); in contrast, both the testes weight and the testes-to-body weight ratio were significantly lower in the shFBXO39 group (Supplementary Fig. 1c, 1d, respectively). Microscopic examination revealed significant sperm morphological defects in the shFBXO39 group, such as tail folding and coiling (Supplementary Fig. 1e), accompanied by a reduction in sperm count (Supplementary Fig. 3c).

Fig. 1.

Validation of FBXO39 knockdown and its role in ferroptosis-mediated reduction of cell viability. (a, b) Validation of FBXO39 knockdown at the mRNA (RT-qPCR) and protein (western blotting) levels in GC-1, GC-2, and TM4 cells. Data were analyzed by Student's t-test. (c) Cell viability measured by MTT assay, and analyzed by two-way ANOVA with Tukey’s multiple comparison test. Differences among groups were analyzed by two-way ANOVA followed by Tukey’s post hoc test. (d) Detection of dead cells by PI staining assay. (e) Proportion of oxidized lipids visualized and quantified. (f) Detection of lipid ROS levels by flow cytometry. (g) Ratio of oxidized to non-oxidized lipids. (h) The accumulation level of MDA. Values represent mean ± SD (n = 3); Statistical significance was determined by one-way ANOVA followed by Tukey’s multiple comparison test (*P < 0.05, **P < 0.01, ***P < 0.001)

We used H&E staining to examine testicular histology. The testicular tissue from NC group exhibited well-organized seminiferous tubules with intact basement membranes and orderly arranged germ cells, along with abundant sperm in the epididymal lumen. Conversely, the seminiferous tubules in the shFBXO39 group were loosely arranged and disordered, displaying prominent vacuolation within the seminiferous epithelium, with widened intercellular spaces, reduced germ cell numbers, thinner epididymal walls, and increased interstitial spaces between epididymal tubules. Besides, a significant decrease in luminal spermatozoa was observed (Supplementary Fig. 1f).

FBXO39 knockdown significantly suppressed both its own protein expression and cellular proliferation (Ki67) in testicular tissue, as evidenced by immunohistochemistry (Supplementary Fig. 1 g, 1j, 1 k, 4, 8e). In NC testes, FBXO39 expression was localized to specific germ cell populations, and Sertoli cells. In contrast, FBXO39 knockdown markedly reduced this expression pattern (Supplementary Fig. 1 g, 4, 8e). Concurrently, the expression of Bcl-2, a key marker of anti-apoptosis, was markedly reduced in the shFBXO39 group, suggesting that the observed testicular damage may be associated with increased apoptosis (Supplementary Fig. 1 h, 1 l, 8e). Following FBXO39 knockdown, the Johnsen score was markedly reduced relative to the NC group, reflecting a severe impairment of spermatogenic function (Supplementary Fig. 1i). These findings were further corroborated by RT-qPCR and western blotting analysis, which demonstrated a significant reduction in FBXO39 mRNA (73.61 ± 0.93%) and protein (81.16 ± 4.78%) levels in the shFBXO39 group (Supplementary Fig. 1 m, 3d), consistent with the immunohistochemistry results. In addition, knockdown of FBXO39 inhibited the expression of FBXO39 in sperm cells (Supplementary Fig. 3 g, 3 h). The results collectively indicated that the knockdown of FBXO39 leads to testicular damage and impaired spermatogenesis in mice.

Knockdown of FBXO39 induced mitochondrial damage and ferroptosis in mouse testes

Mitochondrial function is critically important for the metabolic regulation of spermatogenesis and germ cell maturation (Chang and Miao et al., 2025). Consequently, this study explored the consequences of FBXO39 knockdown on mitochondrial integrity and the induction of ferroptosis in mouse testes. TEM revealed significant mitochondrial deformation in spermatocytes from the shFBXO39 group, characterized by disrupted, blurred, or completely vanished cristae membranes, hallmark features of ferroptosis. Besides, the mitochondria exhibited a marked loss of matrix density and pronounced vacuolization (Supplementary Fig. 2a).

Assessment of mitochondrial function showed that FBXO39 knockdown led to a significant increase in mitochondrial superoxide (Supplementary Fig. 2b), accompanied by decreased ATP levels (Supplementary Fig. 2c) and reduced mtDNA abundance (Supplementary Fig. 2d) in testicular Sample. Similarly, levels of MDA, a key oxidative stress marker and a recognized indicator for assessing ferroptosis, were markedly increased (Supplementary Fig. 2e). Immunohistochemical analysis demonstrated that the shFBXO39 group exhibited decreased expression of three core regulators of mitochondrial function: PGC-1α (biogenesis; Supplementary Fig. 2f, 2 g, 8e), Tfam (transcription/replication; Supplementary Fig. 2f, 2 h, 8e), and TOMM20 (protein import; Supplementary Fig. 2f, 2i, 8e). These deficits collectively impaired mitochondrial dynamics (maintenance of morphology), cargo transport, and apoptosis signaling.

Moreover, the lipid peroxidation product 4-HNE, a critical marker of ferroptosis, was significantly accumulated in the shFBXO39 group (Supplementary Fig. 2j, 2 k, 8e). In addition, key ferroptosis indicators, including GSH, GPX4, and FTH1 levels, exhibited significant downregulation following FBXO39 knockdown (Supplementary Fig. 3e, 3f). Collectively, these findings demonstrated that FBXO39 knockdown induces mitochondrial damage and ferroptosis in mouse testes, providing new insights into its functional role during spermatogenesis.

Knockdown of FBXO39 induced ferroptosis to suppress testicular cell viability

Ferroptosis reduces cell viability by accumulating lipid peroxides, which cause oxidative damage and disrupt cellular functions (Liu and Liu et al. 2025). We further investigated whether FBXO39 knockdown reduces cell viability through ferroptosis. As shown in Fig. 1a, 1b, having confirmed the efficiency of FBXO39 knockdown in GC-1, GC-2 and TM4 cells via RT-qPCR (for mRNA quantification) and western blotting (for protein). As shown in Fig. 1c, the cell viability in the shFBXO39 group was significantly reduced compared to the control group we observed a significant reduction in cell viability in the shFBXO39 group relative to controls. However, treatment with the ferroptosis inhibitor Fer-1 restored cell viability in shFBXO39 cells. Similar results were observed with PI staining, where the decreased cell viability in the shFBXO39 group corresponded to increased fluorescence intensity, which was mitigated by Fer-1 (Fig. 1d, Supplementary Fig. 13a). Consistent with a ferroptotic phenotype, FBXO39 knockdown enhanced lipid peroxidation (increased oxidized lipids, lipid ROS, and oxidized/non-oxidized lipid ratio; Fig. 1e-g, Supplementary Fig. 8c) and depleted key anti-ferroptotic defenses (reduced GPX4, FTH1, and GSH with elevated MDA; Fig. 1h, Supplementary Fig. 5). Notably, these disruptions were rescued by Fer-1 treatment. To establish a direct causal role, FBXO39 expression was reconstituted via lentiviral transduction in the knockdown cells. Overexpression of FBXO39 rescued the phenotypic changes induced by its knockdown, including decreased cell viability, reduced levels of GSH, GPX4, and FTH1, and elevated oxidized lipids (Supplementary Fig. 6, 8c). These findings collectively demonstrated that FBXO39 knockdown reduced cell viability through inducing ferroptosis, as evidenced by increased lipid peroxidation.

Knockdown of FBXO39 promoted ferroptosis by inducing mitochondrial dysfunction

Mitochondrial dysfunction is a key driver of ferroptosis (Guo and Zhou et al. 2022). To investigate whether FBXO39 knockdown impacts cell viability by exacerbating mitochondrial dysfunction-linked ferroptotic cascades, we conducted a series of evaluations. TEM images revealed notable mitochondrial morphological damage in all three cell types following FBXO39 knockdown, with varying degrees of structural abnormalities observed (Fig. 2a). FBXO39 knockdown induced mitochondrial depolarization, as revealed by a reduced JC-1 red/green fluorescence ratio (Fig. 2b, Supplementary Fig. 13c). This loss of membrane potential suggests that mitochondrial dysfunction is a key trigger of apoptosis upon FBXO39 depletion. Besides, mitoSOX fluorescence was markedly elevated in the shFBXO39 group, reflecting increased mitochondrial superoxide levels (Fig. 2c, Supplementary Fig. 13b). MitoTracker staining revealed that mitochondria in the shFBXO39 group predominantly exhibited a fragmented distribution, accompanied by a significant reduction in fluorescence intensity (Fig. 2d, Supplementary Fig. 13d). As mitochondria are the primary site of ATP production, we observed a significant inhibition of ATP generation in the shFBXO39 group (Fig. 3a). Furthermore, mtDNA abundance was markedly reduced in the shFBXO39 group (Fig. 3b). Downregulation of key mitochondrial proteins PGC-1α, Tfam, and TOMM20 was observed in the shFBXO39 group, as shown by Western blot analysis (Fig. 3c). Rescue experiments were conducted to determine the specificity of FBXO39 in mitochondrial function. FBXO39-knockdown GC-1 cells were transduced with FBXO39-encoding lentiviruses to restore protein expression. FBXO39 overexpression in knockdown cells effectively rescued the associated mitochondrial dysfunction, as evidenced by reduced superoxide levels, improved mitochondrial morphology, and recovered expression of the mitochondrial regulators PGC-1α, Tfam, and TOMM20 (Supplementary Fig. 7,Supplementary Fig. 13b, 13d).

Fig. 2.

Role of FBXO39 knockdown in promoting mitochondrial ROS-driven ferroptosis. (a) TEM images of mitochondrial morphology. (b) Assessment of mitochondrial membrane potential in cells by JC-1 staining (red/green fluorescence ratio). (c) Measurement of mitochondrial superoxide levels by mitoSOX fluorescence. (d) Determination of mitochondrial distribution and activity by MitoTracker staining. Values represent mean ± SD (n = 3); Significance was assessed using Student’s t-test (**P < 0.01, ***P < 0.001)

Fig. 3.

Knockdown of FBXO39 promotes ferroptosis by inducing mitochondrial dysfunction. (a) ATP levels quantified via luminescent assay. Significance was assessed using Student’s t-test. (b) MDA accumulation level. Significance was assessed using Student’s t-test. (c) PGC-1α, Tfam, and TOMM20 levels assessed by western blotting in GC-1, GC-2, and TM4 cells post-FBXO39 knockdown. Significance was assessed using Student’s t-test. (d) Proportion of oxidized lipids visualized and quantified. (e) Detection of lipid ROS levels by flow cytometry. (f) Ratio of oxidized to non-oxidized lipids. (g) The accumulation level of MDA. Values represent mean ± SD (n = 3); Statistical significance was determined by one-way ANOVA followed by Tukey’s multiple comparison test (*P < 0.05, ***P < 0.001)

FBXO39 knockdown promoted a lipid peroxidation cascade, as shown through elevated levels of oxidized lipids (Fig. 3d), lipid ROS (Fig. 3e), and the oxidized/non-oxidized lipid ratio (Fig. 3f, Supplementary Fig. 8c). This cascade was significantly mitigated upon inhibition of mitochondrial ROS with Mito-TEMPO (Fig. 3d-f). Concurrently, the level of MDA was significantly elevated, and the protein expression of GPX4 and FTH1, as well as GSH levels, were markedly decreased in the shFBXO39 group, which was partially reversed by Mito-TEMPO co-treatment (Fig. 3g, Supplementary Fig. 7). These findings highlighted mitochondrial ROS as the primary mediator of FBXO39 knockdown-induced lipid peroxidation. Collectively, FBXO39 knockdown disrupts mitochondrial structure and function, contributing to ferroptosis.

FBXO39 promoted KDM5A protein degradation via ubiquitination

FBXO22, a paralog of FBXO39 within the F-box protein family, has been reported to ubiquitinate and degrade KDM5A (Li and He et al. 2023). Given the structural and functional conservation among F-box family members, we speculated that FBXO39 similarly regulates KDM5A degradation via Ubiquitination. The Ubibrowser platform identified KDM5A as a target gene of FBXO39. Previous studies have established that KDM5A promotes mitochondrial dysfunction (Váraljai and Islam et al. 2015) and potentiates ferroptosis (You and Lee et al. 2021), consistent with the phenotypes observed following FBXO39 knockdown. Western blotting and immunohistochemistry revealed a significant upregulation of KDM5A in the shFBXO39 group, consistent with our hypothesis (Fig. 4a, Supplementary Fig. 8a, 8e). This prompted us to examine a potential direct interaction. Using Co-IP and GST pull-down assays, we confirmed that FBXO39 and KDM5A interact at the protein level in GC-1, GC-2, and TM4 cells (Fig. 4b, Supplementary Fig. 8b). Immunofluorescence staining further verified the co-expression of FBXO39 and KDM5A in all three cell types (Fig. 4c, Supplementary Fig. 8d). Moreover, KDM5A protein levels in the NC group decreased in a time-dependent manner with extended CHX treatment. Notably, the marked upregulation of KDM5A following FBXO39 knockdown (Fig. 4d) implies that FBXO39 normally targets KDM5A for degradation, thereby reducing its half-life.The proteasomal inhibitor MG-132 increased KDM5A levels in the NC group, indicating proteasome-dependent degradation. FBXO39 knockdown similarly elevated KDM5A level, suggesting FBXO39 knockdown reduced KDM5A degradation. Notably, shFBXO39 combined with MG-132 caused further accumulation of KDM5A, confirming that proteasome-dependent residual degradation persists despite FBXO39 deficiency (Fig. 4e). To determine whether FBXO39 mediates KDM5A ubiquitination, we performed ubiquitination co-immunoprecipitation assays. FBXO39 knockdown significantly reduced the ubiquitination level of KDM5A (Fig. 4f). These findings collectively suggested that FBXO39 facilitates KDM5A ubiquitination, promoting its proteasome-dependent degradation.

Fig. 4.

FBXO39 promoted KDM5A protein degradation via ubiquitination. (a) Western blotting analysis of KDM5A expression. Significance was assessed using Student’s t-test. (b) Detection of the interaction between FBXO39 and KDM5A by Co-IP. (c) Co-expression of FBXO39 and KDM5A by immunofluorescence staining. (d) Detection of FBXO39 prolongs KDM5A protein half-life by CHX chase assay. Statistical significance was determined by one-way ANOVA followed by Tukey’s multiple comparison test. (e) Evaluation of proteasome-dependent degradation of KDM5A using MG-132 treatment. (f) Ubiquitination assays showing the regulation of KDM5A ubiquitination by FBXO39. Values represent mean ± SD (n = 3); Statistical significance was determined by one-way ANOVA followed by Tukey’s multiple comparison test (**P < 0.01, ***P < 0.001)

The deleterious effects of FBXO39 knockdown on cell viability, mitochondrial homeostasis, and ferroptosis were reversed by concomitant knockdown of KDM5A

Our above experiments demonstrated that FBXO39 facilitates KDM5A protein degradation. To determine whether KDM5A downregulation could mitigate the effects of FBXO39 knockdown in testicular cells, we conducted sequential knockdown experiments and systematically evaluated cellular viability, mitochondrial dysfunction, and ferroptosis.

Western blotting analysis initially revealed that FBXO39 knockdown significantly increased KDM5A expression in GC-1 cells, whereas dual knockdown of FBXO39 and KDM5A restored KDM5A levels (Supplementary Fig. 9a), suggesting a compensatory mechanism. Functionally, FBXO39 knockdown-induced suppression of cell viability (MTT) and apoptosis elevation (PI staining) were restored by KDM5A co-knockdown (Supplementary Fig. 9b, 9c, 13a). ​FBXO39 knockdown-induced​ ferroptosis was attenuated by KDM5A co-knockdown, evidenced by reduced mitochondrial oxidized lipid (Supplementary Fig. 9d), lipid ROS Supplementary Fig. 9e), oxidized-to-non-oxidized lipid ratio (Supplementary Fig. 10a, 8c), MDA levels (Supplementary Fig. 10b), and GPX4 and FTH1 protein expression, as well as GSH levels (Supplementary Fig. 5). In terms of mitochondrial morphology, KDM5A co-knockdown reversed FBXO39 knockdown-induced structural fragmentation (Supplementary Fig. 10c), restored impaired membrane potential (Supplementary Fig.  10d, 13c), lowered superoxide levels (Supplementary Fig. 11a, 13b), preserved mitochondrial structure (Supplementary Fig. 11b, 13d), and recovered ATP production (Supplementary Fig. 11c), while maintaining mtDNA content (Supplementary Fig. 11d) and upregulating mitochondrial proteins PGC-1α, Tfam, and TOMM20 (Supplementary Fig. 11e).

These findings collectively demonstrated that KDM5A knockdown effectively reverses the reduced testicular cell viability, mitochondrial dysfunction, and ferroptosis induced by FBXO39 knockdown.

KDM5A reduces H3K4me3 at the SSBP1 promoter and decreases SSBP1 expression

Integrated analysis of multiple databases (RNAlnter, Chip-atlas, hTFtarget, GTRD, CHEA, ENCODE) identified eight potential KDM5A target genes, including GPATCH3, SNRPC, NDUFAF1, MTX1, NDUFS3, RCE1, DNAJB11, and SSBP1 (Fig. 5a). Notably, current evidence highlights that NDUFS3 and SSBP1 specifically regulate ferroptosis, mitochondrial dysfunction and cell viability (Su and Li et al. 2022; Wang and Hu et al. 2017; Wang and Lv et al. 2024; Yuan and Shen et al. 2024; Feng and Ni et al. 2024), which supports our hypothesis that KDM5A may modulate these biological processes through SSBP1 or NDUFS3. Analysis by ChIP-qPCR revealed that KDM5A binding was specifically enriched at the SSBP1 locus in mouse testicular cell lines (GC-1, GC-2, TM4), demonstrating a direct interaction with its promoter or regulatory regions (Fig. 5b, Supplementary Fig. 6a).

Fig. 5.

KDM5A reduces H3K4me3 at the SSBP1 promoter and decreases SSBP1 expression. (a) Potential KDM5A target genes identified by integrating multiple databases. (b) Detection of the interaction between KDM5A and SSBP1 by ChIP-RT-qPCR analysis. (c) Quantification of H3K4me3 methylation levels at the SSBP1 promoter by ChIP-RT-qPCR analysis. Analysis of the expression of SSBP1 in KDM5A-overexpressing cells by RT-qPCR (d) and Western blotting assay (e). (f) Detection of the abundance of H3K4me3 at the SSBP1 promoter in KDM5A-overexpressing cells. Values represent mean ± SD (n = 3); Significance was assessed using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001)

The WashU Epigenome Browser and Cistrome Data Browser were used to investigate whether SSBP1 is regulated by epigenetic mechanisms. Analysis revealed the presence of H3K4me3, a marker of active transcription, at the SSBP1 promoter region, suggesting that SSBP1 expression may be regulated through histone methylation (Supplementary Fig. 6b). Subsequent ChIP-qPCR analysis confirmed significant enrichment of SSBP1 in the anti-H3K4me3 group relative to controls (Fig. 5c). RT-qPCR analysis of KDM5A-overexpressing cells demonstrated a significant reduction in SSBP1 expression (Fig. 5d), which was corroborated by western blotting (Fig. 5e). Besides, ChIP-RT-PCR analysis in KDM5A-overexpressing cells revealed a marked decrease in H3K4me3 enrichment at the SSBP1 promoter region (Fig. 5f).

In summary, KDM5A transcriptionally represses SSBP1 in testicular cells by demethylating H3K4me3 at its promoter (Supplementary Fig. 12), underscoring the importance of this epigenetic mechanism in SSBP1 regulation.

Reconstitution of SSBP1 expression ameliorated the phenotypic consequences of FBXO39 depletion, restoring cell viability and mitigating mitochondrial dysfunction and ferroptosis

Given the established role of FBXO39 in testicular function, we next examined whether SSBP1 could alleviate the effects of FBXO39 knockdown-mediated cell viability, mitochondrial dysfunction, and ferroptosis. As shown in Fig. 6a-d, Supplementary Fig. 13a, supplementation with SSBP1 ameliorated the consequences of FBXO39 knockdown, concomitantly restoring SSBP1 expression at both transcriptional and translational levels and improving cell viability.

Fig. 6.

Supplementation of SSBP1 ameliorated the detrimental effects of FBXO39 knockdown, restoring cell viability and reducing mitochondrial lipid peroxidation. (a) Quantification of SSBP1 levels by RT-qPCR in GC-1 cells subjected to FBXO39 knockdown with or without SSBP1 overexpression. (b)-(e) Western blotting, MTT assay, PI staining, and flow cytometry assay were performed to evaluate SSBP1 protein expression, cell viability, apoptotic cell death, and mitochondrial oxidized lipid levels, respectively. Values represent mean ± SD (n = 3); Statistical significance was determined by one-way ANOVA followed by Tukey’s multiple comparison test (***P < 0.001). Data from the MTT assay were subjected to two-way ANOVA, with comparisons between specific groups adjusted by Tukey’s multiple comparison test

Mechanistically, SSBP1 attenuated FBXO39 knockdown-induced ferroptosis, as demonstrated by reduced oxidized lipid, lipid ROS, MDA levels, and oxidized/non-oxidized lipid ratios, as well as increased GSH levels, and the protein expression of GPX4 and FTH1 (Fig. 6e-f, Fig. 7a, b, Supplementary Fig. 5, Supplementary Fig. 8c). This attenuation of lipid peroxidation was accompanied by the restoration of FBXO39 knockdown-impaired​mitochondrial ultrastructure, including restored cristae density, membrane potential, and superoxide balance (Fig. 7c, d, 8a, b, Supplementary Fig. 13b-d). SSBP1 further alleviated FBXO39 knockdown-induced impairments by rescuing ATP synthesis capacity and mtDNA abundance (Fig. 8c, d), while sustaining expression of mitochondrial-related proteins PGC-1α, Tfam, and TOMM20 at levels comparable to those in control conditions (Fig. 8e).

Fig. 7.

SSBP1 reversed FBXO39 knockdown-driven mitochondrial dysfunction through ultrastructural and ​lipid peroxidation​ restoration. (a) Fluorescent staining analysis of oxidized to non-oxidized lipid ratio in GC-1 cells treated with FBXO39 knockdown and SSBP1 overexpression. (b)-(d) Colorimetric assay, TEM, and JC-1 probe analysis were performed to assess MDA content, mitochondrial cristae density/membrane integrity, and mitochondrial membrane potential,​​ ​respectively. Values represent mean ± SD (n = 3); Statistical significance was determined by one-way ANOVA followed by Tukey’s multiple comparison test (*P < 0.05, **P < 0.01, ***P < 0.001)

Fig. 8.

The mitochondrial dysfunction caused by FBXO39 knockdown was reversed upon reconstitution of SSBP1. (a) Measurement of mitochondrial superoxide levels by MitoSOX fluorescence in GC-1 cells subjected to FBXO39 knockdown with or without SSBP1 overexpression. (b)-(e) MitoTracker staining, luminescent assay, RT-qPCR, and western blotting assay were performed to measure mitochondrial distribution and activity, ATP synthesis capacity, mitochondrial DNA abundance, and mitochondrial biogenesis markers (PGC-1α, Tfam, TOMM20),​​ ​respectively. Values represent mean ± SD (n = 3); Statistical significance was determined by one-way ANOVA followed by Tukey’s multiple comparison test (*P < 0.05, **P < 0.01, ***P < 0.001)

Taken together, these findings indicated that SSBP1 effectively mitigates the adverse effects of FBXO39 knockdown on testicular cell viability, mitochondrial function, and ferroptosis.

Discussion

With the continued global decline in male fertility, the mechanisms underlying spermatogenesis have become a focal point of reproductive medicine research (Ribeiro and Alves et al. 2021; Chen and Zheng et al. 2018). Growing evidence suggests that mitochondrial dysfunction is a critical driver of ferroptosis and a key factor affecting spermatogenesis (Wang and Yin et al. 2022; Durairajanayagam and Singh et al. 2021; Mai and Yang et al. 2024; Jiao and Ma et al. 2022; Radaelli and Assenmacher et al., 2023). Meanwhile, F-box proteins, which determine substrate specificity for SCF-type E3 ubiquitin ligases, play essential roles in diverse physiological processes, with recent studies suggesting their involvement in multiple stages of spermatogenesis and in regulating sperm production (Wu and Zhang et al. 2022; Chen and Ma et al. 2022). This study focused on the F-box protein family member FBXO39 and systematically explored its critical role in regulating mitochondrial function and ferroptosis in testicular cells. These findings highlighted FBXO39's pivotal mechanisms in spermatogenesis and underscore its potential impact on male fertility.

Our research demonstrated that FBXO39 was essential for maintaining testicular structure and spermatogenic function, as its knockdown in mice led to disorganized testicular structure, impaired spermatogenic function, and abnormal sperm morphology. Furthermore, FBXO39 knockdown induced the expression of pro-apoptotic markers, suggesting that testicular damage may be associated with increased apoptosis. Given that mitochondria are central to apoptosis (Bock and Tait 2020) and play a crucial role in spermatogenesis and germ cell maturation (Zhao and Heng et al. 2022), we further investigated the impact of FBXO39 knockdown on mitochondrial function in mouse testicular cells. Mechanistically, FBXO39 knockdown disrupted mitochondrial homeostasis, as evidenced by severe mitochondrial damage, reduced ATP production, increased lipid peroxidation, and enhanced ferroptosis. These changes contribute to decreased cell viability and can be rescued by the ferroptosis inhibitor Fer-1. Collectively, this study establishes FBXO39 as a critical regulator of spermatogenesis, maintaining mitochondrial homeostasis to inhibit ferroptotic cell death. This work directly links FBXO39 function to the protection of male reproductive health.

This study systematically elucidated the molecular mechanism by which FBXO39 regulates mitochondrial function and ferroptosis in testicular cells via downstream signaling networks. As an F-box protein, FBXO39 mediates substrate ubiquitination to regulate protein stability—a canonical function in protein quality control (Ho and Tsai et al. 2006). Beyond maintaining proteostasis, the ubiquitin–proteasome pathway (UPP) serves as a regulatory hub in processes like lipid peroxidation and cell fate determination by controlling the turnover of mitochondria-associated proteins (Tsakiri and Gumeni et al. 2019; Kodroń and Mussulini et al. 2021).

Through integrative analysis leveraging predictive bioinformatic tools and experimental validation, we identified KDM5A as a key downstream target of FBXO39. Our experimental data demonstrated that FBXO39 facilitated the proteasome-dependent degradation of KDM5A via ubiquitination, significantly affecting its protein stability and epigenetic regulatory function. This regulatory mechanism closely parallels previous findings in which Fbxo22 reduces KDM5A levels or Fbxo24 targets lysine-specific demethylase 1 (LSD1) for ubiquitination-mediated degradation, thereby suppressing breast cancer metastasis and tumorigenesis (Li and He et al. 2023; Dong and Song et al. 2023). Importantly, knockdown of KDM5A markedly reversed the detrimental effects of FBXO39 deficiency, including reduced cell viability, mitochondrial dysfunction, and exacerbated ferroptosis, indicating that KDM5A serves as a central regulatory node in this pathway. As a core member of the JARID1 subfamily, KDM5A utilizes its Jumonji C-terminal (JmjC) domain to exert Fe²⁺/α-ketoglutarate (α-KG)-dependent catalytic activity, specifically demethylating trimethylated histone H3 lysine 4 (H3K4me3), thereby functioning as a critical epigenetic suppressor of gene transcription (Horton and Engstrom et al. 2016; Kirtana and Manna et al. 2020; Petronikolou and Longbotham et al. 2020).

Further mechanistic investigations revealed that KDM5A reduces H3K4me3 at the SSBP1 promoter and decreases SSBP1 expression. Previous studies have shown that SSBP1 acts as a pivotal regulator of mitochondrial and iron metabolism in testicular cells. It plays an essential role in maintaining mitochondrial DNA (mtDNA) integrity, responding to mtDNA damage, modulating ROS levels, and preserving genomic stability (Su and Li et al. 2022; Wang and Hu et al. 2017). Notably, loss-of-function mutations or impairments in SSBP1 significantly reduce mtDNA replication efficiency, disrupt mitochondrial dynamics, and lead to mitochondrial dysfunction (Cha and Lee et al. 2024). In addition, SSBP1 is vital for redox homeostasis and has been identified as a ferroptosis-related biomarker (Su and Li et al. 2022; Lin and Li et al. 2025). Our data demonstrate that restoring SSBP1 expression effectively reversed multiple cellular impairments induced by FBXO39 knockdown, including enhanced cell viability, restoration of mitochondrial ultrastructure, increased ATP production, and significant inhibition of lipid peroxidation and ferroptotic processes. Taken together, these results highlight the critical protective role of SSBP1 within the FBXO39-KDM5A regulatory axis, underscoring its importance in maintaining mitochondrial integrity and cellular function in testicular cells.

This study has several limitations. First, the in vivo knockdown approach may yield mosaic transduction, leading to inconsistent FBXO39 depletion across testicular cells and potentially confounding phenotypic outcomes. Future studies using conditional knockout animal models or cell-type-specific manipulation systems will help more precisely dissect the function of FBXO39 in specific cell populations. Second, given the high structural and functional homology among KDM5 family members (KDM5B, KDM5C, KDM5D), we cannot rule out compensatory mechanisms or redundant interactions in testicular cells, despite our demonstration of KDM5A's pivotal role. Future experiments are needed to further compare the binding specificity of each member to the SSBP1 promoter and their potential roles within this pathway. Besides, the conclusions of this study have not been validated in clinical samples or human-derived cells, and the role of the FBXO39–KDM5A–SSBP1 axis in human male infertility remains to be explored.

In conclusion, our findings delineated a regulatory FBXO39-KDM5A-SSBP1 axis that modulates ferroptosis and mitochondrial homeostasis in testicular cells (Supplementary Fig. 14). This discovery expands the functional repertoire of F-box proteins in reproduction and elucidates a molecular bridge between ubiquitination, histone modification, and cellular metabolism. Understanding this axis may help develop new treatments for male infertility.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

HTJ conceived the study design. TL, KW, and YXC performed *in vitro* experiments and wrote the manuscript. ZCL, SDL, and YZ performed *in vivo* experiments and prepared all the figures. XYZ, HRS, and LG conducted the data analysis. All authors read and approved the final manuscript.

Funding

This work was supported by Hainan Provincial Health Commission Project (No. 20A200360), the start-up funds for talent introduction in the Second Affiliated Hospital of Hainan Medical University (No. KYQDJ-2019–03), Hainan Provincial Doctoral Student Innovation Project (No. Qhyb2023-179).

Data availability

All data that support the findings of this study are available from the corresponding authors upon reasonable request.

Declarations

Ethical approval

The study was approved by the Institutional Animal Care and Use Committee of The Second Affiliated Hospital of Hainan Medical University (No. #JENNIO-IACUC-2024-A065).

Informed consent

Not applicable.

Conflict of interest

The authors declare no competing interests.

Clinical trial number

Not applicable.