Riboflavin activates the Bmp4/Bmpr1α signaling to reverse arsenic-induced male reproductive toxicity: in vivo and in vitro studies

Abstract

As a widely distributed hazardous substance in agriculture and food, arsenic has become a global public health concern. Riboflavin is known for its anti-inflammatory and antioxidant properties, but whether it mitigates arsenic-induced male reproductive damage remains unclear. To clarify arsenic’s reproductive toxicity, network toxicology was first used to identify potential targets. Subsequently, thirty-six 4-week-old male mice were administered arsenic trioxide (5 mg/L in drinking water) and/or riboflavin (10 mg/kg in diet) for 22 weeks to establish a riboflavin intervention model. Network toxicology identified that arsenic induces male infertility via multiple signaling pathways, such as the TGF-β signaling pathway. In vivo, arsenic diminished sperm quality, caused histopathological damage, and impaired both blood-testis barrier integrity and Sertoli cell proliferation, while riboflavin effectively mitigated these detrimental effects. In vitro, TM4 cell experiments further confirmed that riboflavin reversed arsenic-induced decline in cell viability and cytoskeletal damage. This study is the first to demonstrate that the protective effect of riboflavin is mainly attributed to its ability to mediate the binding of Bmp4 to Bmpr1α, thereby reversing arsenic-induced toxicity, providing a new perspective for exploring riboflavin’s role in alleviating arsenic toxicity.

Introduction

Arsenic is a naturally occurring toxic and carcinogenic metalloid element that is extensively present in soil, water, air, and other food and agricultural production^1,2. Chronic exposure to arsenic, primarily through contaminated groundwater and dietary intake, poses serious health risks to both humans and animals^3,4. Data from the World Health Organization (WHO) indicate that numerous regions in the United States, Mexico, China, Canada, Argentina, Bangladesh, Laos, India, and other countries exhibit arsenic contamination levels that surpass the WHO threshold of 10 μg/L⁵. Arsenic exposure induces the development of cancer and other systemic complications such as infertility, skin diseases, lung cancer, liver injury, and cardiovascular disease, which have emerged as a significant worldwide public health concern^6–8.

In the context of the global decline in sperm quality, numerous clinical and animal experiments have identified a correlation between environmental arsenic exposure and male infertility, with the testes being a significant target organ for arsenic toxicity⁹. Studies have shown that oral ingestion of arsenic leads to its accumulation in the testes and epididymis of male animals. Furthermore, male residents living in areas with high-arsenic drinking water (0.05–0.2 mg/L) exhibit significantly elevated arsenic levels in both serum and semen^10,11. This means that arsenic may result in disrupted permeability of the blood-testis barrier (BTB) and subsequent reproductive damage. Animal studies have shown that long-term exposure to arsenic adversely affects testicular morphology, structure, and function. These pathological changes include inhibition of spermatogenesis, reduced sperm motility, altered histone acetylation, and disruption of cell-cell junctions, ultimately leading to male reproductive toxicity^12,13. Despite a growing body of evidence highlighting the reproductive toxicity of arsenic, the complete molecular mechanisms behind its effects have not been fully clarified.

The BTB is formed between testicular Sertoli cells via desmosomes adjacent to the basement membrane, tight junctions (TJs), basal ectoplasmic specialization (bES), and gap junctions (GJs)¹⁴. As a physical barrier to the testes, the BTB not only regulates the exchange of endogenous molecules and external substances but also ensures proper spermatogenesis¹⁵. Additionally, Sertoli cells facilitate germ cell translocation from the basal compartment to the lumen of the seminiferous tubule via cytoskeletal reorganization of microfilaments and microtubules, ultimately releasing mature spermatozoa into the lumen^16,17. Thus, functional abnormalities in Sertoli cells directly impair spermatogenesis. With the deepening of research on the spermatogenesis pathway, multiple evidence point to the role of the transforming growth factor-β (TGF-β)/Smad pathway in spermatogenesis, which can regulate spermatogenesis by influencing the proliferation of TJs, adherens junctions (AJs), and Sertoli cells¹⁸. As a member of the TGF-β superfamily, bone morphogenetic protein 4 (Bmp4) signals via intracellular Smad proteins and is essential for Sertoli cell development¹⁹. Sertoli cells secrete Bmp4, which in turn promotes the proliferation and differentiation of Sertoli cells, thereby maintaining a stable Sertoli cell population²⁰. Bmp4 signaling is mediated by the intracellular Smad proteins. Following the binding of Bmp4 to its receptor, Smads 1/5/8 are phosphorylated, subsequently oligomerize with Smad4, and translocate to the nucleus as a complex, where they function as transcription factors²¹. Arsenic exposure can disrupt the BTB and cause reproductive damage, but whether it can interfere with spermatogenesis through the above signaling pathway and lead to male reproductive dysfunction is unclear.

Riboflavin (Ribo) is a water-soluble vitamin that originates from plants and most microorganisms and is necessary for the growth and reproduction of humans and animals²². As a biosynthetic precursor of flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD), riboflavin serves critical roles in metabolic processes such as mitochondrial energy metabolism, the tricarboxylic acid cycle, electron transport chain activity, redox homeostasis, and DNA repair²³. Riboflavin has been used as a therapeutic agent to treat male infertility and maintain normal sperm function²⁴. Notably, our laboratory previously reported that riboflavin can restore fluoride-induced testicular damage in rat models by regulating IL-17A-mediated classical pyroptosis²⁵. However, the therapeutic potential of riboflavin supplementation in restoring spermatogenesis following arsenic-induced male reproductive injury, as well as its molecular mechanisms, remains poorly characterized.

To investigate this, the study first adopted network toxicology to comprehensively analyze the mechanism by which arsenic induces male infertility. An in vivo mouse model of riboflavin intervention with arsenic exposure was constructed to assess sperm quality, histomorphology, and key biomarkers associated with the BTB, Sertoli cell proliferation, riboflavin metabolism, and the TGF-β/Bmp4/Smad pathway. For mechanistic validation, TM4 cells were used to detect the cytoskeleton in vitro for further verification. The purpose of this work was to explore the mechanism of male reproductive injury caused by arsenic, evaluate riboflavin’s protective effects, and to offer a novel strategy for prevention of arsenic poisoning.

Results

Network toxicology analysis of the potential mechanisms of arsenic-induced male infertility

To determine the potential mechanism of arsenic-induced male infertility, a total of 222 overlapping target genes were retrieved from GeneCards and CTD databases (Fig. 1A). Gene ontology (GO) enrichment analysis revealed that for biological processes (BP), the genes were mainly related to apoptosis, the positive and negative regulation of gene expression, and the proliferation process of germ cell populations. Regarding cellular components (CC), the genes were predominantly located in mitochondria, cell membranes, and perinuclear regions. For molecular functions (MF), the genes were significantly enriched in functions such as ATP binding, enzyme binding, and protein binding (Fig. 1B and Supplement Table 2).

Fig. 1. Network toxicological analysis results of arsenic-induced male infertility.

A Venn diagram for screening arsenic-induced male infertility related genes. B GO enrichment analysis (biological process, cellular component, and molecular function) of arsenic-induced male infertility related genes. C KEGG enrichment results (environmental information processing, cellular processes, and organismal systems) of arsenic-induced male infertility related genes. D PPI network diagram of arsenic-induced male infertility-related genes.

The KEGG results revealed that the pathways related to environmental information processing (e.g., HIF-1, PI3K-Akt, TGF-β, and JAK-STAT signaling pathways), cellular processes (e.g., adherens junction, gap junction, tight junction, apoptosis, and the p53 signaling pathway), and organismal systems (e.g., T cell receptor, Toll-like receptor, and IL-17 signaling) were significantly enriched, which comprehensively covered the pathways related to the biological level of arsenic-induced male infertility (Fig. 1C and Supplement Table 3). The Cytoscape conducted topological analysis on the PPI network of “Mus musculus” and screened out 15 key target genes of the TGF-β signaling pathway based on the node degrees (Fig. 1D). Overall, these findings illustrate that arsenic induces male infertility by disrupting gene regulation, cellular structure, molecular interactions, and multiple signaling pathways, providing a comprehensive molecular framework for understanding the reproductive toxicity process of arsenic.

Establishment of riboflavin intervention and arsenic-induced reproductive injury models in male mice

To investigate the possible mechanisms of arsenic exposure on male reproduction and the alleviating effect of riboflavin, we designed a mouse model of arsenic exposure with riboflavin intervention (Fig. 2A). Analysis of sperm morphology and quality showed that arsenic caused a decrease in sperm motility and sperm vitality and an increase in sperm malformation, which was mitigated by riboflavin intervention. After riboflavin treatment, sperm motility and vitality increased markedly, and sperm malformation was significantly reduced (Fig. 2B, C).

Fig. 2. Preliminary evaluation of riboflavin’s protective effects on arsenic-induced male reproductive toxicity in a mouse model.

A Schematic diagram of mouse model design and treatment. B, C Evaluation of sperm morphology and sperm quality (sperm counts, sperm motility, sperm viability, sperm malformation rate) in mice, the red arrow indicates sperm malformations (n = 5; Sperm malformations included macrocephaly, microcephaly, headlessness, double heads, irregular head shape, neck/midpiece folding, tail coiling, double tails, etc.). D, E Representative images of H&E staining in testicular and epididymal tissue (caput epididymis, corpus epididymis, cauda epididymis) (scale bars = 50 μm and/or 100 μm; Black triangle, cells shed into the lumen; Blue triangle, loss and disordered cell arrangement). F Statistical analysis of testicular tissue (outer diameter of seminiferous tubule, inner diameter of seminiferous tubule, thickness of seminiferous tubules, n = 5). Note: Control: Control group; As: Arsenic trioxide group; Ribo: riboflavin group; As + Ribo: co-administration of arsenic trioxide and riboflavin group. ***P < 0.001, **P < 0.01 and *P < 0.05 vs. Control group; ^###P < 0.001, ^##P < 0.01 and ^#P < 0.05 vs. As group.

Furthermore, the H&E staining showed that spermatogenic cells were abundant and arranged neatly in the testes of mice in the Control group, while the testicular morphology of mice in the As group was destroyed. For example, the distance between spermatogenic tubules widened, and the number of spermatogenic cells decreased. The outer diameter and thickness of spermatogenic tubules decreased markedly, while the inner diameter of the lumen increased significantly. Additionally, the testicular structure loosened, and spermatogenic cells fell off into the lumen (Fig. 2D, F). Meanwhile, the number of sperm in the lumen of the caput and corpus epididymidis was also significantly reduced, while riboflavin intervention ameliorated the arsenic-induced histopathological changes compared with the As group (Fig. 2E). Taken together, these findings indicate that riboflavin intervention plays an effective mitigating role in arsenic-induced male reproductive damage, as it counteracts the decline in sperm quality and the morphological damage to reproductive organs caused by arsenic.

Riboflavin alleviates the damage to the blood-testis barrier integrity induced by arsenic

The expression of key BTB indicators was identified through IF and Western blotting to examine the impact of riboflavin on the BTB induced by arsenic. The IF results revealed that the fluorescence expression of Occludin and CX43 proteins in the testicular tissue of the As group decreased compared to the Control group, while riboflavin intervention increased the fluorescence intensity of Occludin and CX43 proteins after arsenic exposure (Fig. 3A–D). Consistent with the findings described above, the Occludin, CX43, and ZO-1 protein expression were all downregulated in the As group. ZO-1 protein expression was significantly restored in the As + Ribo group relative to the As group (Fig. 3E–G). Notably, no notable disparities were observed in N-Cadherin protein levels in the testes of the four groups (Fig. 3H). These findings indicate that riboflavin can mitigate damage to the BTB induced by arsenic in mice.

Fig. 3. Effect of riboflavin on key indicators of arsenic-induced blood-testis barrier in mice.

A, B Immunofluorescence staining and statistical analysis of Occludin (green) protein in testicular tissue, (Scale bars = 100 μm, n = 5). Nuclei were counterstained with DAPI (blue). C, D Immunofluorescence staining and statistical analysis of CX43 (red) protein in testicular tissue (Scale bars = 100 μm, n = 5). Nuclei were counterstained with DAPI (blue). E–H Western blotting analysis of the protein expression of Occludin, CX43, ZO-1 and N-Cadherin, respectively (n = 3). Note: Control: Control group; As: Arsenic trioxide group; Ribo: riboflavin group; As + Ribo: co-administration of arsenic trioxide and riboflavin group. ***P < 0.001, **P < 0.01 and *P < 0.05 vs. Control group; ^###P < 0.001, ^##P < 0.01 and ^#P < 0.05 vs. As group.

Riboflavin mitigates the disruption of testicular Sertoli cells induced by arsenic

To determine whether arsenic exposure induces reproductive toxicity in male mice by directly affecting the structure and function of testicular Sertoli cells, we examined key indicators associated with testicular Sertoli cell damage. The findings indicated a marked increase in the mRNA levels of Wt1, Ar and Gdnf in the Ribo group, and the mRNA levels of Gdnf and Ar in the As+Ribo group were markedly higher than those in the As group (Fig. 4A–C).To further validate these findings, we examined SOX9 expression in the testis using IHC, showing that riboflavin intervention significantly increased the expression of SOX9 in the testes of mice affected by arsenic (Fig. 4D, E). Consistent with the IHC results, the SOX9 protein expression was significantly decreased in the As group (Fig. 4F). This suggests that riboflavin alleviates the direct damage to mouse testicular Sertoli cells caused by arsenic to some extent.

Fig. 4. Effect of riboflavin on key indicators of arsenic-induced Sertoli cell proliferation.

A–C The mRNA expression level of Wt1, Ar and Gdnf in testicular tissue (n = 5). D, E Immunohistochemistry staining and statistical analysis of SOX9 protein in testicular tissue (Scale bars = 50 μm and 100 μm, n = 5) F Western blotting analysis of the protein expression of SOX9 (n = 3). Note: Control: Control group; As: Arsenic trioxide group; Ribo: riboflavin group; As + Ribo: co-administration of arsenic trioxide and riboflavin group. ***P < 0.001, **P < 0.01 and *P < 0.05 vs. Control group; ^###P < 0.001, ^##P < 0.01 and ^#P < 0.05 vs. As group.

Riboflavin alleviates arsenic-induced damage by enhancing the function of riboflavin transporters

It has been found that riboflavin has the capability to mitigate the harm to the body caused by arsenic exposure²⁶. To ascertain the direct impact of arsenic on riboflavin transport and metabolism, the key proteins of riboflavin transport in the testis were examined. The findings indicated a dramatic increase in FLAD1 and RFK expression in the Ribo group compared with the Control group. Additionally, the mRNA levels of FLAD1 and RFK were notably higher in the As+Ribo group compared to the As group, while there were no remarkable variations in the expression of SLC52A2 and SLC52A3 in testicular tissues among the four groups (Fig. 5A–D). The results of Western blot analysis found that the level of FLAD1 protein expression in the Ribo group was notably increased, and RFK protein expression in the As+Ribo group was remarkably higher than that in the As group (Fig. 5E, F). These results indicate that arsenic exposure at this concentration does not directly alter riboflavin transport or metabolism. Instead, riboflavin supplementation upregulates its own metabolic proteins, likely as a response to increased substrate availability.

Fig. 5. Effect of riboflavin on key indicators of arsenic-induced riboflavin metabolism.

A–D The mRNA expression level of SLC52A2, SLC52A3, FLAD1, and RFK in testicular tissue (n = 5). E, F Western blot analysis of the protein expression of FLAD1 and RFK (n = 3). Note: Control: Control group; As: Arsenic trioxide group; Ribo: riboflavin group; As + Ribo: co-administration of arsenic trioxide and riboflavin group. ***P < 0.001, **P < 0.01 and *P < 0.05 vs. Control group; ^###P < 0.001, ^##P < 0.01 and ^#P < 0.05 vs. As group.

Riboflavin increases the expression of key indicators of the TGF-β/Bmp4/Smad signaling pathway in testicular tissue induced by arsenic

The TGF-β/Bmp4/Smad pathway was crucial in the regulation of arsenic-induced male reproductive damage (Fig. 6A). According to the heatmap results (Fig. 6B), arsenic exposure only dramatically reduced Bmp4 and Bmpr1α mRNA expression, while riboflavin intervention restored the mRNA levels of Bmp4 and Bmpr1α. In addition, the levels of Smad2 and Smad3 in the Ribo group were remarkably increased. The As+Ribo group could decrease TGF-β3 levels and increase Smad1 expression. Subsequently, the proteins of Bmp4 and Bmpr1α in mouse testicular tissue were further detected by IF, and the results were consistent with the mRNA results. The IF results showed a reduction in the levels of Bmp4 and Bmpr1α proteins in the As group compared with the Control group and an increase in the As+Ribo group compared with the As group (Fig. 6C, D). These results suggest that arsenic exposure does not activate the TGF-β/Smad pathway but affects male reproductive toxicity by directly mediating the Bmp4/Smad pathway. Meanwhile, we speculate that Bmp4 may be an important binding target for riboflavin, which alleviates arsenic-induced reproductive toxicity by activating Bmp4/Bmpr1α. The molecular docking results indicated that Bmp4 (−7.5 kcal/mol), Bmpr1α (−9.8 kcal/mol), Smad4 (−6.2 kcal/mol), and Smad5 (−6.3 kcal/mol) all possess multiple binding pockets and can stably bind to riboflavin with low binding energy (Fig. 6E–H and Supplement Table 4).

Fig. 6. Effect of riboflavin on key indicators of arsenic-induced the TGF-β/Smad and Bmp4/Smad signaling pathways.

A Map of pathways associated with TGF-β/Smad and Bmp4/Smad signaling (Created with BioGDP.com). B Heat map of mRNA expression levels of gene TGF-β2, TGF-β3, Smad2, Smad3, Smad4, Serpina5, Bmp4, Bmpr1α, Bmpr2, Bmpr1β, Smad1, Smad5, FGF2 and Kitl in testicular tissue (n = 5). C, D Immunofluorescence staining of Bmp4 (red) and Bmpr1α (red) protein in testicular tissue (Scale bars = 50 μm, n = 5). Nuclei were counterstained with DAPI (blue). E–H Molecular docking results of Bmp4-Riboflavin, Bmpr1α-Riboflavin, Smad4-Riboflavin, and Smad5-Riboflavin, respectively. Note: Control: Control group; As: Arsenic trioxide group; Ribo: riboflavin group; As + Ribo: co-administration of arsenic trioxide and riboflavin group. ***P < 0.001, **P < 0.01 and *P < 0.05 vs. Control group; ^###P < 0.001, ^##P < 0.01 and ^#P < 0.05 vs. As group.

Riboflavin alleviates the damage to TM4 cytoskeleton and the alteration of Bmp4/Bmpr1α signaling induced by arsenic

According to the above results, we found that the effects of arsenic on testicular injury were mainly manifested in Sertoli cells. Therefore, we selected mouse testicular Sertoli cells (TM4 cells) for in vitro experiments. The SOX9 protein has been used as a specific marker for testicular Sertoli cells, and the purchased commercial TM4 cell lines were identified by IF staining to determine cell type and integrity. The results demonstrated that TM4 cells were labeled green with SOX9 protein fluorescence, and the nucleus was stained blue by DAPI (Supplement Fig. 1A). Cell activity was detected with the CCK8 kit to determine the optimal concentration range for As₂O₃, riboflavin, and the co-treatment of As₂O₃ and riboflavin (Supplement Fig. 1B–D). Ultimately, 3 μM As₂O₃ and 4, 8, and 16 μM Ribo were selected to investigate the effects of riboflavin on TM4 cell damage caused by arsenic. To investigate the effect of riboflavin intervention on TM4 cells treated with arsenic, H&E staining was further measured. The results showed that cells exhibited normal morphology and proliferation in the Control group, whereas cells showed sparse intercellular arrangement and significantly inhibited proliferation in the As₂O₃ group. However, the cell growth was improved after the addition of riboflavin (4, 8, and 16 μM) treatment compared to the As₂O₃ group (Fig. 7A). The results further indicate that riboflavin mitigates the suppressive impact of As₂O₃ on the growth of testicular Sertoli cells.

Fig. 7. Effects of riboflavin on arsenic-induced TM4 cytoskeletal proteins and the Bmp4/Bmpr1α pathway.

A Representative image of TM4 cells stained with H&E (Scale bars = 50 μm and 100 μm, n = 5). B, C Rhodamine phalloidin staining and statistical analysis to reveal the F-actin (red) in TM4 cells (Scale bars = 100 μm, n = 5). Nuclei were counterstained with DAPI (blue). D, E Immunofluorescence staining and statistical analysis of Tubulin (green) protein in TM4 cells (Scale bars = 100 μm, n = 5). Nuclei were counterstained with DAPI (blue). F–K Western blotting analysis of the protein expression of Bmp4, Bmpr1α, Bmp4, Bmpr1α, Smad4, and Smad5, respectively (n = 3). Note: ***P < 0.001, **P < 0.01 and *P < 0.05 vs. the no treatment group; ^###P < 0.001, ^##P < 0.01 and ^#P < 0.05 vs. the 3 μM As₂O₃ group.

To explore the impact of riboflavin on the cytoskeleton in TM4 cells following arsenic exposure, F-actin and Tubulin were assessed using fluorescence phalloidin staining and IF assay, respectively (Fig. 7B–E). The diminished fluorescence intensity of F-actin and Tubulin in TM4 cells under arsenic-exposed conditions suggests that arsenic may disrupt the cytoskeleton of Sertoli cells. This disruption was mitigated by riboflavin intervention (4, 8, and 16 μM Ribo) and reduced arsenic’s impairment of the TM4 cells’ barrier function. Subsequently, in order to demonstrate that arsenic causes male reproductive toxicity by inhibiting the Bmp4/Smad signaling pathway, we detected key pathway proteins in TM4 cells of different treatment groups and further validated them using a BMP inhibitor (DMH-1). It was found that the 3 μM As₂O₃ group significantly downregulated the protein expressions of Bmp4 and Bmpr1α in TM4 cells, while the 8 μM Ribo intervention group significantly upregulated the protein expression of Bmp4 (Fig. 7F, G). Based on the above results, 8 μM Ribo was selected for follow-up tests. Additionally, we screened the optimal concentration of DMH-1 in TM4 cells and found that 0.5 μM DMH-1 exerted no cytotoxic effects on cell viability (Supplement Fig. 1E, F). Western blotting further confirmed that Bmp4, Bmpr1α and Smad5 protein expression was lower in the As₂O₃ group than in the Control group. Conversely, Bmp4 protein in the As₂O₃+Ribo group and Smad5 protein in the As₂O₃+Ribo+DMH-1 group were prominently elevated relative to the As₂O₃ group. There was no significant difference in the Smad4 protein among the various groups (Fig. 7H–K). This further corroborates that riboflavin mediates Bmp4/Bmpr1α signaling to alleviate Sertoli cell damage induced by As₂O₃.

Discussion

With the escalating global arsenic pollution, its environmental concentrations in water, air, and food have significantly surpassed safety standards²⁷. Extensive studies have confirmed that arsenic exposure can exert male reproductive toxicity by inducing oxidative stress, activating inflammatory pathways, affecting testosterone synthesis, inhibiting spermatogenesis, and damaging spermatogenic cell functions^13,28. There is a pressing need to elucidate the mechanism of reproductive damage induced by arsenic poisoning in male animals and to identify diagnostic biomarkers and therapeutic targets. To comprehensively assess the reproductive toxicity induced by arsenic and develop effective prevention and control strategies, network toxicology analysis initially revealed that arsenic can interfere with processes such as testicular barrier function, cell proliferation and differentiation, and apoptosis by modulating multiple signaling pathways. Further investigation has found that the TGF-β/Bmp4/Smad pathway may be the potential core mechanism underlying arsenic-induced reproductive toxicity. Specifically, arsenic exposure can abnormally regulate the key components of this pathway, including the Bmp4 ligand and its downstream Smad effectors. This signal imbalance not only directly destabilizes the testicular microenvironment by down-regulating the expression of tight junction and adhesion junction proteins, but also affects the sperm production process by interfering with the nutritional and supportive functions of Sertoli cells for the development of germ cells. Consequently, targeting the TGF-β/Bmp4/Smad pathway may represent a promising therapeutic strategy to mitigate arsenic-induced male infertility.

Approximately 94 to 220 million people are at risk due to high concentrations of arsenic in groundwater, involving numerous countries^5,29. Consumption of arsenic-contaminated water is the primary route of environmental arsenic exposure in humans, directly leading to arsenic poisoning^30,31. In the present study, mice were exposed to arsenic at a dose of 5 mg/L via drinking water, which is approximately 500-fold higher than the WHO guideline value of 10 μg/L, to explore its adverse effects. Given that mice are less sensitive to arsenic than humans, the dose used in this study was several times higher than the arsenic concentrations typically found in highly contaminated groundwater^32,33. Furthermore, long-term low-dose arsenic exposure produces distinct toxic effects compared with acute exposure, and chronic exposure modeling is essential for studying arsenic’s cumulative toxicity³⁴. Based on these studies, a concentration of 5 mg/L arsenic was selected in this experiment to establish a reliable model of arsenic-induced male reproductive toxicity. Riboflavin has been widely studied for its growth-promoting, anti-inflammatory, antioxidant, and immune-enhancing properties and has been proven to have a protective effect against reproductive toxicity³⁵. Importantly, riboflavin is already widely available as a low-cost over-the-counter nutritional supplement, and its safety has been proven, even at high doses³⁶. This makes riboflavin a feasible intervention strategy for areas affected by arsenic pollution. Epidemiological evidence indicates that riboflavin deficiency is highly prevalent in arsenic-endemic regions. This phenomenon is partially attributed to inadequate nutritional status and partially to the potential of arsenic exposure to disrupt riboflavin metabolism³⁷. Consequently, this study reveals the underlying mechanism of arsenic’s effect on the male reproductive system and the protective effect of riboflavin by establishing in vivo and in vitro models of riboflavin intervention for arsenic exposure.

Sperm quality is one of the most direct and critical indicators for evaluating male reproductive capability³⁸. Epidemiological studies have consistently associated arsenic exposure with decreased semen quality, manifested by reduced sperm motility, morphological abnormalities, and compromised chromatin integrity³⁹. The structural and functional integrity of the testes is the foundation of male fertility. Furthermore, the epididymis connects the testicle to the vas deferens and is an important organ responsible for storing sperm, promoting sperm maturation, and assisting in fertilization⁴⁰. Li et al.⁴¹ demonstrated that sub-chronic arsenic exposure significantly decreased testicular and epididymal weight, reduced sperm motility and sperm count, increased the rate of sperm deformity, and caused distinct pathological alterations in the reproductive organs of mice. These observations are in good agreement with our current findings. Our research further confirms that arsenic exposure impairs sperm quality, reduces epididymal sperm count, and induces morphological damage to testicular tissue. Additionally, riboflavin effectively ameliorated arsenic-induced spermatogenic dysfunction, including sperm quality decline and testicular histopathological alterations in the mouse model, providing favorable evidence for riboflavin to reduce the male reproductive toxicity of arsenic.

Maintaining the BTB’s structural integrity is essential for supporting physiological spermatogenesis and preserving male fertility¹⁵. Arsenic exposure undermines the integrity of the BTB, causing progressive bioaccumulation in testicular tissues, which constitutes the primary pathogenic mechanism underlying its reproductive toxicity⁴². TJs, as the predominant junctional complex in the BTB, comprise an intricate network of transmembrane proteins, including Occludin and Claudin-11. These proteins are anchored to the actin cytoskeleton via the intracellular scaffolding protein ZO-1⁴³. GJs are intercellular communication channels composed of connexin hexamers, of which CX43 is a major gap junction protein that demonstrates dynamic regulation in Sertoli cells during spermatogenesis⁴⁴. N-Cadherin, a key component of AJs, facilitates homotypic cell-cell adhesion and is essential for BTB maintenance, spermatogenesis, and sperm release⁴⁵. This study demonstrated through KEGG-based network toxicology that arsenic induces male infertility by affecting adherens junction, gap junction, and tight junction. A key mechanism for this is the arsenic-induced damage to the BTB, resulting in increased permeability and disrupted intercellular communication. The experiment further found that N-Cadherin protein level in the testes showed no significant changes among the groups, indicating that N-Cadherin may exhibit lower sensitivity to arsenic exposure compared to other junction proteins. But interestingly, IF staining and Western blotting found that the levels of CX43, Occludin, and ZO-1 in mouse testes of the As group decreased, while riboflavin mitigated this. The above results are similar to the studies of Niu et al.⁴⁶.

Sertoli cells, as important cells for nutrient supply and structural support of spermatogenesis, are central to the establishment of the BTB⁴⁷. Accumulating toxicological evidence indicates that toxic substances exert adverse effects on male fertility by disrupting Sertoli cell-germ cell communication networks⁴⁸. The results of this research demonstrated that arsenic had no remarkable effect on altering mRNA levels of Sertoli cell proliferation marker genes (e.g., Wt1, Ar, Gdnf) in mouse testes, but riboflavin intervention markedly upregulated their expression. At the same time, the immunohistochemical and Western blotting analysis further revealed that riboflavin could significantly up-regulate the decreased expression of SOX9 induced by arsenic. These findings suggest that arsenic interferes with Sertoli cell proliferation and differentiation, contrasting with Li et al.¹¹, who reported arsenic-induced BTB junctional impairment without direct effects on Sertoli cells. This discrepancy may stem from differences in arsenic dosage and experimental duration. In addition, arsenic exposure markedly reduced the fluorescence intensity of F-actin and Tubulin in TM4 cells, suggesting that arsenic exposure leads to abnormal assembly of the Sertoli cells’ cytoskeleton. The cytoskeleton preserves the structural integrity of the BTB through dynamic recombination during the spermatogenic cycle⁴⁹. F-actin, as a core element of the Sertoli cell cytoskeletal structure, maintains TJ structural integrity through the dynamic interaction of its C-terminal domain with ZO-1, while simultaneously regulating ectoplasmic specialization (ES)-mediated spermatid adhesion processes⁵⁰. Consistent with this, Wang et al.⁵¹ reported that arsenic markedly suppressed the F-actin expression in a rat liver injury model, a phenomenon also corroborated by our findings in male reproductive contexts. These findings emphasize the need for further investigation into the potential mechanisms of arsenic on reproductive health, and it is critical to understand how riboflavin plays a protective role.

Riboflavin is a crucial vitamin necessary for mammals and is transported to cells by riboflavin transporters SLC52A2 and SLC52A3⁵². Riboflavin, serving as the biosynthetic precursor for FMN and FAD, undergoes phosphorylation catalyzed by RFK and FLAD1 to generate these redox-active coenzymes⁵³. Among them, RFK and FLAD1 exert critical regulatory roles in the development and progression of many diseases and serve as potential therapeutic targets for influencing metabolic pathways in the body⁵⁴. Qiao et al.⁵⁵ demonstrated that fluorine exposure disrupts riboflavin metabolism by impairing transporter activity and enzymatic conversion, while IL-17A can mediate the fluoride-induced disturbance of spleen riboflavin metabolism and immunotoxicity to a certain extent. Previous research found that exposure to environmental toxicants, including vanadium, chromium, and thallium, disrupts metabolic homeostasis via oxidative stress, inflammation, and dysregulation of key signaling pathways, consequently elevating the risk of metabolic syndrome, type 2 diabetes, and cardiovascular diseases^56–58. Arsenic also has a similar toxic mechanism. In this study, arsenic exposure did not significantly affect riboflavin transport, this observation may be attributable to the relatively low arsenic dose or the limited sample size. However, riboflavin supplementation significantly upregulated RFK and FLAD1 expression levels. This suggests that exogenous riboflavin may confer protection against arsenic toxicity by enhancing metabolic flux and coenzyme biosynthesis. Therefore, riboflavin supplementation represents a feasible strategy for restoring reproductive homeostasis and alleviating male reproductive damage induced by environmental toxins.

Several studies indicate that the TGF-β/Bmp4/Smad pathway orchestrates spermatogenesis through phosphorylation-dependent modulation of AJs and TJs, while concurrently regulating Sertoli cell proliferative capacity, which may be a key regulatory pathway in the mechanism of arsenic poisoning^18,59. The TGF-β/BMP superfamily signaling crosstalks with cytoskeletal regulators, particularly Rho family GTPases (RhoA, Rac1, and Cdc42), which act as master switches governing actin dynamics and junctional assembly⁶⁰. Lee-Hoeflich ST et al.⁶¹ demonstrated that BMP signaling organizes the actin cytoskeleton via Rho GTPases in diverse cell types, including neurons and endothelial cells. Similarly, arsenic exposure appears to disrupt the functional crosstalk between BMP signaling and cytoskeletal integrity. Although many researchers have tried to explore the effects of arsenic induction on the TGF-β/Bmp4/Smad pathway, the results have been unsatisfactory. Qiu et al.⁶² discovered that exposure of male Sprague-Dawley rats to 5 mg/kg body weight NaAsO₂ orally daily for 9 months upregulated TGF-β1 mRNA expression and p-Smad2/3 protein in liver tissue and hepatic stellate cells (HSCs), ultimately leading to liver fibrosis. Nevertheless, it was also evidence that As₂O₃ suppresses TGF-β/Smad2 pathway activation through the modulation of microRNA-491 and microRNA-155, consequently influencing angiogenesis^63,64. The aforementioned studies indicate that arsenic may play a bidirectional regulatory role in modulating the TGF-β/Smad pathway. Intriguingly, our study found that arsenic exposure had no significant effect on TGF-β expression levels in mouse testicular tissue. As a key gene of the TGF-β signaling pathway, Bmp4 is critical for murine embryonic development and male reproductive function⁶⁵. Yang et al.⁶⁶ observed stage-dependent Bmp4 expression in spermatocytes and Sertoli cells during murine spermatogenesis, with expression levels escalating alongside germ cell maturation. These results imply that Bmp4 plays a pleiotropic role in reproduction. In this study, qRT-PCR and IF results revealed that riboflavin enhanced the downregulation of Bmp4 and Bmpr1α expression levels in mouse testicular tissue and TM4 cells caused by arsenic. The Bmp4, Bmpr1α, Smad4 and Smad5 all demonstrated binding ability with riboflavin, thus the Bmp4/Smad signaling axis identified as the core target pathway. In addition, the use of DMH-1 further confirmed that arsenic reduced the level of Bmp4, Bmpr1α and Smad5 proteins expression in testicular Sertoli cells. This indicates that arsenic may cause reproductive toxicity by interfering with the binding of Bmp4 to its receptor Bmpr1α, in which riboflavin plays a protective role.

In conclusion, our study demonstrates that arsenic impairs sperm quality and the morphological structure of tissues, compromises the BTB and the Sertoli cells’ cytoskeleton, which may together contribute to male reproductive toxicity. Riboflavin supplementation was observed to mitigate these arsenic-induced effects, correlating with upregulation of riboflavin transformation enzymes and activation of Bmp4/Bmpr1α (Fig. 8). Although further studies are needed to establish direct causality, this study provides a new perspective on arsenic-induced male reproductive disorders and the potential protective role of riboflavin.

Fig. 8. Schematic diagram of arsenic-induced male reproductive toxicity and riboflavin protective function via Bmp4/Bmpr1α signaling (Created with BioGDP.com).

Both in vivo and in vitro experiments demonstrated that arsenic exposure induces male reproductive toxicity, primarily characterized by decreased sperm quality, impaired blood–testis barrier integrity, and inhibited proliferation of Sertoli cells or TM4 cells. Mechanistically, riboflavin mediates the binding of Bmp4 to Bmpr1α and participates in riboflavin metabolism, thereby effectively reversing these arsenic-induced adverse effects. Note: Upward red arrows (↑) represent upregulation or enhancement; downward black arrows (↓) represent downregulation or inhibition.

Methods

Reagents and antibodies

Arsenic trioxide (≥99.0% purity, A1010) was purchased from Sigma (MO, USA), and riboflavin (≥99.0% purity, R104137) was obtained from Aladdin (Shanghai, China). Triton X-100 (ST795), 4,6-diamidino-2-phenylindole (DAPI, C1005) and SOX9 antibody (AF2329) were purchased from Beyotime (Shanghai, China). Rhodamine-phalloidin (RM02835), anti-CX43 (A11752), anti-ZO-1 (A28491), anti-N-Cadherin (A3045SP), anti-Bmp4 (A21065), anti-Bmpr1α (A1816), and anti-RFK (A9141) antibodies were provided by ABclonal (Wuhan, China). Anti-FLAD1 (YT6889), anti-GAPDH (YN5697), anti-β-actin (YT0099) and secondary antibodies (Goat Anti-Rabbit IgG or Goat Anti-Mouse IgG) were obtained from Immunoway (Beijing, China). Occludin antibody (WL01996), Smad4 antibody (WL02049), Smad5 antibody (WL00277), and Tubulin antibody (GB15200) were acquired from WanleiBio (Shenyang, China) and Servicebio (Wuhan, China), respectively.

Animals and treatments

A total of thirty-six 4-week-old male Kunming mice (KM, body weight 18–20 g) were procured from SPF (Beijing) Biotechnology Co. Ltd. (Beijing, China) and raised in the animal house of Shanxi Agricultural University under the standard environmental conditions (12 h light/dark cycle, temperature at 22 ± 2 °C, and relative humidity at 55 ± 5%). After a 7-day acclimation period, the mice were assigned randomly to four groups (9 mice/group), which included the As group, Ribo group, As+Ribo group, and Control group. The As group was supplied with deionized water containing 5 mg/L arsenic trioxide (As₂O₃; approximately 3.79 mg/L elemental arsenic) ^67,68 and had normal access to food. The Ribo group received 10 mg/kg riboflavin supplement⁶⁹ and had regular access to water. The As+Ribo group was given deionized water with 5 mg/L arsenic trioxide and a dietary supplement of 10 mg/kg riboflavin. Meanwhile, the Control group did not receive any treatment, drinking water and eating freely.

Throughout the 22-week experiment, all mice had free access to food and water before being humanely euthanized by cervical dislocation⁷⁰ at the end of the study. The left testicle and epididymis tissues of each mouse were fixed in Bouin’s solution for histological observation, and the remaining testicle tissues were collected quickly and stored in a refrigerator at −80 °C after snap-freezing with liquid nitrogen for other tests. All animal experiments were conducted in strict compliance with the guidelines and regulations set forth by the Laboratory Animal Ethics Committee of Shanxi Agricultural University (Approval No: SXAU-EAW-2021M.XB.012019311).

Cell culture and viability analysis

TM4 cells (ZQ0091) were provided by Shanghai Zhong Qiao Xin Zhou Biotech Co., LTD. (Shanghai, China). The cells were cultured in complete medium (1% 100× penicillin-streptomycin solution, 2.5% fetal bovine serum, 5% equine serum and 91.5% DMEM/F12) and placed in an incubator at 37 °C with 5% CO₂. The CCK-8 kit was acquired from Abbkine Scientific Co., Ltd. (Wuhan, Hubei, China) to evaluate TM4 cell proliferation. Briefly, the TM4 cells were subjected to a 96-well plate (1 × 10⁴ cells/well) and cultured for 24 h. Afterwards, the cells were exposed to specific conditions for another 24 h: Control (untreated cells), various concentrations of arsenic trioxide (2, 4, 6, or 8 µM As₂O₃), different doses of riboflavin (0.5, 1, 2, 4, 6, 8, 10, or 16 µM Ribo), or co-treatment with 3 µM As₂O₃ and various concentrations of riboflavin (0.5, 1, 2, 4, 6, 8, 10, or 16 µM Ribo). Finally, the cells were suspended in complete medium containing a 10% CCK-8 solution and cultured at 37 °C with 5% CO₂ for 1–4 h. Each group was repeated three times, and cell viability was evaluated in accordance with the manufacturer’s guidelines.

Network toxicology visualization and molecular docking analysis

A search for potential target genes related to “Arsenic” and “Male infertility” was conducted using the Comparative Toxicogenomics Database (CTD, https://ctdbase.org/; CTD-As inference score ≥ 20; CTD-Male infertility reference score ≥100) and GeneCards (GC, https://www.genecards.org/; GC-As relevance Score ≥ 1; GC-Male infertility relevance Score ≥ 10), and the genes retrieved from the two databases were cross-analyzed to obtain potential target genes for As-induced male infertility. The target genes were imported into the DAVID database (https://davidbioinformatics.nih.gov/) for gene ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Statistical significance was determined using a false discovery rate (FDR) of < 0.05 and P < 0.05. The 10 most enriched GO terms and the 15 most significantly enriched KEGG pathways were listed. Subsequently, the results of analysis by bioinformatics (https://www.bioinformatics.com.cn/) were visually displayed. In addition, a protein-protein interaction (PPI) network was constructed for the screened potential target genes using STRING database (https://cn.string-db.org/) with a confidence score threshold set to 0.4, and the network was visualized using Cytoscape 3.10.0. The structures of riboflavin (CHEMBL1534) and target proteins (Bmp4, P21275; Bmpr1α, P36895; Smad4, P97471; Smad5, P97454) were obtained respectively from the ChEMBL database (https://www.ebi.ac.uk/chembl/) and Uniprot database (https://www.uniprot.org/), and molecular docking was performed through online tool CB-Dock2.

Sperm quality assessment

Sperm quality was evaluated using the standard semen collection methods established in our laboratory^71,72. In brief, the mouse epididymis caudal was taken, sectioned into small pieces, and placed in 1 mL preheated normal saline water for incubation at 37 °C for 15 min to obtain sperm suspension. Subsequently, the sperm suspension was counted using a hemocytometer to determine sperm density and motility. All counts were performed in a double-blinded manner. Meanwhile, 10 μL of the sperm suspension was evenly spread onto a clean, dry microscope slide, allowed to air-dry, and then fixed in methanol for 5 min. After staining with eosin, a total of 200 spermatozoa per sample were randomly examined under an optical microscope (Olympus BX53, Tokyo, Japan) for morphological assessment and counting of normal and abnormal sperm^73,74.

Morphological observation

The testicles and epididymis were immobilized in Bouin’s solution for 24–48 h, washed with running water for 12 h, and then dehydrated with an increased concentration of gradient ethanol solution (70% to 100%). Subsequently, the tissues were made transparent with xylene and then encased in paraffin to prepare 5 μm-thick sections. TM4 cells (1 × 10⁵ cells/mL) were inoculated into each well of the 6-well plates according to different groups, with 3 coverslips placed at the bottom of each well. Following a 24 h incubation, the cover slips were extracted from the wells, washed with PBS, and then placed in 95% ethanol for 20 min. The tissue sections and cell slides were treated with xylene and gradient alcohol, then stained with hematoxylin-eosin (H&E), and observed with an optical microscope (Olympus BX53, Tokyo, Japan) to determine the pathological changes. From each group, three mice were randomly chosen, at least 100 seminiferous tubules were randomly selected from each group of testicular sections, and the outer diameter, inner diameter, and thickness were measured⁷⁴.

Immunohistochemical staining (IHC)

Testicular tissue sections were heated in an incubator at 60 °C for 20 min, dewaxed with xylene, hydrated in a gradient alcohol solution, and permeabilized with 0.1% Triton X-100 for 20 min. After this, they were incubated in a citrate buffer at 98 °C for 15 min before being allowed to cool to room temperature (RT). The endogenous peroxidase activity in the tissue sections was inhibited by treating with 3% H₂O₂ for 40 min, and then incubated for 30 min with normal goat serum as a blocking reagent at 37 °C. They were then incubated with SOX9 antibody (1:200) overnight at 4 °C, while the negative control group was replaced by PBS solution (0.01 mol/L), and the other steps remained the same. Subsequently, the sections were treated with a horseradish peroxidase-conjugated Goat Anti-Rabbit IgG (1:2000) at 37 °C for 1 h, followed by 3, 3’-Diaminobenzidine (DAB) staining and hematoxylin counterstaining. Sections were treated with a gradient alcohol solution and xylene, and then mounted with a neutral adhesive. At least 3 replicates were performed for each tissue, observed with an optical microscope, and randomly selected multiple different view fields for analysis. The positive cells appeared yellow or brownish yellow. The staining intensity was analyzed by Image-Pro Plus software (version 6.0, Media Cybernetics Inc., USA), with the results expressed as the mean optical density value.

Fluorescent phalloidin staining

For the filamentous actin (F-actin) observations, after rinsing TM4 cells with PBS, they were treated with 4% formaldehyde solution and 0.1% Triton X-100 for 15 min and 5 min, respectively. The TM4 cells were then stained with rhodamine-phalloidin (1:100) for 45 min to label F-actin and with DAPI for 15 min to label nuclei. Finally, fluorescence images were recorded using an Olympus BX53 (Tokyo, Japan), and ImageJ software (NIH, USA) analyzed the fluorescence intensity for quantification.

Immunofluorescence (IF)

The TM4 cells and tissue slides were treated with 4% paraformaldehyde for fixation, permeabilized with 0.1% Triton X-100 for 15 min, and sealed with 3% BSA solution for 30 min at RT. The testicular slides were incubated with primary antibodies at 4 °C overnight, including Bmp4 (1:150), Bmpr1α (1:150), Occludin (1:150), and CX43 (1:100), while the TM4 cell cultures were incubated with Tubulin (1:200) overnight at 4 °C. After washing 3 times with PBS, the secondary antibody (1:200) was added, and the slides were incubated for 2 h at RT. Then, nuclei were counterstained with DAPI staining solution. Finally, the results were photographed by a fluorescence microscope (Olympus BX53, Tokyo, Japan). Three mice per group were chosen at random, and at least five fields of view per section were examined. The fluorescence intensity was quantified using ImageJ software (NIH, USA).

Quantitative real-time PCR (qRT-PCR)

Total RNA was extracted from testicular tissue with the Trizol reagent. Reverse transcription of mRNA to cDNA was performed by the TransScript® Uni All-in-One First-Strand cDNA Synthesis SuperMix for qPCR kit (TransGen, China). Then, the expression of the target gene was detected using the PerfectStart® Green qPCR SuperMix kit (TransGen, China) on the MX3000P PCR instrument (Stratagene, USA). Meanwhile, Primer 3.0 Plus designed primer sequences of target genes, which were synthesized by Invitrogen Trading Co., Ltd. (Shanghai, China) and are listed in Supplement Table 1. The relative multiple difference of target gene level between groups was calculated using the 2^−∆∆CT method, with β-actin serving as the calibrator⁷⁵.

Western blot

After extracting total protein from testicular tissue and TM4 cells, the concentration of total protein was determined by the BCA kit (Boster, Wuhan, China). The protein sample solutions of the same concentration (40 μg) were mixed with the loading buffer (Solarbio, Beijing, China) in a ratio of 5:1 proportion, boiled for 10 min to denature, and separated by 10% SDS polyacrylamide gel electrophoresis. Following this, the target proteins were transferred onto a pore size of 0.45 μm nitrocellulose membrane (Boster Biological Technology, China), and the membrane was blocked with 5% fat-free milk at RT for 2 h. After rinsing with TBST three times, the protein membranes were incubated overnight at 4 °C with the following proportions of primary antibodies: Occludin (1:1000), CX43 (1:1000), ZO-1 (1:1000), N-Cadherin (1:1000), SOX9 (1:1000), Bmp4 (1:1000), Bmpr1α (1:1000), RFK (1:1500), FLAD1 (1:1500), Smad4 (1:1000), Smad5 (1:1000), GAPDH (1:8000) and β-actin (1:5000). Before antibody incubation, the membrane was cut into strips for the detection of loading control proteins (β-actin or GAPDH), and the relative band intensities of the target proteins were standardized against β-actin or GAPDH levels. Then, the membranes were treated with Goat Anti-Rabbit IgG (1:5000) for 2 h at RT after washing 3 times with TBST for 10 min each. Finally, the target bands were treated with an enhanced chemiluminescence (ECL, Mylen Biotechnology Co., Ltd., China) and imaged with the FluorChem Q imaging system (Protein Simple, California, USA). Densitometric analysis was performed using ImageJ software (NIH, USA).

Statistics

All experimental group samples had at least three parallel replicates, with the value being mean ± SEM (standard error). A two-tailed Student’s t-test was used for two-group comparisons. For multiple group comparisons, data were first assessed for normality (Shapiro-Wilk test) and homogeneity of variance test (Levene’s test), followed by one-way ANOVA, with Tukey’s HSD post-hoc test applied to control the overall error rate. Statistical mapping was performed using GraphPad Prism 8.0 Software (GraphPad Software Inc., San Diego, CA, USA). p < 0.05 indicates statistical significance between groups. “*” denotes a notable discrepancy between the As group or Ribo group and the Control group, while “#” indicates a notable difference between the As group and the As + Ribo group.

Author contributions

Conceptualization, Y.G.L., J.B., C.L., and J.H.Z.; methodology, Y.X.Z., Y.G.L.,Q.X., X.C.S., and W.Y.; formal analysis, Y.X.Z., C.L., and Y.G.L.; software, Q.X., Y.F.Z., and X.C.S.; investigation, Y.G.L., X.C.S., and W.Y.; data curation, Y.X.Z., Y.G.L., and J.H.Z.; resources, J.B., C.L., M.Y.C., and J.H.Z.; visualization, Y.X.Z., Y.G.L., and Y.F.Z.; validation, X.C.S. and W.Y.; writing—original draft preparation, Y.X.Z. and Y.G.L.; writing—review and editing, Y.X.Z., J.B., C.L., M.Y.C., and J.H.Z.; supervision, J.H.Z.; funding acquisition, J.H.Z.; project administration, J.H.Z. All authors have read and agreed to the published version of the manuscript.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available due to the data related to the application of patented product research and development, but are available from the corresponding author on reasonable request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41538-026-01162-2.