The electroejaculation method was used to collect semen from four healthy male yaks (3–5 years old) at the Longri Breeding Stock Farm in Sichuan Province. Initially, the yaks were guided onto a semen collection platform to maintain an upright posture. Then, electroejaculation was applied to the sacral nerves via an electrode rod positioned at the yak’s tail. Generally, the voltage reaches 24 to 30 volts, yak semen can be effectively collected. The voltage is gradually increased by 2 volts at each step. At each voltage level, the output is continuously maintained for 6 to 8 s, followed by a pause of approximately 4 s. This procedure is repeated twice. The semen collection frequency was limited to a maximum of twice a week, with an average semen yield of 5.2 mL per yak. The semen was collected in centrifuge tubes, initially diluted 5–10 fold with Optidyl® diluent (020996, IMV, France) based on its density, and then transported to the laboratory (22–24).

Artesunate (ART, A800614, Macklin, Shanghai) was initially dissolved in DMSO (D8371, Solarbio, Beijing) as the stock solution. Subsequently, the Optidyl® diluent was employed to precisely dilute the solution to final concentrations of 0.2 mM, 0.5 mM, and 1 mM. All sperm samples were sealed and subjected to a programmed freezing protocol (TYLDP1, Tianyuanaorui, Beijing; Supplementary Figure 1). After freezing, straws were stored in insulated liquid nitrogen tanks. For thawing, straws were rapidly removed from liquid nitrogen and immersed in a 37°C water bath for 30 s. After thawing, the straws were cut open, and the thawed sperm were expelled for further analysis.

Sperm samples were divided into five groups, including fresh group, frozen sperm group (NC, 0 mM ART), and three frozen groups respective with 0.2 mM, 0.5 mM and 1.0 mM ART addition. Equal volumes (10 μL) of fresh and frozen–thawed semen were smeared onto slides for testing (Slide: Sail brand, Guangzhou. Cover glass: Citotest, Nantong). The sperm motility parameters were analyzed using AndroVision® CASA software (Minitube, Germany) and a microscope (BX53, Olympus, Japan). The CASA parameters were presented in Supplementary Figure 2. Each sample was analyzed in at least five fields, with over 1,000 sperm counted per sample. Final motility parameters were calculated as the average of the measurements (25).

Hypotonic swelling solution, used to assess sperm plasma membrane integrity, was prepared by dissolving 0.9 g of fructose and 0.49 g of sodium citrate in 100 mL of distilled water (26). For each group, 100 μL of semen was mixed with 1 mL of hypotonic solution and incubated in a water bath at 37°C for 30 min. Sperm tail morphology was observed under a BX53 microscope (BX53, Olympus, Japan). Sperm exposed to a hypotonic environment will pump water into their plasma membranes. This increases the sperm’s volume-to-surface area ratio, leading to plasma membrane swelling. Intact plasma membranes caused sperm tails to curl due to water absorption, while damaged membranes remained straight (27). At least five random visual fields were selected for each sample, with a total sperm count greater than 1000, and the ratio of curled-tail sperm to the total sperm count was calculated.

The 1:1 (v/v) mixture of 100 μg/mL FITC-PNA (L7381, Sigma, United States) and Hoechst 33258 (C0031, Solarbio, Beijing) was prepared to evaluate the integrity of acrosomal membranes. Each 50 μL of semen was stained with 2.5 μL of the mixture, and smears were prepared on glass slides. The smears were air-dried and then incubated at 37°C for 10 min. Fluorescence images were captured using an Axio Observer microscope (ZEISS, Germany). FITC-PNA binds to acrosomal membrane galactose residues, emitting green fluorescence, while Hoechst 33258 stains nuclei blue (28). At least five random fields were analyzed per sample, with over 1,000 sperm counted. The ratio of intact acrosomal sperm to total sperm was calculated.

Equal volumes (200 μL) of fresh and thawed frozen semen were processed following the instructions of the ROS detection kit (S0033S, Beyotime, Shanghai). Fluorescence intensity was measured using a SpectraMax® ID3 microplate reader (Molecular Devices, Shanghai) with excitation and emission wavelengths set at 488 nm and 525 nm, respectively.

Equal volumes (200 μL) of fresh and thawed frozen semen were processed using assay kits for superoxide dismutase (A001-2, Jiancheng, Nanjing), catalase (A007-1, Jiancheng, Nanjing), glutathione peroxidase (A005-1, Jiancheng, Nanjing), and malondialdehyde (A003-1, Jiancheng, Nanjing). Absorbance values were measured at 550 nm, 405 nm, 412 nm, and 532 nm, respectively, using a SpectraMax® ID3 microplate reader (Molecular Devices, Shanghai). Calculations were performed as per kit instructions. The calculation formulas are as follows:

Note: V_Total, 3.5 mL. V_Sample, 0.2 mL. Cpr, Concentration of protein.

Note: V_Sample, 0.05 mL. T, 60 s. Cpr, Concentration of protein.

Note: T, 5 min. Standard substance concentration, 20 μmol/L.

Note: Standard substance concentration, 10 nmol/mL.

Mitochondrial membrane potential was assessed (250 μL semen samples collected from each group), following the instructions of the mitochondrial membrane potential detection kit (C2006, Beyotime, Shanghai). Stained samples were observed and photographed using an Axio Observer fluorescence microscope (ZEISS, Germany), and fluorescence intensity was quantified using a SpectraMax® ID3 microplate reader (Molecular Devices, Shanghai).

ATP content was determined by lysing semen samples in RIPA buffer (G2002, Servicebio, Wuhan) supplemented with protease inhibitors (M5293, AbMole BioScience, United States). ATP levels were measured at 636 nm using an ATP assay kit (A095-1-1, Jiancheng, Nanjing). The calculation formula is as follows:

Note: Standard concentration, 1000 μmol/L; Cpr, protein concentration.

Total protein was extracted from semen using RIPA buffer (G2002, Servicebio, Wuhan). Protein concentration was quantified using a BCA kit (PD101-500T, Oriscience, Chengdu). Proteins were separated by SDS-PAGE, transferred to PVDF membranes, and probed with primary antibodies (SOD1, A0274, ABclonal; GAPDH, GB12002-100, Servicebio) at 1:1000 dilution. Secondary antibodies (Goat Anti-Rabbit IgG HRP, S0001, Affinity; HRP-conjugated Goat Anti-Mouse IgG, SA00001-1, Proteintech) were used at 1:5000. Bands were visualized using an A32747 gel imager (Invitrogen, United States).

The immunofluorescence staining was performed as described (7). In brief, each sample was fixed in 4% PFA for 20 min and washed with PBS. 100 mM glycine was added to the samples and allowed to stained for 10 min before another PBS wash. The samples were then treated with blocking buffer for 1 h, followed by the addition of primary antibody (SOD1, A0274, ABclonal, Wuhan) at a dilution of 1:200 (v/v), and incubated overnight at 4°C. After washing with PBS, the samples were incubated with secondary antibody goat anti-rabbit IgG (A11011, Invitrogen, United States. 1:1000, v/v) and DAPI for 1 h at 4°C. Finally, the samples were observed under a fluorescence microscope (Axio Observer, ZEISS, Germany).

The 90 μL of semen was taken and placed in a centrifuge tube. The procedure was performed according to the instructions provided by the AO/EB (Acridine Orange/ Ethidium Bromide) staining kit (E607308, Sangon Biotech, Shanghai). The stained samples were then photographed under a fluorescence microscope (Axio Observer, ZEISS, Germany). As described in the instructions, AO has cell permeability and can enter living cells, resulting in uniform green fluorescence in living cells. However, cells with DNA damage exhibit bright green fluorescence due to chromatin clotting and nuclear breakage. For statistical analysis, at least five visual fields were randomly selected from each sample, with a total sperm count of over 1000. The ratio of sperm with intact DNA to the total sperm count was calculated for the selected fields.

Western blot and Immunofluorescence results were analyzed using Image J (version 1.53). Data were presented as Mean ± SEM. Statistical analyses were performed using GraphPad Prism software (version 10) and one-way ANOVA, with p < 0.05 considered significant. The correction method for multiple comparisons is Tukey’s multiple comparisons test. Different uppercase letters on the same shoulder in the same column indicated significant extremely significant difference (p < 0.01). Different lowercase letters on the shoulder indicated significant differences (p < 0.05). Unlabeled letters indicated no significant difference (p > 0.05).

Computer-assisted sperm analysis (CASA) was used to evaluate sperm motility among different groups. Fresh sperm exhibited high motility, with total motility exceeding 80% (Figure 1a) and progressive motility around 75% (Figure 1b), significantly higher than the frozen–thawed control group (p < 0.01). In contrast, untreated frozen sperm showed markedly reduced motility, with approximately 40% total motility (Figure 1a) and 20% progressive motility (Figure 1b). Notably, 0.2 mM ART restored total and progressive motility to ~40% (Figure 1b; p < 0.05; p < 0.01), while significantly reducing the proportion of static sperm (Figure 1d; p < 0.05). Higher ART concentrations (0.5 mM and 1.0 mM) suppressed sperm motility (Figures 1a–d), potentially due to osmotic imbalance caused by excessive ART. These findings indicate that 0.2 mM ART optimally enhances the motility of yak frozen–thawed sperm.

The hypotonic swelling test (HOST) was conducted to evaluate sperm plasma membrane integrity, classifying sperm with curled tails as “intact” and those with straight tails as “damaged” after hypotonic exposure (Figure 2a). Cryopreservation significantly reduced the proportion of sperm with intact plasma membranes (Figure 2b; p < 0.01). However, treatment with 0.2 mM and 0.5 mM ART significantly increased this proportion to over 50% (Figure 2b; p < 0.01). Acrosomal integrity, evaluated via FITC-PNA staining, revealed higher integrity in fresh sperm and in groups treated with 0.2 mM and 0.5 mM ART, whereas severe acrosomal damage was observed in the control and 1.0 mM ART groups (Figure 2c). Statistical analysis confirmed that the fresh, 0.2 mM, and 0.5 mM ART groups exhibited significantly greater acrosomal integrity than the control group (Figure 2d; p < 0.01). These results indicate that appropriate ART concentrations effectively preserve the membrane integrity of frozen–thawed yak sperm.

Reactive oxygen species (ROS) and malondialdehyde (MDA) content, which are key indicators of oxidative stress, were measured via fluorescence probe and thiobarbituric acid (TBA) assays, respectively. Fresh sperm exhibited minimal ROS fluorescence, whereas frozen–thawed sperm showed intense signals, particularly in the head and neck regions (Figure 3a). Quantitative analysis revealed that 0.2 mM ART reduced ROS levels compared to frozen–thawed controls (Figure 3b; p = 0.0516). Conversely, the 1.0 mM ART group showed the highest ROS fluorescence intensity with ~4 times higher than other groups. Oxidative stress-induced lipid peroxidation, assessed via MDA content, was elevated in the control group (Figure 3c; p < 0.01). Notably, 0.2 mM ART treatment markedly decreased MDA levels compared to the control (Figure 3c; p < 0.01). These findings indicate that 0.2 mM ART effectively attenuates oxidative stress in frozen–thawed yak sperm by reducing ROS generation and lipid peroxidation.

The enzyme activities of total superoxide dismutase (T-SOD), catalase (CAT) and glutathione peroxidase (GSH-PX) were detected using the colorimetric method, the ammonium molybdate method and the enzymatic reaction method, respectively. As shown in Figure 4a, freezing resulted in significant downregulation of the activities of three enzymes in sperm (p < 0.01). Notably, treatment with 0.2 mM ART significantly restored T-SOD activity (Figure 4a; p < 0.01), but not that of CAT and GSH-PX (p > 0.05). In particular, ART treatment specifically elevated the activity of SOD1 (Figure 4b; p < 0.01), while SOD2 activity was not significantly affected by ART (Figure 4c; p > 0.05). Western blot and immunofluorescence analysis confirmed a decrease in SOD1 protein in the control group compared to fresh group, and treatment with 0.2 mM ART partially rescued this trend in the sperm (Figures 4d–g; p < 0.01; p < 0.05). In addition, immunofluorescence imaging revealed that SOD1 is primarily localized in the sperm head and neck regions (Figure 4f). Collectively, these data suggest that ART alleviates oxidative stress primarily through maintaining the SOD1 level.

SOD1 is localized in the sperm head, which is implicated in DNA integrity (29), we evaluated it in sperm. AO staining revealed uniform green fluorescence in fresh sperm, whereas frozen sperm exhibited increased bright green fluorescence spots and irregular fluorescence distribution, indicative of DNA damage. ART treatment significantly reduced these fluorescence abnormalities (Figure 5a). Statistical analysis confirmed a significantly lower proportion of sperm with intact DNA in the control group compared to fresh sperm (Figure 5b; p < 0.01). Notably, ART treatment markedly reduced frozen–thawed sperm DNA damage (Figure 5b; p < 0.01). These findings demonstrate that ART exerts antioxidant effects by upregulating SOD1, thereby preserving DNA integrity.

Based on prior evidence that SOD1 is localized to the sperm neck, a region rich in mitochondria (30), we evaluated the mitochondrial status in sperm. JC-1 staining revealed red fluorescence as the dominant signal in the mitochondria of both the fresh group and the ART-treated group (Figure 6a). However, freezing induced a significant increase in green fluorescence within sperm (Figure 6a). Specifically, ART was found to maintain the mitochondrial membrane potential (MMP) in frozen sperm (Figure 6b; p < 0.01). Furthermore, ART treatment resulted in a higher ATP content in sperm compared to the NC group (Figure 6c; p < 0.01). Collectively, these findings suggest that ART exerts antioxidant effects by upregulating SOD1, ultimately preserving mitochondrial integrity and functionality.