Zinc sulfate (ZnSO₄, S22124, Shanghai Yuanye Bio-Technology, Shanghai, China) and zinc gluconate (Z100287, Shanghai Aladdin Biochemical Technology, Shanghai, China) were prepared as stock solutions at a concentration of 20 mmol/L in H₂O and stored at 4 °C. Zinc oxide nanoparticles (ZnO-NPs, 30 nm; Z820772, Macklin, Shanghai, China) were dispersed in phosphate-buffered saline (PBS) to a final concentration of 2.5 mg/mL. The resulting suspension was vortexed for 1 min (VORTEX-5, Kylin-Bell, Haimen, China) and sonicated at 40 kHz for 20 min (SCIENTZ, Ningbo, China), after which it was kept refrigerated (4 °C) until use. Immediately before addition to the cells, the particles were vortexed for 5 min to resuspend them and then brought to the desired working dilutions with culture medium.

The enhanced green fluorescent protein (eGFP)-TSPAN4 [Research Resource Identifier (RRID): Addgene_210323] was supplied by Professor Li Yu from Tsinghua University. The cells were transfected with FuGENE HD (E2311, Promega, Madison, WI, USA) following the supplier’s instructions. Cholesterol (HY-N0322), disulfiram (HY-B0240), and filipin complex (HY-N6716) were purchased from MedChemExpress (Princeton, NJ, USA). Wheat germ agglutinin (WGA) conjugated to fluorescein isothiocyanate (FITC) (RRID: AB_1657336, GTX01502) was supplied by GeneTex (Irvine, CA, USA). Anti-solute carrier family 39 member 6 (SLC39A6) (RRID: AB_11187270, Ab176216) was from Shanghai Aladdin Biochemical Technology (Shanghai, China). N-(6-methoxy-8-quinolyl)-p-toluenesulfonamide (TSQ) (21254, AAT Bioquest, Pleasanton, CA, USA) served as the Zn²⁺ dye.

The human hepatoma line Huh7.5.1 (RRID: CVCL-E049) was supplied by Professor Francis V. Chisari from Scripps Research Institute. Cells were propagated in high-glucose Dulbecco’s modified Eagle’s medium (DMEM) (SH30022.01, GE Healthcare/Hyclone/whatman/Cytiva, Marlborough, MA, USA) with 10% fetal bovine serum (FBS) (FSP500, ExCell Bio, Shanghai, China), 100 U/mL penicillin, and 100 μg/mL streptomycin (SV30010, GE Healthcare/Hyclone/whatman/Cytiva, Marlborough, MA, USA). Cultures were subcultured every 2-3 days after detachment with 0.25% Trypsin-EDTA (25200-056, Gibco, Thermo Fisher Scientific, Waltham, MA, USA), with routine mycoplasma testing. Transfection was performed at about 70% confluence in a 2-cm dish using 1 μg DNA and 2.5 μL FuGENE HD.

Cells were seeded into 96-well plates at 1 × 10⁴ cells/well and allowed to attach for 24 h. The medium was then replaced with zinc gluconate or ZnSO₄ at final concentrations of 0-30 μmol/L, and cells were incubated for a further 24 h. Subsequently, 10 μL CCK-8 reagent (RRID: AB_2314126, Proteintech, Wuhan, China) was added per well, followed by a 1-h incubation. Optical density was read at 450 nm on a Multiskan GO microplate reader (Thermo Fisher, USA).

To visualize migrasomes, cells grown on glass coverslips (YA0350, Solarbio, Beijing, China) were fixed using 4% paraformaldehyde (P1110, Solarbio, Beijing, China) and then treated with WGA (2 μg/mL) or filipin (40 μmol/L) for 30 min. An ECHO RVL2-K microscope (RRID: SCR_026523, USA) was used to capture fluorescence images.

Huh7.5.1 cells were plated in 6-well plates and maintained for 24 h before treatment. The cultures then received 20 μmol/L ZnSO₄ or zinc gluconate for 17 h, with untreated wells included as controls. After being rinsed twice with PBS, cells were exposed for 30 min at 37 °C under 5% CO₂ to a phenol red-free DMEM solution containing 10 μmol TSQ and 2 μg/mL WGA. Cells were then washed three more times with PBS, resuspended in phenol red-free DMEM, and visualized with an ECHO RVL2-K microscope (USA).

Confluent Huh7.5.1 monolayers were prepared in 6-well plates using DMEM supplemented with 10% FBS. Each well was scratched down the center with a P1000 tip, after which the medium was promptly exchanged for serum-free DMEM with zinc gluconate, ZnSO₄, or no supplement (blank). Brightfield images of the identical region of interest (ROI) were captured with an inverted microscope (Novel Optics, Nanjing, China) at 1, 2, and 3 days post-scratch to monitor the leading edge. Migration distance was determined in ImageJ (RRID: SCR_003070) for each time point.

Intracellular ROS levels were assessed with 2’,7’-dichlorodihydrofluorescein diacetate (DCFH-DA). Huh7.5.1 cells were exposed to zinc gluconate or ZnSO₄ for 17 h, rinsed with PBS, and stained for 20 min at 37 °C using the DCFH-DA working solution supplied in a ROS detection kit (RRID: AB_2574932, S0034S, Beyotime, Shanghai, China). After PBS washes, the cells were maintained in DMEM without phenol red. 2’,7’-dichlorofluorescein (DCF) fluorescence was captured at 488 nm excitation on an ECHO RVL2-K microscope (USA).

Huh7.5.1 cultures were exposed to zinc gluconate or ZnSO₄ for 17 h and then rinsed with PBS (pH 7.2-7.4). The cells were scraped into PBS and lysed by repeated freeze-thaw cycles; after centrifugation at 3,000 rpm for 20 min, the supernatant was collected. Aliquots were transferred to the assay wells, sealed with film, and kept at 37 °C for 60 min. Following five wash steps and blotting dry on clean absorbent paper, 100 μL of 3,3’,5,5’-tetramethylbenzidine (TMB) color developer from the sphingomyelinase enzyme-linked immunosorbent assay (SPH-ELISA) detection kit (A101756-48T, Fusheng Industrial, Shanghai, China) was applied, and the plates were held at 37 °C for 15 min. Stop solution (50 μL) was then added to each sample, and absorbance at 450 nm was recorded on a Multiskan GO microplate reader (Thermo Fisher, USA).

All animal work was approved by the Animal Ethics Committee of Shandong First Medical University (Approval No. W202504160624). Eight-week-old male BALB/c mice (RRID: MGI:2161072) were obtained from Jinan Pengyue Experimental Animal Breeding (Shandong, China) and housed under specific pathogen-free (SPF) conditions at 23 ± 1 °C with 60 ± 10% relative humidity. After one week of acclimation, the animals were allocated randomly to three groups. Migrasome counting was performed in 11 mice per group, and platelet counting and bleeding volume measurements were performed in 6 mice per group. At baseline, a 20-μL aliquot of diluted tail blood was combined with 380 μL platelet diluent (R20341, Shanghai Yuanye Bio-Technology, Shanghai, China), and platelets were enumerated in a hemocytometer. Tail blood (3-5 μL) was also used to prepare blood films, which were fixed in methanol for 10 min and then stained with Wright’s solution for 10 min (C0135, Beyotime, Shanghai, China). Zinc gluconate or ZnSO₄ (17 μmol/kg in saline) was then administered by the intraperitoneal route. 24 h later, tail-tip blood was sampled; platelet counts and stained blood films were obtained as at baseline.

Male BALB/c mice (n = 11 per group), eight weeks old, were given an intraperitoneal dose of zinc gluconate (17 μmol/kg), ZnSO₄ (17 μmol/kg), or physiological saline. After 24 h under SPF conditions at 23 ± 1 °C, whole blood was drawn. Migrasomes were then purified using the Migralso^TM Pro Mouse Immune Cells extraction kit (MGS-P007-MS, Migrasome Therapeutics, Beijing, China). The purified migrasomes were resuspended in PBS, stained with 2 µg/mL WGA for 30 min, and spread evenly onto bovine serum albumin (BSA)-precoated slides for imaging.

After isoflurane anesthesia, mice (n = 6 per group) were placed on a flat surface. The distal 5 mm of each tail was amputated. Within 2 min, blood was collected into a capillary tube (10-20 μL; Laixu, Shandong, China) that had been precoated with ethylenediaminetetraacetic acid (EDTA) (ST063, Beyotime, Shanghai, China). The severed tail was then transferred to 50 μL of warm PBS containing 40 mM EDTA and allowed to bleed for 20 min. Finally, the residual blood suspension was placed on a transparent plastic sheet and photographed for documentation.

RNA sequencing (RNA-seq) was performed with three biological replicates per group. TRIzol (Invitrogen) was used for total RNA extraction. Sequencing and bioinformatic analysis were carried out by BGI Genomics (Beijing, China), and RNA quality was checked on an Agilent 2100 Bioanalyzer (RRID: SCR_018043). Sequencing used the DNBSEQ platform (RRID: SCR_025899) in PE150 mode. Differentially expressed genes (DEGs) were identified, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and Gene Ontology (GO) enrichment analyses were conducted on the BGI Dr. Tom platform. Figures were prepared with GraphPad Prism 9 (RRID: SCR_002798; GraphPad Software, Inc., La Jolla, CA, USA). The dataset has been submitted to the Gene Expression Omnibus (GEO) (RRID: SCR_005012) and assigned accession number GSE319187.

Fluorescence intensity was quantified in ImageJ. Adobe Illustrator (RRID: SCR 010279, Adobe Inc., San Jose, CA, USA) was used for figure preparation. Statistical testing was carried out in GraphPad Prism 8 (RRID: SCR_002798, GraphPad Software, Inc., La Jolla, CA, USA). Paired t-tests were applied to platelet count data. Comparisons involving two groups were made with unpaired two-tailed t-tests. When more than two groups were compared, one-way analysis of variance (ANOVA) was used, followed by Tukey-Kramer post hoc testing. Values are presented as mean ± standard error of the mean (SEM). Symbols indicate significance as follows: not significant (NS), P > 0.05; ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; ^****P < 0.0001.

Zinc is an indispensable trace metal for humans that participates in diverse cellular processes and helps maintain ionic balance, with normal blood concentrations of 8-23 μM^[17]. ZnSO₄ and zinc gluconate are commonly used zinc supplements^[18]. Before evaluating their effects on migrasome formation, we exposed Huh7.5.1 cells to 0-30 μM ZnSO₄ or zinc gluconate for 24 h and assessed viability with a CCK-8 assay. Cell viability was unaffected across this range [Figure 1A and B].

We then used WGA, a migrasome marker^[19], to stain Huh7.5.1 cells and examine how zinc concentration affects migrasome formation. Zinc ions inhibited migrasome formation at 10-30 μM [Figure 1C and D], and retraction fiber extension was also suppressed at 10 and 30 μM. At 20 μM, retraction fibers were reduced, though the difference was not statistically significant [Figure 1C and E].

To confirm intracellular zinc ions accumulation, we used TSQ staining. After 17 h treatment with 20 μmol/L zinc gluconate, migrasome numbers were significantly lower than in untreated cells [Figure 1F and G], and TSQ fluorescence showed much higher intracellular zinc ions levels in zinc gluconate-treated cells than in controls [Figure 1F and H]. Similar results were obtained with 20 μmol/L ZnSO₄ [Figure 1I-K]. In scratch assays, cell migration was also inhibited in Huh7.5.1 cells treated with zinc gluconate or ZnSO₄ [Supplementary Figure 1A and B]. These data suggest that zinc ions alter cell migration dynamics, as summarized in the schematic [Figure 1L].

Disulfiram enhances zinc ions efflux through solute carrier family 30A (SLC30A) transporters^[20]. To determine whether the inhibitory effect depends on intracellular zinc ions, we treated Huh7.5.1 cells with disulfiram together with zinc ions supplements. Disulfiram not only increased migrasome numbers on its own but also fully reversed the suppression caused by zinc gluconate [Figure 2A and B]; at the same time, the zinc ions concentration in cytoplasmic was reduced in disulfiram treated cells [Figure 2A and C]. A similar reversal was observed in cells treated with 20 μmol/L ZnSO₄ [Figure 2D-F]. At 17 h, the zinc ions concentration in the culture medium of cells treated with disulfiram plus zinc ions was far higher than that in the medium of cells treated with zinc ions alone [Figure 2G]. A schematic representation of disulfiram-mediated zinc ions efflux is shown in Figure 2H.

Zinc transport between intracellular and extracellular compartments is mediated by the solute carrier family 39A (SLC39A) and SLC30A families^[21]. To prevent extracellular zinc ions from entering cells, we used an antibody against SLC39A6, which competitively blocks zinc ions uptake. Anti-SLC39A6 treatment alone increased migrasome numbers and completely reversed the inhibitory effect of zinc gluconate [Figure 3A and B]. Cytoplasmic zinc ions levels were lower in anti-SLC39A6-treated cells [Figure 3A and C]. The same effect was observed with ZnSO₄ [Figure 3D-F]. After 17 h, zinc ions concentration in the culture medium of cells treated with anti-SLC39A6 plus zinc was much higher than that of cells treated with zinc ions alone [Figure 3G]. The proposed mechanism of antibody-mediated blockade is illustrated in Figure 3H.

To test whether ZnSO₄ affects migrasome numbers in vivo, mice were injected intraperitoneally once with ZnSO₄ (17 μmol/kg) or saline, and whole-blood migrasomes were isolated 24 h later [Figure 4A]. ZnSO₄ treatment significantly reduced circulating migrasome counts relative to saline controls [Figure 4B and C]. Next, we assessed platelet parameters. Blood collected 24 h after injection showed significant decreases in platelet count [Figure 4D] and platelet aggregation [Figure 4E], along with increased bleeding volumes [Figure 4F-H]. As migrasomes have been reported to promote platelet release from neutrophils and to influence platelet counts and function [Figure 4I]^[15], these results suggest that ZnSO₄ impairs platelet production and activation, potentially compromising hemostatic function.

We then tested zinc gluconate under similar conditions. Intraperitoneal injection of 17 μmol/kg zinc gluconate also reduced whole-blood migrasome numbers [Figure 5A and B] and led to lower platelet counts [Figure 5C] and aggregation [Figure 5D], with increased bleeding volume [Figure 5E-G]. Together, these results show that both zinc ions supplements reduce migrasome availability and impair platelet function, thereby potentially affecting hemostasis [Figure 5H].

Next, we examined the transcriptional response to ZnSO₄ and zinc gluconate. RNA-seq of Huh7.5.1 cells incubated with 20 μmol/L of each compound identified 2,693 up-regulated and 2,590 down-regulated genes in ZnSO₄-treated cells relative to controls [Supplementary Figure 2A], and 3,000 up-regulated and 2,846 down-regulated genes in zinc gluconate-treated cells [Supplementary Figure 2B]. Both treatments shared 4,826 regulated genes [Supplementary Figure 2C]. GO and KEGG enrichment analyses revealed significant changes in lipid metabolism [Figure 6A], with most cholesterol metabolism genes down-regulated [Figure 6B] and most cholesterol synthase genes up-regulated [Figure 6C]. Filipin staining confirmed that both zinc ions treatments markedly reduced plasma membrane cholesterol levels after 17 h compared with untreated controls [Figure 6D-G].

To determine whether cholesterol replenishment could reverse this inhibition, 25 μmol/L cholesterol was added to cells exposed to zinc ions. Cholesterol supplementation significantly increased migrasome numbers compared with zinc ions-treated controls [Figure 6H and I], and retraction fibers tended to lengthen, though this change was less pronounced [Figure 6H and J]. These results indicate that cholesterol availability is critical for counteracting zinc ions-induced suppression of migrasome formation [Figure 6K].

Sphingomyelin is critical for migrasome formation, and previous studies show that zinc ions reduce sphingomyelin levels^[22]. In zinc ions-treated cells, we observed downregulation of SGMS1, SPTLC1, SPTLC3, CERS1, CERS2, and NFKBIA, genes associated with sphingomyelin activity [Figure 7A]. Both ZnSO₄ and zinc gluconate reduced sphingomyelin levels [Figure 7B]. To test whether sphingomyelin depletion mediates zinc ions-induced inhibition, we supplemented zinc ions-treated Huh7.5.1 cells with sphingomyelin. Cells supplemented with 10 μmol/L sphingomyelin after zinc ions treatment showed a significant increase in migrasome numbers compared with zinc ions-treated cells [Figure 7C and D], and retraction fibers showed a similar but weaker trend [Figure 7C and E]. These results suggest that zinc ions-induced sphingomyelin reduction inhibits migrasome formation, and sphingomyelin supplementation promotes migrasome development [Figure 7F].

PI(4,5)P2 is also critical for migrasome biogenesis^[23,24]. To examine whether zinc affects PI(4,5)P2 levels, we transfected Huh7.5.1 cells with PH-PLCD1-GFP, a PI(4,5)P2 probe. After 17 h exposure to 20 μM ZnSO₄ or zinc gluconate, immunofluorescence showed no significant change in PH-PLCD1 signal [Figure 7G-J].

Signaling through RhoA and its downstream effector Rho-associated protein kinase 1 (ROCK1)^[25]. RNA-seq suggested that actin cytoskeleton and focal adhesion pathways regulate RhoA GTP activity [Supplementary Figure 2D]. To test this possibility, Huh7.5.1 cells were transfected with eGFP-2×-Rhotekin-GBD, a probe for GTP-bound RhoA. Immunofluorescence analysis after treatment revealed increased signal in cells treated with zinc gluconate [Figure 7K and L], but no significant change in ZnSO₄-treated cells [Figure 7M and N].

RNA-seq also indicated changes in autophagy, autophagosome, and mechanistic target of rapamycin (mTOR) signaling upon zinc ions treatment [Supplementary Figure 2D and E]. Both ZnSO₄ and zinc gluconate reduced intracellular ROS levels and mitigated ROS activation induced by ZnO-NPs [Figure 7O-R].

In summary, our data show that zinc ions regulate migrasome formation via cholesterol, sphingomyelin, and ROS, and that this regulation affects platelet function and thrombosis. These findings establish a connection between trace element homeostasis and migrasome-mediated hemostasis, and point to zinc ion modulation as a potential avenue for therapeutic intervention in thrombotic disorders. They also raise further questions about the interplay between zinc, cholesterol synthesis, and organelle biogenesis that merit additional study.