Skip to main content
NCBI home page
Science Advances logoLink to Science Advances
. 2026 Feb 6;12(6):eaea4782. doi: 10.1126/sciadv.aea4782

In situ coagulation environment regulation–assisted thrombus clearance via hydrogenated silicon-based nanothrombolytics

Ya-Xuan Zhu 1, Zhixin Chen 1, Yanling You 2,3, Yihan Chen 2,3, Wenjie Yu 4, Xue Guan 5, Piao Zhu 1, Jie Yang 1, Min Ge 2,3, Xiaojun Chen 1, Han Lin 1,2,3,*, Jianlin Shi 1,2,3,*
PMCID: PMC12880526  PMID: 41650270

Abstract

Thrombotic disorders remain among the leading causes of global mortality, yet current thrombolytic therapies are limited by poor targeting specificity and inadequate microenvironmental modulation, resulting in suboptimal efficacy and serious side effects. Here, we developed a hydrogen-generating nanothrombolytic agent that enables enzymatic clot dissolution in combination with intelligent microenvironment reprogramming. Specifically, we assembled urokinase, a clinical thrombolytic drug, with hydrogenated silicene (SiH) nanosheet and fibrinogen, a substrate of coagulation reaction, to promote thrombolysis. Functionally, SiH nanosheet plays multiple roles in the nanothrombolytics: blocking the functional sites of urokinase to durably inhibit its activity in circulation to prevent systemic bleeding, followed by urokinase reactivation in response to SiH nanosheet self-degradation and prothrombotic microenvironment regulation through the in situ hydrogen generation, which mitigates the oxidative stress of vascular endothelial cells and inhibits their release of procoagulant factors. This microenvironment-adaptive thrombolysis strategy offers a promising paradigm for the precise management of thrombotic emergencies.

Hydrogen-generating nanothrombolytic enables targeted thrombolysis and microenvironment regulation for a safer thrombosis therapy.

INTRODUCTION

Thrombosis-related diseases, such as cardiovascular diseases, stroke, and pulmonary embolism, represent a substantial health burden worldwide (1). Current clinical approaches to the treatment of thrombosis include surgical intervention, anticoagulation, and thrombolytic therapy (2–5). Among these, thrombolytic therapy serves as the first-line intervention in managing acute thrombotic events using serine proteases, such as tissue plasminogen activator, urokinase (UK), and recombinant plasminogen activators. These therapeutic proteins mediate the dissolution of blood clots by catalytic conversion of plasminogen to plasmin, which subsequently degrades the fibrin matrix in the thrombi. However, the clinical application of the thrombolytics is largely limited by their short half-life in the bloodstream and their low specificity for thrombus dissolution, resulting in a narrow therapeutic window and high risks of systemic hemorrhage (3). To address this problem, various drug delivery systems have been developed to target and respond to the thrombus (6–10). By introducing ligands that specifically bind thrombus-specific biomarkers (e.g., P-selectin) (6, 11–14) or modules that respond to pathological cues (e.g., thrombin concentration gradients) (15–17) or external stimuli (e.g., magnetic field or ultrasound) (18, 19), these systems allow for precise delivery of thrombolytics at the thrombus site while minimizing nonspecific release. Nevertheless, most thrombolytic strategies only eliminate existing thrombi but neglect to reshape the prothrombotic niche at the lesion area, leading to their poor control of repeated thrombus formation.

Extensive evidence has substantiated that the prothrombotic microenvironment is mechanistically driven by thrombus-induced oxidative stress and the inflammatory response within the vascular endothelium (20–23), and attenuation of local oxidative stress is proving effective in the sustained control of thrombosis. Recently, hydrogen gas (H2) has garnered substantial attention in biomedical fields due to its antioxidation and anti-inflammation potency (24, 25). This therapeutic potential and the rapid development of hydrogen-generating materials promote hydrogen to find broad applications in a wide range of oxidative stress–associated diseases, such as inflammatory bowel disease, acute kidney injury, and arthritis (26–37). However, whether hydrogen gas can be used to reshape the prothrombotic microenvironment in local blood vessels remains an unexplored issue. In addition, precisely delivering hydrogen gas to the thrombus site within the highly dynamic and shear stress–dominated environment of blood flow is still an unmet technical challenge.

In this study, we synergize thrombolytic therapy with prothrombotic niche modulation by developing a thrombus-targeting delivery system of UK and hydrogen gas for the treatment of arterial embolism. Specifically, we use a two-dimensional (2D) hydrogenated silicene (SiH) nanosheet (NS) as the hydrogen-generating platform and further engineer the NS surface with fibrinogen (Fib), a thrombus-targeting ligand, and thrombolytic UK (Fig. 1). This system (termed SiH@UK/Fib) is designed to specifically bind thrombus site while maintaining UK inactive during circulation due to the steric shielding of its active site by SiH NS. After rapidly targeting the thrombus site, the NS gradually degrades, unlocking the thrombolytic function of UK and simultaneously generating hydrogen to down-regulate the expression of procoagulation factors in vascular endothelium. We show in the arterial embolism model that this system exhibits superior efficacy in controlling thrombosis compared with conventional thrombolytic therapy and provides mechanistic evidence for hydrogen gas–mediated prothrombotic niche modulation. We envision that the current SiH@UK/Fib system may offer valuable insights into therapeutic strategies for combating thrombosis.

Fig. 1. Schematic illustration showing the targeting, anticoagulation, and thrombolysis mechanisms of SiH@UK/Fib against arterial embolism.

Fig. 1.

Open in a new tab

Created in BioRender, Y. X. Zhu (2025); https://BioRender.com/ggrkvpf.

RESULTS

Fabrication and characterizations of SiH@UK/Fib

SiH NSs, a class of biocompatible 2D nanomaterials, can efficiently generate hydrogen gas by reacting with water and simultaneously undergo self-degradation. To fabricate SiH@UK/Fib, we first synthesized the SiH NS using a previously reported method (26), followed by sequential loading UK and Fib onto the NS surface (Fig. 2A). The transmission electron microscopy (TEM) image shows the sheet-like morphology and elemental distribution of SiH NSs (Fig. 2B). We confirmed the successful synthesis of SiH NSs by atomic force microscopy (AFM), which revealed their 2D morphology with an average thickness of 1.31 nm (Fig. 2, C and D). Leveraging the large specific surface area of the 2D SiH NS, which provides abundant interaction sites for protein binding, we functionalized the NS with UK and Fib. AFM analysis exhibited an increase in NS thickness to 9.92 nm, demonstrating effective surface adsorption of these proteins (Fig. 2, C and D).

Fig. 2. Preparation, characterizations, and thrombolysis activity of SiH@UK/Fib NSs.

Fig. 2.

Open in a new tab

(A) Preparation procedure and antithrombus mechanism of SiH@UK/Fib. (B) TEM result and element mapping of SiH NS. (C) AFM images of SiH and SiH@UK/Fib. (D) Thickness distribution of the SiH and SiH@UK/Fib quantified from (C). (E) Hydrodynamic diameter distributions of SiH@Fib after treated with thrombin for different time periods. (F) SEM results of the thrombus in 10-min incubation with SiH@Fib in PBS and plasma, respectively. (G) Confocal image of tetramethyl rhodamine isothiocynate (TRITC)–labeled thrombus (red) in 10-min incubation with fluorescein isothiocyanate (FITC)–labeled SiH@Fib.

Responsive assembly and thrombus targeting behavior of SiH@UK/Fib

Next, we interrogated the rationale for introducing Fib as the thrombus-targeting ligand. Fib, an essential component in the coagulation cascade, undergoes thrombin-mediated proteolytic cleavage to form fibrin, which serves as the structural scaffold for thrombus formation. To evaluate whether loading Fib on SiH NSs retains thrombin responsiveness, we performed in vitro coagulation assays. TEM and dynamic light scattering analysis revealed that SiH@Fib exhibited rapid thrombin-induced aggregation (Fig. 2E and fig. S1). This thrombin-mediated assembly mechanism enables the NSs to recognize and anchor to nascent fibrin networks. In addition, we demonstrated in thrombin-treated plasma samples that SiH@Fib efficiently accumulated onto blood clots, clearly confirming its specificity to thrombi (Fig. 2, F and G and fig. S2). These data collectively support the rationale of SiH@UK/Fib for targeting thrombus sites in response to thrombin cleavage.

Thrombolytic activity regulation and hydrogen generation

We then investigated whether loading UK onto SiH NSs affected its enzymatic activity. Molecular docking studies revealed that SiH NSs exhibit high binding affinity for multiple critical sites on UK, including His57 and Ser195, which mediate the catalytic activity of UK, Cys45 and Cys58, which dominate protein folding and protein-protein interactions, and Gly216 and Gln192, involved in substrate recognition (Fig. 3A) (38, 39). This result suggests that these key active sites on UK could be sterically blocked by SiH NSs. Next, we compared the thrombolytic activity of native UK and its SiH-loaded counterpart by the chromogenic assay of S-2444, a specific substrate for UK. The activity of UK was completely suppressed after being loaded on SiH NSs (Fig. 3B), supporting the above molecular docking data. These results demonstrate that SiH NSs can inhibit the thrombolytic activity of UK by blocking its key active sites. We then investigated whether the inhibited catalytic activity of UK could be recovered upon SiH degradation. SiH NSs are stable in acidic environments but volatile under alkaline aqueous conditions, resulting in structural decomposition and hydrogen release. We observed that SiH NSs gradually degraded in phosphate-buffered saline (PBS; pH 7.4) within 4 hours of incubation (fig. S3), along with the generation of hydrogen gas (Fig. 3C). This degradation behavior exactly matches the release profiles of UK from SiH NSs and the kinetics of its activity recovery (Fig. 3B and fig. S4), further validating the critical role of SiH NSs in inhibiting the function of UK.

Fig. 3. Thrombolysis activity regulation and hydrogen generation mechanisms of SiH@UK/Fib.

Fig. 3.

Open in a new tab

(A) Molecular docking simulation results of the interaction between UK and SiH. (B) Time-related activity changes of SiH@UK/Fib indicated by the chromogenic reaction of S-2444. a.u., arbitrary units. (C) Hydrogen generation of SiH@UK/Fib in PBS measured by gas chromatography. (D) Heatmap indicating the time-dependent FTIR absorbance changes of SiH. (E) In situ FTIR spectra at different time points. (F) Time-dependent profiles of different characteristic peak areas. (G) Representative photograph of blood clots treated with PBS buffer (pH 7.4) only, UK, and SiH@UK/Fib for different time periods. h, hours.

To further investigate the degradation and hydrogen generation behaviors of SiH, we monitored the dynamic change of the chemical bonds in SiH during decomposition via in situ Fourier transform infrared (FTIR) spectroscopy. To be specific, we dispersed the SiH in D2O, adjusted the pH to 7.4 via NaH2PO4 and Na2HPO4, and monitored the FTIR spectra for 21 min. As shown in Fig. 3 (D to F), the characteristic peak of ν(Si─D) at 2384 cm−1 was first declined indicating the breakage of Si─D on the surface of NSs. Afterward, the peaks at 2603 cm−1 for ν(O─D) and 1201 cm−1 for δ(D─O─D) were declined, indicating the consumption of D2O during the reaction. Besides, the characteristic peaks at 3389 cm−1 for ν(O─H), 1648 cm−1 for δ(H─O─H), and 1448 cm−1 for δ(H─O─D) appeared and were gradually elevated, indicating the transferring of H atom from SiH NS to the environmental D2O. Meanwhile, the areas of the peak at 1072 cm−1 was gradually increased, demonstrating the Si─O─Si bond formation during the decomposition of NSs. To prove that the UK was functionally restored, we compared the thrombolytic efficacies of naked UK and SiH@UK/Fib through an in vitro thrombolysis assay. We noticed that although SiH@UK/Fib exhibited lower efficacies than naked UK within 1 h, possibly due to the incomplete SiH degradation, it reached the full capacity of thrombolysis at 2 hours (Fig. 3G and figs. S5 and S6). Together, these data suggest that the inhibited thrombolytic activity of SiH-loaded UK can be effectively recovered upon SiH degradation.

Oxidative protection and anticoagulation capability of SiH@UK/Fib

Endothelial injury, a critical factor inducing thrombus formation, is intricately linked to cellular oxidative stress. The oxidative milieu not only disrupts endothelial cell integrity, exposing subendothelial components that facilitate platelet activation, but also up-regulates the expression of prothrombotic factors such as von Willebrand factor (vWF) and plasminogen activator inhibitor–1 (PAI-1). Therefore, we next assessed whether hydrogen gas could protect the vascular endothelium from oxidative injury and the underlying mechanisms of hydrogen in anticoagulation regulation (Fig. 4A). Using human umbilical vein endothelial cells (HUVECs) as the model, we observed that SiH@Fib significantly increased the expression of phosphorylated ‌nuclear factor erythroid-derived 2-like 2‌ (p-Nrf2) (Fig. 4, B and C), an essential regulator in cellular antioxidant defense (40, 41), which suggests the activation of endogenous antioxidant pathways. To prove this, we then profiled the expression levels of heme oxygenase–1 (HO-1) and catalase (CAT), two downstream antioxidant enzymes whose expression is transcriptionally regulated by p-Nrf2, and found that their mRNA levels were significantly increased in SiH@Fib-treated HUVECs (Fig. 4, D and E). The activation of antioxidant pathways in HUVECs was further validated by their enhanced total antioxidant capability (Fig. 4F). Next, we investigated the ability of HUVECs to withstand H2O2-mediated oxidative stress following pretreatment with SiH@Fib. Notably, HUVECs pretreated with SiH@Fib demonstrated a significant reduction in reactive oxygen species (ROS) levels and improved cell viability upon exposure to H2O2 (Fig. 4G and fig. S7), highlighting their enhanced antioxidant defense. These findings collectively underscore that SiH@Fib can mitigate cellular oxidative damage by activating endogenous antioxidant pathways.

Fig. 4. Hydrogen-mediated antioxidation and anticoagulation mechanisms of SiH@UK/Fib NSs.

Fig. 4.

Open in a new tab

Antioxidation and anticoagulation efficiency of SiH@Fib. (A) Anticoagulation mechanism of hydrogen. Created in BioRender, Y. X. Zhu (2025); https://BioRender.com/ggrkvpf. (B) Levels of Nrf2 and p-Nrf2 on HUVECs after SiH@Fib treatment (SiH: 20 μg/ml, 4 hours) indicated by Western blot assay. (C) Quantified results of (B). Expression levels of (D) HO-1 and (E) CAT on HUVECs after SiH@Fib treatment (SiH: 20 μg/ml, 4 hours) measured by quantitative real-time polymerase chain reaction (qPCR) assay. (F) Total antioxidation ability of HUVECs after treated with different concentrations of SiH@Fib. TEAC, Trolox Equivalent Antioxidant Capacity. (G) Confocal images of HUVECs stained with 2′,7′-dichlorodiydrofluorescein diacetate (DCFH-DA) after various treatments (H2O2: 400 μM and SiH: 20 μg/ml, 4 hours). Expression levels of (H) ICAM-1, (I) PAI-1, and (J) vWF on HUVECs after different treatments (H2O2: 400 μM and SiH: 20 μg/ml, 6 hours) measured by qPCR assay. (K) Immunofluorescence staining of HUVECs after different treatments. (L) Expression levels of ICAM-1 and PAI after different treatments (H2O2: 400 μM and SiH: 20 μg/ml, 6 hours) indicated by Western blot assay. (M and N) Quantified results of (L). (O) PAF secretion levels of HUVECs after different treatments measured by enzyme-linked immunosorbent assay (H2O2: 400 μM and SiH: 20 μg/ml, 6 hours). Statistical data are presented as means ± SD and analyzed by one-way analysis of variance (ANOVA) (***P < 0.001 and ****P < 0.0001). ns, nonsignificance.

Next, we examined whether SiH@Fib could down-regulate the expression of procoagulant factors by vascular endothelium. Quantitative real-time polymerase chain reaction (qPCR) results revealed that H2O2-induced oxidative damage increased the expression of intercellular adhesion molecule–1 (ICAM-1), PAI-1, and vWF in HUVECs (Fig. 4, H to J), which all play key roles in platelet activation and adhesion involved in coagulation processes. In contrast, SiH@Fib treatment effectively reduced the expression of these factors to levels comparable to those in the control group. Immunofluorescence imaging confirmed that vWF exhibited significant accumulation on the plasma membranes of H2O2-treated HUVECs, a phenomenon indicating endothelial cell activation, which was efficiently reversed by SiH@Fib treatment (Fig. 4K). The modulatory effect of SiH@Fib on PAI-1 and ICAM-1 was further validated by Western blot analysis, which is consistent with the qPCR data (Fig. 4, L to N). In addition, we further characterized the expression of platelet-activating factor (PAF), a proinflammatory mediator released by endothelial cells and involved in platelet activation. Similarly, SiH@Fib significantly inhibited PAF release from HUVECs following H2O2 treatment (Fig. 4O). Moreover, the SiH@Fib exhibited no significant toxicity toward HUVECs, promising its good biosafety for further application (fig. S8). Together, these results suggest that SiH@Fib has the potential to reshape the prothrombotic microenvironment by reducing the release of procoagulant factors by injured endothelial cells.

Moreover, hydrogen has the potential to promote the repair of damaged endothelial cells and prevent the exposure of collagen, a known procoagulant factor. As shown in Fig. 5A, SiH@Fib effectively alleviates H2O2-induced cytoskeletal damage, and treatment with SiH@Fib significantly reduces the gaps between endothelial cells (Fig. 5B). Besides, the expression of vascular endothelial (VE)–cadherin was restored after SiH@Fib treatment, suggesting the reestablishment of endothelial wall integrity (fig. S9). To further assess endothelial healing, we performed a wound scratch assay, and the SiH@Fib-treated group exhibited accelerated wound closure, highlighting the beneficial role of hydrogen in endothelial repair (Fig. 5, C and D). Next, we examined collagen exposure following different treatments. SiH@Fib significantly reduced collagen exposure in 4 hours after H2O2 treatment (Fig. 5E). Overall, SiH@Fib treatment reduced platelet adhesion and activation, likely due to the decreased release of procoagulant factors and the reduced exposure of collagen (Fig. 5F).

Fig. 5. Inhibition of collagen-induced blood clotting by endothelial repair.

Fig. 5.

Open in a new tab

(A) Procoagulating mechanism of endothelial impairment–induced collagen exposure. Created in BioRender, Y. X. Zhu (2025); https://BioRender.com/ggrkvpf. (B) Confocal images of H2O2-treated or H2O2 + SiH@Fib (H2O2: 400 μM and SiH: 20 μg/ml)–treated HUVECs (red, cytoskeleton; blue, cell nuclei). (C) The scratch wound healing assay results of HUVECs after different treatments. (D) Quantitative results of the healing distances of HUVECs from (C). (E) Contents of collagen exposure from HUVEC layer after different treatments. (F) Confocal images of HUVECs seeded over collagen layer after different treatments, followed by incubation with blood plasma (blue, cell nuclei; red, cytoskeleton; green, platelets). (G) Volcano plots showing the identified up-regulated and down-regulated genes in different groups. (H) Gene set enrichment analysis (GSEA) result highlights the significant inhibition of inflammatory response after SiH@Fib treatment. (I) GSEA result highlights the significant inhibition of complement activation after SiH@Fib treatment. (J) Sankey diagram showing biological functions of the differentially expressed genes (DEGs) of H2O2-treated versus H2O2 + SiH@Fib–treated cells. NES, normalized enrichment score.

The anticoagulant mechanism of SiH@Fib was systematically evaluated using transcriptomics. As shown in Fig. 5G, key genes such as SERPINC1, vWF, C3, and PROZ were significantly down-regulated following SiH@Fib treatment, all of which are closely associated with blood coagulation. Gene Ontology analysis revealed that these alterations were enriched in pathways related to the plasma membrane, which may provide active sites for platelet and neutrophil adhesion, thus contributing to blood coagulation (fig. S10). Furthermore, the down-regulation of apoptosis, inflammation, and complement and coagulation pathways was confirmed through gene set enrichment analysis (GSEA) (Fig. 5, H and I). The differentially expressed genes (DEGs) were primarily enriched in pathways related to oxidative stress response, inflammatory response, growth factor activity, angiogenesis, cell adhesion, blood coagulation, and complement activation, further supporting our hypothesis regarding the anticoagulant mechanisms of SiH@Fib (Fig. 5J).

In vivo thrombus targeting and in situ coagulation microenvironment regulation

Inspired by the efficient thrombolytic and anticoagulant capabilities of SiH@UK/Fib, we next investigated its in vivo targeting and accumulation behavior to evaluate its therapeutic potential (Fig. 6A). The prolonged circulation time of UK in the SiH@UK/Fib group supports the possibility of targeted accumulation at the thrombus site (Fig. 6B), providing sufficient time to dissolve the thrombus and regulate the coagulation environment through hydrogen generation. Benefiting from the rapid response of Fib in the coagulation reaction, SiH@UK/Fib efficiently assembles at the thrombus site within 30 min after injection (Fig. 6C) and remains at the site for about 1 hour to release UK and hydrogen for thrombus resolution. During the thrombolysis process, SiH@UK/Fib begins to clear from the thrombus site within 2 hours after injection, with complete clearance occurring in 8 hours, signaling the termination of the coagulation cascade. The efficient thrombus-targeting behavior of SiH/UK@Fib was further confirmed by scanning electron microscopy (SEM) imaging of the endovascular lining, which showed the presence of SiH NSs at the thrombus site (Fig. 6D). In addition, oxidative stress levels were reduced within 2 hours, indicating effective alleviation of oxidative damage following SiH@UK/Fib treatment (Fig. 6E). As expected, the expression of procoagulant factors such as vWF and PAI was significantly reduced, which, in turn, inhibited platelet activation and adhesion (Fig. 6, F and G). The expression of VE-cadherin also demonstrated the effective protection of endothelial cell barriers by SiH@UK/Fib (fig. S11). Furthermore, the tumor necrosis factor–α (TNF-α) level was also decreased following SiH@UK/Fib treatment, confirming the down-regulation of the inflammatory response (Fig. 6H). Collectively, these findings demonstrate that SiH@UK/Fib can prolong the circulation time of UK, rapidly accumulate at thrombus sites in response to the procoagulant environment, and release UK and hydrogen within ~2 hours, achieving a synergistic thrombolytic and anticoagulant therapeutic effect.

Fig. 6. In vivo thrombus targeting and anticoagulation capability of SiH@UK/Fib NSs.

Fig. 6.

Open in a new tab

(A) Scheme illustrating the design of the animal experiment. The diagram was created using BioRender. (B) Time-dependent blood concentration alterations of UK and SiH@UK/Fib indicated by fluorescence intensity of blood (UK was labeled with TRITC before injection). (C) Accumulation of SiH@UK/Fib (labeled with FITC) at the thrombus site at different time periods after intravenous injection. (D) SEM images of endovascular lining 1 hour after intravenous injection with SiH@UK/Fib. (E) ROS levels indicated by Dihydroethidium (DHE) staining after different treatments. (F) vWF and CD41 levels of blood vessels from mice after different treatments. DAPI, 4′,6-diamidino-2-phenylindole. (G) PAI-1 levels of blood vessels from mice after different treatments. Statistical data are presented as means ± SD (n = 3) and analyzed by one-way ANOVA (*P < 0.05 and **P < 0.01). (H) Concentrations of TNF-α in blood serum from mice after different treatment.

In vivo therapeutic effect toward carotid artery embolism

The antithrombus effect of SiH@UK/Fib was further investigated by monitoring the blood flow of the carotid artery. After SiH@UK/Fib treatment, blood flow at the thrombus site was fully restored, whereas the UK and SiH@Fib groups only exhibited partial recovery at the same time point (Fig. 7, A and B). The hematoxylin and eosin (H&E) staining of blood vessels further substantiates the therapeutic efficacy of SiH@UK/Fib, as the thrombus is nearly undetectable following treatment (Fig. 7C). The reduction in myeloperoxidase (MPO) expression also points to the anti-inflammatory properties of SiH, which are intricately linked to the coagulation response (Fig. 7D). Notably, the UK-only group demonstrated suboptimal thrombolytic effects, likely attributable to the rapid clearance and limited specificity of UK. In contrast, the endovascular wall in the SiH@UK/Fib-treated group displayed a smooth morphology similar to that of the control group, with no residual fibrin, indicating complete thrombolysis (Fig. 7E). The efficient clearance of thrombus was further evidenced by the decreased levels of d-dimer (fig. S12), Fib (Fig. 7F) and thrombin-antithrombin complex (Fig. 7G). In addition, platelet activation was assessed by flow cytometry (figs. S13 and S14) and β-thromboglobulin (β-TG) (Fig. 7H) measurement, which collectively demonstrated a significant inhibition of platelet activation following SiH@UK/Fib treatment. Together, these findings highlight the robust antithrombotic efficacy of SiH@UK/Fib through efficient clot dissolution, vascular protection, and suppression of inflammation and platelet activation.

Fig. 7. In vivo thrombolysis outcome and safety evaluations.

Fig. 7.

Open in a new tab

(A) Representative color-coded laser speckle images of the carotid artery in mice after different treatments. (B) Time-dependent quantified blood flow of mice 1 day after different treatments. The SiH@Fib, UK, and SiH@UK/Fib (SiH: 5 mg/kg and UK: 0.6 mg/kg) were intravenously injected 1 hour after modeling. (C) H&E and (D) MPO-staining of the carotid artery from mice after different treatments. (E) SEM images of endovascular lining from mice after different treatments. (F) Fib, (G) thrombin-antithrombin (TAT) complex, and (H) β-TG concentrations of plasma from mice after different treatments. (I) Hemolysis results of SiH@UK/Fib (Si concentration:10, 20, 40, 60, 80, and 100 μg/ml) incubation with red blood cells at 37°C for 2 hours. (J) PT and APTT values of mice 30 min after UK or SiH@UK/Fib injection. NC, no coagulation; SC, slight coagulation. (K) Bleeding amount and (L) bleeding time of mice after intravenous injection with UK or SiH@UK/Fib. Statistical data are presented as means ± SD [n = 5 for (G); n = 3 for (F), (H), and (L)] and analyzed by one-way ANOVA (*P < 0.05, **P < 0.01, and ***P < 0.001).

Biosafety evaluations of SiH@UK/Fib

Building on the efficient thrombolytic and anticoagulant actions of SiH@UK/Fib, we next assessed the biosafety of this nanosystem to evaluate its potential for clinical application. Hemolysis assays demonstrated that SiH@UK/Fib did not exhibit noticeable hemolytic activity at tested concentrations (Fig. 7I), indicating its good hemocompatibility. Thrombolytic treatments typically activate the fibrinolytic system, which can lead to a significant risk of acute bleeding—one of the major limitations of conventional thrombolytic therapies. Because of the effective inhibition of UK activity by SiH NSs during systemic circulation, SiH@UK/Fib prevents inadvertent plasminogen activation, thereby mitigating the bleeding risk associated with UK. As shown in Fig. 7J, UK-treated mice exhibited marked disturbances in blood coagulation, whereas mice treated with SiH@UK/Fib showed prothrombin time (PT) and activated partial thromboplastin time (APTT) values comparable to those of the control group. Furthermore, both bleeding time (Fig. 7K) and blood loss (Fig. 7L) were within normal ranges in the SiH@UK/Fib-treated group, indicating the favorable biosafety profile of the nanosystem. Moreover, the long-term biosafety evaluations further revealed that PT, APTT, complete blood counts, and serum biochemical markers remained within normal ranges and showed no significant deviations from those of healthy mice (figs. S15 to S17).

DISCUSSION

Thrombosis therapy demands strategies enabling efficient thrombolysis and localized anticoagulation without disturbing systemic coagulation machinery. In this study, we addressed this challenge by developing a 2D SiH NS system with triple functionalities for effective and safe thrombolysis. First, the therapeutic safety could be ensured through the reversible activity locking of the loaded thrombolytic agent, UK, via the interaction between the enzyme and SiH surface. Second, the therapeutic efficacy could be guaranteed by the rapid reactivation of UK through thrombus-responsive SiH degradation, resulting in efficient and safe clot dissolution. Third, the in situ hydrogen generation of SiH@UK/Fib at the thrombus site reprograms the coagulation environment to inhibit the release of procoagulant factors and collagen exposure by alleviating oxidative stress and promoting endothelial repair, thereby achieving improved anticoagulation outcomes. Mechanistically, because the activation of this nanosystem is closely coupled to thrombin and platelet activity, it is more responsive to the hypercoagulable milieu of active clots, while aged clots with reduced enzymatic activity may show weaker responsiveness. While the current design preferentially targets active thrombi, future efforts may focus on optimizing targeting ligands to broaden therapeutic applicability to a wider spectrum of thrombus types. This synergistic strategy establishes a platform for next-generation precision thrombolytics that dynamically harmonize clot dissolution with microenvironmental reprogramming.

MATERIALS AND METHODS

Materials

Calcium silicide (CaSi2, tech-95, 95%, powder, Gelest) was purchased from Alfa Aesar (MA, USA). Hydrochloric acid (HCl) was purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). Ethanol was bought from Shanghai Lingfeng Reagent Chemical Co. Ltd. (Shanghai, China). UK, S-2444, Fib, collagen, cyanine-3 (Cy3) N-hydroxysuccinimide ester, fluorescein isothiocyanate (FITC), and tetramethyl rhodamine isothiocyanate (TRITC) were purchased from Shanghai Yuanye Bio-Technology Co. Ltd. (Shanghai, China). CaCl2 and FeCl3 were purchased from Aladdin Chemistry Co. Ltd. (Shanghai, China). Rabbit anti-Nrf2 antibody, rabbit anti–p-Nrf2 antibody, rabbit anti-CD41 antibody, rabbit anti–PAI-1 antibody, rabbit anti–collagen I antibody, FITC-labeled goat anti-rabbit immunoglobulin G (IgG) antibody, and Cy3-labeled goat anti-rabbit IgG antibody were obtained from Bioss Biotechnology Co. Ltd. (Beijing, China). Rabbit anti–ICAM-1 antibody and rabbit anti-vWF antibody were purchased from ABclonal Biotechnology Co. Ltd. (Wuhan China). Rabbit anti–glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody was purchased from Boster Biological Technology Co. Ltd. (Wuhan, China). Mouse anti–VE-cadherin antibody, mouse anti–β-actin antibody, horseradish peroxidase (HRP)–labeled goat anti-rabbit antibody, and HRP-labeled goat anti-mouse antibody were obtained from Proteintech Group Inc. (Wuhan, China). Anti–CD61-Allophycocyanin (APC) and anti–CD62p-phycoerythrin (PE) antibodies were purchased from BioLegend Inc. (San Diego, CA, USA). Thrombin was purchased from Shanghai Yingxin Laboratory Equipment Co. Ltd. (Shanghai, China). Actin-Tracker Red-Rhodamine, Hoechst 33342, and Total Antioxidant Capacity Assay Kit with 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) method were purchased from Beyotime Biotechnology Co. Ltd. (Shanghai, China).

Fabrication and characterizations

To synthesize SiH NSs, 1 g of CaSi2 was dispersed in 200 ml of HCl precooled under argon protection and kept stirring for 7 days at −20°C. Then, the SiH sheets were centrifuged (12,000 rpm for 10 min) and sequentially washed two times with acetone and ethanol. The product was dispersed in 200 ml of ethanol and subjected to 8-hour tip sonication (600 W; on, 2 s; off, 4 s) under ice bath to obtain SiH NSs. Then, 1 ml of UK and 1 ml of Fib aqueous solution were added in to the SiH dispersion (2 mg/ml; in ethanol) under vigorous stirring subsequently. After 5-min stirring, the mixture was centrifuged (8000 rpm for 5 min) and washed three times by deionized water to obtain the SiH@UK/Fib. SiH@Fib were also prepared via the similar methods. To obtain FITC-labeled Fib, FITC was first dissolved in N,N′-dimethylformamide (2 mg/ml) and then mixed with the aqueous solution of Fib (10 mg/ml) at the molar ratio Fib:FITC = 1:3. Then, the pH was adjusted to 9.5 and kept reaction for 12 hours under 4°C. Then, the FITC-labeled Fib was dialyzed [molecular weight (MWCO), 7 kDa] against water for 2 days under 4°C. To obtain TRITC-labeled UK, TRITC was first dissolved in DMF (2 mg/ml) and then mixed with the aqueous solution of UK (0.5 mg/ml) at the molar ratio UK:TRITC = 1:1.5. Then, the pH was adjusted to 9.5 and kept reaction for 12 hours under 4°C. Then, the TRITC-labeled UK was dialyzed (MWCO, 7 kDa) against water for 2 days under 4°C. AFM (Dimension ICON, Bruker) was applied to measure the thickness of SiH and SiH@UK/Fib NSs by Bruker Dimension ICON. The morphology of SiH@UK/Fib was characterized via TEM (JEM-2100F, JEOL). The size of SiH@UK/Fib was measured by a zetasizer (Nano AS90, Malvern Instrument). The quantitative elemental analysis was detected by inductively coupled plasma optical emission spectrometry (Agilent 700 Series, Agilent Technologies, USA).

Hydrogen generation evaluations

To evaluate the hydrogen generation property, SiH or SiH@UK/Fib NSs were first dispersed in deionized water at a Si concentration of 2 mg/ml. Then, 500 μl of the dispersion was added to a sealed 10-ml round-bottom flask containing 2 ml of PBS, which was filled with nitrogen. Next, 1 ml of gas in the flask was extracted using a syringe at different time points and detected by gas chromatography (GC2010, Shimadzu).

In vitro thrombus targeting ability investigations

To evaluate the thrombin-responsive aggregation of SiH@Fib, SiH@Fib was first dispersed in PBS (200 μg/ml), and then 2 μl of thrombin (1 kU/ml) was added to the SiH@Fib dispersion and incubated at 37°C for 10 min. TEM evaluations were carried out before or after thrombin addition. To prepare the blood clots, the whole blood from BALB/c mice was mixed up with clotting mixture [containing 66 mM tris-HCl, 130 mM NaCl, 45 mM CaCl2, and thrombin (0.2 U/ml)] and incubated for 30 min at 37°C. Then, the clots were incubated with SiH@Fib (200 μg/ml in PBS) for 10 min at 37°C. Then, the thrombus was fixed by 4% glutaraldehyde overnight and then dehydrated in ethanol for SEM (S-4700, Hitachi) observation. For confocal laser scanning microscopy experiments, the Fib and the blood clots were labeled with FITC and Cy3, respectively.

Acknowledgments

We thank the staff members of the BL01B beamline (https://cstr.cn/31129.02.NFPS.BL01B) at the National Facility for Protein Science in Shanghai (https://cstr.cn/31129.02.NFPS) for providing technical support and assistance in data collection and analysis.

Funding:

This work was supported by National Natural Science Foundation of China (T2495260-T2495263, 22422510, 52372276, 22575175, 22205166, 82203681, and 32201140), Youth Innovation Promotion Association of the Chinese Academy of Sciences (2023262), Natural Science Foundation of Shanghai (23ZR1472300), and CAMS Innovation Fund for Medical Sciences (2021-I2M-5-012).

Author contributions:

Conceptualization: J.S., H.L., and Y.-X.Z. Methodology: Y.-X.Z., Z.C., Y.Y., W.Y., and Y.C. Validation: Y.-X.Z., W.Y., and X.C. Formal analysis: Y.-X.Z. Investigation: Y.-X.Z., P.Z., X.G., and J.Y. Visualization: Y.-X.Z. and M.G. Resources: Y.-X.Z. Supervision: J.S. and H.L. Writing—original draft: Y.-X.Z. Writing—review and editing: J.S. and H.L. Project administration: J.S. and H.L. Funding acquisition: J.S., H.L., Y.-X.Z., X.G., and P.Z.

Competing interests:

The authors declare that they have no competing interests.

Data and materials availability:

The study involved no new materials preparation. All data and code needed to evaluate and reproduce the results in the paper are present in the paper and/or the Supplementary Materials.

Supplementary Materials

This PDF file includes:

Supplementary Text

Figs. S1 to S17

sciadv.aea4782_sm.pdf (1MB, pdf)

REFERENCES

  • 1.Lippi G., Franchini M., Targher G., Arterial thrombus formation in cardiovascular disease. Nat. Rev. Cardiol. 8, 502–512 (2011). [DOI] [PubMed] [Google Scholar]
  • 2.Mackman N., Triggers, targets and treatments for thrombosis. Nature 451, 914–918 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Mackman N., Bergmeier W., Stouffer G. A., Weitz J. I., Therapeutic strategies for thrombosis: New targets and approaches. Nat. Rev. Drug Discov. 19, 333–352 (2020). [DOI] [PubMed] [Google Scholar]
  • 4.Zheng M., Guo J., Li Q., Yang J., Han Y., Yang H., Yu M., Zhong L., Lu D., Li L., Sun L., Syntheses and characterization of anti-thrombotic and anti-oxidative gastrodin-modified polyurethane for vascular tissue engineering. Bioact. Mater. 6, 404–419 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Gawaz M., Geisler T., Borst O., Current concepts and novel targets for antiplatelet therapy. Nat. Rev. Cardiol. 20, 583–599 (2023). [DOI] [PubMed] [Google Scholar]
  • 6.Zhang H., Qu H., He Q., Gao L., Zhang H., Wang Y., Zhang Z., Hou L., Thrombus-targeted nanoparticles for thrombin-triggered thrombolysis and local inflammatory microenvironment regulation. J. Control. Release 339, 195–207 (2021). [DOI] [PubMed] [Google Scholar]
  • 7.Zhang H., Wang J., Wu H., Wang Y., Zhang S., Sun J., He Z., Luo C., On-site self-penetrating nanomedicine enabling dual-priming drug activation and inside-out thrombus ablation. ACS Nano 18, 34683–34697 (2024). [DOI] [PubMed] [Google Scholar]
  • 8.Su M., Dai Q., Chen C., Zeng Y., Chu C., Liu G., Nano-medicine for thrombosis: A precise diagnosis and treatment strategy. Nanomicro. Lett. 12, 96 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Zhou Z., Chen W., Cao Y., Abdi R., Tao W., Nanomedicine-based strategies for the treatment of vein graft disease. Nat. Rev. Cardiol. 22, 255–272 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Chen W., Schilperoort M., Cao Y., Shi J., Tabas I., Tao W., Macrophage-targeted nanomedicine for the diagnosis and treatment of atherosclerosis. Nat. Rev. Cardiol. 19, 228–249 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Cheng J., Zhang S., Li C., Li K., Jia X., Wei Q., Qi H., Zhang J., Functionally integrating nanoparticles alleviate deep vein thrombosis in pregnancy and rescue intrauterine growth restriction. Nat. Commun. 13, 7166 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Huang Y., Gu B., Salles-Crawley I. I., Taylor K. A., Yu L., Ren J., Liu X., Emerson M., Longstaff C., Hughes A. D., Thom S. A., Xu X. Y., Chen R., Fibrinogen-mimicking, multiarm nanovesicles for human thrombus-specific delivery of tissue plasminogen activator and targeted thrombolytic therapy. Sci. Adv. 7, eabf9033 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Xu J., Zhang Y., Xu J., Liu G., Di C., Zhao X., Li X., Li Y., Pang N., Yang C., Li Y., Li B., Lu Z., Wang M., Dai K., Yan R., Li S., Nie G., Engineered nanoplatelets for targeted delivery of plasminogen activators to reverse thrombus in multiple mouse thrombosis models. Adv. Mater. 32, e1905145 (2020). [DOI] [PubMed] [Google Scholar]
  • 14.Hu Q., Qian C., Sun W., Wang J., Chen Z., Bomba H. N., Xin H., Shen Q., Gu Z., Engineered nanoplatelets for enhanced treatment of multiple myeloma and thrombus. Adv. Mater. 28, 9573–9580 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Wang Z., Zhao Y., Hou Y., Tang G., Zhang R., Yang Y., Yan X., Fan K., A Thrombin-activated peptide-templated nanozyme for remedying ischemic stroke via thrombolytic and neuroprotective actions. Adv. Mater. 36, 2210144 (2024). [DOI] [PubMed] [Google Scholar]
  • 16.Zhang H., Zhang H., Pei Y., Gao L., He Q., Zhang H., Zhu L., Zhang Z., Hou L., Shear force responsive and fixed-point separated system for targeted treatment of arterial thrombus. Nano Today 38, 101186 (2021). [Google Scholar]
  • 17.Yin J., Wang S., Wang J., Zhang Y., Fan C., Chao J., Gao Y., Wang L., An intelligent DNA nanodevice for precision thrombolysis. Nat. Mater. 23, 854–862 (2024). [DOI] [PubMed] [Google Scholar]
  • 18.Xu L., Luo Y., Du Q., Zhang W., Hu L., Fang N., Wang J., Liu J., Zhou J., Zhong Y., Liu Y., Ran H., Guo D., Xu J., Magnetic response combined with bioactive ion therapy: A RONS-scavenging theranostic nanoplatform for thrombolysis and renal ischemia-reperfusion injury. ACS Nano 17, 5695–5712 (2023). [DOI] [PubMed] [Google Scholar]
  • 19.Wang S., Guo X., Xiu W., Liu Y., Ren L., Xiao H., Yang F., Gao Y., Xu C., Wang L., Accelerating thrombolysis using a precision and clot-penetrating drug delivery strategy by nanoparticle-shelled microbubbles. Sci. Adv. 6, eaaz8204 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Everett L. A., Cleuren A. C., Khoriaty R. N., Ginsburg D., Murine coagulation factor VIII is synthesized in endothelial cells. Blood 123, 3697–3705 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Su F., Zhang C., Zhang Q., Shen Y., Li S., Shi J., Zhu Y. X., Lin H., He B., Multifaceted immunomodulatory nanocomplexes target neutrophilic-ROS inflammation in acute lung injury. Adv. Sci. 12, 2411823 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Furie B., Furie B. C., Mechanisms of thrombus formation. N. Engl. J. Med. 359, 938–949 (2008). [DOI] [PubMed] [Google Scholar]
  • 23.Stark K., Massberg S., Interplay between inflammation and thrombosis in cardiovascular pathology. Nat. Rev. Cardiol. 18, 666–682 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Zhou G., Goshi E., He Q., Micro/nanomaterials-augmented hydrogen therapy. Adv. Healthc. Mater. 8, e1900463 (2019). [DOI] [PubMed] [Google Scholar]
  • 25.Mi F., Zhao N., Jin L., Zhang Z., Wang X., Fang X., Li W., Liu Z., Shu P., Zhang X., Wu C., Conjugated polymers as photocatalysts for hydrogen therapy. BME Mat 3, e12126 (2025). [Google Scholar]
  • 26.You Y., Zhu Y. X., Jiang J., Wang M., Chen Z., Wu C., Wang J., Qiu W., Xu D., Lin H., Shi J., Water-enabled H2 generation from hydrogenated silicon nanosheets for efficient anti-inflammation. J. Am. Chem. Soc. 144, 14195–14206 (2022). [DOI] [PubMed] [Google Scholar]
  • 27.Xu Z., Yang Y., Li X., Wang J., Chen S., An T., Hu C., Deng C., Zhou F., Xiang L., Qu Y., Man Y., A visible-light photocatalysis/hydrolysis hydrogen-generating nanoplatform for dynamic inflammation management via immune metabolism orchestration during wound repair. ACS Appl. Mater. Interfaces 17, 24918–24939 (2025). [DOI] [PubMed] [Google Scholar]
  • 28.Gong F., Xu J., Liu B., Yang N., Cheng L., Huang P., Wang C., Chen Q., Ni C., Liu Z., Nanoscale CaH2 materials for synergistic hydrogen-immune cancer therapy. Chem 8, 268–286 (2022). [Google Scholar]
  • 29.Yang N., Yang X., Cheng S., Gao X., Sun S., Huang X., Ge J., Han Z., Huang C., Wang Y., Cheng C., Cheng L., Magnesium implants with alternating magnetic field-enhanced hydrogen release and proton depletion for anti-infection treatment and tissue repair. Bioact. Mater. 38, 374–383 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Chen S., Yu Y., Xie S., Liang D., Shi W., Chen S., Li G., Tang W., Liu C., He Q., Local H2 release remodels senescence microenvironment for improved repair of injured bone. Nat. Commun. 14, 7783 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Hu R., Dai C., Dong C., Ding L., Huang H., Chen Y., Zhang B., Living macrophage-delivered tetrapod PdH nanoenzyme for targeted atherosclerosis management by ROS scavenging, hydrogen anti-inflammation, and autophagy activation. ACS Nano 16, 15959–15976 (2022). [DOI] [PubMed] [Google Scholar]
  • 32.Xu M., Wu G., You Q., Chen X., The landscape of smart biomaterial-based hydrogen therapy. Adv. Sci. 11, e2401310 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Zhu Y.-X., You Y., Chen Z., Xu D., Yue W., Ma X., Jiang J., Wu W., Lin H., Shi J., Inorganic nanosheet-shielded probiotics: A self-adaptable oral delivery system for intestinal disease treatment. Nano Lett. 23, 4683–4692 (2023). [DOI] [PubMed] [Google Scholar]
  • 34.You Y., Jiang J., Zheng G., Chen Z., Zhu Y. X., Ma H., Lin H., Guo X., Shi J., In situ piezoelectric-catalytic anti-inflammation promotes the rehabilitation of acute spinal cord injury in synergy. Adv. Mater. 36, 2311429 (2024). [DOI] [PubMed] [Google Scholar]
  • 35.Ji P., Qiu S., Huang J., Wang L., Wang Y., Wu P., Huo M., Shi J., Hydrolysis of 2D nanosheets reverses rheumatoid arthritis through anti-inflammation and osteogenesis. Adv. Mater. 37, 2415543 (2025). [DOI] [PubMed] [Google Scholar]
  • 36.Lin Z., Chen Z., Chen Y., Yang N., Shi J., Tang Z., Zhang C., Lin H., Yin J., Hydrogenated silicene nanosheet functionalized scaffold enables immuno-bone remodeling. Exploration 3, 20220149 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Zhao Y., Xie R., Yodsanit N., Ye M., Wang Y., Gong S., Biomimetic fibrin-targeted and H2O2-responsive nanocarriers for thrombus therapy. Nano Today 35, 100986 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Sulimov V. B., Katkove E. V., Oferkin I. V., Sulimov A. V., Romanov A. N., Roschin A. I., Beloglazova I. B., Plekhanova O. S., Tkachuk V. A., Sadovnichiy V. A., Application of molecular modeling to urokinase inhibitors development. Biomed. Res. Int. 2014, 625176 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Djouad S.-E., Berredjem M., Hadjadj Aoul F. Z., Bouchareb F., Guefi M., Hadda T. B., Aissaoui M., Belhani B., In silico drug design and molecular docking of novel amidophosphonates and sulfamidophosphonates as inhibitors of urokinase-type plasminogen activator. J. Indian Chem. Soc. 99, 100650 (2022). [Google Scholar]
  • 40.Ngo V., Duennwald M. L., Nrf2 and oxidative stress: A general overview of mechanisms and implications in human disease. Antioxidants (Basel) 11, 3345 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Nguyen T., Nioi P., Pickett C. B., The Nrf2-antioxidant response element signaling pathway and its activation by oxidative stress. J. Biol. Chem. 284, 13291–13295 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Text

Figs. S1 to S17

sciadv.aea4782_sm.pdf (1MB, pdf)

Data Availability Statement

The study involved no new materials preparation. All data and code needed to evaluate and reproduce the results in the paper are present in the paper and/or the Supplementary Materials.

Articles from Science Advances are provided here courtesy of American Association for the Advancement of Science

RESOURCES