Platelet-Targeted Self-Amplifying mRNA for Safe, Long-Acting Thromboprophylaxis Authors: Yuyang Song, Thanh Hung Nguyen, Pengkai Shi, Ahmed Refaat, Cuong Viet Pham, Bangyan Xu, Shulei Ren, Julia I. Skoraczynski, Anne Nguyen, Haikun Liu, Alexander Klesen, Sylvain Trépout, Mark Louis P. Vidallon, Karlheinz Peter, Xiaowei Wang Published: 2026 ABSTRACT BACKGROUND:Thrombosis is a major contributor to morbidity and mortality in cardiovascular diseases. Although current antiplatelet therapies reduce thrombotic complications, they are associated with systemic bleeding complications. We previously developed a single-chain variable fragment (scFv) that specifically binds and blocks the activated platelet integrin αIIbβ3 (GPIIb/IIIa [glycoprotein IIb/IIIa]), preventing thrombosis without impairing hemostasis. We hypothesized that an mRNA-based therapeutic encoding scFvαIIbβ3 would offer prolonged expression and enhanced therapeutic durability while maintaining a favorable safety profile.METHODS:We designed and synthesized both conventional mRNA and self-amplifying mRNA constructs encoding scFvαIIbβ3. The constructs were first validated in vitro by confirming their expression of a functional scFvαIIbβ3. For in vivo experiments, mRNAs were encapsulated in a novel lipid nanoparticle formulation comprising of γ-oryzanol and DLin-KC2-DMA (OryKL) for systemic delivery. Their biosafety was assessed by a series of biochemical and histological examinations. The murine FeCl3-induced arterial thrombosis model was used to assess the preventive effect of scFvαIIbβ3-encoding mRNAs on thrombus formation.RESULTS:In vitro assays confirmed the efficient expression and secretion of a functional scFvαIIbβ3 that selectively bound to and blocked activated GPIIb/IIIa, thereby inhibited platelet aggregation. In mice with induced thrombosis, OryKL-delivered scFvαIIbβ3 mRNA significantly prolonged occlusion time at 24 hours posttreatment, while self-amplifying mRNA provided sustained thromboprotection for over 7 days. Tail bleeding times were unchanged across all groups, and no systemic toxicity or histopathologic abnormalities were observed.CONCLUSIONS:Our findings demonstrate that scFvαIIbβ3 mRNA and self-amplifying mRNA therapeutics enable safe, effective, and long-acting antithrombotic protection in vivo, offering a promising strategy for thromboprophylaxis that prevents the bleeding risks associated with current antiplatelet therapies. Novelty and Significance What Is Known? Current antiplatelet therapies reduce thrombosis but are associated with a high risk of bleeding complications, also restricting their use in patients with preexisting high bleeding risk. Current antiplatelet therapies reduce thrombosis but are associated with a high risk of bleeding complications, also restricting their use in patients with preexisting high bleeding risk. Activation-specific scFvs (single-chain antibody fragment variants) that bind to the active conformation of GPIIb/IIIa (glycoprotein IIb/IIIa), but not nonactivated GPIIb/IIIa on resting platelets, can selectively block fibrinogen binding and platelet aggregation, while preserving hemostasis. Activation-specific scFvs (single-chain antibody fragment variants) that bind to the active conformation of GPIIb/IIIa (glycoprotein IIb/IIIa), but not nonactivated GPIIb/IIIa on resting platelets, can selectively block fibrinogen binding and platelet aggregation, while preserving hemostasis. mRNA therapeutics enable in vivo production of therapeutic proteins, but their use for sustained delivery of activation-specific platelet inhibitors has not been established. mRNA therapeutics enable in vivo production of therapeutic proteins, but their use for sustained delivery of activation-specific platelet inhibitors has not been established. What New Information Does This Article Contribute? This study establishes conventional and self-amplifying mRNA therapeutics encoding an activated platelet–specific anti–GPIIb/IIIa scFv as a novel in vivo antithrombotic strategy. This study establishes conventional and self-amplifying mRNA therapeutics encoding an activated platelet–specific anti–GPIIb/IIIa scFv as a novel in vivo antithrombotic strategy. Intravenous OryKL (lipid nanoparticles) delivery enables systemic expression of the scFv with a favorable in vivo safety profile. Intravenous OryKL (lipid nanoparticles) delivery enables systemic expression of the scFv with a favorable in vivo safety profile. Self-amplifying mRNA extends protection against thrombosis beyond 7 days, supporting a safe, long-acting thromboprophylactic approach. Self-amplifying mRNA extends protection against thrombosis beyond 7 days, supporting a safe, long-acting thromboprophylactic approach. mRNA delivery of an activated platelet–specific anti–GPIIb/IIIa scFv enables in vivo antithrombotic protection without affecting hemostasis. By extending this effect beyond 7 days, self-amplifying mRNA proves to be a safe and long-acting platform for thromboprophylaxis. Meet the First Author, see p e000759 Editorial, see Article by Becker Thrombotic complications remain a major cause of mortality and morbidity worldwide, causing acute coronary syndromes, stroke, venous thromboembolism, and sudden cardiac death. Current antithrombotic therapies, including aspirin, P2Y12 inhibitors, direct oral anticoagulants, and GPIIb/IIIa (glycoprotein IIb/IIIa) inhibitors, have improved outcomes1,2 but are limited by the systemic bleeding risk and the need for daily or continuous dosing. Thrombosis arises from disruption of vascular hemostasis, which depends on a tightly regulated balance between various circulating elements, such as platelets, procoagulants, and anticoagulant factors, and an intact endothelial lining.3 On vascular injury or atherosclerotic plaque rupture, subendothelial matrix proteins, particularly collagen, become exposed, triggering platelet adhesion and activation.4 A key downstream event is the conformational change of the platelet integrin αIIbβ3 (GPIIb/IIIa), which is transformed into a high-affinity receptor for fibrinogen/fibrin. This interaction cross-bridges adjacent platelets and mediates platelet aggregation,4,5 a pivotal step in thrombus formation. Given its central role in the final common pathway of platelet activation, GPIIb/IIIa has long been a validated target for antiplatelet therapy. However, GPIIb/IIIa inhibitors such as abciximab, tirofiban, and eptifibatide are associated with increased bleeding risk and are unsuitable for prophylactic applications.6–11 Despite significant resources devoted to its treatment, thrombosis continues to drive healthcare utilization and poor outcomes globally.12 Therapeutic success often hinges on timely intervention in well-equipped facilities, an approach not always feasible in resource-limited or prehospital settings.13 Thus, new therapies for thrombosis are urgently needed. We previously engineered a single-chain variable fragment (scFv) that selectively binds to and blocks the activated conformation of the GPIIb/IIIa receptor (scFvαIIbβ3). In murine models of arterial thrombosis, intravenous administration of this scFvαIIbβ3 inhibited platelet aggregation and reduced thrombus burden without interfering with hemostasis.13–15 However, its practical application is hindered by the requirement for high dosing and by its short systemic half-life. Recent advances in mRNA technology offer a compelling alternative to direct protein administration.16–18 mRNA enables transient, host-driven expression of therapeutic proteins with proper folding, posttranslational modification, and intracellular localization. mRNA therapies avoid the immunogenicity and manufacturing complexity of recombinant proteins and can be rapidly scaled and adapted.16–19 Moreover, self-amplifying mRNA (samRNA) enables extended expression from a single dose through intracellular RNA replication, potentially reducing dosing requirements and extending therapeutic duration. In the present study, we investigated the potential application of lipid nanoparticle (LNP)–formulated mRNA and samRNA constructs, encoding for scFvαIIbβ3 that binds and blocks activated GPIIb/IIIa receptors. We show that a single dose of conventional mRNA conferred protection for up to 24 hours, whereas samRNA extended this effect to 7 days and beyond. Both platforms significantly delayed thrombus formation in vivo without causing bleeding or systemic toxicity. These findings support mRNA-based GPIIb/IIIa inhibition as a viable and safe approach for thromboprophylaxis. Methods Data Availability The data that support the findings of this study are available from the corresponding authors on reasonable request. Detailed description of the methodology is provided in the Supplemental Material. Additional data will be shared on reasonable request to the corresponding authors. Please see the Major Resources Table in the Supplemental Material. mRNA and samRNA Constructs mRNA and samRNA encoding scFvαIIbβ314,20 were produced by in vitro transcription and purified before use. A scFv that is a nonbinding mutant (scFvmut) was generated as control. Cell Culture and Expression Chinese hamster ovary (CHO), human hepatocellular carcinoma G2 (HepG2), and human hepatocellular carcinoma 3B (Hep3B) cells were transfected with control, scFvmut, or scFvαIIbβ3 mRNA. Secreted proteins from the collected cell supernatants were assessed posttransfection by Western blotting and flow cytometry. Platelet Binding and Function Assays Human blood was collected from healthy volunteers with approval from the Alfred Human Ethics Committee (Project No. 627/17). The binding of scFvαIIbβ3 to activated human platelets was assessed by flow cytometry in a dose-dependent manner. Inhibition of platelet aggregation was evaluated using aggregometry21 and under flow via collagen-coated microfluidic assays. LNP Formulation and Characterization mRNA was encapsulated in OryKL22 or Onpattro LNP formulations. Particle size, surface charge, and encapsulation efficiency were assessed by dynamic light scattering and RiboGreen assay. Particle morphology was evaluated by electron microscopy. Animal Studies All animal procedures were approved by the Alfred Plus Alliance Animal Ethics Committee (E/10630/2024, E/11691/2025), conducted and reported in accordance with institutional and Animal Research: Reporting of In Vivo Experiments, update 2.0 guidelines, respectively. Mice received intravenous administration of controls or mRNA-loaded LNPs. Hairless and immunocompetent mouse strain (SKH1) mice were used for biodistribution studies and were assessed by In Vivo Imaging System (IVIS) of Nanoluciferase expression. C57BL/6 mice were used for thrombosis, bleeding, and safety studies. FeCl3-induced carotid artery induced injury model was monitored using nano-Doppler flow probe. Bleeding risk was evaluated via tail bleeding time and blood loss assays. Histology and Safety Analyses Major organs were collected for assessment of organ weight and histology. Blood was collected for hematologic analysis, serum biochemistry, and cytokine profiling. Statistical Analyses Statistical analyses were performed using Graphpad Prism version 10.0. Data are presented as mean±SD. Normality was assessed using Shapiro-Wilk testing. For comparisons between 2 groups, Student t tests were used for normally distributed data, whereas Mann-Whitney U tests were used for non-normally distributed data. For multiple group comparisons, 1-way ANOVA followed by Tukey post hoc multiple comparisons test was used for normally distributed data, whereas the Kruskal-Wallis test followed by Dunn multiple comparisons test was applied when normality was not reached. For comparisons involving 2 factors (eg, treatment and time), 2-way ANOVA followed by Holm-Sidak multiple comparisons was performed for normally distributed data. Survival curves were compared using the log-rank (Mantel-Cox) test. A value of P<0.05 was considered statistically significant. Specific tests are provided in the figure legends. Results Successful Design, Optimization, and Expression of scFvαIIbβ3 mRNA The recombinant scFvαIIbβ3, which selectively targets the activated conformation of GPIIb/IIIa, has been previously developed and validated.14,20 A nonbinding scFvmut was generated as control. For mRNA-based expression, both constructs were optimized for stability, translation efficiency, and secretion (Figure 1A). The open reading frame was codon-optimized and fused with the murine IgK signal peptide at its N-terminal to facilitate extracellular secretion and to a C-terminal 6×His tag for detection and purification. The constructs also included optimized 5′ and 3′ untranslated regions and a 143-nucleotide poly(A) tail. Both scFvαIIbβ3 and scFvmut mRNA constructs were synthesized in vitro with N1-methylpseudouridine in place of uridine and cotranscriptionally capped with an antireverse cap analogue to enhance stability and translation (Figure 1B). Transfection of CHO cells with the scFvαIIbβ3 mRNA led to robust expression and secretion of the antibody into the culture medium, as confirmed by Western blot analysis. Expression of scFvαIIbβ3 mRNA was also confirmed using HepG2 and Hep3B in vitro (Figure S1). Peak levels were observed around 24 hours posttransfection and remained detectable up to 72 hours (Figure 1C; Figure S2). These results confirm that the mRNA constructs are efficiently translated and secreted in vitro. Design and expression of the scFvαIIbβ3 (single-chain variable fragment targeting the activated glycoprotein IIb/IIIa receptor) mRNA targeting activated platelets (AP). A, Schematic representation of the scFvαIIbβ3 mRNA construct, including the 5′ untranslated region (UTR), the coding region containing a secretion signal peptide, and the 3′ UTR, followed by a poly (A) tail (143 adenosines). B, Assessment of the sizes (both 1647 bp) and integrity of N1-methylpseudouridine-modified scFvαIIbβ3 and nonbinding mutant (scFvmut) mRNA visualized by agarose gel electrophoresis. C, Expression and secretion of scFvαIIbβ3 and scFvmut proteins (45 kD) in mRNA-transfected Chinese hamster ovary cells, detected by Western blotting using a peroxidase-coupled anti-His-tag-antibody. D, Specific binding of scFvαIIbβ3-containing supernatant to APs in platelet-rich plasma (PRP), measured by flow cytometry using anti-His Alexa Fluor488 antibody; bar graphs show mean fluorescence intensity (MFI). No binding was observed for control supernatant (without [w/o] mRNA or scFvmut mRNA). E, Inhibition of scFvαIIbβ3 binding to APs when pretreated with eptifibatide. F, Competitive binding between scFvαIIbβ3 and PAC-1 (antibody recognizing an epitope on the glycoprotein IIb/IIIa)–fluorescein isothiocyanate (FITC). Nonactivated or activated PRPs preincubated with control, scFvαIIbβ3, or scFvmut, followed by staining with PAC-1-FITC. G, Concentration-dependent reduction in competitive binding of PAC-1-FITC to ADP-APs, with complete inhibition of PAC-1 binding observed at higher concentrations (lower dilution factors). H. Kinetic ADP-induced platelet aggregation curves over 20 minutes; inhibition of ADP-induced platelet aggregation by scFvαIIbβ3. I. Bar graph quantifying percentage inhibition. Eptifibatide (Ept) was used as a control to inhibit platelet aggregation. Data are presented as mean±SD (n=3 biological replicates in D, F through I and n=5 biological replicates in E). Statistical analyses were performed using paired Student t test (D and E), 1-way ANOVA, followed by Tukey post hoc multiple comparisons test (F, G, and I), and 2-way ANOVA with Holm-Sidak multiple comparisons (H). 5′m7G indicates 5′ 7-methylguanosine cap; a.u., arbitrary units; L, ladder; mut, mutant; and NAP, nonactivated platelets. mRNA-Expressed Protein scFvαIIbβ3 Binds to Activated Platelets and Inhibits Platelet Aggregation In Vitro The functional activity of mRNA-expressed scFvαIIbβ3 and scFvmut was evaluated using human platelet-rich plasma. Platelet activation was confirmed by PAC-1 (antibody recognizing an epitope on the activated GPIIb/IIIa) staining (Figures S3 and S11). Flow cytometry demonstrated that scFvαIIbβ3 bound specifically to activated platelets with no detectable binding to resting (nonactivated) platelets, whereas the scFvmut showed no binding to either (Figure 1D; Figure S4A). Pretreatment with eptifibatide blocked ligand binding to GPIIb/IIIa and abolished the binding of scFvαIIbβ3 to activated platelets (Figure 1E). Pretreatment with scFvαIIbβ3-containing supernatant significantly reduced PAC-1 binding, whereas functional blockade of the receptors was not achieved with scFvmut (Figure 1F; Figure S4B). Titration of the scFvαIIbβ3-containing supernatant demonstrated a concentration-dependent, competitive reduction in PAC-1, with complete inhibition at lower dilutions (Figure 1G). These data demonstrate that mRNA-expressed scFvαIIbβ3 selectively targets and competitively inhibits the activated GPIIb/IIIa integrin. Because ligand engagement of GPIIb/IIIa is critical for platelet aggregation via fibrinogen/fibrin,23 we next examined whether mRNA-expressed scFvαIIbβ3 inhibits platelet aggregation. Light transmission aggregometry revealed that mRNA-expressed scFvαIIbβ3 markedly inhibited ADP-induced platelet aggregation compared with control supernatant (Figure 1H and 1I; Figure S5), confirming functional blockade of GPIIb/IIIa-mediated platelet crosslinking. Both eptifibatide and scFvαIIbβ3 effectively blocked platelet aggregation, with no significant difference between treatments (P=nonsignificant). To model dynamic shear conditions, we used a microfluidic flow-chamber assay using collagen-coated channels and citrated whole blood. This assay allows for in vitro modeling of in vivo thrombus formation under controlled conditions.21 Platelets were labeled with 3,3′-dihexyloxacarbocyanine iodide dye, and clot formation was visualized under both 2-dimensional and 3-dimensional conditions. Although control supernatant had no effect, perfusion with supernatant containing mRNA-expressed scFvαIIbβ3-treated blood significantly reduced platelet adhesion and thrombus formation in collagen-coated microchannels. After flow, surface-bound scFvαIIbβ3 was detected using an anti-His Alexa Fluor647-conjugated antibody, confirming its binding to activated platelets and effective blockade of GPIIb/IIIa under shear conditions, thereby preventing aggregation and clot formation (Figure 2A and 2B; Videos S1 and S2). These findings were consistent across repeated experiments. Together, these data demonstrate that mRNA-expressed scFvαIIbβ3 selectively binds to activated GPIIb/IIIa, effectively blocks ligand binding, and inhibits platelet aggregation and thrombus formation under both static and flow conditions in vitro. scFvαIIbβ3 (single-chain variable fragment targeting the activated glycoprotein IIb/IIIa receptor) prevents thrombus formation under in vitro flow conditions. A, Representative images from a microfluidic flow-chamber assay showing thrombus formation on collagen-coated surfaces using citrated whole blood treated with vehicle control or mRNA-expressed scFvαIIbβ3 treatment; bright-field images show clots and immobilized collagen fibers, while green fluorescence (3,3′-dihexyloxacarbocyanine iodide) depicts thrombi and collagen fibers, and indicates platelet accumulation, and red fluorescence (anti-His Alexa Fluor647 antibody) marks scFvαIIbβ3 binding to platelets/thrombi. B, Quantification of thrombus area and C, Green mean fluorescence intensity from control and mRNA-expressed scFvαIIbβ3-treated samples. Data are presented in B and C as mean±SD (n=5 biological replicates). Statistical analysis was performed using paired Student t tests. a.u. indicates arbitrary units; dm2 square decimeter; and w/o, without. OryKL-Encapsulated scFvαIIbβ3 mRNA Prevents Arterial Thrombosis In Vivo For in vivo studies, scFvαIIbβ3 mRNA was delivered using OryKL, a LNP formulation, which has previously demonstrated safety and efficiency for systemic mRNA delivery.22 Cryogenic transmission electron microscopy revealed homogeneous spherical particles with highly ordered internal structures, whereas dynamic light scattering confirmed a uniform size distribution (114±2 nm; polydispersity index <0.2; Figure 3A). Encapsulation efficiency exceeded 90% (Figure S6). mRNA-loaded γ-oryzanol and DLin-KC2-DMA comprising lipid nanoparticles (OryKL) mediates hepatic mRNA expression after intravenous delivery. A, Schematic diagram representing mRNA encapsulation into OryKLs (left); representative cryogenic transmission electron microscopy images of formulated particles (middle); size and polydispersity index of scFvαIIbβ3 (single-chain variable fragment targeting the activated glycoprotein IIb/IIIa receptor) mRNA/OryKL (right, n=3; with 3 technical replicates). B, Representative IVIS bioluminescence images of hairless and immunocompetent (SKH1) mice at day 1 and day 2 after intravenous administration of PBS or NanoLuc (nLuc [nanoluciferase]) mRNA–loaded OryKLs (10 μg mRNA per mouse). The accompanying bar graph shows total flux measurements. C, Representative immunofluorescence images of liver sections collected 24 hours after intravenous delivery of scFvαIIbβ3 mRNA–loaded OryKLs or control treatment. Nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole; blue), and scFvαIIbβ3 expression was detected using anti-His Alexa Fluor 647 (red). Data in B are presented as mean±SD (n=4). Statistical analysis was performed using paired Student t tests. d, diameter (hydrodynamic); max, maximum; min, minimum; and PDI, polydispersity index. Created in BioRender. Wang, X. (2026) https://BioRender.com/5pbhsnd. Biodistribution of nanoluciferase mRNA delivered by OryKL was assessed in SKH1 hairless mice. Bioluminescence imaging revealed a strong, localized signal in the hepatic region at 24 hours after intravenous administration, with minimal signal detected in extrahepatic tissues, indicating predominant liver expression (Figure 3B). Liver sections from mice treated with scFvαIIbβ3 mRNA–loaded OryKLs showed robust anti-His immunofluorescence at 24 hours, whereas no signal was detected in control animals (Figure 3C), confirming successful hepatic expression of scFvαIIbβ3 mRNA after systemic delivery. To assess the antithrombotic effect of scFvαIIbβ3 mRNA in vivo, we utilized the FeCl3-induced carotid artery injury model. Mice were injected intravenously with either PBS, eptifibatide, or OryKL-encapsulated eGFP (enhanced green fluorescent protein), scFvmut, or scFvαIIbβ3 mRNA. Carotid injury was induced 24 hours postinjection, and blood flow was measured via a Doppler flow probe (Figure 4A and 4B). Control groups (PBS, eGFP, or scFvmut mRNA) developed rapid thrombus formation and complete arterial occlusion in around 7 to 9 minutes (7.9±3.3 minutes, n=12, 8.6±2.5 minutes, n=10, and 7.3±2.8 minutes, n=12, respectively). Eptifibatide-treated mice exhibited significantly delayed (5 in 10 mice) or no occlusion (5 in 10 mice), with occlusion times between 16 and 30 minutes (n=10; Figure 4B through 4D), consistent with prior reports.14,20 Preventive effect of mRNA-expressed scFvαIIbβ3 (single-chain variable fragment targeting the activated glycoprotein IIb/IIIa receptor) on thrombosis in vivo. A, Schematic diagram depicting the treatment regimen in a FeCl3-induced thrombosis mouse model. B, Effects of scFvαIIbβ3 mRNA and controls, including vehicle, 10 μg/g body weight eptifibatide (Ept), 50 μg of nonbinding mutant (scFvmut) or eGFP (enhanced green fluorescent protein) mRNA/OryKL on thrombosis; representative traces of carotid artery blood flow were recorded using a nano-Doppler flow probe (left); quantification of blood flow over time shown as area under the curve (AUC; right). C, Violin plot of the carotid artery occlusion times in each treatment group. D, Kaplan-Meier survival curve indicating the proportions of mice without occlusion after FeCl3-induced thrombosis. Data are presented as mean±SD (n=9–12 mice per group; ∆male; □female). Statistical analyses were performed using 1-way ANOVA, followed by Tukey post hoc multiple comparisons test (B), Kruskal-Wallis tests with Dunn multiple comparison (C), and Mantel-Cox tests (D). 24 h b indicates 24 hours before; a.u., arbitrary units; Ket, ketamine; LNP, lipid nanoparticle; mut, scFvmut; and Xyl, xylazine. Created in BioRender. Wang, X. (2026) https://BioRender.com/srdth7l. In contrast, animals receiving 50 μg of scFvαIIbβ3 mRNA demonstrated a significant delay in thrombus formation (24.0±8 minutes, n=11, P<0.05), as compared with all 3 controls (PBS, scFvmut and eGFP mRNA). Mice treated with 25 μg of scFvαIIbβ3 mRNA (16.2±6.8 minutes, n=11) also showed a significant difference in occlusion time compared with scFvmut mRNA control (P<0.05); however, the 10 μg group (11.0±5.5 minutes, n=11) had no statistical difference from negative controls (Figure 4C). Interestingly, 1 in 11 mice (9.1%) in the 25 μg group and 6 in 11 (54.5%) in the 50 μg group exhibited a complete absence of occlusion during the 30-minute monitoring period (Figure 4B through 4D). These observations were supported by quantitative analyses of the area under the curve, which measures blood flows and percentages of nonoccluded vessels over time.24 Both parameters were significantly increased in the 50 μg group, reflecting sustained arterial patency and reduced thrombus burden, as compared with all 3 controls (PBS, eGFP, and scFvmut mRNA, Figure 4B; right). Notably, animals receiving 25 μg of scFvαIIbβ3 mRNA also exhibited a significant increase in area under the curve compared with PBS and eGFP mRNA, with a trend (but nonsignificant) against scFvmut mRNA controls, suggesting a subtle effect of scFvαIIbβ3 mRNA that was not detected by occlusion time measurements. These results demonstrate that mRNA-encoded scFvαIIbβ3, when delivered via OryKLs, selectively targets activated platelets in vivo and effectively prevents thrombosis in a well-established model of arterial injury. scFvαIIbβ3 mRNA Preserves Hemostasis and Exhibits Favorable Safety Profile Physiological hemostasis was assessed using a tail bleeding time assay 24 hours after intravenous injection. As expected, eptifibatide-treated mice exhibited prolonged bleeding, with no cessation within the 30-minute observation period. In contrast, mice treated with 10, 25, or 50 μg of scFvαIIbβ3 mRNA showed no significant difference in bleeding time compared with PBS, scFvmut, or eGFP mRNA controls (Figure 5A). Quantification of blood loss by hemoglobin measurement confirmed significantly increased bleeding in eptifibatide-treated mice. All mRNA–treated groups remained comparable to PBS (Figure 5B). These results indicate that mRNA-expressed scFvαIIbβ3 selectively inhibits activated platelets while retaining hemostasis.Systemic tolerability was evaluated 24 hours postinjection using hematologic and biochemical analyses. Red blood cell counts and hemoglobin levels remained unchanged across groups (Figure 6A). Liver enzyme levels remained within normal ranges, with a mild elevation in ALT (alanine aminotransferase) at higher scFvαIIbβ3 mRNA doses (Figure 6B). Markers of renal function (creatinine and urea) were unaffected (Figure 6C). Organ weights (spleen, liver, kidney, and lung) of mice treated with mRNA/OryKL were not significantly different compared with PBS control (Figure 6D). Histological analysis of tissue sections revealed no evidence of hemorrhage, necrosis, or inflammatory infiltration in any major organ 24 hours after treatment (Figure 6E). These observations confirm that systemic delivery of OryKL-encapsulated scFvαIIbβ3 mRNA is well-tolerated and does not cause overt tissue toxicity. Effects of scFvαIIbβ3 (single-chain variable fragment targeting the activated glycoprotein IIb/IIIa receptor) mRNA on hemostasis assessed by tail bleeding. A, Tail bleeding time was measured after tail transection in each treatment group. B, Quantification of blood loss based on absorbance (Abs) at 550 nm after red blood cell lysis of collected blood. Data are presented as mean±SD (n=11–12 mice per group). Statistical analyses were performed using Kruskal-Wallis tests with Dunn multiple comparison (A) and 1-way ANOVA, followed by Tukey post hoc multiple comparisons test (B). The P values shown indicate comparisons with the positive eptifibatide (Ept) control only. eGFP indicates enhanced green fluorescent protein; and Mut, mutant. In vivo safety evaluation of scFvαIIbβ3 (single-chain variable fragment targeting the activated glycoprotein IIb/IIIa receptor) mRNA therapy in mice. A, Hematologic parameters 24 hours after intravenous (IV) injection of controls or various doses of scFvαIIbβ3 mRNA, including red blood cell count (RBC, left) and HGB (hemoglobin) content (right); B, Liver function markers: ALP (alkaline phosphatase; left), ALT (alanine aminotransferase; middle), and AST (aspartate aminotransferase; right). C, Kidney function markers: serum creatinine (left) and urea (right). D, Weights of major organs (spleen, liver, kidney, and lung) after treatment across groups. E, Representative hematoxylin and eosin–stained sections of spleen, liver, kidney, and lung from mice treated with control or 50 μg of scFvαIIbβ3 mRNA. Images were taken at ×20 magnification using the Zeiss Axioscan 7 microscope. Scale bars: 100 μm. Data are shown as mean±SD (n=9–14 mice per group). Statistical analysis was performed using 1-way ANOVA, followed by Tukey post hoc multiple comparisons test (A through C and D spleen, liver, and kidney) and Kruskal-Wallis with Dunn multiple comparisons (D lung). Ept indicates Eptifibatide; Mut, mutant; and scFv, single-chain antibody fragment variant. These safety and histopathologic findings corroborate the functional hemostasis data and support the use of scFvαIIbβ3 mRNA encapsulated in OryKLs as a safe antithrombotic strategy with minimal off-target effects. Prolonged Thromboprophylactic Efficacy of scFvαIIbβ3 mRNA Delivered via samRNA Although scFvαIIbβ3 mRNA effectively delayed thrombus formation in vivo, its duration of protection was limited to ≈24 hours after injection. This observation aligns with our in vitro expression data, which showed a peak in scFvαIIbβ3 protein secretion around 24 hours, followed by a decline by 72 hours (Figure 1C). To extend the therapeutic window, we explored the use of samRNA, which encodes an RNA-dependent RNA polymerase and self-replicating elements that enable prolonged protein expression from a single dose.25,26 Unlike DNA-based vectors, samRNA remains cytoplasmic and does not pose a risk of genomic integration. This long-lasting expression offers an unprecedented opportunity to convert our scFvαIIbβ3 mRNA into a clinically viable, sustained-release, and long-acting antithrombotic drug. Therefore, we generated a samRNA construct encoding for scFvαIIbβ3 (Figure 7A and 7B). In CHO cells, scFvαIIbβ3 protein expression from samRNA was comparable to conventional mRNA at 24 hours (Figure S7) but continued to increase, peaking at ≈1 week and remaining detectable for up to 2 weeks (Figure 7C; Figure S7). samRNA-expressed scFvαIIbβ3 selectively bound to activated GPIIb/IIIa and inhibited PAC-1 binding (Figure 7D). It also exhibited a strong inhibitory effect on ADP-induced aggregation, as compared with control or conventional mRNA-expressed scFvαIIbβ3 collected at 1 week post-mRNA transfection (Figure 7E), confirming prolonged expression by the samRNA. Importantly, the inhibitory effect of samRNA-expressed scFvαIIbβ3 is similar to the positive control, Eptifibatide (P=nonsignificant). Design and evaluation of scFvαIIbβ3 (single-chain variable fragment targeting the activated glycoprotein IIb/IIIa receptor) self-amplifying mRNA (samRNA) for prolonged antithrombotic protection. A, Schematic drawing of scFvαIIbβ3 samRNA construct. B, Size and integrity of the in vitro synthesized samRNA (9060 bp) visualized by agarose gel electrophoresis. C, Time-course analysis of scFvαIIbβ3 expression and secretion from samRNA-transfected cells over 15 days, detected by Western blotting. D, Competitive binding of samRNA-expressed scFvαIIbβ3 vs PAC-1 (antibody recognizing an epitope on the glycoprotein IIb/IIIa) in activated human platelet-rich plasma (PRP) by flow-cytometry assay; representative histograms show minimal PAC-1–fluorescein isothiocyanate (FITC) staining in nonactivated PRP (nonactivated platelets [NAP], far left), strong PAC-1 binding in activated human PRP (middle), and reduced PAC-1–FITC signal in activated PRP preincubated with samRNA-expressed scFvαIIbβ3 protein (right in purple), indicating effective competitive binding; bar graph quantifying FITC mean fluorescence intensity (MFI; far right). E. Inhibition of aggregation of human platelets by scFvαIIbβ3 protein harvested 1 week after mRNA or samRNA transfection, measured by light transmission aggregometry (595 nm, 10 minutes after ADP addition). Eptifibatide (Ept) was included as a control for the aggregation assay. F, Violin plot of carotid artery occlusion times in mice treated with control or samRNA-loaded lipid nanoparticles (LNPs), using either the Onpattro (ONP) or γ-oryzanol and DLin-KC2-DMA comprising LNP (OryKL) formulation. G, Area under the curve (AUC) analysis of carotid artery blood flow after FeCl3-induced thrombosis. H, Kaplan-Meier survival curve showing percentage of mice without (w/o) occlusion. I, Tail bleeding time 1 week after treatment with control or scFvαIIbβ3 samRNA–loaded LNPs. J, Body weight measurements before and after 1 week of treatment with control or scFvαIIbβ3 samRNA–loaded LNPs. Data are presented as mean±SD (D and E, n=3 biological replicates; F through J, n=8–14 mice per group; ∆male; □female). Statistical analyses were performed using 1-way ANOVA, followed by Tukey post hoc multiple comparisons test (D, E, and G), Kruskal-Wallis with Dunn multiple comparisons (F and I), Mantel-Cox tests (H), and 2-way ANOVA with Holm-Sidak multiple comparisons (J). a.u. indicates arbitrary units; AP, activated platelets; L, ladder; nsP, nonstructural protein; UTR, untranslated region; VH, variable domain of the heavy chain; and VL, variable domain of the light chain. To evaluate the in vivo thromboprophylactic efficacy, 50 μg of scFvαIIbβ3 samRNA, encapsulated in OryKL or the benchmark Onpattro formulation, was injected intravenous before injury. In PBS-treated control, complete occlusion occurred rapidly (8.3±2.1 minutes; n=12). At day 7, mice treated with Onpattro formulation demonstrated significantly prolonged vessel patency (16.9±9.6 minutes; n=14). Mice treated with OryKL showed similarly prolonged vessel patency (20.1±9.8 minutes; n=14; P=nonsignificant), where 6 in 14 animals (42.9%) maintained partial or full blood flow for the entire 30-minute monitoring period (Figure 7F). These results were corroborated by significantly higher area under the curve values (Figure 7G) and increased nonocclusion percentages (Figure 7H), confirming sustained thromboprophylactic activity 1 week after a single dose. The therapeutic effect of samRNA occurs by day 1 postinjection, and occlusion time remained significantly prolonged at day 10. By day 14, occlusion times were indistinguishable from PBS-treated animals, indicating loss of functional, platelet-inhibiting activity at this time point (Figure S8). To further evaluate safety, tail transection assays were performed and showed no significant prolongation of bleeding time or increase in hemoglobin loss after scFvαIIbβ3 samRNA treatment (Figure 7I), indicating preserved physiological hemostasis. Mice were also monitored over 7 days for weight changes and general health (Figure 7J). The weights of major organs were unchanged after samRNA treatment, with the exception of a modest increase in spleen weight for both OryKL and Onpattro formulation delivery platforms (Figure S9A). Hematologic analysis revealed no significant differences in red blood cell counts or HGB (hemoglobin) levels between control and samRNA–treated groups (Figure S9B). Serum markers of liver and renal function, including ALP (alkaline phosphatase), ALT (alanine aminotransferase), AST (aspartate aminotransferase), creatinine, and urea, were not altered by treatment with samRNA-loaded OryKLs (Figure S10A and S10B). In addition, serum cytokine analysis showed no evidence of systemic inflammatory activation after samRNA administration (Figure S10C through S10H). These findings demonstrate that scFvαIIbβ3 delivered via samRNA maintains its functional antiplatelet activity while offering significantly extended thromboprophylaxis without significantly compromising safety. Discussion In this study, we developed and validated novel mRNA-based antithrombotic therapies by expressing a scFv that selectively blocks the activated platelet integrin GPIIb/IIIa and thereby inhibits platelet aggregation. Using a well-established FeCl3-induced carotid artery thrombosis model, we demonstrate that scFvαIIbβ3 mRNA systemically delivered via LNPs significantly delayed thrombus formation in a dose-dependent manner, with a favorable safety profile. Comparable protective efficacy was achieved using both conventional and samRNA formats, although the antithrombotic effect was notably prolonged in mice treated with samRNA, persisting for 7 days and beyond. Importantly, treatment with scFvαIIbβ3 mRNA and samRNA did not increase tail bleeding time or total blood loss, in contrast to the clinically used GPIIb/IIIa inhibitor eptifibatide. There was also no evidence of organ damage or systemic toxicity, and treated animals maintained stable body weight and normal behavior. These findings indicate that mRNA-expressed scFvαIIbβ3 is an effective antiplatelet approach that promises to avoid the bleeding complications seen with clinically used GPIIb/IIIa inhibitors. To our knowledge, this is the first study to apply mRNA technology for the expression of a therapeutic antiplatelet biological drug aimed towards antithrombotic therapy. The antithrombotic profile of scFvαIIbβ3 mRNA and samRNA observed in vivo is consistent with its known mechanism of action, its selective binding and inhibition of the activated conformation of GPIIb/IIIa.14,20,21,27 Our scFvαIIbβ3 is specifically engineered to recognize this activated conformational epitope and to bind to it under both static and shear flow conditions. This selective binding blocks fibrinogen from binding to the activated GPIIb/IIIa, thereby blocking the fibrinogen-mediated crosslinking of platelets and thus inhibiting platelet aggregation. Crucially, scFvαIIbβ3 does not interact with the resting, low-affinity conformation of GPIIb/IIIa present on circulating platelets. This preserves platelet adhesion to immobilized fibrinogen and early hemostatic plug formation at vascular injury sites.28 As a result, scFvαIIbβ3 preserves physiological hemostasis, as shown by the absence of prolonged bleeding time in vivo.14,15,20,21,28 This contrasts sharply with conventional antiplatelet agents, such as aspirin and clopidogrel, which inhibit platelet activation broadly, and with GPIIb/IIIa inhibitors like eptifibatide, which block both nonactivated and activated GPIIb/IIIa receptors and thereby all of these antiplatelet drugs increase bleeding risk.6 These bleeding liabilities are significantly restricting the clinical utility of antiplatelet drugs. Consequently, a strategy that can achieve potent platelet inhibition while sparing resting platelets and maintaining hemostasis has the potential to fundamentally expand the patient populations eligible for antiplatelet therapy. Using a conventional mRNA platform, platelet inhibition is confined to a transient 24- to 72-hour time window. This approach appears suitable, for example, for elective, high-risk percutaneous coronary interventions or surgeries that are prone to acute platelet activation, such as coronary artery bypass grafting. In these perioperative, acute prothrombotic scenarios, the temporary expression of scFvαIIbβ3 offers robust antithrombotic protection without bleeding risk associated with long-acting and nonactivation-specific antiplatelet agents.29 Conversely, the samRNA platform, which provides an extended therapeutic window, holds significant potential for medium-term prophylactic applications. This sustained expression is particularly advantageous for patients facing prolonged thrombotic risk, such as those undergoing chemotherapy or experiencing extended periods of immobility, where effective antithrombotic effects without bleeding risk are required. Especially the use of scFvαIIbβ3, beyond its antiplatelet effects, as clot-targeting moiety for anticoagulants (as discussed below) is highly attractive. Longer-term secondary prophylaxis, for example, in patients after myocardial infarction also represents a valid aspirational goal, particularly with further expected improvements of the samRNA platforms, regarding the achievable time periods of protein expression. The therapeutic potential of scFvαIIbβ3 mRNA demonstrated in this study primarily relates to arterial thrombosis, where platelet activation clearly plays the dominant role. Importantly, broader clinical opportunities exist in using scFvαIIbβ3 as an activated platelet–specific targeting module within an mRNA-based platform. In this context, scFvαIIbβ3 could be used for site-directed delivery of additional antithrombotic effectors to sites of thrombus development. For example, mRNA constructs encoding fusions of scFvαIIbβ3 with anticoagulant, fibrinolytic or anti-inflammatory proteins, such as TAP (tick anticoagulant peptide),30 scuPA (single-chain urokinase plasminogen activator),21 or CD39 (ectonucleoside triphosphate diphosphohydrolase 1),31,32 could enable localized delivery to thrombus-developing sites while minimizing systemic exposure. In such settings, this targeting strategy, rather than the isolated antiplatelet activity of scFvαIIbβ3 itself, may be relevant to broader clinical indications, including thromboprophylaxis for patients with atrial fibrillation or venous thromboembolism. For the current proof-of-concept study, we used unmodified samRNA, as until recently, samRNA was typically synthesized with canonical nucleotides due to the replicase complex’s intolerance towards chemically modified nucleotides.33,34 However, recent studies have demonstrated successful substitution of uridine and cytosine by chemically modified nucleotides, such as 5-hydroxymethylcytidine, 5-methyluridine, and 5-methylcytidine, without compromising replicase function.33,34 These modifications help prevent innate immune recognition and degradation, and thereby extend in vivo protein expression. Additionally, structural engineering of mRNA with multicapped 5′ ends using locked N7-methylguanosine analogs and incorporation of nondegradable RNA oligos into branched poly(A) tails have been shown to significantly enhance expression up to 10-fold in vivo35 and to prolong the duration of protein expression 2- to 3-fold.36 In summary, future systematic evaluation of modified samRNA constructs holds potential to further extend the therapeutic window while also improving tolerability. In general, exogenous pharmacological control is an important consideration in the development of antithrombotic drugs. In the present study, the platelet activation-specific targeting with scFvαIIbβ3 is designed to preserve hemostasis, thereby mitigating the need for acute reversal agents. Nevertheless, the modular nature of the samRNA platform enables exogenous control of antithrombotic activity, as recently demonstrated using the small-molecule inhibitor ML336.37 Such off-switch functionality is particularly valuable when adapting the platform to therapeutic payloads with narrower safety margins. Optimizing LNP delivery systems remains a critical determinant of successful mRNA-based therapeutics. In this study, OryKLs were selected based on favorable performance in our preliminary screening assays.22 Both conventional mRNA and samRNA achieved robust antithrombotic effects when formulated with OryKLs, underscoring the functional compatibility of this LNP delivery platform with different RNA modalities. It has been reported that samRNA encapsulation does not universally translate into effective protein expression with all LNPs, whereas OryKLs consistently support expression and efficacy, highlighting the importance of matching RNA constructs with optimal delivery systems.38 Although the translational potential of OryKLs is supported by the in vivo efficacy observed, the precise intracellular delivery efficiency and translation kinetics of scFvαIIbβ3 mRNA/samRNA remain to be fully characterized. Prior studies have estimated that cytosolic release of RNA from endocytosed LNPs can be as low as 2%, potentially constraining protein expression39–42 but this also highlights the opportunity and the potential scope to enhance delivery performance. Strategies such as improving endosomal escape, increasing cellular uptake, and refining particle composition were suggested to augment both expression and reproducibility.43 Importantly, these refinements must be achieved while maintaining low immunogenicity, high biocompatibility, and shelf stability to meet the demands of clinical drug development. Across all in vivo studies, treatment with scFvαIIbβ3 mRNA or samRNA was not associated with relevant off-target pharmacology. A modest increase in spleen weight observed in the absence of changes in other organ weights, hematologic parameters, serum cytokines, or liver and renal function assays is consistent with transient innate immune sensing rather than pathological inflammation. A slight increase in ALT level was observed after administration of the highest dose of scFvαIIbβ3 mRNA, but values remained within the normal physiological range. Similar transient ALT elevations have been reported for polyethylene glycol–containing LNPs in both preclinical and clinical studies of approved mRNA-based therapies.38,44–47 The polyethylene glycol–lipid content of our LNP formulation is comparable to that of clinically approved platforms, supporting translational relevance. The transient splenic responses and mild hepatic enzyme elevations associated with polyethylene glycol–containing LNPs are attributed to short-lived innate immune activation rather than hepatocellular injury.44–46 Overall, these findings support a favorable safety profile of the described mRNA platforms. Bleeding risk is a major determinant of the success of antithrombotic drug development/use; this is particularly important as patients who require such treatments are often at risk for bleeding as well. Our platelet activation-specific scFvαIIbβ3 is unique in its preservation of hemostasis, as we have shown in tail bleeding time and blood loss assays, which are considered a standard in the preclinical assessment of hemostatic safety. To strengthen the important safety profile, additional complementary approaches, including humanized platelet function assays, microfluidic vascular injury platforms, saphenous vein scalpel injury, and integrated profiling of coagulation and endothelial responses, will be valuable in future studies toward completing the preclinical drug development program.48–50 An unexplored area in this study is the potential influence of sex differences on the response to the presented scFvαIIbβ3 mRNA therapy. Sex-dependent differences in platelet biology have been reported, with platelets from female mice exhibiting greater agonist-induced activation and aggregation in ex vivo assays compared with platelets from male mice.51 However, reports on the impact of sex on in vivo arterial thrombosis models, such as FeCl3-induced injury, are inconsistent, with reports of shorter, similar, or longer occlusion times in female mice.52–54 Future studies specifically designed and powered to evaluate sex as a biological variable in mRNA-based antiplatelet therapies will be important to determine whether sex influences the efficacy or safety of mRNA-based therapeutic antiplatelet strategies. There are several important challenges regarding the clinical development of antithrombotic mRNA therapeutics. Regulatory pathways for mRNA-based therapeutic platforms have yet to be fully defined, and the clinical development of novel antithrombotic agents is intrinsically costly. Such drug development programs and their associated clinical trials must benchmark new candidates against established standards of care while simultaneously meeting safety expectations, with particular attention to bleeding risk. Despite recent clinical successes and its overall promises, mRNA therapy in general comes with limitations as discussed above and summarized here: Although LNPs protect mRNA and enable cellular uptake, only a fraction of internalized RNA escapes endosomes to reach the cytoplasm, limiting effective protein expression.16,40–42,55–57 LNP-based delivery is also predominantly directed towards the liver, which, in the case of the production of scFvαIIbβ3, is highly beneficial, but may require alternative tissue targeting for other applications. Both mRNA and its LNP carriers, including the polyethylene glycol component of the LNPs, may provoke innate immune sensing, particularly at higher doses, requiring careful balancing of efficacy and tolerability. In addition, mRNA expression is inherently transient, which enhances safety but limits suitability for chronic indications and may necessitate repeat dosing, which has not yet been systematically explored. However, recent technical progress in the field and the multitude of research groups and pharmaceutical companies heavily engaged in this attractive field of research hold enormous promise for successful clinical translation of mRNA therapeutics, including in cardiovascular medicine. In summary, this study provides compelling preclinical evidence supporting mRNA- and samRNA-based delivery of scFvαIIbβ3 as a novel thromboprophylactic strategy. The innovative mRNA-based therapeutics combine activated platelet specificity, extended thromboprophylaxis, and safety, and represent a promising direction for preventive antithrombotic therapy. Further studies are warranted to optimize delivery, extend expression duration, and ensure long-term safety in preparation for clinical translation. ARTICLE INFORMATION Acknowledgments The authors acknowledge the facilities and assistance of the BASE mRNA Facility at the University of Queensland, supported by Therapeutic Innovation Australia, part of the National Collaborative Research Infrastructure Strategy (NCRIS) of the Australian Government. The authors also acknowledge Microscopy Australia at the Ramaciotti Centre for Cryo-Electron Microscopy, Monash University, enabled by NCRIS. Schematic illustrations in Figures 3 and 4 were created with BioRender.com. Author Contribution Y. Song, T.H. Nguyen, X. Wang, and K. Peter conceptualized the idea and designed the study; Y. Song, T.H. Nguyen, P. Shi, A. Refaat, S. Ren, C.V. Pham, B. Xu, J.I. Skoraczynski, A. Nguyen, H. Liu, A. Klesen, M.L.P. Vidallon, and S. Trépout performed experiments; Y. Song, T.H. Nguyen, and X. Wang analyzed and explained the data and wrote the manuscript; K. Peter revised the manuscript. Disclosures Baker Heart and Diabetes Institute has filed 2 patent applications (Application Nos. 2025903067 and 2025903069) related to the materials described in this manuscript. The authors report no conflicts. Supplemental Material Supplemental Methods Figures S1–S11 Videos S1–S2 Major Resources Table Unedited Western Blots Supplementary Material