Endothelium-targeted and untargeted nanoparticle therapies in thromboinflammatory vascular diseases Authors: Micheal O. Avoseh, Hua Jin, Oscar Moreno, Andrea T. Obi, Peter K. Henke, Souvik Sen, Michael A. Matthay, You-Yang Zhao, Colin E. Evans Published: 2026 ABSTRACT AbstractEndothelial cells (ECs) line the vasculature and integrate hemodynamic, inflammatory, and metabolic cues that are crucial in maintaining vascular health. Endothelial dysfunction is a driver of thromboinflammatory conditions including atherosclerosis, thrombosis, stroke, and sepsis-induced multiorgan dysfunction syndrome. Therapies for these conditions that are given systemically fail to target the site of disease and can be associated with off-target side effects. EC-targeted nanoparticles represent a novel method of concentrating the drug- or gene-based therapy in ECs at the site of disease. Here, we describe studies of EC-directed nanoparticles in thromboinflammatory vascular diseases. We discuss challenges in the development of EC-targeted nanoparticle therapies, including issues with biosafety, delivery efficiency, scalability, and clinical translation. To target ECs, nanoparticle surfaces have been coated with ligand-directed antibodies or peptides against EC markers. EC targeting has also been achieved by coating nanoparticles with cell-mimicking membranes to leverage native homing pathways to sites of vascular injury. Stimuli-responsive nanoparticles have used EC or disease-specific cues to trigger drug release, for instance via reactive oxygen species or shear stress stimuli. EC-targeted nanoparticles extend the therapy circulation time, enhance drug accumulation at target sites, improve therapeutic efficacy, and reduce toxicity. In preclinical studies, EC-targeted nanoparticles have stabilized atherosclerotic plaques, degraded venous thrombi, and reduced cerebral infarct volume. Although EC-targeted nanoparticle therapies have not yet been approved for clinical use, advances in the design and safety of nanoparticles are narrowing the gap between experimental promise and clinical reality. Introduction Endothelial cells (ECs) line the inner wall of blood vessels and play a pivotal role in regulating vascular homeostasis, permeability, blood flow, coagulation, and inflammation.1,2 Endothelial dysfunction is a characteristic of thrombotic vascular diseases including sepsis, atherosclerosis, stroke, and venous thromboembolism.1,3 Pathological activation of ECs leads to increased vascular inflammation, oxidative stress, vascular permeability, and thrombosis.1,3, 4, 5, 6 These effects are amplified by the increased release of inflammatory cytokines and EC adhesion to leukocytes; EC dysfunction in patients with sepsis can culminate in tissue hypoperfusion and multiorgan failure.6 In patients with atherosclerosis, activated ECs express adhesion molecules such as vascular cell adhesion molecule 1 (VCAM-1) and intercellular adhesion molecule 1 (ICAM-1), which increase monocyte recruitment and migration into the intima, where they differentiate into lipid-engulfing macrophages that are characteristic of early atherosclerotic plaques.7 The disrupted endothelial barrier also allows oxidized low-density lipoprotein cholesterol to infiltrate the lesion, further stimulating inflammation.5 In advanced atherosclerosis, plaque rupture and acute thrombosis cause myocardial infarction or stroke.8 In ischemic stroke, neurovascular inflammation activates the blood-brain barrier endothelium and induces the expression of cell adhesion molecules that recruit leukocytes into the ischemic penumbra, contributing to the ongoing blood-brain barrier disruption and vasogenic edema.9,10 In venous thromboembolism, endothelial activation promotes a prothrombotic phenotype through the release of von Willebrand factor from Weibel-Palade bodies and overexpression of E-/P-selectins on the EC surface, which increases leukocyte and platelet adhesion to the endothelium.3,4 Current treatments for thromboinflammatory vascular diseases, for example, anti-inflammatory, anticoagulant, and thrombolytic agents, act systemically and do not specifically localize to sites of endothelial injury.11 As a result, high doses are often required to achieve therapeutic efficacy, which can increase the risk of toxicity and off-target effects such as bleeding and immune suppression. The need for targeted delivery of vascular therapeutics has led to the development of EC-targeted nanoparticle therapies that aim to enhance therapeutic delivery and efficacy to the injured ECs.12 As described in this review, nanoparticles have been engineered with targeting ligands, biomimetic coatings that recognize diseased ECs, or as stimuli-responsive nanoparticles that selectively internalize into ECs and release their cargo in the pathological microenvironment. EC-directed nanoparticles are being developed to deliver diverse therapeutic cargos, including small-molecule drugs, RNAs, plasmid DNA, and recombinant proteins. EC-targeted nanoparticles, including those targeting the endothelial marker ICAM-1 or VCAM-1, are in preclinical development.9,13 In this review, we describe endothelium-targeted nanoparticle therapies for the inhibition or treatment of thromboinflammatory vascular diseases, which can be categorized as ligand-directed nanoparticles, biomimetic cell membrane-coated nanoparticles, or stimuli-responsive nanoparticles. We will also discuss the knowledge gaps and therapeutic challenges for the development of EC-targeted nanoparticle therapies in thromboinflammatory vascular diseases. Ligand-directed endothelium-targeted nanoparticle therapies Nanoparticle surfaces have been engineered with ligands that directly bind to the surface markers present on activated endothelium.14 ECs have numerous receptors and adhesion molecules that are upregulated on their surface during thromboinflammation, which can act as a unique “address” for EC-targeting nanoparticles.15 Targets of these nanoparticles include adhesion molecules such as E-selectin, P-selectin, ICAM-1, and VCAM-1, which are expressed at low levels in healthy ECs but are highly upregulated on inflamed ECs.9,16, 17, 18 Nanoparticles coated with antibodies or peptides targeting these molecules can bind directly to the inflamed endothelium even under physiological flow conditions.19 This allows nanoparticles to target diseased sites on the blood vessel wall and deliver therapies locally, without deeply penetrating into the tissue or organ. In mouse models of sepsis-induced organ dysfunction and in cultured ECs, co-coating of unloaded polymer nanoparticles with anti–ICAM-1 antibody and CD47 results in increased lung EC targeting and retention time, and decreased degradation.20,21 Kim et al20 found that co-coating of CD47 on anti–ICAM-1-coated nanoparticles reduces macrophage phagocytosis by ∼50%, and increases lung EC targeting in vitro by >85% compared with nanoparticles coated with anti–ICAM-1 alone. In mice, co-coating of CD47 on anti–ICAM-1-coated nanoparticles reduced uptake into the liver and spleen by 30% to 50% and increased lung targeting by approximately twofold compared with nanoparticles coated with anti–ICAM-1 alone.20 IV delivery of an anti-inflammatory drug (berberine) in these nanoparticles safely improves treatment efficacy in the lung of sepsis-challenged mice vs free berberine or berberine loaded in naked nanoparticles (Figure 1).Figure 1.Sepsis-induced lung injury and endothelium-targeted nanoparticle (NP) interventions. (A) Graphical representation of the generation of BBR-loaded polymer NPs (blue) coated with CD47 (red) to resist phagocytosis, and anti–ICAM-1 (green) to target inflammatory ECs. (B) BBR-loaded lung EC-targeting NPs were generated as described,20,21 and scanning electron microscopy of the co-coated NPs was performed as described.22 Scale bar, 200 nm. (C) To image the NP organ distribution, coumarin 6-loaded lung EC-targeting NPs were generated as described20,21 and administered IV to C57BL/6J mice aged 3 to 4 months at 3 hours after intraperitoneal (IP) liposaccharide (3 mg/kg). At 8 hours after NP administration, organs were excised to visualize NP distribution by fluorescence imaging as described.23 Images are representative of n = 3 to 5 per group; dashed lines denote separate images within a composite. Co-coating of polymer NPs with CD47 and anti–ICAM-1 resulted in increased uptake in the lung, as demonstrated by fluorescence imaging of coumarin 6-loaded NPs. (D) C57BL/6J mice (aged 3-4 months) received IP liposaccharide (3 mg/kg) then after 3 hours were treated with IV vehicle or NPs. At 24 hours after lipopolysaccharide (LPS), lungs were perfused and imaged for Evans blue-conjugated albumin retention as described.23 Images are representative of n = 3 to 5 per group. Co-coating of berberine-loaded NPs with CD47 and anti–ICAM-1 reduced lung injury, as demonstrated by reduced vascular permeability to Evans blue-conjugated albumin (blue). BBR, berberine; EDC, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide; NHS, N-hydroxysuccinimide; PBS, phosphate-buffered saline. Sepsis-induced lung injury and endothelium-targeted nanoparticle (NP) interventions. (A) Graphical representation of the generation of BBR-loaded polymer NPs (blue) coated with CD47 (red) to resist phagocytosis, and anti–ICAM-1 (green) to target inflammatory ECs. (B) BBR-loaded lung EC-targeting NPs were generated as described,20,21 and scanning electron microscopy of the co-coated NPs was performed as described.22 Scale bar, 200 nm. (C) To image the NP organ distribution, coumarin 6-loaded lung EC-targeting NPs were generated as described20,21 and administered IV to C57BL/6J mice aged 3 to 4 months at 3 hours after intraperitoneal (IP) liposaccharide (3 mg/kg). At 8 hours after NP administration, organs were excised to visualize NP distribution by fluorescence imaging as described.23 Images are representative of n = 3 to 5 per group; dashed lines denote separate images within a composite. Co-coating of polymer NPs with CD47 and anti–ICAM-1 resulted in increased uptake in the lung, as demonstrated by fluorescence imaging of coumarin 6-loaded NPs. (D) C57BL/6J mice (aged 3-4 months) received IP liposaccharide (3 mg/kg) then after 3 hours were treated with IV vehicle or NPs. At 24 hours after lipopolysaccharide (LPS), lungs were perfused and imaged for Evans blue-conjugated albumin retention as described.23 Images are representative of n = 3 to 5 per group. Co-coating of berberine-loaded NPs with CD47 and anti–ICAM-1 reduced lung injury, as demonstrated by reduced vascular permeability to Evans blue-conjugated albumin (blue). BBR, berberine; EDC, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide; NHS, N-hydroxysuccinimide; PBS, phosphate-buffered saline. Adhesion molecules such as E-selectin, P-selectin, ICAM-1, and VCAM-1 are also highly expressed and can thus be targeted on inflamed ECs in atherosclerosis (Figure 2). In atherosclerosis and thrombosis, in which activated platelets and the subendothelial matrix become exposed, nanoparticles can bind to target molecules such as collagen and von Willebrand factor.8,19 Nanoparticles can also be targeted to the endothelium using organ-specific markers. For example, a peptide that binds with high affinity to vascular endothelial growth factor receptor 3 (Fms-related tyrosine kinase 4), which is expressed on the Schlemm’s canal endothelium, has been used to target nanoparticles to the eye for glaucoma therapy, with endothelial-targeted nanoparticles localizing to diseased tissues more so than untargeted nanoparticles.24 In addition to vascular endothelial growth factor receptor 3/Fms-related tyrosine kinase 4, ligand-directed nanoparticles have been designed to selectively permeate by targeting the arginine-glycine-aspartate motif and the vitronectin receptor, also known as αvβ3 integrin,25,26 enabling organ-specific delivery to the brain10 and eye.24 Nong et al10 developed VCAM-1–targeted lipid nanocarriers that bind adhesion molecules on the inflamed blood-brain barrier in a transient middle cerebral artery occlusion model; these VCAM-1 nanocarriers achieved nearly twofold greater brain delivery than untargeted carriers and reduced infarct volume by 35% to 62%.Figure 2.Atherosclerosis and endothelium-targeted NP interventions. As blood vessels progress from a healthy state to advanced atherosclerosis, vascular changes include endothelial activation and dysfunction, which is associated with subendothelial retention of apoB-containing lipoproteins and leukocyte accumulation. Recruited monocytes differentiate into macrophages and become lipid-laden foam cells. Prolonged inflammation, smooth muscle cell migration and proliferation, and extracellular matrix deposition eventually lead to the formation of a fibrous cap over a lipid-rich necrotic core, followed by plaque rupture and thrombosis. Anti-atherosclerotic NP therapies can be targeted to activated endothelium to deliver anti-inflammatory and plaque-stabilizing drugs, thereby reducing leukocyte recruitment and slowing or reversing plaque progression. Figure created with biorender.com. Avoseh MO. (2026) https://biorender.com/klhjkw4. Atherosclerosis and endothelium-targeted NP interventions. As blood vessels progress from a healthy state to advanced atherosclerosis, vascular changes include endothelial activation and dysfunction, which is associated with subendothelial retention of apoB-containing lipoproteins and leukocyte accumulation. Recruited monocytes differentiate into macrophages and become lipid-laden foam cells. Prolonged inflammation, smooth muscle cell migration and proliferation, and extracellular matrix deposition eventually lead to the formation of a fibrous cap over a lipid-rich necrotic core, followed by plaque rupture and thrombosis. Anti-atherosclerotic NP therapies can be targeted to activated endothelium to deliver anti-inflammatory and plaque-stabilizing drugs, thereby reducing leukocyte recruitment and slowing or reversing plaque progression. Figure created with biorender.com. Avoseh MO. (2026) https://biorender.com/klhjkw4. Antibody- and membrane-coated nanoparticles have also been used to target the brain (Figure 3). Marcos-Contreras et al14 conjugated anti–VCAM-1 antibodies to lipid nanoparticles loaded with thrombomodulin messenger RNA (mRNA) to restore this anticoagulant protein in inflamed brain endothelium. In a mouse model of tumor necrosis factor α (TNF-α)-induced brain inflammation, brain uptake was greater for VCAM-1–targeted nanoparticles than nontargeted nanoparticles.14 The nanoparticle-delivered thrombomodulin gene was expressed in the cerebral endothelium, alleviating vascular inflammation and reducing brain edema.14 Polymer nanoparticles functionalized with anti–ICAM-1 antibodies and E-selectin–binding peptides have also been tested in a physiological flow chamber, where they adhered to activated ECs.27 EC binding of targeted nanoparticles was greater than that of untargeted nanoparticles, as was nanoparticle internalization into ECs, leading to improved delivery of therapeutic cargo.27,28 This approach suggests that shear-resistant targeting is advantageous, given that blood flow can dislodge nanoparticles that bind less effectively.9,28 Examples of ligand-directed nanoparticles that exploit the molecular “zip codes” of dysfunctional endothelium are shown in Table 1.Figure 3.Stroke and endothelium-targeted NP interventions. Top left box: as cerebral blood vessels shift from healthy to injured states, endothelial dysfunction drives disease through VCAM-1–mediated leukocyte recruitment, blood-brain barrier disruption, and vasogenic edema in TNF-α-induced inflammation. Top right box: NPs coated with neural stem cell membranes have been used to target ischemic brain lesions, concentrating treatments in the stroke penumbra to promote neuroprotection and decrease infarct size. Bottom row: in ischemic stroke, arterial occlusion damages cerebral endothelium, triggers hypoxic signals, and creates a proinflammatory environment. Biomimetic and EC-targeted NP therapies can precisely target injured brain tissue vasculature: VCAM-1–conjugated lipid NPs have been used to deliver thrombomodulin mRNA to resolve inflammation and reduce brain edema. Figure created with biorender.com. Moreno O. (2026) https://biorender.com/zluc9w7. BBB, blood-brain barrier; NSC, neural stem cell.Table 1.Ligand-directed endothelium-targeted nanoparticles in thromboinflammatory diseasesTargetTypeModelOutcomeTherapyReferenceVCAM-1PolymericInflamed endotheliumTargeting improved EC uptakeInterleukin-10 plasmid DNA29ICAM-1 and VCAM-1Antibody-coated polystyrene nanospheresInflamed endotheliumCombination targeting improved EC capture and adhesionRecombinant acid sphingomyelinase30αvβ3 integrinPolymericHUVECs under static and flow conditionsTargeting enhanced EC association and uptakeFluorescent label28VEGFR3/FLT4MicellesGlaucomaTargeting enhanced homing to ECsLatrunculin A24Vascular ECsIonizable-lipid encapsulatingDeep vein thrombosisTargeting increased local tPA expression and safely reduced thrombosis burdentPA mRNA31FLT4, Fms-related tyrosine kinase 4; HUVEC, human umbilical vein endothelial cells; tPA, tissue plasminogen activator; VEGFR3, vascular endothelial growth factor receptor 3. Stroke and endothelium-targeted NP interventions. Top left box: as cerebral blood vessels shift from healthy to injured states, endothelial dysfunction drives disease through VCAM-1–mediated leukocyte recruitment, blood-brain barrier disruption, and vasogenic edema in TNF-α-induced inflammation. Top right box: NPs coated with neural stem cell membranes have been used to target ischemic brain lesions, concentrating treatments in the stroke penumbra to promote neuroprotection and decrease infarct size. Bottom row: in ischemic stroke, arterial occlusion damages cerebral endothelium, triggers hypoxic signals, and creates a proinflammatory environment. Biomimetic and EC-targeted NP therapies can precisely target injured brain tissue vasculature: VCAM-1–conjugated lipid NPs have been used to deliver thrombomodulin mRNA to resolve inflammation and reduce brain edema. Figure created with biorender.com. Moreno O. (2026) https://biorender.com/zluc9w7. BBB, blood-brain barrier; NSC, neural stem cell. Table 1.Ligand-directed endothelium-targeted nanoparticles in thromboinflammatory diseasesTargetTypeModelOutcomeTherapyReferenceVCAM-1PolymericInflamed endotheliumTargeting improved EC uptakeInterleukin-10 plasmid DNA29ICAM-1 and VCAM-1Antibody-coated polystyrene nanospheresInflamed endotheliumCombination targeting improved EC capture and adhesionRecombinant acid sphingomyelinase30αvβ3 integrinPolymericHUVECs under static and flow conditionsTargeting enhanced EC association and uptakeFluorescent label28VEGFR3/FLT4MicellesGlaucomaTargeting enhanced homing to ECsLatrunculin A24Vascular ECsIonizable-lipid encapsulatingDeep vein thrombosisTargeting increased local tPA expression and safely reduced thrombosis burdentPA mRNA31FLT4, Fms-related tyrosine kinase 4; HUVEC, human umbilical vein endothelial cells; tPA, tissue plasminogen activator; VEGFR3, vascular endothelial growth factor receptor 3. Ligand-directed endothelium-targeted nanoparticles in thromboinflammatory diseases FLT4, Fms-related tyrosine kinase 4; HUVEC, human umbilical vein endothelial cells; tPA, tissue plasminogen activator; VEGFR3, vascular endothelial growth factor receptor 3. Biomimetic endothelium-targeted nanoparticle therapies Biomimetic nanoparticles are cloaked in natural cell membranes from platelets, leukocytes, or erythrocytes to mimic cell surface biology.32, 33, 34, 35, 36, 37 This cloaking process typically involves isolating the source-cell membrane, forming membrane vesicles by sonication or extrusion, and then fusing these vesicles with preformed nanoparticle cores to generate membrane-coated core-shell particles that can retain source-cell membrane proteins.38, 39, 40 The coated nanoparticles directly interact with the endothelium.41,42 For vascular targeting, platelet and leukocyte coatings commonly interact with injured or inflamed endothelium.36,43 The addition of cell membranes extends the circulation time of the nanoparticles by reducing clearance by the immune system.33 In the brain, Ma et al44 used a biomimetic strategy in which nanoparticles were coated with the membrane of engineered neural stem cells; these particles homed to ischemic tissue, which concentrated the drugs in the injured vasculature in an experimental stroke model. Gao et al45 engineered a platelet-membrane–coated nanoparticle carrying angiopoietin-like 4 and Fe3O4, which is a reactive oxygen species (ROS) scavenger. The platelet coating localized the nanoparticles to the damaged endothelium in the ischemic heart; thrombin expressed at the lesion cleaved the linker to release angiopoietin-like 4 and stabilize vascular endothelial cadherin junctions, while internalized Fe3O4 scavenged endothelial ROS.45 In mice with myocardial ischemia-reperfusion injury, the nanoparticles reduced endothelial leakage, leukocyte extravasation, intracardial hemorrhage, and tissue damage, and improved cardiac function compared with untargeted nanoparticle formulations.45 Biomimetic nanoparticles cloaked with platelet membranes46 have been used in animal models of atherosclerosis.47 Platelet-coated poly(lactic-co-glycolic) acid nanoparticles carrying rapamycin can be used to target atherosclerotic plaque and the damaged endothelium; following ultrasound-targeted microbubble destruction, these nanoparticles deposit the drug locally and inhibit plaque growth in ApoE−/− mice.47 Nanoparticles cloaked in neutrophil or monocyte membranes also localize to inflamed endothelium.48,49 Neutrophil-membrane–coated nanoparticles retain β2-integrins (eg, CD18) that bind ICAM-1 on activated endothelium, driving accumulation in inflamed atherosclerotic lesions and suppressing plaque inflammation.48 Macrophage-membrane–camouflaged carriers likewise exploit endothelial activation via VCAM-1 to home to plaques and improve lipid metabolism and inflammation in ApoE−/− mice.50 Erythrocyte membranes extend the circulation time of nanoparticle-guided treatments. For example, when paired with shear- and ROS-responsive micelles, this biomimetic system enriches nanoparticle deposition in atherosclerotic plaques and releases simvastatin, which reduces plaque burden.51 The biodistribution and therapeutic efficacy of biomimetic nanoparticles outperformed those of uncoated nanoparticles by increasing their accumulation in the diseased vasculature.45,47,51 Examples of biomimetic endothelium-targeted nanoparticles and their outcomes are summarized in Table 2.Table 2.Biomimetic endothelium-targeted nanoparticles in thromboinflammatory diseasesTargetTypeModelOutcomeTherapyReferenceInjured myocardiumSilica: mesoporous silica nanoparticlesMyocardial ischemia: reperfusion (rat)Targeted injured myocardium; reduced MI/R injurySS31 peptide46Atherosclerotic plaquePLGA polymericAtherosclerosis (ApoE−/− mice)Plaque targeting achieved high drug deposition in plaques, inhibiting plaque growthRapamycin47ICAM-1 on inflamed endotheliumMetal-organic frameworksAtherosclerosis (inflamed endothelium)Targeting improved plaque homing; gene silencing reduced inflammationAnti-miR-15548Atherosclerotic plaquesPolymericAtherosclerosis (high ROS plaques)Targeted release prolonged circulation time and reduced plaque burdenSimvastatin prodrug51MI/R, myocardial ischemia/reperfusion; PLGA, poly(lactic-co-glycolic) acid; SS31, mitochondria-targeted antioxidant tetrapeptide (elamipretide). Table 2.Biomimetic endothelium-targeted nanoparticles in thromboinflammatory diseasesTargetTypeModelOutcomeTherapyReferenceInjured myocardiumSilica: mesoporous silica nanoparticlesMyocardial ischemia: reperfusion (rat)Targeted injured myocardium; reduced MI/R injurySS31 peptide46Atherosclerotic plaquePLGA polymericAtherosclerosis (ApoE−/− mice)Plaque targeting achieved high drug deposition in plaques, inhibiting plaque growthRapamycin47ICAM-1 on inflamed endotheliumMetal-organic frameworksAtherosclerosis (inflamed endothelium)Targeting improved plaque homing; gene silencing reduced inflammationAnti-miR-15548Atherosclerotic plaquesPolymericAtherosclerosis (high ROS plaques)Targeted release prolonged circulation time and reduced plaque burdenSimvastatin prodrug51MI/R, myocardial ischemia/reperfusion; PLGA, poly(lactic-co-glycolic) acid; SS31, mitochondria-targeted antioxidant tetrapeptide (elamipretide). Biomimetic endothelium-targeted nanoparticles in thromboinflammatory diseases MI/R, myocardial ischemia/reperfusion; PLGA, poly(lactic-co-glycolic) acid; SS31, mitochondria-targeted antioxidant tetrapeptide (elamipretide). Stimuli-responsive endothelium-targeted nanoparticle therapies Stimuli-responsive nanoparticles are engineered to release their therapeutic payload only in the presence of specific pathological triggers that are abundant at disease sites, allowing the nanoparticles to circulate in an inert state until activation.52, 53, 54 Common stimuli for EC-directed nanoparticles include enzymes, pH, redox conditions, and shear force.51,53,55 Stimuli-responsive nanoparticles are designed to minimize off-target effects by acting locally, for example, in an inflamed vessel segment or thrombus.56,57 One such enzyme-responsive system has been used to target tumor endothelium using γ-glutamyl transpeptidase-driven liposomes. Wang et al58 designed these liposomes that remain stable until they encounter γ-glutamyl transpeptidase on the luminal surface of tumor endothelium, which triggers caveola-mediated transcytosis across the endothelium into the tumor. This nanoparticle modification increased drug penetration and efficacy compared with nonresponsive nanoparticles.58 Another pathological stimulus that has been leveraged for nanoparticle-guided therapy is oxidative stress. Lesions in atherosclerosis and chronic inflammation exhibit elevated levels of ROS.59 Stimuli-responsive nanoparticles can leverage this scenario by releasing drug under ROS-rich conditions, which concentrates the therapy at diseased sites while limiting off-target exposure.60 For example, erythrocyte-coated, shear stress-, and ROS-responsive micelles carrying a simvastatin prodrug demonstrated prolonged circulation time, preferential plaque accumulation, ROS-triggered drug release, and reduced plaque inflammation and size.51 Platelet-membrane–coated poly(γ-glutamic acid) nanoparticles target pathological endothelium and disassemble in the presence of ROS, triggering cathepsin-B–mediated GWAS release, and enabling plaque imaging and therapy in atherosclerosis.60 Guo et al31 used an endothelial ionizable lipid nanoparticle approach for IV mRNA delivery. In a mouse deep vein thrombosis model, delivery of tissue-type plasminogen activator mRNA increased local tissue plasminogen activator expression and reduced the incidence and formation of venous thrombi (Figure 4). Thrombin-activated release of the mRNA via thrombin-triggered cleavage of a thrombin-sensitive peptide linker in platelet-membrane–coated nanoparticles improved cardiac function and reduced vascular leakage and apoptosis in a model of ischemia/reperfusion, compared with nonresponsive/untargeted nanoparticles.45 Examples of stimuli-responsive endothelial-targeting nanoparticles are shown in Table 3.Figure 4.Thrombosis and endothelium-targeted NP interventions. Thrombus formation involves endothelial activation and the aggregation of platelets, erythrocytes, and fibrin in the intact circulation. Endothelium-targeted lipid NPs have been used to deliver tissue plasminogen activator (tPA) mRNA to activated ECs, thereby increasing local tPA expression, enhancing fibrinolysis, and reducing thrombus burden.Table 3.Stimuli-responsive endothelium-targeted nanoparticles in thromboinflammatory diseasesTargetTypeModelOutcomeTherapyReferenceγ-Glutamyl transpeptidase on tumor endotheliumLiposomesHepatocellular carcinoma and pancreatic ductal carcinomaGGT-triggered transcytosis increased tumor accumulation and penetration; reduced tumor progression in both modelsDoxorubicin58Acidic pH (injury site) and disulfide cleavagePeptide-drug conjugateMyocardial ischemia-reperfusionpH-triggered assembly restored redox levels and improved I/R outcomes4-Octyl itaconate and elamipretide61ROS (tumor microenvironment)PolymericH1299 lung cancer cells; subcutaneous H1299 xenografts (mouse)ROS cleavage enabled p53 delivery; photodynamic therapy improved tumor inhibitionP53 mRNA and indocyanine green62ROS and ischemiaMicellesMCAO/R-induced neuroinflammationROS-triggered release reduced infarct size and neuroinflammationEdaravone63GGT, γ-glutamyl transpeptidase; I/R, ischemia/reperfusion; MCAO/R, middle cerebral artery occlusion and reperfusion. Thrombosis and endothelium-targeted NP interventions. Thrombus formation involves endothelial activation and the aggregation of platelets, erythrocytes, and fibrin in the intact circulation. Endothelium-targeted lipid NPs have been used to deliver tissue plasminogen activator (tPA) mRNA to activated ECs, thereby increasing local tPA expression, enhancing fibrinolysis, and reducing thrombus burden. Table 3.Stimuli-responsive endothelium-targeted nanoparticles in thromboinflammatory diseasesTargetTypeModelOutcomeTherapyReferenceγ-Glutamyl transpeptidase on tumor endotheliumLiposomesHepatocellular carcinoma and pancreatic ductal carcinomaGGT-triggered transcytosis increased tumor accumulation and penetration; reduced tumor progression in both modelsDoxorubicin58Acidic pH (injury site) and disulfide cleavagePeptide-drug conjugateMyocardial ischemia-reperfusionpH-triggered assembly restored redox levels and improved I/R outcomes4-Octyl itaconate and elamipretide61ROS (tumor microenvironment)PolymericH1299 lung cancer cells; subcutaneous H1299 xenografts (mouse)ROS cleavage enabled p53 delivery; photodynamic therapy improved tumor inhibitionP53 mRNA and indocyanine green62ROS and ischemiaMicellesMCAO/R-induced neuroinflammationROS-triggered release reduced infarct size and neuroinflammationEdaravone63GGT, γ-glutamyl transpeptidase; I/R, ischemia/reperfusion; MCAO/R, middle cerebral artery occlusion and reperfusion. Stimuli-responsive endothelium-targeted nanoparticles in thromboinflammatory diseases GGT, γ-glutamyl transpeptidase; I/R, ischemia/reperfusion; MCAO/R, middle cerebral artery occlusion and reperfusion. Endothelium-targeted nanoparticle therapies for delivery of nucleic acids Some nanoparticles can target vascular endothelium without an external antibody, peptide, or targeting ligand. Zhang et al23 reported a formulation of polymer nanoparticles that directly target vascular endothelium for delivery of plasmid DNA comprising the CRISPR system, which achieved robust EC-specific knockout of multiple genes in adult mice in <5 to 7 days. This technology has demonstrated efficacy in mouse models of sepsis-induced lung injury and thrombosis-associated inflammatory lung injury by improving the survival or proliferation of the pulmonary endothelium.64,65 Bolte et al66 also developed an endothelium-targeted nanoparticle system for delivery of plasmid DNA and showed that nanoparticle delivery of plasmid DNA expressing Foxf1 and Foxm1 stimulates lung angiogenesis and alveolarization during recovery from neonatal hyperoxic injury. Dahlman et al67 developed an endothelium-targeted polymer nanoparticle system for delivery of siRNA, which showed preferential targeting of the pulmonary vascular endothelium. These findings together highlight the potential for nanoparticle-guided EC-targeted genetic editing for the treatment of thromboinflammatory vascular diseases. Comparisons between endothelium-targeted and untargeted therapies As described above, endothelium-targeted nanotherapies use specialized nanoparticles to deliver drugs directly to diseased vascular endothelium, achieving site-specific drug accumulation.49 In contrast, untargeted treatments including systemically administered drugs lack targeting ability and often fail to localize at the site of disease, resulting in suboptimal efficacy and off-target effects.68 In a TNF-α–induced brain-inflammation model, anti–VCAM-1-targeted lipid nanoparticles achieved ∼7.6% brain uptake vs ∼0.1% in the group treated with untargeted isotype control lipid nanoparticles; the same anti–VCAM-1 lipid nanoparticles carrying thrombomodulin mRNA restored endothelial thrombomodulin and reduced vascular leakage/edema vs untargeted control lipid nanoparticles.14 In the same TNF-α–induced acute brain-inflammation mouse model, anti–VCAM-1-coated liposomes showed ∼27-fold higher brain uptake than nanoparticles coated with antitransferrin receptor, and approximately eightfold higher than nanoparticles coated with anti–ICAM-1.14 In intracerebral hemorrhage, Reyes-Esteves et al69 showed that collagenase-induced intracerebral hemorrhage produces only a short window of vascular leak, and that brain accumulation of untargeted immunoglobulin G (IgG) and untargeted polyethylene glycol nanoparticles is consistently lower than endothelium-targeted agents: anti–platelet endothelial cell adhesion molecule 1-coated nanoparticles achieved approximately eightfold higher levels in the brain than control IgG-coated nanoparticles, and anti–VCAM-1-targeted nanoparticles delivered the antioxidant enzyme, superoxide dismutase, ∼4.5-fold more efficiently than free superoxide dismutase. Another advantage of both untargeted and targeted nanocarriers is their ability to carry complex or fragile payloads (eg, miRNA, small interfering RNA, and peptides) that would otherwise be degraded or cleared if delivered freely.14,29,69 By encapsulating and protecting these therapies, nanoparticles have facilitated the delivery of mitochondrial antioxidant peptides such as SS31 (elamipretide).29,46,69 Nanoparticles can also carry diagnostic imaging agents or multiple drugs within the same platform.70,71 Nanoparticles that bind to ICAM-1 on inflamed endothelium can transport an anti-inflammatory drug and exert an anti-inflammatory effect through binding to ICAM-1 and blocking leukocyte adhesion.9 In the future, endothelial-targeted nanoparticle therapies may complement rather than replace conventional therapies. Table 4 summarizes the strengths and weaknesses of endothelium-targeted vs untargeted nanotherapies.Table 4.Comparisons between untargeted and endothelium-targeted nanoparticle therapiesFeatureUntargeted therapyReferencesEC-targeted therapyReferencesTarget specificityLimited fraction of the therapy reaches diseased endothelium5Optimized fraction of the therapy accumulates at the site of endothelial injury5,60Payload versatilityLimited to stable therapies19Able to deliver multiple, labile therapies19,72EfficacyEffectiveness is limited by low accumulation at the diseased site; therapeutic efficacy might require high doses5,19Higher uptake in target tissues resulting in stronger efficacy30,60Side effectsSignificant risk if high dose is required to achieve efficacy41,73Reduction in systemic exposure can reduce off-target toxicity19,72FunctionalityEach therapy typically has 1 primary pharmacologic role5,73Nanoparticles can deliver multiple drugs, each with different functions19,60PharmacokineticsPharmacokinetics and dosing schedules are well characterized73Nanoparticle biodistribution and clearance may not be well characterized and likely depend on nanoparticle size, surface charge, and material19,41Regulatory statusNumerous agents are FDA-approved5,73Only a small number of vascular nanomedicines have been FDA-approved19,41FDA, US Food and Drug Administration. Table 4.Comparisons between untargeted and endothelium-targeted nanoparticle therapiesFeatureUntargeted therapyReferencesEC-targeted therapyReferencesTarget specificityLimited fraction of the therapy reaches diseased endothelium5Optimized fraction of the therapy accumulates at the site of endothelial injury5,60Payload versatilityLimited to stable therapies19Able to deliver multiple, labile therapies19,72EfficacyEffectiveness is limited by low accumulation at the diseased site; therapeutic efficacy might require high doses5,19Higher uptake in target tissues resulting in stronger efficacy30,60Side effectsSignificant risk if high dose is required to achieve efficacy41,73Reduction in systemic exposure can reduce off-target toxicity19,72FunctionalityEach therapy typically has 1 primary pharmacologic role5,73Nanoparticles can deliver multiple drugs, each with different functions19,60PharmacokineticsPharmacokinetics and dosing schedules are well characterized73Nanoparticle biodistribution and clearance may not be well characterized and likely depend on nanoparticle size, surface charge, and material19,41Regulatory statusNumerous agents are FDA-approved5,73Only a small number of vascular nanomedicines have been FDA-approved19,41FDA, US Food and Drug Administration. Comparisons between untargeted and endothelium-targeted nanoparticle therapies FDA, US Food and Drug Administration. Challenges to clinical translation and potential solutions Biosafety and toxicity EC-targeted nanoparticles will encounter ECs in organs beyond the site of disease.19,41 Some nanomaterials, including certain inorganic particles and cationic polymers, can cause EC injury or inflammation,19,41 which can also depend on nanoparticle features such as size and zeta potential.74 For this reason, biocompatible materials such as biodegradable polymers, lipid-based formulations, and natural cell membrane coatings are preferred.19,41,75 Surface modifications like PEGylation form a hydrophilic stealth corona that decreases nonspecific uptake by endothelial and immune cells.75 Current strategies to optimize nanoparticle composition by eliminating toxic components and adjusting size and charge aim to prevent unintended activation or damage to healthy endothelium. In addition, biomimetic coatings made from platelet and erythrocyte membranes are generally well tolerated.60,72 In vivo delivery efficiency Once administered, nanoparticles encounter a series of physiological barriers. After administration, nanoparticles are opsonized by plasma proteins, which promotes recognition and clearance by the mononuclear phagocyte system (of note hepatic Kupffer cells and splenic macrophages).19,52 Conversely, particles with a hydrodynamic diameter (ie, the nanoparticle size in solution, including the particle core, surface coating, and surrounding hydration layer) below the renal filtration cutoff are prone to rapid elimination through glomerular filtration, reducing systemic residence time and biodistribution. Hemodynamic and physicochemical parameters collectively govern nanoparticle fate and targeting efficiency. High shear environments in arterial circulation can mechanically inhibit ligand-receptor interactions unless the surface ligands possess rapid kinetic on-rates and/or are presented multivalently to increase effective avidity.19,41,75 Therefore, nanoparticle physicochemical properties are optimized to balance clearance and tissue access. Hydrodynamic diameters in the 5 to 8 nm range are cleared renally, whereas larger nanoparticles in the 10 to 250 nm range accumulate in the reticuloendothelial system or in macrophages.76 Neutral to slightly negative surface properties help reduce nonspecific interactions with serum proteins and cells19; and soft, deformable constructs (eg, liposomes, polymeric micelles) exhibit enhanced capillary extravasation and tissue penetration relative to rigid nanoparticles.72,75 Negative surface charges can also activate the intrinsic coagulation pathway.77 Surface modification with hydrophilic polymers (eg, PEGylation) or biomimetic coatings (eg, macrophage membrane coatings) reduced opsonization and hepatic/splenic clearance, thereby extending systemic residence time.72,75 Although PEGylation can decrease nanoparticle aggregation, opsonin adsorption, and mononuclear phagocyte system-mediated clearance, anti-PEG humoral immunity represents an important translational barrier for PEGylated nanomedicines.78,79 Pre-existing or treatment-emergent anti-PEG IgM and IgG antibodies can bind PEGylated nanoparticles and promote the accelerated blood clearance phenomenon through antibody-dependent opsonization, complement activation, and enhanced hepatic and splenic uptake.78,80 This immune recognition may alter pharmacokinetics, reduce target-site exposure, impair pharmacodynamic efficacy, and contribute to complement activation-related pseudoallergy or hypersensitivity.78,81 Therefore, preclinical and clinical development of PEGylated endothelial-targeted nanoparticles should include assessment of anti-PEG antibody titers, complement activation markers, hemocompatibility, biodistribution, and repeat-dose pharmacokinetics.78,80,81 In endothelial-targeting systems, ligands directed against selectin- or ICAM-1–mediated adhesion pathways are commonly used, and multivalent ligand architectures have been shown to enhance adhesion under flow,30,72,75,82 whereas inadequate shear-resistant binding may be resolved by increasing ligand density or valency.30,72,82 Future studies could use ex vivo perfused human organs to test the efficacy of endothelial-targeted or untargeted nanoparticle-guided therapies. For example, an ex vivo perfused human lung system has been used to assess the efficacy of novel treatments in pneumonia- or sepsis-induced inflammatory lung injury.83 These studies could provide a pathway for development of nanoparticle-based therapies from the bench to translational studies and clinical trials. Manufacturing and scalability Nanoparticles with specialized targeting ligands or membrane coatings are complex multicomponent systems, raising challenges for large scale, reproducible manufacturing.41,52 Consistency is crucial: small variations in nanoparticle size, surface ligand density, or purity between batches could affect safety and efficacy.41,52 To achieve scalability, researchers are adopting techniques like microfluidic formulation, which allow precise control over nanoparticle self-assembly and mixing at larger volumes.84 Membrane-coated nanoparticles add an extra step: isolating cell membranes and cloaking the nanoparticle core. This step must be optimized for uniformity, for example, by using extrusion methods to produce similar coated vesicles.36 Characterization tools are equally important: advanced electron microscopy and chromatography methods should be used to ensure each batch meets specifications for size, charge, ligand density, and sterility. Regulatory agencies will require extensive quality control data. Processes such as lyophilization (freeze-drying) are being explored to improve nanoparticle shelf-life and stability for storage and transport.85 Scalability is a potential problem, but progress in lipid nanoparticle manufacturing (as seen with the mRNA vaccine production during COVID-19) provides a template for how nanoparticle medicines can be produced in large scale with high quality. Regulatory and translational hurdles Regulatory bodies have understandably strict requirements: demonstration of safety (including long-term biocompatibility and clearance), consistent manufacturing, and clear evidence of clinical benefit over existing therapies.41 To accelerate translation, some groups are conducting comprehensive preclinical studies not only on efficacy but also on pharmacokinetics to assess how long the nanoparticles circulate and where they accumulate, as well as toxicology tests to assess major organs for nanoparticle-induced damage.29,49,86 Artificial intelligence and computational modeling are also being used to predict nanoparticle behavior and optimize nanoparticle design before entering into expensive human trials.87 Regulatory initiatives including the US Food and Drug Administration guidance on liposome drug products have begun to establish standards for nanomedicine evaluation.88,89 In addition to scientific and regulatory challenges, clinician and patient acceptance could pose an issue. Physicians and patients will need to be convinced of the therapeutic value of complex nanoparticle therapies over existing therapies. Transparent communication and rigorous testing could help advance nanoparticle therapies into the clinic. Interdisciplinary studies that bridge nanotechnology, vascular biology, and clinical research could ultimately narrow the gap between preclinical promise and clinical reality. Conclusions Nanoparticle therapies for thromboinflammatory vascular diseases include untargeted nanoparticle-directed therapies and ligand-targeted, biomimetic, stimuli-responsive, and ligand-free formulations that aim to improve drug delivery to diseased vascular endothelium. Endothelium-targeted nanoparticle therapies aim to direct the treatment to the pathogenic endothelium, which is the nexus of inflammation, coagulation, and barrier dysfunction. The goal of these treatments is to improve therapeutic efficacy and safety compared with untargeted nanoparticles or systemic drugs.90 Preclinical studies demonstrate that EC-targeted nanoparticles localize to injured vessels and restore vascular integrity in sepsis; suppress plaque inflammation and stabilize lesions in atherosclerosis; and reduce injury and inflammation in stroke.29,49,50,82,86,91,92 Moving forward, progress will depend on continued innovation in ligand-dependent, stimuli-responsive, and biomimetic nanoparticle technologies, as well as interdisciplinary collaborations between biologists, chemists, engineers, imaging experts, and clinicians. If scientific and translational hurdles are overcome, EC-targeted nanotherapies could become a valuable tool for the treatment of thromboinflammatory vascular diseases including stroke, thrombosis, and atherosclerosis by using the diseased endothelium as an address for drug delivery. Conflict-of-interest disclosure: The authors declare no competing financial interests. The current affiliation for H.J. is School of Pharmacology, Guangdong Medical University, Dongguan, China.