Endothelial Cells at the Crossroad of Atherosclerosis and Metabolic Dysfunction–Associated Steatotic Liver Disease

Highlights

• •MASLD and atherosclerosis frequently coexist, increasing disease burden.
• •Endothelial dysfunction links hepatic and vascular disease.
• •Endothelial pathways may enable dual-disease therapies.
• •Biomarkers may guide endothelial-targeted therapies.

Abstract

Atherosclerosis and metabolic dysfunction–associated steatotic liver disease (MASLD) are 2 highly prevalent conditions that frequently coexist and amplify each other's clinical burden. MASLD not only increases cardiovascular disease incidence but also promotes atherosclerosis independently of traditional risk factors, including obesity, diabetes, and dyslipidemia. Despite this strong clinical association, the biological mechanisms linking these diseases remain incompletely understood. Emerging evidence identifies endothelial dysfunction as a unifying pathogenic axis underlying their crosstalk. In atherosclerosis, vascular endothelial dysfunction facilitates lipoprotein retention, leukocyte recruitment, and plaque formation. Similarly, in MASLD, liver sinusoidal endothelial cells lose their capacity to regulate lipid trafficking, immune tolerance, and hepatic homeostasis, thereby promoting steatosis, inflammation, and fibrosis. This review highlights endothelial dysfunction as a common denominator in both diseases, emphasizing shared molecular mechanisms such as impaired nitric oxide production, oxidative stress, and cytokine-driven inflammation. We further discuss how metabolic disturbances including insulin resistance and dyslipidemia exacerbate endothelial activation, lipid accumulation, and immune cell responses across vascular and hepatic beds. By examining the interplay between MASLD and atherosclerosis at the endothelial level, we propose endothelial cells as a promising therapeutic target for both conditions. Emerging strategies, such as targeted cell therapies and nanomedicine, offer the potential to simultaneously address endothelial dysfunction in the liver and vasculature. Future research should prioritize therapies that address this shared endothelial axis, with the potential to reduce the morbidity and mortality associated with these interconnected diseases.

Central Illustration

Cardiometabolic diseases, including cardiovascular disease (CVD), metabolic dysfunction–associated steatotic liver disease (MASLD), chronic kidney disease, obesity, diabetes, and stroke, constitute a major global health burden.1, 2, 3 These conditions are mechanistically linked through metabolic dysregulation, oxidative stress, and chronic low-grade inflammation.^4,5 Atherosclerosis, the primary driver of CVD, remains the leading cause of death worldwide and is driven by dyslipidemia and persistent arterial inflammation, resulting in plaque formation, luminal narrowing, and the risk of rupture leading to myocardial infarction or stroke.6, 7, 8, 9, 10 MASLD, formerly known as nonalcoholic fatty liver disease, is the most prevalent chronic liver disease globally, affecting approximately 38.2% of adults.11, 12, 13 Although primarily a hepatic disorder, MASLD has profound systemic consequences, particularly on cardiovascular health,14, 15, 16, 17, 18, 19 with a 2022 meta-analysis of over 11 million individuals reporting a 1.5-fold increased risk of heart failure.²⁰ Accordingly, MASLD is now recognized as an independent risk factor for atherosclerosis development and progression, even in the absence of traditional cardiovascular risk factors.14, 15, 16, 17, 18, 19^,21, 22, 23 MASLD encompasses a disease spectrum ranging from simple steatosis to metabolic dysfunction–associated steatohepatitis (MASH), characterized by hepatocellular injury, inflammation, and fibrosis.24, 25, 26 Patients with MASH exhibit markedly elevated cardiovascular risk, with an approximately 2.6-fold higher incidence of CVD,²⁷ and MASLD-related hepatocellular carcinoma has become the fastest-growing cause of liver cancer worldwide.²⁶ The rising prevalence of MASLD parallels increasing rates of obesity and type 2 diabetes,²⁸ highlighting the importance of early detection and targeted interventions addressing metabolic dysfunction and inflammation.^29,30 Importantly, the prevalence, progression, and cardiometabolic consequences of MASLD and atherosclerosis may differ according to sex, influenced by hormonal status, body fat distribution, and immune-metabolic regulation.^24,31, 32, 33, 34, 35

Atherosclerosis and MASLD share common pathological mechanisms centered on endothelial dysfunction, chronic inflammation, dysregulated lipid metabolism, and fibrosis.^21,36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46 In atherosclerosis, vascular endothelial dysfunction promotes lipid accumulation, inflammatory cell recruitment, and fibroproliferative remodeling, driving plaque formation within the arterial wall.47, 48, 49 Similarly, in MASLD, liver sinusoidal endothelial cells (LSECs) undergo capillarization and functional impairment, disrupting lipid handling and promoting hepatic steatosis, inflammation, and fibrosis.^50,51 Both diseases are characterized by reduced nitric oxide (NO) bioavailability, oxidative stress, and cytokine-driven inflammatory signaling, culminating in progressive tissue injury and fibrotic remodeling52, 53, 54 (Figure 1). As regulators of vascular tone, permeability, leukocyte adhesion, and hemostasis, endothelial cells serve as central gatekeepers of vascular homeostasis.^55,56 Under metabolic stress, they shift from a protective to a dysfunctional phenotype that amplifies inflammation and tissue damage.⁵⁶

Figure 1.

Shared Progression of Atherosclerotic CVD and MASLD

Both diseases are initiated by dysregulated lipid metabolism, leading to lipid accumulation within the arterial intima or hepatocytes. This triggers inflammatory responses that promote disease progression. In atherosclerosis, immune cells adhere to the endothelium, infiltrate the vessel wall, and contribute to plaque formation. In MASLD, hepatic steatosis can progress to metabolic dysfunction–associated steatohepatitis (MASH), characterized by immune cell infiltration and hepatic inflammation. Chronic inflammation activates collagen-producing cells, including vascular smooth muscle cells in atherosclerotic plaques and hepatic stellate cells in the liver, driving fibrotic remodeling. In advanced disease, vulnerable plaques may rupture and cause thrombosis or acute cardiovascular events, whereas progressive hepatic fibrosis may lead to cirrhosis and liver failure. CVD = cardiovascular disease; MASLD = metabolic dysfunction–associated steatotic liver disease.

Beyond shared metabolic risk factors, increasing evidence supports a broader interorgan communication network linking the gut, liver, and cardiovascular system. The liver functions not only as a metabolic organ but also as a major source, clearance site, and bioactivation hub for circulating mediators that influence vascular biology, including lipoproteins, cytokines, coagulation factors, vasoactive molecules, and oxidative stress-related metabolites.57, 58, 59, 60, 61, 62, 63, 64, 65, 66 During MASLD progression, disruption of these pathways can reshape systemic endothelial signaling and promote vascular inflammation.^15,67, 68, 69, 70, 71, 72 In parallel, gut-derived endotoxemia, microbial metabolites, and bile acid signaling further modulate endothelial phenotype across hepatic and vascular beds.73, 74, 75, 76, 77, 78 Importantly, this axis is bidirectional: Hepatic inflammation can aggravate systemic vascular dysfunction, while vascular endothelial dysfunction and impaired perfusion may further exacerbate liver injury.^66,79, 80, 81, 82, 83 These observations reinforce endothelial cells as a central mechanistic crossroad connecting liver and cardiovascular disease.

Although endothelial dysfunction has been extensively studied in atherosclerosis and MASLD independently, fewer studies have addressed the shared endothelial mechanisms linking these conditions or explored endothelial cells as a dual therapeutic target. This review focuses on endothelial dysfunction as a unifying pathological axis between atherosclerosis and MASLD and discusses the potential of endothelial-targeted strategies for dual disease treatment (Central Illustration). To understand how endothelial dysfunction links these diseases, it is first necessary to consider the structural and functional heterogeneity of endothelial cells across vascular beds.

Central Illustration.

Endothelial Dysfunction: A Shared Mechanism and Therapeutic Target in Atherosclerosis and MASLD

Systemic cardiometabolic stressors, including dyslipidemia, metabolic imbalance, inflammation, and gut-derived signals, promote endothelial activation characterized by reduced nitric oxide (NO) bioavailability, increased reactive oxygen species (ROS), and upregulation of adhesion molecules. In the vascular wall, these changes facilitate low-density lipoprotein (LDL) infiltration, oxidation into oxidized LDL (oxLDL), and recruitment of inflammatory cells such as macrophages (Mφ), contributing to plaque formation and progression. In the liver, similar endothelial alterations lead to liver sinusoidal endothelial cell (LSEC) dysfunction, promoting activation of HSC, inflammation, and fibrogenesis. Collectively, these shared endothelial mechanisms identify endothelial cells as a central mechanistic crossroad linking vascular and hepatic disease and a promising therapeutic target for the dual treatment of atherosclerosis and MASLD.

General Overview of Endothelial Cells

The vascular endothelium constitutes the largest autocrine, paracrine, and endocrine organ in the human body, forming a dynamic interface between circulating blood and surrounding tissues.^84,85 Once viewed as a passive barrier, endothelial cells are now recognized as active regulators of vascular tone, inflammation, immune surveillance, and thrombosis.⁸⁶ Endothelial cells also function as immune sentinels by sensing inflammatory stimuli, presenting signals to circulating leukocytes, and coordinating tissue-specific immune cell trafficking. Importantly, endothelial cells exhibit marked structural and functional heterogeneity depending on their vascular bed and organ context, enabling tissue-specific regulation of vascular homeostasis.^87,88 Early electron microscopy studies in the 1960s identified 3 major capillary phenotypes: continuous, fenestrated, and sinusoidal endothelium.89, 90, 91, 92 Continuous endothelium, present in arteries, veins, and organs such as the brain, lungs, and heart, forms a tightly regulated barrier supported by an intact basement membrane, limiting macromolecular passage.⁹⁰ Fenestrated endothelium, found in tissues with high exchange demands such as the kidney and intestinal mucosa, contains membrane-covered pores that facilitate rapid transport of water and small solutes.⁹³ In contrast, sinusoidal endothelium, located in the liver, spleen, and bone marrow, is characterized by wide intercellular gaps and a discontinuous or absent basement membrane, permitting the passage of large molecules.⁹⁰ LSECs represent a highly specialized subset uniquely adapted to the hepatic microenvironment.⁹⁴ Unlike classical vascular endothelial cells, LSECs lack a basement membrane and exhibit fenestrations organized into sieve plates, enabling efficient exchange of lipoproteins, hormones, and metabolites between blood and hepatocytes⁹⁴ (Figure 2). In addition, LSECs possess scavenger and antigen-presenting properties that contribute to hepatic immune tolerance and inflammation regulation.⁹⁵ Through coordinated interactions with hepatocytes, Kupffer cells, and hepatic stellate cells, LSECs play a central role in maintaining liver homeostasis but can actively drive pathology when dysfunctional.^94,95 In addition to structural heterogeneity, endothelial cells display distinct metabolic programs adapted to their local environment.⁹⁶ Although endothelial cells rely predominantly on glycolysis for ATP production, mitochondrial metabolism, fatty acid handling, and redox balance also regulate angiogenesis, barrier integrity, and inflammatory activation.96, 97, 98, 99, 100, 101, 102, 103, 104 These metabolic features may be reprogrammed during disease. In atherosclerosis, arterial endothelial cells exposed to disturbed flow and lipid excess develop oxidative and inflammatory metabolic stress, whereas in MASLD, LSECs are influenced by the unique nutrient-rich hepatic milieu, lipotoxicity, and altered sinusoidal oxygen gradients. Thus, endothelial metabolism represents another shared yet context-dependent layer linking vascular and liver disease.^37,105, 106, 107 This endothelial diversity is particularly relevant when comparing arterial endothelial cells and LSECs, as their distinct baseline phenotypes shape responses to shared cardiometabolic stressors.

Figure 2.

Distinct Endothelial Structure and Lipid Handling in the Liver and Coronary Arteries

(A) Under physiological conditions, liver sinusoidal endothelial cells (LSECs) contain fenestrations that permit chylomicron remnants to enter the space of Disse, where hepatocytes process lipids for very–low density lipoprotein (VLDL) synthesis. Newly formed VLDL is then released back into the circulation. Endothelial nitric oxide synthase (eNOS)–derived nitric oxide (NO) produced by LSECs contributes to the regulation of intrahepatic vascular tone. (B) Coronary arteries are lined by continuous endothelial cells lacking fenestrations. In this setting, low-density lipoprotein (LDL) crosses the endothelial barrier primarily through vesicular transcytosis and accumulates in the subendothelial space. Retained LDL undergoes oxidative modification to form oxidized LDL (oxLDL), which promotes endothelial activation, immune cell recruitment, and early atherogenesis. Cardiac endothelial eNOS-derived NO helps maintain vascular tone, redox balance, and tissue perfusion. EC = endothelial cell; SMC = smooth muscle cell; TG = triglyceride.

Link Between Atherosclerosis and MASLD: The Role of Endothelial Cells

When atherosclerosis and MASLD coexist, shared systemic stressors such as dyslipidemia, chronic inflammation, oxidative stress, and altered hepatic-derived mediators may simultaneously impair endothelial function in both arterial and hepatic vascular beds.^37,66,67,72,108, 109, 110, 111

Early endothelial activation

The concept of endothelial activation emerged in the 1960s after observations that endothelial cells undergo dynamic morphological and functional changes during inflammatory responses.¹¹² This paradigm shift established endothelial cells as active participants in vascular inflammation rather than passive barriers.^113,114 This concept was later formalized by Pober in the 1980s to describe endothelial responses to proinflammatory cytokines such as tumor necrosis factor-α and interleukin-1β (IL-1β), characterized by the upregulation of leukocyte adhesion molecules including vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1), and E-selectin.¹¹⁵ Closely related, the concept of endothelial dysfunction emerged following the discovery of endothelium-derived relaxing factor by Furchgott and Zawadzki,¹¹⁶ later identified as NO, which established endothelial cells as central regulators of vascular tone and inflammation. Endothelial NO, produced by endothelial nitric oxide synthase (eNOS), maintains vascular homeostasis by promoting vasodilation and suppressing leukocyte adhesion, platelet activation, and inflammatory signaling.117, 118, 119, 120 Endothelial dysfunction arises when NO bioavailability is reduced and endothelial cells adopt a proinflammatory, prothrombotic phenotype. This state is a hallmark of cardiometabolic diseases, including atherosclerosis, hypertension, diabetes, obesity, and MASLD, and impaired flow-mediated dilation (FMD) is a strong predictor of cardiovascular risk.120, 121, 122, 123, 124 Reduced NO bioavailability is further exacerbated by oxidative stress, as reactive oxygen species (ROS) generated by mitochondrial dysfunction or nicotinamide adenine dinucleotide phosphate (NADPH) oxidase activity scavenge NO, amplifying endothelial injury and inflammatory signaling.^123,124 These endothelial responses may also be influenced by sex hormones, particularly estrogen, which can enhance eNOS activity and exert anti-inflammatory vascular effects. In atherosclerosis, endothelial activation and dysfunction represent early and sustained drivers of disease initiation and progression, promoting leukocyte recruitment, lipid accumulation, and chronic arterial inflammation.¹²⁵ Analogously, endothelial dysfunction is an early initiating event in MASLD, where structural and functional alterations in LSECs are among the earliest detectable pathological changes.¹²⁶ Under physiological conditions, LSECs express low levels of classical vascular adhesion molecules and maintain an immunotolerant hepatic environment. With chronic liver injury, LSECs upregulate VCAM-1 and ICAM-1, facilitating proinflammatory monocyte recruitment.^72,127 Beyond passive recruitment, endothelial cells actively shape immune responses through cytokine and chemokine secretion, antigen presentation, and regulation of leukocyte adhesion and transmigration. In turn, activated immune cells release inflammatory mediators and reactive species that further amplify endothelial dysfunction, creating a self-reinforcing inflammatory loop across vascular and hepatic tissues.^70,72,110,128, 129, 130, 131, 132, 133, 134, 135, 136, 137 Genetic or pharmacologic blockade of VCAM-1 reduces monocyte infiltration and fibrogenesis in experimental models, underscoring a functional adhesion role closely paralleling that of arterial endothelium in atherosclerosis.⁷² Recent evidence further demonstrates that lipotoxic stress induces epigenetic activation of ICAM-1 in LSECs, promoting myeloid cell recruitment and fibroinflammatory progression in MASH.¹¹⁰ Importantly, ICAM-1 neutralization or endothelial-specific epigenetic suppression attenuated hepatic inflammation and fibrosis in vivo, highlighting endothelial adhesion pathways as actionable therapeutic targets.¹¹⁰ Together, these observations identify early endothelial activation as a shared pathogenic mechanism linking vascular and metabolic liver disease and redefining endothelial control of lipid transport. Beyond inflammatory activation, endothelial dysfunction also directly alters lipid trafficking, a central event in both atherogenesis and hepatic steatosis.

Lipid transport

A crucial early event in atherogenesis is the entry of low-density lipoprotein (LDL) into the subendothelial space.138, 139, 140 Contrary to early assumptions, LDL does not enter the intima through intercellular gaps; instead, it crosses the endothelium predominantly via vesicular transcytosis.¹⁴¹ In arterial endothelial cells, this process is largely mediated by caveolae, flask-shaped invaginations enriched in caveolin-1 (Cav1), cholesterol, and sphingolipids.¹⁴¹ Caveolae internalize native LDL at the luminal surface, transport it across the endothelial cytoplasm, and release it basolaterally into the intima.¹⁴² Caveolin-1 knockout (Cav1^−/−) mice exhibit dramatic protection from atherosclerosis,¹⁴² and Cav1 overexpression accelerates atherosclerosis.¹⁴³ Beyond Cav1, additional regulators of endothelial LDL transport have been identified. Endothelial SCARB1/SR-BI promotes LDL delivery to the subendothelial space.¹⁴⁴ Genetic or pharmacologic disruption of SR-BI or its downstream effector guanine nucleotide exchange factor dedicator of cytokinesis 4 (DOCK4) markedly reduces LDL transcytosis and attenuates atherosclerotic lipid deposition.¹⁴⁵ Likewise, endothelial myosin-9 (MYH9) localizes to lipid rafts and supports efficient LDL transcytosis by promoting vesicle trafficking and basal exocytosis.¹⁴⁶ Loss or inhibition of endothelial MYH9 impairs LDL passage into the intima and diminishes early plaque formation in mouse models¹⁴⁶ (Figure 3). High-density lipoprotein (HDL) also plays an important counter-regulatory role in vascular lipid homeostasis. Beyond its function in reverse cholesterol transport, HDL exerts antioxidant, anti-inflammatory, and endothelial-protective effects by promoting NO bioavailability, limiting LDL oxidation, and reducing adhesion molecule expression.147, 148, 149, 150 Conversely, impaired HDL quantity or function, frequently observed in MASLD and metabolic syndrome, may further contribute to endothelial dysfunction and atherogenic risk.^151,152

Figure 3.

Shared Endothelial Mechanisms Linking Atherosclerosis and MASLD

Obesogenic diet and physical inactivity promote metabolic and vascular alterations that connect steatotic liver disease with atherosclerotic plaque formation. In the liver, excess fatty acids enhance triglyceride (TG) and cholesterol synthesis, leading to increased very–low density lipoprotein (VLDL) production and hyperlipidemia. Liver sinusoidal endothelial cells (LSECs) undergo capillarization and reduced endothelial nitric oxide synthase (eNOS) activity, resulting in lower nitric oxide (NO) bioavailability and impaired sinusoidal function. In the vasculature, elevated VLDL and low-density lipoprotein (LDL) promote lipid deposition within the arterial wall. Oxidized LDL (oxLDL) induces endothelial activation and the expression of adhesion molecules, including ICAM-1 and VCAM-1, which facilitate monocyte recruitment. Monocyte-derived macrophages internalize oxLDL and amplify inflammation through cytokine release. Smooth muscle cells (SMCs) contribute to fibrous cap formation through extracellular matrix production. Together, these processes illustrate how endothelial dysfunction links MASLD with atherosclerotic plaque development. FFA = free fatty acids; MASLD = metabolic dysfunction–associated steatotic liver disease.

In the liver, capillarization of LSECs is a hallmark of early-stage MASLD and has been widely observed in both animal models and human patients.153, 154, 155 Reduced LSECs permeability prevents the efficient passage of chylomicron remnants from the sinusoidal circulation into the space of Disse, where hepatocytes internalize them to support very–low density lipoprotein (VLDL) synthesis.¹⁵⁶ This impaired transport mechanism may lead to the retention of triglycerides and cholesterol within hepatocytes, which combined with enhanced de novo lipid synthesis, worsens hepatic steatosis.^156,157 Additionally, Fraser et al¹⁵⁸ proposed that LSECs capillarization hinders the inhibitory effect of dietary cholesterol on hepatocyte HMG-CoA reductase, the rate-limiting enzyme in cholesterol biosynthesis, further promoting lipid accumulation. This suggests that LSEC capillarization may play a direct role in the progression of MASLD and metabolic syndrome. Genetic models further underscore the importance of fenestration. For instance, plasmalemma vesicle-associated protein (PLVAP) knockout mice, which cannot form fenestrae, spontaneously develop hepatic steatosis and hyperlipoproteinemia under normal conditions, highlighting the notion that impaired chylomicron remnant passage may hinder VLDL synthesis and promote triglyceride accumulation in hepatocytes¹⁵⁶ (Figure 3). Yet, how circulating VLDL levels remain elevated¹⁵⁹ despite LSEC defenestration remains incompletely understood. One possibility is that alternative lipoprotein export routes or compensatory overproduction mechanisms by hepatocytes counterbalance reduced exchange across the sinusoidal endothelium. This paradox underscores the complexity of hepatic-vascular lipid trafficking in MASLD and suggests that capillarization may differentially affect intrahepatic lipid retention and systemic lipoprotein release, contributing simultaneously to steatosis and atherogenic dyslipidemia. Thus, altered hepatic lipoprotein handling may represent a key mechanism through which liver dysfunction directly contributes to systemic atherogenic risk.¹⁶⁰ Animal studies confirm that capillarization can occur as early as 1 week following a choline-deficient l-amino acid–defined diet in mice¹⁶¹ or 3 weeks after high-fat feeding in rats.¹⁶² Conversely, fasting for 48 hours can restore fenestrae diameter.¹⁶³ Recent findings indicate that aging contribute to the onset of LSEC capillarization. During aging, deficiency in endothelial C-kit disrupts the functional balance of LSECs and is associated with elevated expression of profibrotic (TIMP-1, collagen type I, PDGF-β) and proinflammatory markers (IL-6, and IL-1β).¹⁶⁴ The gut-liver axis also contributes to LSEC remodeling. Rats administered a single injection of endotoxin exhibited reductions in both the number and diameter of LSEC fenestrae,¹⁶⁵ and LSEC fenestrae were inversely associated with the abundance of Bacteroidetes and positively correlated with Firmicutes.¹⁶⁶ Additionally, Desroches-Castan et al¹⁶⁷ demonstrated that bone morphogenetic protein 9 (BMP9) plays a critical paracrine role in maintaining LSEC differentiation. In BMP9-knockout mice, LSECs exhibited a marked reduction in fenestrae, while treatment with BMP9 in culture restored fenestration and upregulated key differentiation markers such as GATA4 and PLVAP, underscoring its importance in preserving LSEC phenotype.¹⁶⁷ Together, these findings highlight endothelial cells as active regulators of lipid flux, in which enhanced lipoprotein transcytosis in arteries and impaired fenestration in the liver jointly promote subendothelial lipid retention, hepatic steatosis, and atherogenic dyslipidemia across cardiometabolic disease. Once retained or dysregulated, lipoproteins become substrates for oxidative modification, further amplifying endothelial injury.^24,31, 32, 33, 34, 35

LDL oxidation

Atherogenesis is initiated when accumulating LDL becomes oxidized (oxLDL) within the intima.168, 169, 170 Recognized by endothelial cells as a danger signal, oxLDL triggers the expression of adhesion molecules, including VCAM-1, ICAM-1, and E-selectin, which facilitate the attachment of circulating monocytes to the endothelial surface.^168,169 These monocytes subsequently transmigrate into the intima, differentiate into macrophages, and internalize oxLDL via scavenger receptors, leading to foam cell formation and fatty streak development.^171,172

In the context of MASLD, systemic dyslipidemia and hepatic oxidative stress generate conditions that favor LDL oxidation, suggesting that oxLDL may similarly contribute to endothelial dysfunction and inflammatory signaling within the liver.¹⁷³ Consistent with this concept, exposure to oxLDL reduces fenestrae diameter and porosity in human primary liver sinusoidal endothelial cells, indicating that oxidized lipoproteins can directly impair LSEC structure and function.¹⁷⁴ Together, these observations position oxLDL as a key molecular link between dysregulated lipid metabolism, endothelial activation, and inflammatory amplification across vascular and metabolic liver disease.

Flow patterns

Blood flow exerts a variety of hemodynamic forces on the vascular endothelium, including normal (transmural) stress, circumferential stress, and tangential shear stress. While transmural pressure and circumferential stress primarily affect and regulate medial vascular smooth muscle cells,^86,175 fluid shear stress has a profound impact on endothelial cells, potently regulating their function.¹⁷⁶ In atherosclerosis, plaques preferentially develop at arterial regions exposed to disturbed flow, such as bifurcations, branch points, and inner curvatures, whereas regions subjected to steady laminar shear stress, typically straight arterial segments, are protected.¹⁷⁷ Under laminar flow, endothelial cells adopt a quiescent, atheroprotective phenotype characterized by low turnover, enhanced NO production, reduced ROS, and anti-inflammatory gene expression.178, 179, 180, 181, 182 These effects are mediated in part through induction of transcription factors such as Krüppel-like factor 2 (KLF2) and KLF4, which inhibit nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) signaling and adhesion molecule expression.¹⁸³ In contrast, disturbed flow promotes an atheroprone endothelial state marked by reduced NO bioavailability, increased oxidative stress, heightened endothelial permeability, and upregulation of proinflammatory and prothrombotic genes such as ICAM-1 and VCAM-1.^178,179 These regions exhibit elevated basal NF-κB activity and exaggerated inflammatory responses to systemic stimuli such as hypercholesterolemia or lipopolysaccharide exposure.^184,185 Targeted inhibition of NF-κB signaling, either pharmacologically, or genetically, via endothelial-specific expression of dominant-negative IκBα, markedly reduces vascular inflammation, leukocyte adhesion, and plaque burden, reinforcing the critical role of endothelial NF-κB activation in atherogenesis.¹⁸⁶

Importantly, the hemodynamic context in MASLD is fundamentally distinct from the disturbed shear stress paradigm observed in arteries. Rather than oscillatory or low shear stress, MASLD is characterized by alterations in hepatic microcirculation and sinusoidal perfusion, which impair oxygen and nutrient delivery and disrupt LSECs function.187, 188, 189 Reduced hepatic microcirculatory blood flow has been consistently observed in both experimental models and patients with MASLD and correlates inversely with disease severity.^189,190 In high-fat diet–fed Wistar rats, hepatic microcirculatory blood flow is reduced by approximately 30%, a defect that can be partially reversed by dietary interventions and further improved by pyridoxamine supplementation.¹⁹¹ These findings indicate that, unlike atherosclerosis, MASLD-associated endothelial dysfunction arises primarily from impaired sinusoidal perfusion and microvascular resistance rather than disturbed shear stress, yet converges on similar endothelial outcomes, including reduced NO signaling, inflammatory activation, and loss of endothelial homeostasis.

Oxidative stress

Oxidative stress is a central mechanism linking atherosclerosis and MASLD, and its pathological impact is largely mediated through endothelial dysfunction. Whereas physiological levels of ROS regulate endothelial gene expression, proliferation, and signaling, excessive or chronic ROS production drives distinct pathological endothelial phenotypes, including impaired NO signaling, cellular senescence, endothelial-to-mesenchymal transition (EndMT), and glycocalyx disruption.^192,193 These phenotypic changes collectively contribute to loss of endothelial homeostasis and vascular integrity.^192,193 In atherosclerosis, disturbed flow increases multiple ROS species, including superoxide (O₂•⁻), hydrogen peroxide (H₂O₂), and peroxynitrite (ONOO⁻), which lead to NO scavenging, eNOS uncoupling, and oxidative modification of LDL.^194,195 These events trigger endothelial activation characterized by increased expression of adhesion molecules, enhanced leukocyte recruitment, and proinflammatory signaling. Sustained oxidative stress further promotes endothelial senescence, EndMT, and microvascular rarefaction, contributing to plaque progression and instability.196, 197, 198, 199, 200, 201 Multiple enzymatic sources contribute to ROS production under disturbed flow, including uncoupled eNOS, mitochondrial electron transport enzymes, cyclooxygenases, xanthine oxidase, and NADPH oxidases (Nox), the latter representing the major endothelial ROS source.^202,203 Consistent with this, oxidative stress mediates the association between metabolic syndrome markers, such as the triglyceride-glucose index, and cardiovascular disease, highlighting its role in shaping endothelial dysfunction.²⁰⁴

In MASLD, lipid overload and mitochondrial dysfunction drive excessive production of ROS and reactive nitrogen species, leading to hepatocellular injury and inflammation.^50,205 Importantly, these oxidative mechanisms also directly affect LSECs, where ROS impairs NO signaling, promotes capillarization, and increases endothelial permeability.^206,207 These phenotypic alterations impair lipid exchange between blood and hepatocytes and contribute to steatosis progression.²⁰⁸ In addition, oxidative stress promotes LSEC dysfunction–driven activation of hepatic stellate cells, thereby linking endothelial injury to fibrogenesis.^71,209 Similar to the vascular context, oxidative stress in the liver may also induce endothelial senescence and phenotypic reprogramming, reinforcing a proinflammatory and profibrotic microenvironment.^37,137 Liver-derived oxidative and inflammatory mediators may further amplify endothelial dysfunction in distant vascular beds.³⁸ Together, these findings highlight oxidative stress as a key driver of endothelial phenotypic transitions across vascular beds, linking atherosclerosis and MASLD through shared mechanisms of endothelial injury and dysfunction.^50,205 Persistent oxidative stress also interacts closely with intracellular stress pathways, particularly endoplasmic reticulum stress.

Endoplasmic reticulum stress

Oxidative stress and endoplasmic reticulum (ER) stress are tightly interconnected in endothelial cells, as sustained ROS production disrupts ER redox homeostasis, leading to protein misfolding and activation of the unfolded protein response.^210,211 In atherosclerosis, endothelial ER stress has been linked to the upregulation of adhesion molecules such as VCAM-1²¹² and the induction of prothrombotic factors,²¹³ thereby coupling cellular stress responses to plaque progression. However, ER stress in endothelial cells can also arise through ROS-independent mechanisms driven by extracellular matrix (ECM) remodeling and integrin signaling. Disturbed flow rapidly alters the endothelial ECM, shifting it from a laminin- and collagen IV–rich basement membrane toward a provisional, fibronectin-enriched matrix, one of the earliest structural changes in atheroprone regions.214, 215, 216, 217 This remodeled ECM modifies endothelial responsiveness to both shear stress and oxLDL by engaging specific integrin heterodimers that mediate context-dependent signaling.^215,218, 219, 220, 221, 222 Integrin signaling exhibits heterodimer-specific functions. αvβ3 primarily mediates shear-induced endothelial activation, whereas α5β1 contributes to both shear- and oxLDL-induced inflammatory signaling via NF-κB.^218,223 Mechanistically, fibronectin-α5β1 engagement suppresses the anti-inflammatory cAMP/PKA pathway through recruitment of the phosphodiesterase PDE4D5, lowering intracellular cAMP and relieving PKA-dependent inhibition of NF-κB.²²⁴ Notably, α5β1 signaling also promotes fibronectin deposition, creating a feed-forward loop that amplifies downstream αvβ3-mediated activation.²¹⁹ Consistent with a division of labor between integrins, genetic deletion of α5β1 or αvβ3 reduces endothelial ER stress, with α5β1 mediating ER stress in response to oxLDL and αvβ3 mediating ER stress induced by disturbed flow.²²⁵ This integrin-dependent, mechanosensitive ER stress response operates independently of classical ROS or protein-misfolding triggers and represents a novel axis linking matrix remodeling, endothelial activation, and metabolic stress across vascular beds.²²⁵ Although this pathway is well characterized in arterial endothelium, its contribution to LSECs dysfunction remains incompletely defined. Whether ER stress initiates LSEC capillarization and lipid dysregulation or instead arises secondary to these alterations remains an open question, suggesting the presence of reinforcing stress-injury loops. Sustained activation of fibronectin-integrin–dependent inflammatory and stress pathways not only perpetuate endothelial dysfunction but also primes endothelial cells for phenotypic reprogramming toward a mesenchymal-like state, underscoring the therapeutic relevance of targeting ECM-integrin-ER stress signaling in cardiometabolic disease.^219,225,226

Endothelial-to-mesenchymal transition

EndMT is a phenotypic reprogramming process in which endothelial cells progressively lose endothelial identity while acquiring mesenchymal and fibroblast-like characteristics.²²⁷ This transition is marked by downregulation of endothelial junctional proteins, such as VE-cadherin, alongside induction of mesenchymal markers including fibronectin, α-smooth muscle actin (α-SMA), N-cadherin, vimentin, and transcriptional regulators such as Snail, Slug, and Twist1.²²⁸ Through these changes, EndMT contributes to fibrotic remodeling, intimal thickening, and plaque destabilization.²²⁸ In atherosclerosis, lineage-tracing approaches and single-cell transcriptomic analyses have demonstrated that EndMT-derived cells constitute a substantial fraction of fibroblast-like cells within advanced human plaques, particularly unstable lesions, where they associate with increased matrix metalloproteinase expression and collagen degradation, linking EndMT to plaque vulnerability.^229,230 Hemodynamic and inflammatory cues are key drivers of EndMT in the arterial wall. Disturbed flow promotes EndMT through activation of transforming growth factor (TGF)-β signaling pathways, a process facilitated by suppression of the flow-sensitive transcription factor KLF4 and reduced expression of extracellular matrix stabilizers such as tenascin-X.²³¹ Disturbed flow also induces endothelial expression of collagen VIII (COL8A1), which appears to function as a compensatory matrix component that helps preserve endothelial identity and limit excessive EndMT and NF-κB activation.²³² In parallel, proinflammatory cytokines, including interleukin-1β and tumor necrosis factor-α, synergize with TGF-β/Smad signaling to promote inflammatory EndMT programs within atheroprone regions, reinforcing endothelial dysfunction and maladaptive vascular remodeling.²³³

In the liver, chronic injury similarly induces phenotypic plasticity in LSECs, giving rise to mesenchymal-like cells that contribute to fibrotic progression. Rather than maintaining their specialized fenestrated and antifibrotic phenotype, subsets of LSECs exposed to sustained inflammatory, metabolic, and oxidative stress signals downregulate canonical endothelial markers, such as vWF and VE-cadherin, and adopt transcriptional and functional features characteristic of myofibroblasts, such as α-SMA, vimentin, and collagens type I and III.234, 235, 236 Single-cell RNA sequencing studies of fibrotic liver tissue reveal that LSECs progressively shift toward mesenchymal and ECM-producing states, with transitional populations retaining residual sinusoidal markers while acquiring fibrogenic gene expression profiles.²³⁶ Importantly, emerging evidence identifies microRNA-dependent regulation as a critical upstream modulator of LSEC plasticity.²³⁷ For example, miR-27b-3p has been shown to reverse mesenchymal features in cirrhotic LSECs and restore key endothelial characteristics, highlighting EndMT as a dynamic and potentially reversible process.²³⁷ Collectively, these findings position EndMT as a shared maladaptive response to chronic vascular and metabolic stress, contributing to fibrosis and disease progression in both atherosclerosis and MASLD, while also identifying endothelial plasticity as a promising therapeutic target.^111,226 These phenotypic transitions have major structural consequences, culminating in tissue remodeling and fibrosis.²¹

Fibrosis

Fibrosis is a defining structural feature of both atherosclerosis and MASLD, yet it serves fundamentally different roles across these conditions.²¹ In atherosclerosis, fibrotic remodeling within the arterial wall is largely protective. Vascular smooth muscle cell–derived ECM deposition gives rise to the fibrous cap, a collagen-rich structure that stabilizes plaques and limits rupture and thrombotic events.²³⁸ Endothelial cells indirectly influence this process by regulating the vascular microenvironment that governs vascular smooth muscle cell phenotypic modulation, matrix organization, and plaque architecture. Consequently, plaque vulnerability is primarily associated with fibrous cap thinning or degradation rather than excessive fibrosis per se.^229,239,240

In contrast, fibrosis is the strongest predictor of liver-related morbidity and mortality in MASLD and is primarily driven by activation of hepatic stellate cells (HSCs), leading to excessive ECM deposition and progressive liver dysfunction.²⁴¹ LSECs play a critical upstream role in modulating this fibrogenic response. Under physiological conditions, LSECs maintain HSC quiescence and preserve a nonfibrogenic ECM environment.^242,243 During MASLD progression, structural and phenotypic alterations in LSECs, most notably capillarization, disrupt endothelial-stromal crosstalk and promote HSC activation. Dysfunctional LSECs facilitate fibrogenesis by reinforcing ECM deposition and profibrotic signaling within the hepatic niche, including through endothelial-derived adhesion and signaling molecules such as vascular adhesion protein (VAP)-1, whose inhibition attenuates liver fibrosis in experimental models.²⁴⁴ In addition, paracrine pathways activated during hepatocellular injury, including Hedgehog signaling, further couple LSEC dysfunction to HSC activation and progressive fibrotic remodeling.^245,246 Together, these observations highlight how endothelial cell–stromal interactions shape divergent fibrotic outcomes across cardiometabolic diseases: Fibrosis acts as a stabilizing adaptation in atherosclerotic plaques, whereas in MASLD it represents a maladaptive, progressive process driving organ dysfunction.

Endothelial Cell Responses to Therapeutic Strategies

Because atherosclerosis and MASLD are systemic cardiometabolic diseases, this section discusses therapeutic strategies based on their effects on endothelial dysfunction across vascular beds, with particular emphasis on therapies for which evidence exists in both systemic endothelial cells and LSECs. Importantly, current evidence is heterogeneous, and for many therapies, LSEC-specific mechanisms remain incompletely characterized.

Statins

Statins, best known for their lipid-lowering properties, offer important vascular benefits that extend to the management of cardiovascular complications in MASLD patients.²⁴⁷ Statins remain central to the prevention and treatment of atherosclerosis, as endorsed by the 2026 American College of Cardiology/American Heart Association/Multisociety guideline on the management of dyslipidemia. These guidelines recommend risk-based statin therapy, with at least moderate-intensity treatment for intermediate-risk individuals and high-intensity therapy for high-risk and most secondary prevention populations, while endorsing explicit LDL cholesterol and non–HDL cholesterol treatment goals that become progressively lower with increasing atherosclerotic CVD risk.^248,249 Mechanistically, statins inhibit HMG-CoA reductase, reducing endogenous cholesterol synthesis and circulating LDL cholesterol levels.²⁴⁸ However, accumulating evidence indicates that their cardiovascular protection also derives from several pleiotropic endothelial effects. Statins promote eNOS transcription by modulating KLF2, a transcription factor known to promote endothelial homeostasis, through myocyte enhancer factor 2 (MEF2) activation, increase eNOS mRNA stability, and enhance eNOS phosphorylation via activation of the PI3K/Akt pathway.^197,250,251 These actions result in enhanced NO bioavailability and improved endothelium-dependent vasodilation.^197,250,251 Experimental evidence supports these mechanisms. For example, simvastatin reduced vascular leakage and aortic cholesterol accumulation in hyperlipidemic rabbits and apoE^−/− mice, and inhibited EndoMT in human umbilical vein endothelial cells by suppressing TGF-β/Smad signaling and oxidative stress252, 253, 254 (Figure 4). Additional mechanisms include reduced caveolin-1 expression (which inhibits eNOS),²⁵⁵ suppression of NF-κB–driven inflammation, inhibition of Wnt/β-catenin signaling, and upregulation of protective noncoding RNAs such as mechanosensitive antisense transcript of NOS3 interacting with SRF (MANTIS).^256,257 Statins also exert antiapoptotic effects, further preserving endothelial integrity.²⁵⁸

Figure 4.

Endothelial Responses to Selected Pharmacological and Nanotherapeutic Strategies in Atherosclerosis and MASLD

Schematic illustration of therapeutic approaches with direct or indirect endothelial relevance across the liver and vasculature. Pharmacological therapies, including statins and glucagon-like peptide-1 receptor agonists (GLP-1 medicines), may improve endothelial homeostasis, reduce vascular leakage and endothelial-to-mesenchymal transition (EndoMT), restore hepatic microvascular architecture, and attenuate portal hypertension, inflammation, and fibrosis. Nanotherapeutic approaches, including simvastatin-loaded liposomes, collagen-targeted nanoparticles, hyaluronic acid-based nanoparticles, titanium dioxide nanoparticles, αvβ3-targeted nanoparticles, selenium-containing nanoparticles, and siRNA-containing nanoparticles, are shown to modulate endothelial permeability, preserve liver sinusoidal endothelial cell (LSEC) fenestrations, suppress inflammatory and fibrogenic signaling, improve vascular function, and reduce atherosclerotic lesion development or progression. Question marks indicate mechanisms that remain incompletely defined or require further validation. EC = endothelial cell; HSC = hepatic stellate cell; LSEC = liver sinusoidal endothelial cell; MASLD = metabolic dysfunction–associated steatotic liver disease; NO = nitric oxide; SMC = smooth muscle cell.

Beyond systemic vascular protection, statins also influence hepatic microvascular biology in MASLD. In particular, they have been shown to improve LSEC function through mechanisms involving the KLF2-eNOS-NO signaling axis.⁸⁰ KLF2 expression in LSECs exerts protective effects by enhancing NO production and mediating paracrine signals that suppress hepatic stellate cell activation, contributing to the attenuation of hepatic inflammation and fibrosis²⁵⁹ (Figure 4). Among statins, simvastatin has been the most extensively studied in the context of hepatic microvascular and LSEC biology, particularly in models of portal hypertension and cirrhosis where activation of the KLF2-eNOS-NO pathway improves intrahepatic vascular resistance.^80,259, 260, 261, 262, 263, 264, 265, 266, 267 Several preclinical and translational studies demonstrated that simvastatin improves LSEC function, restores hepatic microvascular architecture, and attenuates portal hypertension. In bile duct ligation and CCl₄-induced cirrhotic rat models, both conventional and nanoparticle-based simvastatin increased KLF2-dependent eNOS expression, reduced hepatic stellate cell activation, and improved hepatic perfusion.260, 261, 262 In patients with cirrhosis, short-term simvastatin treatment enhanced hepatic NO production and reduced vascular tone.²⁶⁸ Additional studies in diet-induced and toxin-induced models revealed improved LSEC differentiation, decreased collagen I and iNOS expression, and reversal of microvascular remodeling.^80,263, 264, 265 In parallel, statins activate peroxisome proliferator-activated receptor alpha (PPARα), a transcription factor involved in regulating vascular tone, oxidative stress, and fibrogenic gene expression, all of which are implicated in the pathogenesis of MASLD and the development of portal hypertension.^266,267 In contrast, for other statins such as pravastatin or fluvastatin, MASLD-related evidence mainly derives from observational studies demonstrating safety and clinical associations, while direct LSEC-specific mechanistic studies remain limited (Table 1).

Table 1.

Statins as Endothelial Modulators Linking Vascular and Hepatic Microcirculatory Dysfunction in Atherosclerosis and MASLD

Statin	Atherosclerosis Endothelial Effects	MASLD/MASH LSEC Effects	Key Endothelial Shared Mechanisms
Simvastatin	↑eNOS activation (Akt/AMPK) →↑NO bioavailability ↓Rho/ROCK signaling ↓oxidative stress and adhesion molecules271, 272, 273, 274, 275, 276	↑KLF2-eNOS-NO signaling in LSECs ↓stellate cell activation improved sinusoidal microcirculation and portal pressure80,259, 260, 261, 262, 263, 264, 265, 266, 267	eNOS-NO restoration and anti-inflammatory endothelial signaling linking vascular and sinusoidal protection
Atorvastatin	↑eNOS/NO via PI3K/Akt ↓Rho/ROCK and endothelial inflammation277, 278, 279, 280	Improves portal hypertension and fibrosis ↑ KLF2 signaling in LSECs ↓ intrahepatic vascular resistance281	KLF2-NO axis and improved endothelial repair across vascular beds
Lovastatin	↑eNOS expression ↓oxidative stress and endothelial dysfunction282,283	LSEC-specific data lacking in MASLD models	No evidence
Pravastatin	Improves endothelial-dependent vasodilation ↓oxidative stress and adhesion molecules284, 285, 286	LSEC-specific data lacking in MASLD models	No evidence
Fluvastatin	↓vascular inflammation and leukocyte–endothelium adhesion ↓oxidative stress287	LSEC-specific data lacking in MASLD models	No evidence

Overview of statin-associated endothelial mechanisms across systemic vascular endothelium and LSECs. In addition to lipid-lowering effects, statins improve endothelial homeostasis through pathways such as eNOS-NO signaling, KLF2 activation, and suppression of oxidative and inflammatory signaling. In the liver, these mechanisms contribute to improved sinusoidal function, reduced hepatic stellate cell activation, and decreased portal hypertension in experimental and translational models of MASLD/MASH.

eNOS = endothelial nitric oxide synthase; LSEC = liver sinusoidal endothelial cell; MASH = metabolic dysfunction–associated steatohepatitis; MASLD = metabolic-associated steatotic liver disease; NO = nitric oxide.

Although statins are generally well tolerated, clinicians should remain cautious of their potential adverse effects, especially in patients with existing liver disease. Reported side effects include elevations in liver enzymes, an increased incidence of new-onset type 2 diabetes, and musculoskeletal symptoms ranging from mild myalgia to rare cases of rhabdomyolysis.^269,270 Therefore, balancing therapeutic efficacy with safety is crucial. Optimizing dosing regimens and exploring novel statin-based or statin-synergistic therapies may enhance outcomes in both MASLD and atherosclerosis, particularly where endothelial dysfunction is central to disease progression.

PCSK9 inhibition

Proprotein convertase subtilisin/kexin type 9 (PCSK9) has emerged as an important regulator of lipid metabolism and vascular inflammation, and may represent a key mediator linking hepatic metabolic dysfunction with systemic endothelial pathology.288, 289, 290 PCSK9 is primarily produced by hepatocytes and is best known for promoting degradation of the LDL receptor, thereby increasing circulating LDL cholesterol levels.²⁹¹ However, accumulating evidence indicates that PCSK9 also exerts direct effects on vascular cells that extend beyond lipid metabolism. Experimental studies demonstrate that PCSK9 promotes endothelial dysfunction, oxidative stress, and inflammatory signaling.²⁹² In primary human and murine endothelial cells, recombinant PCSK9 impairs efferocytosis and increases markers of endothelial senescence, whereas genetic or pharmacological inhibition of PCSK9 restores endothelial function and reduces oxidative and inflammatory signaling.^292,293 PCSK9 also enhances endothelial inflammatory activation by promoting expression of adhesion molecules and cytokines, partly through interactions with oxidized LDL signaling pathways such as LOX-1.²⁹⁴ In experimental atherosclerosis models, PCSK9 deletion reduces endothelial expression of inflammatory genes including ICAM-1, CCL2, and IL-6 and markedly attenuates plaque development.²⁹⁵ Beyond endothelial cells, PCSK9 can influence thrombosis by activating platelets through CD36-dependent ROS/MAPK pathways, further contributing to vascular inflammation and microvascular injury.²⁹⁶ Although direct evidence for PCSK9-mediated LSEC dysfunction in MASLD remains limited, alterations in PCSK9 signaling may indirectly contribute to hepatic inflammation and steatosis through dysregulated lipid metabolism. Consistent with this concept, PCSK9 inhibition has demonstrated substantial cardiovascular risk reduction in large clinical trials and is increasingly explored as a therapeutic strategy with potential benefits extending beyond LDL lowering.288, 289, 290 Together, these findings position PCSK9 as a promising mechanistic and therapeutic node linking hepatic lipid metabolism with systemic endothelial and cardiovascular dysfunction, although further studies are needed to clarify whether PCSK9 directly modulates LSEC phenotype in MASLD.

GLP-1 receptor agonists

Glucagon-like peptide-1 receptor agonists (GLP-1RAs), including semaglutide, have emerged as important cardiometabolic therapies with beneficial effects on vascular endothelial function (Figure 4). In atherosclerosis, GLP-1R signaling improves endothelial homeostasis through activation of AMPK/Akt-dependent eNOS phosphorylation, increasing NO bioavailability and improving endothelium-dependent vasodilation.297, 298, 299 In patients with obesity and cardiovascular disease, semaglutide modulates the endocrine function of epicardial adipose tissue, leading to increased secretion of antithrombotic proteins such as gelsolin and a reduction in the adipokine fatty acid binding protein 4 (FABP4), a molecule linked to endothelial activation and coronary artery disease severity.³⁰⁰ Moreover, semaglutide was shown to alter neutrophil phenotype and reduce their adhesion to human aortic endothelia cells likely through downregulation of adhesion molecules such as ICAM-1 and VCAM-1.³⁰⁰ Beyond these endothelial mechanisms, recent integrative preclinical and clinical evidence indicates that GLP-1RAs also exert broader antiatherosclerotic and anti-inflammatory effects.³⁰¹ In normoglycemic, nonobese rabbits, liraglutide reduced plaque progression, cathepsin activity, macrophage accumulation, and circulating C-reactive protein levels.³⁰¹ In a large clinical cohort, GLP-1RA use was associated with lower inflammatory biomarkers and fewer major adverse cardiovascular events, irrespective of glycated hemoglobin or body mass index.³⁰¹

In patients with MASH, semaglutide not only facilitates weight loss and improves insulin sensitivity but also appears to exert direct hepatoprotective effects independent of weight reduction.302, 303, 304 A recent phase III randomized controlled trial demonstrated that once-weekly semaglutide significantly increased rates of MASH resolution without worsening fibrosis, representing the first pharmacologic agent to achieve such histologic improvement in a large, placebo-controlled cohort.¹¹ These pivotal results led to Food and Drug Administration approval of semaglutide as the first treatment for MASH in 2025.³⁰⁵ A complementary retrospective study involving over 400 patients demonstrated significant improvements in liver enzymes (aspartate aminotransferase and alanine aminotransferase) and noninvasive fibrosis scores (NFS, FIB-4, APRI) after prolonged semaglutide treatment, irrespective of baseline body mass index or degree of weight loss.³⁰⁶ Importantly, recent mechanistic work from the Drucker laboratory provides direct evidence that some weight-loss–independent hepatic benefits of semaglutide are mediated through intrahepatic GLP-1 receptors expressed on LSECs.³⁰⁷ In experimental MASH, endothelial-specific deletion or knockdown of GLP-1R substantially attenuated the beneficial effects of semaglutide despite preserved weight loss, while transcriptomic analyses showed reversal of stress-associated LSEC phenotypes and restoration of injury-repair signaling networks.³⁰⁷ These findings identify LSECs as active mediators of semaglutide action and strengthen the concept that GLP-1RAs may directly modulate hepatic endothelial biology in addition to their systemic metabolic effects.³⁰⁷ Clinically, current MASLD guidelines recommend GLP-1RAs for their approved indications (type 2 diabetes and obesity) and consider them safe in MASLD, although they are not yet recommended solely as MASH-targeted therapy outside clinical trials.²⁴ Collectively, these findings position GLP-1RAs as dual-action therapies capable of improving systemic endothelial function while also engaging intrahepatic endothelial pathways relevant to MASH progression.

SGLT2 inhibitors

Sodium glucose cotransporter-2 (SGLT2) inhibitors, originally developed for the treatment of type 2 diabetes mellitus, have emerged as important cardiometabolic therapies with beneficial effects on vascular endothelial function and systemic inflammation.^308,309 In the context of atherosclerosis, SGLT2 inhibitors reduce circulating inflammatory mediators such as C-reactive protein and IL-6, decrease oxidative stress, and improve endothelial NO bioavailability, collectively contributing to improved endothelial function and arterial compliance.^308,309 Preclinical studies further demonstrate that SGLT2 inhibition can directly ameliorate endothelial dysfunction through anti-inflammatory and antioxidant signaling pathways, supporting a class effect on vascular homeostasis.^308,309 In MASLD, SGLT2 inhibitors such as empagliflozin and dapagliflozin improve metabolic parameters by reducing hyperglycemia, enhancing insulin sensitivity, and decreasing hepatic lipid accumulation.^21,310 These metabolic improvements are associated with reductions in liver inflammation and fibrosis in both experimental models and clinical studies.^21,310 Randomized trials and observational cohorts have reported improvements in liver enzymes, hepatic steatosis, and histologic endpoints following SGLT2 inhibitor treatment, while large real-world studies also suggest reduced risks of hepatic and cardiovascular complications in patients with MASLD and diabetes.311, 312, 313, 314, 315 Despite these encouraging findings, direct evidence that SGLT2 inhibitors specifically restore LSEC structure or reverse sinusoidal capillarization in MASLD/MASH is currently lacking, as available studies have not included LSEC-specific phenotyping or sinusoidal endothelial functional measurements. The hepatic benefits observed with SGLT2 inhibition therefore appear to be mediated primarily through metabolic and immune mechanisms rather than direct modulation of LSEC biology.

Nanotherapy

Nanotherapy has emerged as a highly adaptable strategy for modulating endothelial dysfunction and inflammation in both atherosclerosis and MASH, and several nanoplatforms demonstrate relevance across both diseases by targeting shared endothelial pathways³¹⁶ (Figure 4). Statin-based nanotherapies provide a clear example of this cross-disease potential: Simvastatin-loaded liposomes (LIPOSTAT) reduce endothelial activation, monocyte adhesion, and inflammatory cytokine expression within atherosclerotic lesions,³¹⁷ while parallel formulations (PM127-simv) are preferentially internalized by LSECs, where they activate KLF2-NO signaling, improve endothelial phenotype, and exhibit lower toxicity than free drug.³¹⁷ Moreover, advanced nanoplatforms have been engineered to address the structural and biochemical barriers to drug delivery posed by liver sinusoidal capillarization and extracellular matrix deposition.²⁴² Zhang et al³¹⁸ developed a dual-function nanodrug system consisting of fenestrae-repairing, LSEC-targeted nanoparticles (HA-NPs/SMV) loaded with simvastatin, which restored sinusoidal fenestration. These particles facilitated the entry of a second nanoplatform, codecorated collagen-degrading, HSC-targeted nanoparticles (CV-NPs/siCol1α1) into the space of Disse. This 2-step approach successfully degraded fibrotic ECM and increased nanoparticle accumulation in activated HSCs, as confirmed by scanning electron microscopy showing recovery of liver sinusoidal structure.³¹⁸

In contrast, titanium dioxide nanoparticles (TiO₂ NPs) exemplify the critical importance of endothelial specialization in determining nanoparticle outcome. In the arterial circulation, TiO₂ NPs exacerbate atherogenesis by promoting endothelial inflammation and dyslipidemia, thereby accelerating plaque development and highlighting the potential vascular toxicity of poorly targeted inorganic nanoparticles.^319,320 Paradoxically, in the hepatic microcirculation, TiO₂ NPs exert beneficial effects by inducing controlled sinusoidal leakiness in LSECs, restoring permeability, enhancing drug uptake, and facilitating hepatic recovery in MASH models.³²¹ This divergence reflects fundamental differences between continuous arterial endothelium, where barrier disruption is pathogenic, and fenestrated sinusoidal endothelium, where increased permeability can be therapeutically advantageous.^242,322 Another instructive example of shared targeting with divergent cellular outcomes is αvβ3 integrin–directed nanotherapy. In atherosclerosis, αvβ3 integrin is upregulated on activated vascular endothelial cells within inflamed plaques, enabling selective delivery of therapeutics to diseased endothelium.³²³ Cathepsin K–responsive, αvβ3-targeted nanoparticles (RAP@T/R NPs) have been shown to release anti-inflammatory payloads specifically within protease-rich plaques, reducing oxLDL uptake, suppressing inflammatory cytokine production, and significantly attenuating plaque burden in Apoe^−/− mice.³²³ In contrast, in MASH, αvβ3 integrin targeting is primarily exploited on activated HSC, rather than LSECs, to address fibrogenesis. Ligand-functionalized liposomes incorporating cyclic RGD peptides (eg, cRGDyK) preferentially internalize into αvβ3-expressing HSC while sparing LSECs, enabling cell-type–specific antifibrotic delivery.³²⁴ Thus, αvβ3 integrin represents a shared molecular target across both diseases, but its therapeutic utility depends on tissue context and the pathogenic cell population being targeted, activated endothelium in atherosclerosis vs activated stellate cells in MASH.

Selenium-containing nanoparticles further illustrate how endothelial-targeted nanotherapies can improve vascular function in atherosclerosis while leaving their relevance to hepatic endothelium unresolved.³²⁵ Xiao et al³²⁵ showed that oral administration of selenium-containing nanoparticles (CS SeNPs and Na₂SeO₃) increased NO bioavailability and reduced adhesion molecule expression in high-fat diet–fed Apoe^−/− mice, leading to improved endothelial function and attenuation of atherosclerotic lesion formation. However, these effects were evaluated exclusively in the arterial vasculature, and whether selenium-based nanoparticles are taken up by liver sinusoidal endothelial cells or modulate LSECs dysfunction in MASLD/MASH remains unknown.

Along the same lines, polymeric nanoparticles composed of low-molecular-weight polyamines and lipids have been shown to deliver siRNAs to endothelial cells with high efficiency in vivo, enabling simultaneous silencing of multiple endothelial genes.³²⁶ Notably, this platform achieves durable endothelial gene silencing while largely sparing hepatocytes and immune cells, even at doses required for vascular targeting.³²⁶ Using this approach, endothelial-targeted delivery of siRNAs against transforming growth factor-β receptors (Tgfbr1/2) in Apoe^−/− mice fed a high-cholesterol diet resulted in approximately 50% reductions in aortic lipid deposition, marked decreases in brachiocephalic plaque size, and regression of established atherosclerotic lesions following dietary normalization.³²⁷ However, while this platform effectively targets vascular endothelium, its uptake, specificity, and functional effects in liver sinusoidal endothelial cells were not directly examined, and its applicability to MASLD/MASH-associated sinusoidal dysfunction therefore remains to be determined.

In addition to these in vivo platforms, several nanomaterials have shown endothelial-protective effects in vitro, providing mechanistic proof-of-concept for endothelial modulation. Amphiphilic polysiloxane nanoparticles enhanced NO production in human aortic endothelial cells via caveolae-mediated endocytosis, contributing to improved endothelial function.³²⁸ In a complementary approach, chitosan-based hydrogels embedded with nanocurcumin and arginine promoted eNOS phosphorylation and antioxidant activity in endothelial cells under hypoxic conditions.³²⁹ Although these studies support the feasibility of nanomaterial-driven endothelial reprogramming, their relevance to disease-specific endothelial subtypes, including LSECs, and their translational potential in atherosclerosis or MASLD/MASH remain to be established in vivo.

Overall, while several cardiometabolic therapies improve systemic endothelial function, direct evidence for modulation of LSEC phenotype remains limited, with statins, particularly simvastatin and atorvastatin, representing the most mechanistically supported class across both vascular and hepatic endothelium.

Biomarkers for Endothelial Dysfunction

Translating these mechanistic insights into clinical practice requires robust biomarkers that reflect endothelial injury and therapeutic response across organs. Identifying such markers is essential for the early detection and management of both atherosclerosis and MASLD. Promising candidates include microRNAs and circulating endothelial microparticles. MicroRNAs play a significant role in regulating endothelial function and could serve as novel biomarkers.330, 331, 332 Specific microRNAs have been implicated in the modulation of endothelial cell behavior, including inflammation and angiogenesis, making them potential targets for therapeutic intervention.333, 334, 335, 336, 337 Developing noninvasive biomarkers to assess endothelial health is critical for early diagnosis and monitoring treatment efficacy in both MASLD and atherosclerosis. Techniques like FMD and pulse-wave analysis are used to evaluate endothelial function directly and represent some of the most clinically translatable functional vascular measures currently available.^338,339 Additional measures such as arterial stiffness and endothelial-dependent vasoreactivity may also serve as shared vascular endpoints in future MASLD–atherosclerosis trials.^67,340 Circulating biomarkers offer a more accessible and less invasive approach for routine monitoring.^24,341

Endothelial microparticles (EMPs) are small vesicles released from activated or apoptotic endothelial cells and have been recognized as indicators of endothelial dysfunction.³⁴² They are involved in various pathological processes, including atherosclerosis, by promoting oxidative stress and vascular inflammation. Studies have shown that EMP levels are elevated in patients with CVDs and can serve as biomarkers for disease status and progression.^339,343 Because standardization of isolation and quantification methods remains limited, EMPs are currently best considered promising research biomarkers rather than routine clinical tools.344, 345, 346 More broadly, extracellular vesicles, including endothelial-derived vesicles, are also emerging as potential biomarkers in MASLD, where their circulating cargo may reflect hepatic inflammation, fibrosis, and vascular injury.^66,347,348 Among the classic markers, soluble intercellular adhesion molecule-1 (sICAM-1) and soluble vascular cell adhesion molecule-1 (sVCAM-1) have long been recognized as indicators of endothelial activation and inflammation.349, 350, 351, 352, 353, 354 These adhesion molecules are shed into the circulation upon endothelial stimulation and reflect the early inflammatory response of the endothelium.349, 350, 351, 352, 353, 354 Elevated plasma levels of sICAM-1 and sVCAM-1 have been consistently associated with cardiovascular risk, as well as with liver inflammation and fibrosis in MASLD, making them among the biomarkers closest to current clinical translation for endothelial risk stratification.349, 350, 351, 352, 353, 354

Future clinical trials evaluating therapies in MASLD patients with increased cardiovascular risk should incorporate endothelial endpoints alongside liver-specific outcomes.³⁵⁵ A combined framework including circulating biomarkers (eg, sICAM-1, sVCAM-1), vascular functional measures (eg, FMD, pulse-wave analysis), and hepatic endpoints may better capture dual liver-vascular therapeutic benefit. Sex-specific baseline differences and treatment responses should also be considered when selecting and interpreting endothelial biomarkers. Overall, the identification and validation of these biomarkers will enhance our ability to diagnose and manage endothelial dysfunction in these diseases effectively. Collectively, these emerging and established biomarkers not only allow earlier detection of endothelial injury but also provide measurable endpoints to evaluate the impact of novel therapeutic strategies on endothelial repair and function.

Current Gaps and Challenges

Despite growing progress, several conceptual and technical challenges still limit endothelial-targeted translation. Researching endothelial cells poses several challenges given their significant heterogeneity across different vascular beds. This heterogeneity complicates the development of generalized therapeutic approaches, as endothelial cells in various organs exhibit distinct molecular signatures and respond differently to environmental cues.^356,357 For instance, endothelial cells in the brain and heart express classical inflammatory adhesion molecules like E-selectin and P-selectin, whereas those in the lung upregulate chemokines such as Cxcl1 and Cxcl9 in response to systemic inflammation.³⁵⁶ An additional major challenge lies in the limited ability to selectively and efficiently genetically target specific endothelial subpopulations, particularly vascular endothelial cells or LSECs, which has constrained mechanistic studies and slowed the development of cell-type–specific therapeutic strategies.³⁵⁸ Another challenge lies in the limitations of current animal models for MASLD and atherosclerosis. These models only partially replicate the complex human pathophysiology, which limits the translatability of preclinical findings to human diseases.^359,360 For example, although high-fat diet and Western diet models effectively simulate metabolic aspects of MASLD, they often result in milder histological phenotypes compared with human disease.³⁵⁹ Choline-deficient dietary models rapidly induce fibrosis but fail to accurately model the metabolic burden of MASLD.³⁵⁹ Recent sex-based multiomics analyses have further underscored these limitations, revealing that commonly used dietary models recapitulate concurrent MASH and atherosclerosis in a sex- and diet-dependent manner, with no single model fully capturing disease co-occurrence across sexes. In Ldlr^−/− mice, modified choline-deficient high-fat diets induce both MASH and atherosclerosis in males and females, whereas Western diet–based models are largely effective only in males, highlighting a critical gap in modeling female cardiometabolic disease.³⁵⁴ Furthermore, endothelial cells exhibit context-dependent phenotypic changes in response to metabolic stress, which adds complexity to their study. These changes are influenced by the extracellular environment and can be epigenetically fixed, making it challenging to capture their full range of behaviors in vitro.³⁵⁷ The removal of endothelial cells from their native tissue for culture can lead to phenotypic drift, further complicating the interpretation of experimental results.³⁵⁷ Another important challenge is the limited integration of sex as a biological variable in preclinical and clinical studies, despite evidence that endothelial biology, disease progression, and therapeutic responses may differ between males and females. Addressing these challenges will be crucial for developing effective therapeutic strategies targeting endothelial cells in MASLD and atherosclerosis.

Conclusions

Endothelial cells play a pivotal role as a common pathological link between atherosclerosis and MASLD. By targeting endothelial dysfunction, it is possible to address the root causes of both cardiovascular and liver disease. This approach offers a promising strategy for managing the interrelated pathologies of atherosclerosis and MASLD, which are increasingly recognized as components of a broader metabolic syndrome. Future research should focus on developing endothelial-specific therapies that can effectively restore endothelial function and reduce inflammation. This includes exploring novel therapeutic agents such as NO donors, anti-inflammatory drugs, and nanomedicine-based delivery systems that can selectively target endothelial cells. Additionally, refining biomarkers to monitor endothelial health in MASLD and atherosclerosis is crucial for early diagnosis and assessing treatment efficacy. Circulating endothelial microparticles, markers of oxidative stress, and microRNAs involved in endothelial function are potential candidates for these biomarkers. The ultimate goal is to develop comprehensive, endothelial-targeted treatments that address the dual burden of metabolic and cardiovascular disease. By doing so, we can improve outcomes for patients worldwide by reducing the risk of liver fibrosis, cardiovascular events, and other complications associated with atherosclerosis and MASLD. This integrated approach underscores the importance of interdisciplinary research and collaboration between hepatologists, cardiologists, and vascular biologists to advance our understanding and management of these interconnected diseases.

Funding Support and Author Disclosures

This work was supported by the American Heart Association Postdoctoral Fellowship (26POST1566114) and the Center for Cardiovascular Diseases and Sciences Malcolm Feist Postdoctoral Fellowship (to Dr Ben Dhaou); by the following National Institutes of Health awards: HL098435, HL133497, HL173972, and GM121307 (to Dr Orr); DK136685, DK134011, and HL150233 (to Dr Rom); and by the National Science Foundation grant 2537597 (to Dr Rom and Dr Orr). Dr Rom serves as a scientific advisor to Diapin Therapeutics LLC. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.