Research Progress on Nanoformulations Based on Active Components from Traditional Chinese Medicine for MASLD

Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD), characterized by hepatic steatosis with cardiometabolic risk factors, exhibits a rising global trend and currently affects approximately 38% of the population.^1,2 MASLD is closely associated with systemic metabolic dysfunction and commonly coexists with multiple metabolic risk factors.³ A subset of patients may progress to metabolic dysfunction-associated steatohepatitis (MASH), which represents a critical stage in disease progression.⁴ Aligned with the updated international consensus, the terms MASLD and MASH are used herein, superseding the previous designations of non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatitis (NASH).³

The pathogenesis of MASLD involves a complex network characterized by the interplay of multidimensional mechanisms. Multiple factors, such as overnutrition, obesity, and insulin resistance (IR), induce an excessive influx of free fatty acids into the liver, leading to hepatic lipid deposition and lipotoxicity.^5,6 This, in turn, exacerbates hepatic IR to form a vicious cycle, further triggering hepatocyte injury via endoplasmic reticulum stress, mitochondrial dysfunction, oxidative stress, impaired lipophagy, and ferroptosis.^7,8 Concurrently, gut microbiota dysbiosis, compromised intestinal barrier integrity, and disrupted bile acid metabolism further aggravate hepatic injury via the gut–liver axis.^9,10 These multifaceted insults collectively promote hepatic inflammation and immune microenvironment remodeling, eventually resulting in hepatic stellate cells (HSCs) activation and fibrotic progression.^7,11,12

Although pharmacological development for MASLD has advanced in recent years, currently available therapies remain limited.¹³ Obeticholic acid, once considered a promising candidate, did not receive U.S. Food and Drug Administration (FDA) approval for MASLD/MASH because of concerns over its overall benefit–risk profile.¹³ By contrast, Resmetirom was approved by the FDA in 2024 as the first approved therapy for adults with non-cirrhotic MASH and moderate to advanced fibrosis.^14,15 However, the response rate in the Phase 3 Resmetirom trial was only about 30%,¹⁵ while treatment cost remains high and long-term safety issues have yet to be fully resolved.¹⁶ These data highlight the imperative to explore safer and more effective therapeutic strategies.

TCM has attracted increasing attention in the treatment of MASLD because many of its bioactive constituents, such as terpenoids, glycosides, polyphenols, and flavonoids, can act on multiple pathogenic processes involved in the disease.^17–20 Owing to their multitarget properties, these compounds exhibit multidimensional therapeutic potential against MASLD by simultaneously intervening in its interconnected metabolic, inflammatory, and fibrotic mechanisms.²¹ This therapeutic relevance, together with the growing body of MASLD-related studies on TCM-derived compounds, highlights TCM as a particularly relevant therapeutic domain in MASLD research.^19,20,22

However, the translational application of many TCM active compounds remains hindered by poor water solubility, low chemical stability, inadequate bioavailability, and insufficient tissue specificity.²³ Nanotechnology provides a practical bridge between the pharmacological potential of these compounds and the drug delivery demands of MASLD, as it can both enhance their therapeutic potential and improve delivery efficiency.²⁴ By increasing drug loading, protecting unstable compounds, enhancing targeting, and enabling controlled or stimuli-responsive release, nanoformulations may substantially improve the therapeutic efficacy of TCM active compounds.²⁵

Several recent reviews have addressed nanomedicine for MASLD/MASH or nanocarrier-based delivery of TCM active compounds.^23,26–29 However, these studies have generally focused either on broad nanotechnology strategies for MASLD or on TCM nanocarriers across multiple diseases, rather than specifically on TCM-derived nanoformulations in MASLD. A focused synthesis of MASLD-oriented TCM nanoformulations, particularly one integrating platform classification, mechanistic relevance, and translational considerations, therefore remains lacking. Accordingly, this review, informed by a structured search strategy, summarizes current advances in TCM-derived nanoformulations for MASLD from the perspectives of nanoplatform types, design features, representative applications, and translational challenges.

Survey Methods

This review was conducted as a narrative review supported by a structured literature search. PubMed, Web of Science, Scopus, and Embase were searched for English-language publications published between January 2015 and December 2025, with additional manual screening of the reference lists of relevant articles. The search strategy combined controlled vocabulary and free-text terms related to nanotechnology, MASLD-related diseases, and active components derived from traditional Chinese medicine. Detailed inclusion and exclusion criteria are provided in Supplementary Document S1. Through the structured screening process, 45 original studies were included in the focused synthesis to summarize mechanistic evidence, delivery or targeting rationale, formulation innovation, therapeutic significance, and translational value of nanoformulations based on traditional Chinese medicine-derived active compounds in MASLD. The search and selection process is summarized in Figure 1, and the detailed search strategies are provided in Supplementary Table S1.

Figure 1 Flowchart of the structured literature search. (A total of 151 references were ultimately cited in this narrative review. Among them, 45 original studies identified through the structured screening process were included in the focused synthesis. Figure created by the authors).

Barrier-Informed Design and Targeting Logic of MASLD-Oriented TCM Nanoformulations

The delivery performance of TCM nanoformulations in MASLD is determined not only by carrier composition but also by the disease-specific barrier landscape and the corresponding targeting requirements. As summarized in Figure 2, administration route and MASLD-specific barriers collectively determine how nanocarrier properties such as particle size, surface charge, and ligand decoration influence hepatic exposure, cell selectivity, and gut–liver-axis intervention.

Figure 2 MASLD-specific delivery barriers and design implications for TCM nanoformulations. (Figure created by the authors).

MASLD-Specific Delivery Barriers and Nanocarrier Design Considerations

The design of TCM nanoformulations for MASLD should be interpreted against a disease-specific transport landscape rather than generic liver accumulation alone. In MASLD, liver sinusoidal endothelial cells (LSECs) undergo capillarization and fenestrae loss from early disease stages, and these changes become more pronounced with fibrosis, thereby reducing trans-sinusoidal exchange and limiting nanoparticle access to the space of Disse and hepatocytes.^30,31 At the same time, Kupffer cells and LSECs constitute a major hepatic clearance barrier, so liver accumulation does not necessarily indicate efficient delivery to the intended target cells.³² As MASLD progresses toward MASH and fibrosis, fenestrae loss and excessive extracellular matrix deposition may further hinder nanoparticle penetration.^31,33

These pathological changes have direct implications for nanocarrier design. Particle size influences whether nanoparticles can pass through sinusoidal fenestrae and gain access to the space of Disse.³² Surface charge can also influence how nanoparticles are distributed among different hepatic cell populations.³² Negatively charged nanoparticles tend to interact more readily with scavenger receptor-rich LSECs and Kupffer cells, whereas positively charged particles often show stronger membrane interactions and endocytosis, which may favor uptake by hepatocytes in some settings.³² Surface chemistry also affects protein adsorption; compared with neutral or hydrophilic coatings, charged or hydrophobic surfaces are generally more likely to adsorb serum proteins, which may in turn modify opsonization and alter in vivo fate.^32,34 For orally administered systems, mucus and epithelial barriers must also be considered, because nanoparticle diffusion in mucus is strongly influenced by particle size and surface properties.³⁵ Therefore, nanocarrier design in MASLD should be aligned with the intended route, target cell population, and dominant pathological barrier in a given therapeutic scenario.³²

Targeting Strategies of TCM Nanoformulations for MASLD: Passive, Active, and Gut–Liver-Axis-Oriented Delivery

Targeting strategies in MASLD-oriented TCM nanoformulations can be differentiated according to their principal therapeutic level of action. Passive targeting primarily aims to enhance hepatic exposure through nanoparticle physicochemical properties, whereas active targeting seeks to improve cell-type selectivity within the liver; by contrast, gut–liver-axis-oriented delivery acts mainly through extrahepatic regulation.^36,37

After intravenous administration, passive liver-oriented delivery mainly arises from the intrinsic tendency of circulating nanoparticles to accumulate in the liver. However, such accumulation does not necessarily translate into efficient hepatocyte delivery, because a substantial fraction of nanoparticles is intercepted by Kupffer cells and LSECs.³² After oral administration, by contrast, passive liver-oriented delivery is influenced more strongly by gastrointestinal barriers and transport processes, including mucus penetration, epithelial uptake/transcytosis, and, for some lipid-based systems, intestinal lymphatic transport, which may partially circumvent first-pass loss.^38,39

Active targeting complements passive hepatic exposure by promoting nanoparticle retention and internalization in specific hepatic cell populations.^36,40 Among hepatocyte-targeted strategies, asialoglycoprotein receptor (ASGPR) remains the best-established receptor because it is predominantly expressed on hepatocytes and recognizes ligands bearing terminal galactose or GalNAc residues; accordingly, galactose-, GalNAc-, and lactobionic acid (Lac)-based modifications have been widely used to enhance hepatocyte uptake.⁴⁰ Beyond ASGPR, other hepatocyte surface receptors may also be exploited for active targeting, including scavenger receptor class B type 1 (SR-B1), which can mediate apolipoprotein A1 (Apo-A1)-guided uptake, as well as lectin-like carbohydrate-recognition pathways that may be engaged by polysaccharide ligands such as pullulan.^41,42 In contrast, negatively charged nanoparticles are more readily recognized by scavenger receptor-rich LSECs and Kupffer cells, which may compromise hepatocyte selectivity but can be advantageous when therapeutic modulation of non-parenchymal cells is desired.³² Fibrosis-related targeting usually focuses on activated HSCs and the fibrotic microenvironment; in this context, hyaluronic acid (HA)/CD44-based targeting systems have shown enhanced uptake in activated HSCs.⁴³

In addition, some MASLD nanoformulations are not designed for direct hepatocyte targeting but instead aim to intervene through the gut–liver axis. Oral systems with colon-targeting, mucus-adapted delivery, or microbiota-modulating functions may help restore intestinal barrier integrity and reshape microbial- or bile-acid-related signaling, thereby indirectly improving hepatic steatosis and inflammation.³⁷

Overall, the key distinction among these strategies lies less in the mere introduction of a targeting ligand than in whether the principal therapeutic objective is to enhance hepatic exposure, improve intrahepatic cell selectivity, or regulate extrahepatic drivers of MASLD.^36,37,40 Nevertheless, in vivo efficacy remains limited by cross-platform factors, including protein adsorption/opsonization and corona formation, mononuclear phagocyte system clearance, and off-target accumulation/sequestration, which may uncouple apparent liver accumulation from true target-cell delivery.^32,44

Engineering Nanocarriers to Deliver Active Components of TCM

Lipid-Based Nanocarriers

Liposomes

Liposomes are spherical vesicles formed by an amphiphilic phospholipid bilayer surrounding an aqueous core, typically ranging in size from 20 to 1000 nanometers.⁴⁵ This unique structure enables them to simultaneously encapsulate both hydrophilic and hydrophobic drugs.⁴⁶ As a mature nanodelivery platform, liposomes offer good biocompatibility and modifiability;⁴⁷ for example, cholesterol can improve membrane stability, whereas modification with polyethylene glycol (PEG) may prolong circulation.^47,48 These properties make liposomes a versatile platform for improving drug solubility and enabling surface engineering.

In recent years, liposome engineering has increasingly focused on overcoming conventional loading limitations and improving delivery performance. For example, Tanshinone IIA (TSIIA) suffers from extremely poor water solubility.⁴⁹ To address this defect, Cai et al developed an innovative nanocrystal liposome technology (TNC@Lipo) Instead of the conventional loading of TSIIA directly into the lipid bilayer, this strategy involved the initial preparation of TSIIA into highly soluble nanocrystals (TNC) via anti-solvent precipitation combined with ultrasonication. Subsequently, the TNCs were encapsulated within the liposomal core to form a unique “crystal core-lipid shell” structure. This approach addressed the common limitation of low drug loading associated with conventional liposomes for hydrophobic drugs, while also significantly enhancing the physical stability of TSIIA. The system not only reduced the cytotoxicity of the free drug but also significantly enhanced the anti-fibrotic efficacy of TSIIA by increasing drug exposure at the lesion site.⁵⁰

Furthermore, liposomal platforms can support co-delivery and multi-target intervention through flexible formulation and surface design. Chen et al designed a nano-liposome (DEAE-DEX@LSDBC) for the co-delivery of berberine (BER) and curcumin (CUR) This system achieved enhanced flexibility and stability by incorporating sodium deoxycholate into the lipid bilayer to form bilosomes. Surface coating with the cationic polymer diethylaminoethyl dextran (DEAE-DEX) further enhanced mucosal stability and oral absorption, which was associated with increased hepatic accumulation after oral administration. From a mechanistic perspective, the therapeutic effects of this system were associated with coordinated modulation of oxidative stress and inflammatory signaling, particularly the Nrf2-related antioxidant response and the NF-κB/TXNIP–NLRP3 axis.⁵¹

Moreover, similar liposome engineering strategies have been successfully applied to the delivery of flavonoid compounds, such as Chrysin,⁵² Vitexin,⁵³ and baicalin,⁵⁴ significantly enhancing their therapeutic efficacy against MASLD by improving their pharmacokinetics.

Despite their versatility, liposomes in MASLD still face route-dependent translational limitations. After parenteral administration, plasma protein adsorption and mononuclear phagocyte system clearance may reduce effective hepatocyte exposure.^32,55 Although modification with PEG can prolong circulation, it does not fully prevent protein corona formation and may be further complicated by anti-PEG antibodies and accelerated blood clearance (ABC phenomenon) upon repeated dosing.^55,56 For oral delivery, conventional liposomes remain constrained by gastrointestinal instability and limited epithelial permeability, which explains why bile-salt-containing or other surface-engineered variants are often preferred.⁵⁷

Nanoemulsions, Microemulsions, and SMEDDS

Nanoemulsions, microemulsions, and self-microemulsifying drug delivery systems (SMEDDS) are closely related oral lipid-based carriers for hydrophobic TCM active ingredients. All three rely on oil-surfactant systems, often with a co-surfactant or cosolvent, to improve drug solubilization, but differ in how the dispersed state is generated and stabilized.^58,59 Nanoemulsions are preformed aqueous dispersions that generally require external energy input during preparation and remain kinetically stable, whereas microemulsions are spontaneously formed single-phase systems that are thermodynamically stable By comparison, SMEDDS are anhydrous preconcentrates that generate fine droplets only after contact with gastrointestinal fluids.^58–60

To address the poor aqueous solubility and low stability of the flavonoid dihydromyricetin (DMY) for the treatment of MASLD,⁶¹ Li et al developed a hybrid nanoemulsion (DMY-hNE) with a biomimetic “rigid-soft” core-shell structure⁶² In this system, polylactic acid loaded with DMY forms the rigid core, while the outer layer consists of flexible lipids composed of phospholipids and cholesterol. This chylomicron-mimetic architecture facilitates intestinal mucus penetration and supports passive liver-oriented delivery through physiological lipid transport pathways.⁶² By comparison, Lyu et al used a DMY-loaded SMEDDS constructed from D-α-tocopherol polyethylene glycol 1000 succinate and saponin, which improved oral absorption by enhancing drug solubilization and generating finely dispersed droplets after aqueous dilution.⁶³

He et al further developed a curcumin/DHA-rich algal oil microemulsion for co-delivery. In PA-injured WRL-68 and LX2 cells, this formulation showed hepatoprotective and anti-fibrotic potential, while in HFD-fed mice it alleviated liver injury and lowered serum triglyceride and low-density lipoprotein cholesterol levels, accompanied by reduced expression of fatty acid synthase (FAS), 5-LOX, and cPLA2. A separate rat pharmacokinetic study further suggested improved oral exposure to the loaded cargos.⁶⁴

Nanoemulsion-based systems can also intervene in MASLD through the gut–liver axis. In the case of BZEP-NE, therapeutic benefit was associated with remodeling of gut microbial composition, restoration of intestinal barrier function, and activation of the intestinal LXR–ABCA1–HDL pathway, through which high-density lipoprotein cholesterol (HDL-C) may neutralize and clear gut-derived lipopolysaccharide in portal circulation, thereby limiting downstream inflammatory activation.⁶⁵

Despite their oral-delivery advantages, nanoemulsions, microemulsions, and SMEDDS present distinct translational trade-offs. Nanoemulsions remain only kinetically stable, so long-term storage and process robustness require careful control, whereas SMEDDS improve storage stability as anhydrous preconcentrates but may still face precipitation after dilution or dosage-form compatibility issues.⁶⁶ Microemulsions, although thermodynamically stable and relatively easy to form, share with the other two systems a frequent reliance on relatively high surfactant/co-surfactant levels. Across these systems, excessive excipient burdens may perturb intestinal epithelial function, making minimization of surfactant load an important design principle.^58,67 For translational development, the selected excipients should also be assessed for oral regulatory acceptability, such as GRAS status or prior inclusion in the FDA Inactive Ingredient Database.⁶⁷

Nanostructured Lipid Carriers (NLC)

To overcome the limitations of solid lipid nanoparticles (SLN), such as low drug loading capacity and drug leakage during storage, nanostructured lipid carriers have been developed as a new generation of lipid-based nanocarriers.⁶⁸ NLCs are formed by mixing liquid lipids with solid lipids in the presence of surfactants, generating a less ordered or partially amorphous lipid matrix with an “imperfect crystal” arrangement. Compared with SLN, this less ordered matrix can generally accommodate more drug molecules and reduce the risk of drug expulsion during storage.^69,70

NLC technology shows significant advantages in delivering hydrophobic TCM active components. For instance, to address the extremely low oral bioavailability of Naringenin (NGN) (approximately 15%),⁷¹ Hu et al developed an NGN-loaded NLC system (NGN-NLC)⁷² This system utilized stearic acid and glyceryl monostearate as the solid matrix, with oleic acid serving as the liquid lipid. It was prepared using emulsion evaporation and low-temperature solidification techniques. The resulting nanoparticles achieved a particle size of about 163 nm and a high drug loading capacity of 22.5%. Moreover, they successfully increased the drug’s solubility by 115-fold. Mechanistically, inhibitor studies suggested that the enhanced intestinal absorption of NGN-NLC involved clathrin-mediated epithelial uptake together with reduced P-glycoprotein-related efflux. These mechanisms dually promote the intestinal epithelial absorption of NGN. Consequently, NGN-NLC demonstrated superior lipid-lowering and hepatoprotective efficacy compared to the free drug in MASLD mouse models.⁷²

Furthermore, a study developed NLCs co-loaded with astaxanthin and kaempferol. This co-delivery system not only improved the stability and water dispersibility of the ingredients but also enhanced their overall bioactivity compared with astaxanthin-loaded NLCs alone. Specifically, the co-delivery system exhibited superior performance in antioxidant activity by reducing H₂O₂-induced cytotoxicity, and in anti-inflammatory activity by significantly inhibiting lipopolysaccharides (LPS) -induced NO production and the gene expression of inflammatory factors, such as iNOS, TNF-α, and IL-6. In terms of lipid metabolism, this co-loaded system inhibited the transcription of FAS and stearoyl-CoA desaturase 1 by upregulating Insig-2a gene expression and subsequently blocking SREBP-1c activation. It was also associated with activation of the LXRα/CYP7A1 axis to promote cholesterol conversion to bile acids and enhancement of PPARα/CPT-1-mediated fatty acid oxidation, thereby improving hepatocellular lipid homeostasis.⁷³

However, the critical issue for NLCs is whether their delivery advantages can be consistently reproduced in robust and scalable formulations. Because NLC performance is highly dependent on formulation composition and internal structure, especially the solid/liquid lipid balance and surfactant system, small formulation or process variations may alter drug retention and other critical quality attributes, thereby affecting storage stability and batch reproducibility.^74,75 Therefore, for MASLD therapy, further development of NLCs should prioritize storage stability, formulation simplicity, and manufacturability.⁷⁵

Polymer-Based Nanocarriers

Polymeric Micelles

Polymeric micelles are nanocarriers with a core-shell structure (typically 5–100 nm) formed by the self-assembly of amphipathic polymers in an aqueous phase.^76,77 Among these, the most commonly used amphiphilic polymers are block copolymers composed of two or more hydrophilic and hydrophobic polymer chains.^76,78 The hydrophobic core of polymeric micelles effectively enhances the solubility of hydrophobic drugs, while the hydrophilic shell contributes to colloidal stability and prolonged circulation, making them a useful platform for improving the pharmacokinetics of poorly soluble active ingredients from TCM.^76,77 Furthermore, polymeric micelles possess the advantage of ease of engineering, enabling active targeting and intelligent drug release via functional modification.

For instance, to address the limitation of the low bioavailability of resveratrol (Res) for the treatment of MASLD,⁷⁹ two independent studies adopted similar hepatocyte-targeting strategies. Teng et al constructed Gal-OSL/Res micelles using galactose-modified oxidized starch-lysozyme,⁸⁰ whereas Li et al developed Gly-LA-Lac/Res micelles using lactobionic-acid-modified natural glycogen.⁸¹ In both systems, galactose-containing ligands promoted hepatocyte-directed delivery through ASGPR recognition. Gly-LA-Lac/Res reduced hepatic lipid accumulation and oxidative stress partly via modulation of the TLR4/NF-κB pathway,⁸¹ whereas Gal-OSL/Res was shown to activate the AMPK/SIRT1 axis and improve hepatic insulin signaling by decreasing hepatic insulin receptor substrate-1 phosphorylation.⁸⁰ A distinct hepatocyte-targeting strategy was later reported for oral celastrol micelles (OPDEA-PCL/CEL).⁴² Rather than relying on a preinstalled hepatocyte ligand, this system first achieved efficient oral absorption and then selectively adsorbed plasma high-density lipoprotein cholesterol (HDL) after entering the circulation, thereby exposing Apo-A1 and promoting SR-B1-mediated hepatocyte uptake. Mechanistically, OPDEA-PCL/CEL alleviated steatosis by activating AMPK and suppressing SREBP1/FAS-driven lipogenesis, while also reducing hepatic inflammation and mitigating celastrol-associated toxicity.⁴²

Furthermore, polymeric micelles can be engineered as “smart” systems integrating diagnosis and treatment. Wang et al constructed innovative polymeric micelles (PGOD NPs) for the delivery of Glycyrrhetinic acid (GA) In this system, GA, PEG chains, and fluorescent probes are covalently linked via reactive oxygen species (ROS)-sensitive oxalate bonds. Upon activation in the high-ROS environment of MASLD lesions, the cleavage of oxalate bonds initially consumed ROS to alleviate oxidative stress. Subsequently, fluorescent probes with aggregation-induced emission characteristics were activated, enabling the real-time visualization of lipid droplets. Finally, the released GA activated the SIRT1/AMPK pathway, promoting lipophagy and reducing excess triglycerides. This system also increased GSH and GPX4 activity while lowering MDA accumulation, thereby alleviating hepatic lipotoxicity through multiple mechanisms. Notably, PGOD NPs showed concentration- and time-dependent H₂O₂-responsive fluorescence recovery, strong lipid-droplet colocalization (Pearson’s r = 0.96), and a 30.6% reduction in intracellular TG; however, formal quantitative evaluation of imaging sensitivity and image–therapy correlation remained limited.⁸²

Clinically, a trial of SinaCurcumin^® (nanomicellar curcumin) reported reduced serum transaminase levels in patients with MASLD, providing preliminary human support for the micellization strategy.⁸³

However, a practical limitation of polymeric micelles is that critical micelle concentration (CMC)-related dilution can compromise micellar integrity after administration; when polymer concentration falls below the CMC, micellar disassembly and premature drug release may occur, although this also depends on kinetic stability and drug–polymer interactions.⁷⁷ Clinical translation is also limited by reproducibility and scale-up, because commonly used preparation methods such as dialysis and thin-film hydration are process-sensitive and may lead to batch-to-batch inconsistency during large-scale production.⁸⁴

Polymeric Nanoparticles

Polymeric nanoparticles (PNPs) are solid-state nanocarriers made from natural or synthetic polymers, primarily existing as nanospheres or nanocapsules.^85–87 PNPs possess advantages such as good biocompatibility, controllable drug release kinetics, and easy functional modification. By selecting different polymer materials and fabrication strategies, PNPs can achieve diverse design goals, from drug solubilization and efficacy enhancement to precise drug targeting.^88,89

Gut microbiota dysbiosis plays a critical role in the progression of MASLD. Consequently, delivering drugs precisely to the colon to modulate the microecology has emerged as an effective therapeutic approach. A prominent example is the colon-targeted delivery system for Luteolin.⁹⁰ Researchers encapsulated luteolin within an mPEG-PLGA core and subsequently coated the surface with the pH-sensitive anionic polymer Eudragit S100 to form core-shell nanoparticles (Lu-NPs). The Eudragit S100 shell minimized drug release in the gastric (pH 1.2) and small-intestinal (pH 6.8) environments but dissolved under colonic pH (pH 7.4) conditions, thereby enabling colon-directed release. This strategy modulated the gut microbiota and repaired the intestinal barrier, producing therapeutic effects through the gut–liver axis.^90,91

In addition, alternative strategies focus on achieving efficient hepatic accumulation of drugs to address the core injury. Yu et al developed a multifunctional nanosystem (GCNp-Cur NPs) for the delivery of Curcumin (Cur).⁹² This system utilized chitosan oligosaccharide as the backbone and integrated NAC and GA. Specifically, GA promoted hepatocyte-targeted liver accumulation of the system through GA-receptor recognition, whereas the introduction of NAC endowed the carrier itself with antioxidant properties. Thus, the carrier itself participated in therapy rather than serving only as a passive matrix. This system exhibited particular value in ferroptosis-oriented intervention. NAC not only served as a precursor of GSH but also directly chelated excessive intracellular Fe²⁺, thereby suppressing the Fenton reaction and ROS burst at the source, while curcumin enhanced GPX4 expression. Direct evidence for this ferroptosis-oriented effect was provided by restoration of the GPX4/GSH axis together with reduced iron burden and oxidative/lipid-peroxidative stress, as well as protection against erastin-induced ferroptotic injury, supporting inhibition of ferroptosis-related hepatocellular damage in MASLD.⁹²

Natural-polymer-based nanocomplexes have also been explored. For example, nanocomplexes constructed from gum arabic and xanthan gum improved curcumin solubility and prolonged intestinal retention through mucoadhesion, thereby enhancing absorption. These systems were associated with downregulation of hepatic CD36 expression and suppression of HMGB1-related inflammatory signaling, helping to limit progression from steatosis toward fibrosis.^93,94

An important limitation of biodegradable PNPs is that degradation kinetics can affect safety as well as release behavior; in PLGA-based systems, hydrolysis may generate an acidic microenvironment that accelerates degradation and has been associated with reduced cell viability and pro-inflammatory macrophage responses.⁹⁵ Therefore, material safety issues need to be considered in translational applications.

Other Polymer-Based Nanocarriers

Other polymeric nanostructures, such as nanogels and polymeric vesicles (polymersomes), have also demonstrated unique advantages in improving the delivery of active ingredients from TCM due to their distinct structural characteristics.

Polymeric nanogels are swollen, three-dimensional network structures formed by the physical or chemical cross-linking of hydrophilic polymers.⁹⁶ This hydrated network supports high drug loading and controlled release. During delivery, network swelling enlarges aqueous channels and promotes drug diffusion, whereas matrix degradation further releases entrapped cargo. Because both processes are modulated by cross-link density and the ionic environment, they directly affect intracellular release kinetics and delivery efficiency.⁹⁷ For example, Mauri et al constructed PEG/PEI nanogels (HT-NGs) for delivering hydroxytyrosol (HT), with a particle size of approximately 250 nm and a drug loading of 83%. The PEI component conferred a positive charge that enhanced membrane interaction and intracellular delivery, while the nanogel network enabled sustained intracellular HT release for up to 24 h, resulting in greater efficacy against hepatocyte lipid accumulation and lipotoxicity than free HT.⁹⁸

Polymersomes are hollow spheres formed by the self-assembly of amphiphilic block copolymers. They possess a bilayer membrane structure similar to liposomes and can simultaneously encapsulate both hydrophilic and hydrophobic molecules, while typically exhibiting greater mechanical stability.^96,99 To address the extremely poor solubility of oridonin (ORI), Zhang et al first complexed ORI with hydroxypropyl-β-cyclodextrin to resolve its solubility issues and then co-assembled the inclusion complex with H9 peptide and miltefosine to form nanovesicles. This system achieved a high drug loading of 63.60%, and the introduction of HePC significantly enhanced the cellular endocytosis of the vesicles. In vivo experiments confirmed that these nanovesicles effectively accumulated in the liver, significantly reducing the formation of lipid droplet vacuoles and collagen deposition in MASLD models.^100,101

Despite their structural versatility, further development of these polymer-based carriers for MASLD still requires reproducible formulation quality, evaluation of polymer-related safety, and regulatory assessment of physicochemical characterization, batch consistency, stability, and manufacturing standardization.¹⁰²

Inorganic Nanocarriers

Inorganic nanocarriers, including metal/metal oxide nanoparticles (such as CeO₂ and ZnO), carbon-based materials (such as graphene), and elemental nanoparticles (such as selenium and silicon), have attracted significant attention due to their high specific surface area, controllable pore size, and unique physicochemical properties.^103–105However, their application in MASLD treatment requires particular caution. Some materials, such as certain metal-based and carbon nanotubes may aggravate hepatic oxidative stress or inflammatory injury.^106–108 Beyond intrinsic toxicity, incomplete biodegradation or long-term tissue retention may further create chronic safety uncertainty.¹⁰⁹ Therefore, the biosafety of the carriers is the primary prerequisite for their application in this field. Currently, researchers have developed two main strategies to maximize advantages while mitigating risks.

One strategy involves the modification of potentially toxic carriers via “green synthesis”. For example, Abdelmoneim et al developed naringenin-loaded reduced graphene oxide nanosheets (Nar-RGO), in which naringenin served both as the therapeutic cargo and as a green reducing agent to convert GO into lower-toxicity RGO. Owing to the high surface area and interaction capacity of RGO, the system achieved a drug loading of 68% and showed stronger suppression of SREBP-1c/FAS-driven lipogenesis and inflammatory cytokines than free naringenin in MASLD models.¹¹⁰

Another strategy is to exploit intrinsically bioactive inorganic carriers so that both the carrier and the drug contribute to therapy. For example, selenium is an essential human trace element and a powerful antioxidant.¹¹¹ Based on this, galactose-modified mesoporous selenium nanoparticles (GA-MSe) were constructed to deliver arctiin (AR). The galactose moiety was introduced to promote ASGPR-mediated hepatocyte targeting, while the mesoporous selenium structure improved AR loading; after cellular uptake, GA-MSe@AR showed transient lysosomal accumulation followed by efficient escape into the cytoplasm. In addition, selenium release promoted antioxidant selenoprotein-related activity and, together with AR, reduced intracellular ROS, protected mitochondrial integrity, and inhibited the IGF1/PI3K/AKT axis, thereby suppressing oxidative stress and excessive lipid synthesis.¹¹¹

A related carrier-drug cooperative strategy was also observed in luteolin-zinc oxide nanoparticles (Lut/ZnO NPs), in which zinc ions exerted insulin-mimetic effects and, together with luteolin, upregulated the phosphorylation levels of the hepatic insulin signaling pathway (IRS/PI3K/AKT) and suppressed FoxO1-driven gluconeogenesis, while also reducing SREBP-1c-mediated lipogenesis.¹¹² Furthermore, CeO₂ hollow mesoporous nanocarriers leverage the cyclic redox conversion between Ce³⁺/Ce⁴⁺ to scavenge ROS, thereby augmenting the therapeutic effect of resveratrol against MASLD.¹¹³

Bio-Derived and Bionic Nanoscale Systems

In addition to engineered nanocarriers, an emerging frontier in MASLD nanomedicine involves utilizing or mimicking nature’s own transport systems. These bio-derived and biomimetic strategies typically exhibit superior biocompatibility, inherent bioactivity, and low immunogenicity.^114–116

Endogenous Carriers and Natural Nanovesicles

Among these bio-derived strategies, the most direct approach is to utilize naturally occurring transport components. These mainly include endogenous biomolecules (such as proteins) and cell-secreted nanovesicles.

Endogenous Protein Carriers

Among the many endogenous components, serum albumin (especially human serum albumin and bovine serum albumin) has emerged as an ideal drug delivery platform. As early as 2005, the albumin-based nanodrug Abraxane^® (albumin-bound paclitaxel) was approved by FDA for clinical use.¹¹⁷ As the most abundant protein in plasma, albumin possesses excellent biocompatibility, biodegradability, and low immunogenicity. Furthermore, it inherently functions to transport hydrophobic molecules, such as fatty acids and hormones, as well as drugs.^118,119 Its three-dimensional structure provides high-affinity binding sites, such as Sudlow sites I and II, for hydrophobic drugs. However, these binding interactions are not unlimited, because albumin contains only a finite number of major drug-binding pockets and different cargos show unequal binding affinities.¹²⁰ Meanwhile, the recycling mechanism mediated by the neonatal Fc receptor confers a long circulatory half-life of approximately 19 days. Furthermore, albumin can specifically bind to the gp60 receptor and SPARC protein. This capability allows it to naturally accumulate in inflamed and diseased tissues.^117,118

To address the bottlenecks hindering the clinical translation of celastrol in MASLD treatment, specifically its poor water solubility and systemic toxicity,¹²¹ Fan et al designed a lactobionic acid-modified bovine serum albumin (Lac-BSA) nanosystem.¹²² This system exploited the natural hydrophobic drug-carrying capacity of albumin while achieving ASGPR-mediated liver targeting through surface lactobionic acid ligands. As a result, it enhanced the anti-steatotic and anti-fibrotic effects of celastrol while reducing non-target exposure, and was associated with activation of the AMPK/SIRT/FAS/SREBP1c axis together with upregulation of fatty-acid-oxidation-related genes such as Acox-1 and Cpt-1.¹²²

Similarly, the albumin-based nanomedicine of ginsenoside compound K (nabCK) achieved selective hepatic accumulation by leveraging the natural distribution characteristics of albumin. Multi-omics analysis suggested that nabCK alleviated lipotoxicity mainly through mTOR inhibition, thereby reducing FASN-associated lipogenesis, limiting lipid storage, and enhancing APOB-related hepatic lipid export. This remodeling of lipid homeostasis attenuated steatosis and fibrosis and was accompanied by an improved systemic lipid profile.¹²³

Despite these advantages, albumin-based loading is constrained by the finite number of major binding sites and by cargo-dependent binding affinity, so loading capacity is not universally high across different drugs.¹²⁰ Moreover, because endogenous ligands and loaded drugs may compete for or perturb albumin binding, high loading efficiency does not necessarily guarantee stable drug retention, which can limit formulation flexibility for some cargos.¹²⁴

Biogenic Nanovesicles

Bio-derived nanovesicles (BNVs) are natural lipid bilayer vesicles actively secreted by the living cells of organisms, such as plants, mammals, or microorganisms. They primarily include exosomes and other extracellular vesicles (EVs).^125,126 They carry bioactive substances derived from parent cells, such as proteins, lipids, and nucleic acids. These contents constitute the foundation of inter-kingdom cellular communication.¹²⁷ These natural nanosystems possess excellent biocompatibility and the capability for efficient transport across biological barriers. Moreover, their inherent bioactive contents often exhibit defined pharmacological activities. Consequently, they represent promising bioactive nanosystems for MASLD.

Nanovesicles derived from various TCM sources, such as tea,¹²⁸ Artemisia capillaris,¹²⁹ garlic,¹³⁰ Pericarpium Citri Reticulatae,¹³¹ honey,¹³² and honeysuckle,¹³³ have shown therapeutic activity in preclinical MASLD models, demonstrating the unique advantage of multi-target intervention in the MASLD pathological network. For example, tangerine-derived nanovesicles (TNVs) not only remodeled gut microbiota and restored intestinal barrier integrity but also promoted short-chain fatty acids (SCFAs) production and activated intestinal Farnesoid X receptor (FXR). This leads to increased FGF15/19 release and upregulated hepatic SHP expression to inhibit CYP7A1 activity, coupled with the regulation of bile acid transporters (eg, BSEP and NTCP), thereby comprehensively restoring bile acid homeostasis In parallel, TNVs downregulated gluconeogenic and lipogenic genes while enhancing fatty-acid-oxidation-related programs, ultimately alleviating IR and hepatic lipid accumulation.¹³¹

Furthermore, certain TCM-derived vesicles target the hepatic inflammatory and immune microenvironment. For instance, Xu et al isolated nanovesicles (ACDEs) from the traditional hepatoprotective herb Artemisia capillaris, and these ACDEs exhibited dual activity in MASLD models. Besides downregulating key lipogenic genes, they effectively inhibited NF-κB activation in hepatocytes and macrophages, as reflected by reduced phosphorylation of IKKβ and p65 and decreased production of TNF-α, IL-6, and IL-1β, thereby alleviating hepatic inflammatory burden. These effects may be related to the abundant small RNAs carried by ACDEs, such as gma-miR5368.¹²⁹ Honey vesicle-like nanoparticles (H-VLNs) exhibited a deeper anti-inflammatory mechanism. With natural tropism toward hepatic Kupffer cells, their bioactive cargos, including miR-5119, miR-5108, and luteolin, coordinately suppressed C-JUN/NF-κB signaling and NLRP3 inflammasome activation, thereby limiting the chronic inflammation that drives fibrosis at its source. This was accompanied by reduced expression of fibrosis-related genes such as Col1a1, Timp1, and Mmp13, together with decreased HSC activation and attenuation of fibrosis.¹³² Furthermore, Garlic-derived exosomes (GDEs) demonstrated a distinct immunometabolic mechanism. After uptake by macrophages, vesicular miR-396e directly inhibited the mRNA expression of the glycolytic enzyme PFKFB3, thereby reducing energy supply required for excessive M1-type inflammatory responses. As a result, IL-1β and TNF-α release was decreased, and hepatocyte lipogenic genes such as FAS and ACC1 were indirectly downregulated through macrophage-hepatocyte crosstalk.¹³⁰

Beyond their native forms, BNVs can be further engineered to enhance therapeutic efficacy. A distinctive engineered example is pullulan-modified exosomes loaded with naringenin (Pul-Exos@NGN). In this design, pullulan served not only as a hepatocyte-directed ligand through lectin-like receptor recognition, but also increased vesicle size, thereby favoring caveolae-mediated uptake after intravenous administration and helping reduce Kupffer-cell-dominated sequestration. Mechanistically, Pul-Exos@NGN promoted lipid-droplet turnover by enhancing ubiquitin–proteasome-system-dependent removal of the lipid-droplet coat protein PLIN2 together with lipophagy flux, thereby alleviating hepatic lipid accumulation.⁴¹ Fucoxanthin-loaded Lactobacillus paracasei-derived EVs (LpEVs) likewise utilized the bacteria-derived lipid membrane structure to improve the stability of the hydrophobic cargo and, after glycyrrhetinic-acid modification, achieved liver-targeted delivery with downregulation of the SREBP-1/ACC1/FAS pathway.¹³⁴ Another strategy is donor-cell preprogramming: exosomes derived from curcumin-pretreated mesenchymal stem cells (MSCs/Exo-Cur) showed an altered cargo profile and were confirmed to downregulate the ASK-1/JNK/BAX apoptotic pathway, inhibit HSC activation, and exert sustained anti-MASH effects for up to 3 months.¹³⁵

Despite their biological advantages, BNVs remain constrained by marked source- and process-related heterogeneity. Vesicles derived from different organisms cells can differ substantially in composition, bioactivity, and tropism, while commonly used isolation methods such as ultracentrifugation, precipitation, size-exclusion chromatography, or tangential-flow filtration may yield major differences in purity, recovery, production yield, and batch reproducibility.^136,137 Moreover, dosing metrics are not standardized across studies, because EV preparations are variably reported by particle number, protein amount, and/or other abundance-related measures, which complicates cross-study comparison and quality control.¹³⁶ Accordingly, further translation will require clearer characterization standards, GMP-compatible scalable manufacturing, and more explicit regulatory classification.^137,138

Bionic Assembly and TCM-Derived Nanodrugs

In contrast to directly utilizing natural components, another advanced strategy involves constructing “carrier-free” nanodrugs through biomimetic self-assembly or the direct nanonization of TCM ingredients. These approaches minimize the use of exogenous materials and can improve delivery efficiency while preserving the intrinsic bioactivity of TCM-derived components.^139,140 One important strategy is the spontaneous self-assembly of TCM ingredients into nanoparticles through the cooperative balance of non-covalent interactions, including hydrophobic interactions, hydrogen bonding, electrostatic forces, π–π stacking, and van der Waals interactions. In aqueous systems, assembly is favored when these interactions drive the molecules into a more stable state, often by clustering hydrophobic regions away from water.^115,139

For example, celastrol and HA were reported to spontaneously co-assemble into nanoparticles (CHNPs) in aqueous solution through supramolecular interactions. This HA-templated assembly improved the solubility and stability of celastrol and, because HA is a natural ligand of CD44, enabled targeting of hepatic non-parenchymal cells. In experimental models, CHNPs inhibited LPS-induced M1 macrophage polarization, significantly alleviated hepatic lipid deposition and fibrosis, and also improved obesity, IR, and leptin resistance. Proteomic analysis further suggested involvement of the PPAR signaling pathway and CD36 regulation.¹⁴¹

Furthermore, TCM components themselves can be directly converted into functional nanomaterials with intrinsic therapeutic activity. Researchers synthesized highly water-soluble carbon dots (HCDs) directly from the TCM Hawthorn using a green hydrothermal method. HCDs inherited and amplified the bioactivity of Hawthorn. Consequently, they exhibited significant intrinsic antioxidant and anti-inflammatory capabilities. In high-fat diet-induced obese mouse models, it directly alleviates hepatic lipid accumulation and improves glucose metabolism. Simultaneously, it effectively remodels the gut microbiota structure, thereby it exerts therapeutic effects indirectly by regulating the gut-liver axis.¹⁴⁰

Despite their carrier-free design, further translation of these systems remains challenging, because the final nanostructure can be affected by assembly conditions and source materials, making precise composition characterization, structural homogeneity, and batch reproducibility during scale-up difficult to standardize.^115,139 (Table 1)

Table 1 Overview of Literature Information on TCM-Derived Nanoparticles for MASLD Treatment

Conclusion and Future Perspectives

This review summarizes current advances in TCM-derived nanoformulations for MASLD by integrating nanoplatform classification with a pathology-oriented overview of therapeutic mechanisms. As shown in Figure 3, current nano-TCM systems span engineered, bio-derived, and carrier-free platforms with distinct design logics. Figure 4 further maps representative mechanistic findings from these studies onto the major pathological processes of MASLD, providing an overview of how current nanoformulations may intervene in hepatic lipid dysregulation and hepatocyte stress, inflammatory and immune remodeling, gut–liver-axis dysfunction, fibrosis, and broader metabolic disturbances.

Figure 3 Classification of traditional Chinese medicine nanoparticles for the treatment of MASLD. (Figure created by the authors.).

Figure 4 Multi-target mechanisms of TCM-derived nanomedicine in treating MASLD. (This schematic illustrates the therapeutic mechanisms of various TCM-derived nanoformulations across the pathological progression of MASLD, highlighting both shared and distinct pathways. For clarity, all nanoformulations are represented by a uniform spherical symbol rather than their specific nanostructures. Green upward arrows indicate promotion, enhancement, or upregulation of biological processes/targets, while red downward arrows denote inhibition, reduction, or downregulation. Panel (A) Regulation of hepatic lipid metabolism and alleviation of hepatocyte stress and death. Panel (B) Remodeling of hepatic inflammation and the immune microenvironment. Panel (C) Modulation of gut microbiota and intervention in the gut-liver axis. Panel (D) Inhibition of liver fibrosis and improvement of systemic metabolic disorders. Figure created by the authors.).

Taken together, different nanoplatforms appear to offer distinct but complementary strengths. Engineered systems are advantageous when improved drug loading, controlled release, or active targeting is required,^47,88,89 whereas bio-derived, biomimetic, and carrier-free systems may offer benefits in biocompatibility, intrinsic bioactivity, microenvironmental modulation, and reduced reliance on exogenous excipients.^114,115

However, their translation remains limited by unresolved issues in long-term safety, in vivo fate, immunogenicity, scalable manufacturing, batch consistency, and quality control, while most current evidence still comes from small-animal models that do not fully reflect the heterogeneity and stage complexity of human MASLD.^{106,110,137,138}

Looking ahead, further development of nano-TCM strategies for MASLD remains important, because these systems can improve the utilization efficiency of TCM-derived active ingredients, overcome limitations in solubility, bioavailability, and tissue specificity, and thereby more effectively intervene in the multiple interconnected pathological processes of MASLD. Future research should therefore focus on developing more precise and clinically relevant nanoformulations. In this context, a “reverse design” strategy should be advanced, in which nanoplatforms are selected or tailored according to clinically relevant requirements, such as disease stage, dominant pathological process, target cell population, route of administration, safety, and manufacturability. At the same time, advanced tools such as organoids and spatial transcriptomics should be used to deepen mechanistic studies and clarify the cell-specific and immunometabolic actions of nano-TCM systems in MASLD.¹⁵⁰ In addition, interdisciplinary collaboration should be strengthened to connect platform design with standardized production and clinical evaluation.¹⁵¹

Overall, TCM-derived nanoformulations show broad mechanistic potential in MASLD, but their clinical value will depend on translating that potential into safe, standardized, and clinically relevant therapies.