Introduction

Hepatic fibrosis represents a pivotal pathological stage in the progression from chronic hepatitis to cirrhosis and hepatocellular carcinoma.1 This condition stems from abnormal proliferation of connective tissue in response to sustained liver injury, with major etiological factors including viral infections,2 alcohol/drug toxicity,3 metabolic disorders,4 non-alcoholic fatty liver disease (NAFLD),5 and autoimmune conditions.6 Liver fibrosis occurs during nearly all forms of persistent hepatic injury. If left untreated, chronic fibrosis progresses irreversibly to cirrhosis, with subsequent risk of hepatocarcinogenesis.7 Epidemiologically, cirrhosis affects 1%-2% of the global population, accounting for approximately one million deaths annually.8 As a highly prevalent complication of chronic liver disease, most untreated hepatitis cases eventually develop fibrotic manifestations.9 Clinically, hepatic fibrosis is characterized by insidious onset, potential early-stage reversibility, and multifactorial pathogenesis involving hepatic stellate cells (HSCs) activation, extracellular matrix (ECM) dysregulation, oxidative stress, inflammatory pathways, and aberrant angiogenesis.10,11 Recent studies have demonstrated that key molecules, including galectin-3 and discoidin domain receptor 1 (DDR1), play pivotal roles in liver fibrosis pathogenesis.12,13 Given this complex mechanistic interplay, research into fibrogenic triggers and therapeutic interventions remains profoundly important. These molecules activate HSCs and drive fibrotic progression by modulating immune-inflammatory responses. Notably, targeted inhibition of galectin-3 or genetic ablation of DDR1 attenuates hepatic inflammation and fibrosis, suggesting their potential as therapeutic targets.

The emergence of nano-medicine has brought the dawn of liver fibrosis treatment. Many researches have been devoted to using nanotechnology to address the treatment barriers in liver fibrosis.14,15 MNMs are becoming a research hotspot because of their unique physical and chemical properties.16,17 However, there is a dual role in the in the treatment of liver fibrosis.18,19 On one hand, high surface-to-volume ratios, tunable morphology, size controllability, and versatile surface functionalization- that render them highly advantageous for liver fibrosis diagnosis and treatment.20 As a prominent subset of nanotechnology, these materials demonstrate exceptional utility across diverse applications in energy, electronics, catalysis and biomedicine due to their unique photonic, electronic, and biological properties.21–23 Within hepatic fibrosis therapy, they enable targeted diagnostics, precision drug delivery, gene therapy, and photothermal interventions.24–27 However, metallic nanomaterials present a concerning duality.28 When introduced systemically (such as, via intravenous or oral routes), exogenous nanomaterials like titanium dioxide nanoparticles (TiO2 NPs),29 copper nanoparticles (CuNPs),30 and nickel oxide nanoparticles (NiONPs)31 can trigger fibrogenesis in healthy murine models by activating transforming growth factor-beta (TGF-β1) pathways. More critically, in hepatitis-compromised livers with impaired detoxification capacity, certain nanomaterials may accelerate fibrotic progression. For instance, surface-modified gold nanorods (AuNRs) exacerbate fibrosis in hepatitis models by inducing macrophage polarization,32 while gold nanoparticles (AuNPs) and silver nanoparticles (AgNPs) aggravate hepatic injury via oxidative stress and pro-inflammatory responses, ultimately promoting fibrogenesis.33,34 Consequently, elucidating the dual modulation of hepatic fibrosis by MNMs is imperative for optimizing their therapeutic design and clinical translation.

This review systematically evaluates the unique properties of MNMs and their dual roles- both therapeutic/diagnostic and fibrogenic- in hepatic fibrosis progression. We comprehensively analyzed the physical and chemical properties (including size, morphology, chemical composition and surface modification) of MNMs that determine their dual effects. And then, the different applications of MNMs for the treatment of liver fibrosis and the mechanism of MNMs-induced liver fibrosis were discussed. Finally, the current status of clinical translation of MNMs, the barriers facing them, and potential strategies to optimize their clinical use in fibrosis management are highlighted.

MNMs

MNMs are defined as metallic or composite structures with at least one dimension confined to the 1–100 nm scale. As a foundational domain of nanotechnology, these materials exhibit exceptional potential across diverse fields including energy systems, electronics, catalysis, and biomedicine, owing to their unique physicochemical and biological properties (Figure 1)18,35 The defining characteristics of MNMs predominantly stem from nanoscale-induced phenomena:36

Figure 1 Typical MNMs and their applications.

Enhanced Surface Reactivity

The nanoscale dimensions confer extraordinarily high specific surface areas (typically hundreds to thousands of square meters per gram), with surface atoms comprising 40%-80% of total atoms in 2 nm particles.37 This atomic configuration, characterized by reduced coordination numbers and abundant unsaturated bonds, substantially improves catalytic efficiency and adsorption capacity.38

Quantum Confinement Effects

Discrete electronic energy states emerge as continuous bands transition to quantized levels, producing distinctive optical (such as, surface plasmon resonance in AuNPs), magnetic (such as, single-domain ferromagnetic behavior), and electronic (such as, quantum tunneling resistance) properties.39

Size-Dependent Thermodynamic Properties

Nanoscale materials exhibit depressed melting points (such as 600 K for 2 nm gold particles compared to 1337 K for bulk gold) while simultaneously displaying superior mechanical characteristics, including exceptional fracture strength (1300 MPa in nanostructured steel) and ductility (>500% elongation in aluminum nanocomposites).40–42

Tunable Material Behavior

Precise property modulation is achievable through compositional engineering, morphological control (such as, facet-specific catalysis), and surface functionalization strategies.43–45

Multifunctionality Integration

These materials uniquely combine properties such as high electrical/thermal conductivity (such as, copper nanopastes for interconnects), magnetic responsiveness (such as, iron oxides for data storage), and biological targeting capability (such as, ligand-conjugated AuNPs), enabling sophisticated multifunctional designs.46–48

MNMs are systematically categorized into four classes based on composition: pure metals, metal oxides, metal sulfide/selenide/telluride, and metal composite nanomaterials, each exhibiting unique physicochemical properties.

Metallic nanomaterials, particularly noble metal nanoparticles (such as, Au, Ag, Pt, Pd), demonstrate exceptional electrical conductivity and surface plasmon resonance, with optical properties tunable via quantum confinement effects.49 These traits originate from discrete electronic states, crystalline lattice arrangements, and under coordinated surface atoms that provide catalytically active sites.50

Metal oxide nanomaterials feature ionic-bonded crystalline structures where nanoscale confinement generates oxygen vacancies (such as, TiO2 defect sites), enhancing photocatalytic activity.51 Lattice strain (such as, in ZnO nanowires) further promotes charge carrier separation, optimizing optoelectronic performance.52 Such characteristics enable applications ranging from photocatalytic degradation (such as, TiO2-mediated pollutant breakdown) to biomedical imaging (such as, Fe3O4-based MRI contrast enhancement).51,53

Metal sulfide/selenide/telluride, which is a compound formed by metal and chalcogenide elements (S, Se, Te). It is mainly a semiconductor with significant quantum confinement effect, and is mostly used in photoelectric and catalytic fields.54

Metal composite nanomaterials are composed of two or more metal/metal compounds (alloys, heterostructures), which have both component characteristics and synergistic effects, and strong structural controllability.55–57

Each class differentially influences hepatic fibrosis progression via distinct mechanisms. The following sections will explore the mechanisms underlying the dual effects of MNMs from the perspective of the physicochemical properties, and critically evaluate the dual roles of these nanomaterials in fibrosis theranostics and pathogenesis.

The Dual Roles of MNMs: Mechanistic Regulation via Physicochemical Properties

The dual effect is primarily governed by the intrinsic physicochemical properties of nanomaterials, including size, morphology, chemical composition, and surface modifications. These properties modulate cellular uptake,58 subcellular localization,59 biological activity,60 and toxicity61 —factors that collectively dictate whether nanomaterials exert therapeutic or adverse effects in the hepatic fibrotic microenvironment.

Size

The “size effect” of nanomaterials serves as a fundamental determinant of their catalytic performance. At the nanoscale, materials exhibit a dramatic increase in specific surface area, leading to a substantial rise in the proportion of surface atoms.62 These surface atoms display unsaturated coordination states with elevated surface energy and reactivity, thereby forming catalytic active sites.63,64 Systematic studies confirm that—within a defined size range—metal-based nanozymes with smaller dimensions possess greater numbers of exposed active sites and generally demonstrate enhanced catalytic activity. This catalytic activity plays an important role in the dual role in liver fibrosis.

The size of nanomaterials is a critical determinant of their in vivo biodistribution and clearance mechanisms. Studies show that nanoparticles smaller than 10 nm (especially < 5 nm) typically exhibit a short circulation half-life due to rapid renal clearance, as their small size enables evasion of the reticuloendothelial system (RES) and permeation through glomerular filtration.65 While this minimizes long-term accumulation risks, it may also limit sustained enrichment at hepatic lesion sites.66 Conversely, larger nanoparticles (> 200 nm) are prone to phagocytic uptake—particularly by liver-resident Kupffer cells (KCs)—leading to non-specific hepatic accumulation.67 If this accumulation surpasses the metabolic capacity of hepatocytes, it may induce persistent inflammation, activating quiescent HSCs and subsequently exacerbating fibrogenesis (Figure 2).68

Figure 2 MNMs’ size, morphology, composition, and surface modification regulate the dual roles in liver fibrosis.

Nanoparticles sized 10–200 nm represent an optimal range for liver-targeted therapies.69 This size range enables both accumulation in injured liver tissue via the enhanced permeability and retention (EPR) effect and efficient uptake by hepatocytes and activated HSCs.70 However, this selective uptake presents dual consequences: while facilitating targeted delivery of therapeutic agents (eg, antifibrotic drugs), excessive intracellular accumulation—particularly when nanomaterials resist degradation in endosomal/lysosomal compartments—can induce lysosomal membrane permeabilization.71 Subsequent lysosomal rupture releases proteolytic enzymes and acidic contents into the cytosol, activating the NLRP3 inflammasome pathway, triggering pyroptotic cell death, and stimulating proinflammatory cytokine (eg, IL-1β) release.72 This cascade has been identified as a critical mechanism linking nanomaterial exposure to profibrotic inflammatory responses.73 Consequently, nanomaterials design must carefully balance targeting efficacy with avoidance of cellular overload.74 However, until now, quantitative comparative studies examining how different sizes of MNMs affect HSCs activation and fibrosis progression in animal models remain scarce. Establishing precise size-effect-toxicity correlations and defining the therapeutic window for specific nanomaterials represent critical scientific challenges in this field.

Morphology

The morphology of nanomaterials (eg, spherical, rod-like, plate-like, or flower-like structures) directly governs the relative exposure of different crystal facets.75 Each facet exhibits unique atomic arrangements, surface atomic densities, and surface energies, resulting in substantial variations in catalytic performance.76,77 For instance, nanocrystals enclosed by high-index facets typically demonstrate superior catalytic activity compared to those with low-index facets, attributable to their higher density of steps, edges, and other under coordinated atomic sites.78 Furthermore, nanostructures with porous or hollow architectures (eg, nanoflowers or nanocages) not only offer enlarged specific surface areas but also contain internal cavities that mimic the active pockets of natural enzymes, concentrating substrate molecules and consequently enhancing catalytic efficiency.79

In addition, compared with spherical nanoparticles, anisotropic structures such as nanorods, nanowires, or two-dimensional nanosheets/nanotubes with a high aspect ratio often have a stronger interaction with cell membranes due to their larger contact area and unique edge effects.80 This strong interaction can significantly enhance cellular uptake efficiency, but it also greatly increases the risk of directly physically damaging the cell membrane or lysosomal membrane.80 For instance, sharp edges may puncture membrane structures like “needles”, or exert greater mechanical tension on the endosomal membrane during endocytosis, leading to membrane integrity damage, triggering calcium ion influx and a series of cellular stress repair responses, which themselves are strong HSCs activation signals.81 In contrast, spherical nanoparticles with smooth surfaces typically exhibit lower cell membrane perturbation capabilities and better blood compatibility.82 Their internalization process is more gentle and mainly relies on the classical endocytic pathway, thereby reducing the risk of fibrosis directly induced by physical damage.

Chemical Composition

While nanomaterial size and morphology govern their cellular entry mechanisms, chemical composition dictate their subsequent biological interactions.83 These intrinsic chemical properties fundamentally determine their therapeutic efficacy and potential toxicity.

The elemental composition of MNMs constitutes the fundamental determinant of their biochemical activities.84–86 Noble metal nanoparticles (eg, AuNPs, AgNPs) are particularly valuable for photothermal therapy and bioimaging applications due to their chemical inertness and distinctive surface plasmon resonance (SPR).87 The superparamagnetic properties of iron oxide nanoparticles represent their most fundamental physical characteristic, enabling unique diagnostic applications.88 Their exceptional stability minimizes nonspecific biomolecular interactions, conferring high biocompatibility.

Furthermore, enzyme-mimicking (“nanozyme”) activity demonstrated by multivalent MNMs—particularly those containing redox-active elements such as cerium (Ce), iron (Fe), and manganese (Mn)—has emerged as a critically important therapeutic mechanism in nanomedicine.89,90 These materials can emulate natural enzymatic functions (eg, superoxide dismutase (SOD), catalase (CAT), peroxidase (POD)), enabling direct modulation of the oxidative stress microenvironment – a central pathological feature in hepatic fibrosis.89 For instance, cerium oxide nanoparticles (CeO2NPs) exhibit dynamic Ce3+/Ce4+ redox cycling, demonstrating potent reactive oxygen species (ROS) scavenging capacity and validated anti-fibrotic efficacy across preclinical liver fibrosis models.91 However, this enzymatic activity demonstrates marked environmental dependency, representing its crucial dual functionality. The same nanozyme may exert opposing effects in distinct biological microenvironments.92,93 Under physiological conditions (pH~7.4), they typically manifest antioxidant activity (eg, SOD-/CAT-like effects). However, when localized in acidic lysosomal compartments (pH 4.5–5.0) or inflammatory sites with elevated H2O2 concentrations, they can undergo functional switching to peroxidase-/Fenton-like activity, catalyzing hydroxyl radical (•OH) production.94 This dramatic transition from ROS elimination to ROS generation can rapidly intensify oxidative cellular damage, directly activating HSCs and triggering apoptotic/necrotic pathways – consequently converting potential therapeutic benefits into fibrotic exacerbation.95

For non-inert MNMs, their biodegradation behavior and metal ion release kinetics represent critical determinants of long-term biosafety.96 Ideally, biodegradable nanomaterials should degrade into either non-toxic constituents or metabolites that can be safely excreted upon therapeutic function completion.97 However, uncontrolled metal ion release may induce “ion toxicity”, constituting a primary mechanism for nanomaterial-induced secondary toxicity.98 Numerous studies demonstrate that biologically active metal ions (eg, Cu2+, Ag+, Cd2+, Ni2+) readily bind to crucial protein functional groups, particularly sulfhydryl (-SH) groups, leading to enzyme inactivation and protein denaturation.99 Copper ions (Cu2+) exhibit particularly pronounced effects in this context.100 Substantial evidence confirms that Cu2⁺ released from copper-based nanomaterials can directly activate hepatic stellate cells (HSCs) and exacerbate fibrogenesis by triggering canonical fibrotic pathways, including TGF-β/SMAD signaling.101 Furthermore, these unregulated metal ions disrupt mitochondrial electron transport chains, causing mitochondrial dysfunction and excessive ROS generation.102 This establishes a vicious cycle of “ion toxicity-oxidative stress,” ultimately resulting in cellular necrosis, robust inflammatory responses, and accelerated fibrosis progression. Consequently, precise modulation of nanomaterial degradation kinetics is paramount.103

Current research focuses on developing stimulus-responsive “smart” nanomaterials capable of microenvironment- triggered degradation (eg, in acidic or high glutathione (GSH) conditions). Concurrently, accurate ion release quantification has become essential for safety assessment. However, a critical challenge persists: the absence of standardized protocols for simulating hepatic microenvironments in degradation and ion release studies. This methodological inconsistency impedes cross-study comparisons and complicates comprehensive safety evaluations.

Surface Modification

The surface serves as a critical interface mediating interactions between nanomaterials and biological systems. Through controlled surface chemical modification, researchers can systematically engineer nanomaterial biological identity, enabling precise modulation of their interactions with proteins, cells, and tissues.104 These engineered modifications ultimately determine whether nanomaterials elicit therapeutic effects or toxic responses.105

Surface charge represents a critical parameter governing nanomaterial stability and cellular interactions.106 Cationic surfaces exhibit strong electrostatic attraction to negatively charged cell membranes, facilitating cellular internalization.107 However, such highly positive charges may concurrently induce membrane destabilization, lysosomal disruption, and pronounced cytotoxicity, eliciting detrimental inflammatory responses that exacerbate liver fibrosis progression.108 In contrast, neutral or slightly anionic surfaces typically minimize nonspecific protein adsorption, thereby reducing immune recognition.109 Among surface modification strategies, polyethylene glycol (PEG)ylation remains the gold standard technique.110 PEG chains form a dense hydrophilic corona that sterically shields the nanomaterial core, substantially attenuating protein corona formation.111 This stealth effect prolongs blood circulation half-life while decreasing hepatic and splenic sequestration, thereby enhancing targeted delivery to fibrotic lesions.112 Such surface engineering effectively mitigates inflammation and fibrogenesis risks associated with nonspecific phagocytosis.113

To achieve selective delivery of antifibrotic agents to activated HSCs—the primary effector cells in fibrosis—various active targeting moieties have been explored. These typically include ligands against receptors overexpressed on activated HSCs, such as retinol (vitamin A) receptors,114 platelet-derived growth factor receptor β (PDGFRβ),115 integrin αvβ3,116 hyaluronic acid receptor (CD44)117 and chemokine receptor CXCR4.118 Emerging evidence demonstrates the therapeutic potential of targeted nanocarriers. Ligand-conjugated nanoparticles have successfully delivered pathway inhibitors to HSCs, significantly suppressing key fibrotic markers including α-SMA and Col1A1.102 Notably, engineered iron oxide nanoparticles (eg, FGF2-SPIONs) exhibit dual functionality—specific HSCs targeting coupled with direct TGF-β1 pathway modulation, thereby addressing fibrosis pathogenesis at the molecular level.119 Nevertheless, this targeted approach entails inherent risks. Active targeting strategies markedly increase nanomaterial accumulation within specific cell population. If material toxicity factors (eg, uncontrolled ion release or paradoxical enzymatic activity) remain unmitigated, or if drug payload exceeds therapeutic thresholds, such targeted accumulation may precipitate cellular overload. This could paradoxically induce apoptosis/necrosis in precisely targeted cells, exacerbating localized inflammation and tissue damage – outcomes diametrically opposed to therapeutic objectives.120

In conclusion, the dual nature of MNMs is not a stochastic phenomenon, but rather results from dynamic interactions between their physicochemical properties and the complex hepatic microenvironment. Comprehensive understanding and precise regulation of these properties are essential for developing safe and effective metal-based nanotherapeutics for fibrosis treatment.

MNMs for Hepatic Fibrosis Therapy and Diagnosis

Owing to their structural diversity and exceptional physicochemical properties, MNMs show significant potential for hepatic fibrosis therapeutics.121 The role of MNMs in the treatment of liver fibrosis can be summarized into four points: 1.Targeted Drug Delivery; 2. Natural Product-Metal Nanomaterial Synergistic Systems; 3. Nanozyme-Mediated Microenvironment Remodeling; 4. Diagnosis and Theranostic Platforms.

Targeted Drug Delivery Systems for Fibrosis Inhibition

Targeted drug delivery systems are crucial for improving therapeutic outcomes in liver fibrosis while minimizing off-target effects. Nanocarriers can achieve passive accumulation through the EPR effect and RES uptake, or active targeting via ligand-receptor interactions.122,123

Passive Target

The liver possesses a distinctive physiological architecture that enables passive nanoparticle accumulation via the EPR effect and KCs uptake.124,125 NPs’ size is a crucial determinant of passive targeting. The fenestrated endothelial structure of hepatic sinusoids (with 100–200 nm pores) and discontinuous basement membranes create a natural filtration system, enabling nanoparticles (10–200 nm) to extravasate into the Space of Disse.126 While optimal phagocytic efficiency of KCs occurs within the 70–200 nm range, as this size minimizes renal clearance (< 5 nm) while avoiding splenic sequestration (>200 nm), which primarily results from particle aggregation and subsequent macrophage uptake.127 Surface properties further modulate this interaction. Neutral or negatively charged nanoparticles exhibit preferential KCs uptake, while positive charges promote vascular endothelial binding.128,129 Moderate hydrophilicity balances opsonin adsorption and phagocytic efficiency without requiring active targeting ligands.129 Therefore, the size and surface charge of nanomaterials can be precisely tuned to enable passive targeting to the liver. MNMs are particularly advantageous for such applications, as they possess tunable physicochemical properties (eg, controlled size and surface charge), making them highly suitable for passive targeted delivery in hepatic fibrosis treatment (Table 1). For example, AuNPs exhibit natural tropism for KCs, enabling passive targeting that offers therapeutic advantages for fibrosis treatment.130 Carvalho and colleagues demonstrated that neutral AuNPs (7.4 ± 1.6 nm) preferentially accumulate in murine livers, where they inhibit the abnormal activation of KCs and HSCs by regulating the PI3K/AKT signaling pathway and MAPK signaling pathway, reduce the release of pro-inflammatory cytokines (IL-1β, TNF-α), and block the inflammation-fibrosis cascade.34 Moreover, AuNPs serve as effective drug carriers that enhance hepatic drug accumulation via passive targeting mechanisms, consequently improving therapeutic outcomes in liver fibrosis. Silymarin-conjugated gold nanoparticles (SGNPs) with a diameter of 16–20 nm and strong negative surface charge (−38.9 mV) demonstrated significantly enhanced hepatic accumulation compared to free silymarin. The drug delivery efficiency increased by 2–5 fold, achieving “high efficacy at low dosage” - the therapeutic dose of SGNPs was only 1/7 of free drug while showing superior therapeutic outcomes.119 The same enhanced delivery effect was observed for both 20 nm131 and 10 nm132 AuNPs in silymarin delivery systems. Similarly, 19 nm AuNPs effectively deliver Cornus sanguinea L. polyphenols to the liver, enhancing their therapeutic efficacy against liver fibrosis.133 Chitosan-curcumin-coated silver nanoparticles (AgNPs) measuring 23.7 ± 0.8 nm in diameter enable passive hepatic targeting of curcumin.134 Curcumin demonstrates comparable passive targeting efficacy when delivered via platinum nanoparticles.135 200 nm MnO2@PLGA/Ssb1 NPs can also be used to delivery saikosaponin b1 (Ssb1) to liver by passive targeting effect.29 Administration of MnO2@PLGA/Ssb1 significantly enhanced liver accumulation of Ssb1, showing a 38.7% increase compared with free Ssb1 and higher hepatic distribution than other organs.

Table 1 Passive Targeting of MNMs for Liver Fibrosis Therapy

Active Targeting

Compared to passive targeting, active targeting enables nanoparticle accumulation in specific organelles, cells, tissues, or organs through molecular recognition mechanisms.126 MNMs exhibit distinct advantages over other nanomaterial systems, including polymeric nanoparticles, liposomes, silica, and carbon-based materials. Notably, their well-defined surface chemistry—particularly in AuNPs (through Au–S bonding) and Fe3O4 (via coordination or hydroxyl functional groups)—facilitates efficient conjugation with a variety of targeting ligands, such as arginine-glycine-aspartic acid (RGD) peptides, glycyrrhetinic acid, and ligands specific to HSCs.136 Moreover, these nanomaterials support a high density of ligand immobilization, enabling the formation of a densely functionalized surface layer that enhances targeting precision and efficacy (Figure 3A).137 As a result, MNMs hold significant promise for active targeted delivery in the treatment of hepatic fibrosis (Table 2).

Table 2 Active Targeting of MNMs for HSCs and Hepatocytes in Liver Fibrosis

In hepatic fibrosis, activated HSCs represent the primary therapeutic target as they are the main ECM producers and undergo activation into myofibroblast-like cells upon injury.148,149 During liver injury, quiescent HSCs undergo activation, exhibiting significant phenotypic alterations and upregulation of multiple receptors (Figure 3B). PDGFRβ is among the most distinctive surface markers expressed on activated HSCs.150 Studies have shown that functionalizing nanocarriers with PDGFRβ-targeting peptides or aptamers can substantially improve drug delivery efficiency to HSCs.151 Ribera and colleagues achieved targeted delivery to HSCs by functionalizing the surface of AuNRs with anti-PDGFRβ ligands. The incorporation of PDGFRβ- specific ligands enabled a precise shift in delivery specificity from the organ level to the cellular level, resulting in a seven-fold increase in cellular uptake and a doubling of accumulation in the fibrotic liver.24 Concurrently, therapeutic efficacy was synergistically enhanced, with anti-fibrotic, hepatoprotective, and anti-inflammatory effects significantly surpassing those observed with non-targeted formulations or monotherapies. Furthermore, the optimized ligand density—approximately four ligands per AuNRs-struck an effective balance between specific binding affinity and preservation of nanomaterial functionality, thereby supporting robust targeting performance. This study demonstrates that surface functionalization of gold nanomaterials with targeting ligands is a pivotal strategy for enhancing both delivery efficiency and therapeutic outcomes, and that precise control over ligand density is essential for ensuring consistent and reliable targeting performance.24

Figure 3 Multiple strategies for achieving targeted delivery of functionalized MNMs in the context of liver fibrosis treatment, with a focus on the targeted interaction between MNMs and specific hepatocyte types. (A) The key components of MNM design, including the metallic core. (B) Multiple strategies for MNMs targeting activated HSCs in liver fibrosis. (C) Dual targeting of activated HSCs and LSECs via TfR1. (D) MNMs modified with galactose or GA targets toward hepatocytes.

Upon activation and transformation into myofibroblasts, HSCs exhibit significant upregulation of αvβ3 integrin expression, representing an optimal therapeutic target. The RGD peptide demonstrates high binding affinity for αvβ3 integrins, enabling active targeting when conjugated to nanomaterial systems. For example, the SPIO@SiO2–ICG–RGD probe demonstrates significantly prolonged hepatic retention following RGD conjugation, with near-infrared (NIR) signals persisting beyond 72 hours.139 Histological analysis reveals markedly stronger iron staining intensity in fibrotic livers compared to healthy controls, confirming the probe’s specific enrichment within fibrotic regions. Moreover, following RGD modification, the cellular uptake of PLGA nanoparticles by HSCs was significantly enhanced, thereby improving the therapeutic and diagnostic efficacy of encapsulated Fe3O4 and ferulic acid (FA).140 Cui et al developed an RGD-modified, metformin-loaded ferritin–platinum–manganese nanoplatform (FNMMR) for the precise targeting of activated HSCs.144 Fluorescence imaging revealed significantly stronger Cy5.5 signal intensity in LX-2 cells treated with FNMMR (RGD-modified) compared to its unmodified counterpart (FNM). Furthermore, late-stage fibrotic murine livers exhibited substantially higher fluorescence intensity than early-stage fibrotic or healthy controls, demonstrating enhanced activated HSCs-specific accumulation in progressive disease.

During the activation and transformation of HSCs into myofibroblast-like cells, the expression of hyaluronic acid (HA) receptors—such as CD44—markedly increases.152 Consequently, HA receptors represent another critical target for HSC-specific drug delivery.153 Lu et al synthesized carnosic acid (CAR)-hyaluronic acid (HA)-conjugated gold nanostars (CAR-HA-AuNS) via non-covalent interactions (eg, electrostatic and hydrophobic binding) to enable targeted CAR delivery to HSCs.138 The targeted uptake efficiency of HSCs in fibrotic liver towards AuNS@CAR-HA was 2–3 folds higher than that in healthy mice, while hepatic tissue enrichment efficiency was 1.5–2 folds greater. Compared to monotherapy using AuNSs or CAR-HA alone, AuNS@CAR-HA exhibited a 1.2–2 folds improvement in overall therapeutic efficacy, as evidenced by reduced fibrosis markers and improved liver function.

HSCs serve as the body’s primary reservoirs for vitamin A storage and exhibit high expression levels of STRA6, the specific receptor responsible for vitamin A uptake. This receptor enables selective recognition and internalization of vitamin A and its complexes.154 Leveraging this biological feature, vitamin A derivatives can function as targeting ligands to facilitate precision delivery to HSCs. For example, vitamin A -conjugated superparamagnetic iron oxide (SPIO) micelles was used targeted delivery of both miRNA-29b and miRNA-122 to HSCs.141 Vitamin A-modified probes (T-SCR) exhibited 33-fold greater fluorescence intensity (4.6 × 104 a.u.) in activated HSCs compared to unmodified probes (N-SCR). Expression levels of COL1A1, ɑ-SMA, and TIMP1 were reduced by 65.2%, 62.1%, and 79.4%, respectively, demonstrating superior efficacy compared to the non-targeted control group. Pirfenidone (PFD) was encapsulated into ZIF-8 via a one-pot synthesis method and subsequently coated with a 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) phospholipid bilayer and vitamin A to target HSCs.142 Under acidic conditions (pH~5.0), 76% cumulative drug release was achieved after 72 hours, which is significantly higher than the 9.45% observed in physiological environments. This system reduced collagen deposition by 80% and hydroxyproline (Hyp) content by 72%. Co-loading ZIF-8 with PFD and TGF-β1 siRNA further enhanced therapeutic efficacy by dually inhibiting HSCs activation.143

In a liver fibrosis model, transferrin receptor 1 (TfR1) expression was significantly upregulated on both liver sinusoidal endothelial cells (LSECs) and activated HSCs (Figure 3C). Leveraging ferritin’s natural high affinity for TfR1. Xu et al developed ferritin-based nanoparticles (FMP) encapsulating platycodin D (PLD), and manganese dioxide (MnO2) nanozymes, enabling precise passive targeting to fibrotic liver tissue.145 The Cy7-labeled FMP nanoparticles demonstrated progressive hepatic accumulation in CCl4-induced fibrotic mice following tail vein injection, with fluorescence signals detectable between 4–24 hours post-injection and peaking at 8 hours.

Hepatocytes, while not the primary drivers of fibrosis, play a pivotal role in its initiation through cellular damage. Protecting these cells and delivering anti-apoptotic or antioxidant agents represents an upstream therapeutic strategy to inhibit fibrosis progression (Figure 3D).155,156 Hepatocytes abundantly express the asialoglycoprotein receptor (ASGPR) on their surface, a well-characterized target that selectively binds ligands terminating in galactose or N-acetylgalactosamine—establishing the classical hepatocyte-targeting pathway.157 Galactose-functionalized AuNRs (Gal-ARPs) with high aspect ratios enhanced hepatocyte-specific CRISPR/Cas9 plasmid delivery by 10% compared to unmodified counterparts, enabling targeted Fas gene editing that reduced fibrotic progression.146 In addition, GA (glycyrrhetinic acid) as a liver-targeted molecule can specifically bind to hepatocyte, which can enhance the diagnostic role of AuNSs in liver fibrosis.147

Natural Product-Metal Nanomaterial Synergistic Systems

Many natural compounds (such as curcumin, silymarin, and resveratrol) possess excellent anti-inflammatory, antioxidant and anti-fibrotic activities.158,159 However, their poor water solubility, low bioavailability, and rapid metabolism in the body greatly limit their clinical application. Metal-based nanoparticles with sizes ranging from 10 to 100 nm, enhance drug accumulation through passive hepatic targeting via the EPR effect.160 These nanoparticles also exhibit inherent biological activities, including anti-inflammatory properties and ROS scavenging capabilities (Table 3).161,162 This synergistic effects represents an efficient strategy for liver fibrosis treatment, although its synergistic mechanisms and clinical translation potential warrant further systematic investigation.

Table 3 Synergistic Systems of Natural Products and MNMs for Liver Fibrosis Therapy

Integration Approaches of MNMs with Natural Products

The integration of nanomaterials with natural products primarily involves four approaches.

The first approach utilizes natural substances during the concurrent synthesis of metal nanoparticles, where they serve dual roles as reducing agents and stabilizers, or as structure-directing agents, ultimately becoming incorporated into the final product (Figure 4A). Many natural products (eg, polyphenols, flavonoids, and alkaloids) possess both reducing properties and metal coordination capabilities.170 These compounds can reduce metal ions (eg, Ag+, Au3+) to their zerovalent nanoparticle states while simultaneously encapsulating the particles to prevent agglomeration. For example, GNPs-silymarin system used tetrachloroauric acid (HAuCl4) as the precursor, silymarin (a flavonolignan complex) performs a tri-functional role, serving simultaneously as: (i) a reducing agent, (ii) a stabilizer, and (iii) the active pharmaceutical compound.119,131,132 AgNPs - plant extracts can also be synthesized by plant-mediated reduction. In this approach, silver nitrate precursors are reduced by plant extracts including Cymbopogon citratus (rich in citral), Rhizophora apiculata extract (high in tannins and flavonoids), Andrographis paniculata extract (containing and rographolide) and Helianthemum lippii extract (containing flavonoids and phenolic acid components), which simultaneously facilitate drug loading.163–165,171 PtNPs-curcumin used H2PtCl6 as the precursor, curcumin acts as a reducing agent, stabilizer, and targeting molecule.135

Figure 4 Integration approaches of natural products with MNMs including (A) One-pot green synthesis. (B) Synthesis-then-modification. (C) Natural product core add metal shell. (D) Nanomedicine co-delivery system.

The second approach involves synthesizing MNMs with well-defined morphology and dimensions (as either bare cores or with preliminary coatings) before functionalizing their surfaces through the controlled immobilization of natural products via physicochemical processes (Figure 4B). For example, Streptomyces metabolites (bearing reductases) reduce silver nitrate to form AgNPs. Subsequent curcumin loading is achieved through hydrogen bonding, while chitosan modification improves targeting capability.134 In addition, the integration of natural products with nanomaterials employs a “covalent conjugation–physical composite” dual strategy, yielding hydrogels with combined biocompatibility and stimuli-responsive functionality.168 First, mesoporous MnO2NPs or MnO2 nanosheets were dispersed in protocatechuic dialdehyde/hyaluronic acid (PD/HA) aqueous solution to form the MnO2@PD/HA composite. Next, the composite was combined with polyvinyl alcohol (PVA) and chitosan-gallic acid (CHI-GA) solution to fabricate the experimental hydrogel (Exp-gel). The core functionalities of all components were preserved in the final construct.

The third approach entails synthesizing the natural product core followed by coating or modifying with MNMs (Figure 4C). For example, Yang et al employed thin-film hydration to encapsulate oleanolic acid (OA) within liposomes, generating drug-loaded liposomes (Lipo@OA). Subsequently, hyaluronic acid/cerium nanoparticles (HA/CeNPs) were conjugated to the Lipo@OA surface via electrostatic interactions, yielding the final HCOL nanotherapeutic formulation.167 The load rate of OA is about 3.34 ± 0.03%, as well as the Encapsulation Efficiency (EE%) of OA is 83.60 ± 0.83%. In addition, PLGA nanoparticles encapsulating Ssb1 (PLGA/Ssb1) were synthesized via emulsion solvent evaporation, and then an MnO2 shell was subsequently deposited on the surface through in situ redox reactions, enabling dual functionality of drug delivery and antioxidant protection.166 Encapsulation efficiency PLGA/Ssb1and MnO2@PLGA/Ssb1 is 65% and 52.5% respectively.

The fourth approach employs nanoformulation- based co-delivery systems to simultaneously transport therapeutic MNMs and natural products to target organs, enabling synergistic therapeutic effects (Figure 4D). The biocompatible ferritin nanocage serves as a co-delivery platform for platycodonin D and MnO2NPs, with pH-dependent assembly ensuring structural stability.145 At a PLD: ferritin molar ratio of 1: 4, the encapsulation efficiency reaches 92.22 ± 2.01%. A hyaluronic acid-dopamine/methyl acrylate (HA-DA/AMA) matrix can be used to encapsulate MoS2 nanozymes and Chlorella photosensitizers for combined treatment of hepatocellular carcinoma.172 This formulation consisted of MoS2 nanoparticles at 200 μg mL−1, and Chlorella sp. (microalgae) at a concentration of 5×106 cells mL−1.

Synergistic Mechanism in Liver Fibrosis

The synergy between natural products and nanotechnology represents more than a simple “additive effect” of therapeutic outcomes. Rather, it achieves enhanced therapeutic efficacy (“1+1>2” effect) through multidimensional interactions at molecular, cellular, and tissue levels (Table 3). The core mechanisms can be categorized into three principal aspects:

Enhanced Safety or Biocompatibility Through Strategic Modifications

While MNMs (eg, AgNPs, AuNPs) offer therapeutic benefits, their potential toxicity risks - including cytotoxicity (eg, Ag⁺-induced DNA damage) and organ accumulation - must be addressed.161,162 Plant-mediated green synthesis and surface modifications provide dual protective mechanisms (Figure 5A). First, plant extracts serve as biocompatible reducing agents, eliminating toxic chemical residues such as sodium borohydride. For example, cymbopogon citratus decreasing AgNPs cytotoxicity by 30–50%.134 Pure ZnO-NPs/Ag-NPs may be cytotoxic when used alone, but the biocompatibility of the nanoparticles was improved when combined with purslane extract.173 Second, phytochemical surface coatings enhance nanoparticle dispersion (improving AgNPs dispersion index from 0.5 to 0.8),165 which minimizes cellular membrane damage and reduces metal ion leaching (eg, decreasing Ag+ release from 10 μg/mL to 2 μg/mL).161 These modifications collectively mitigate organ accumulation risks.134 In addition, natural products (HA, GA, CHI) were integrated with nanomaterials (mesoporous MnO2 nanoparticles and MnO2 nanosheets) to construct a multifunctional hydrogel nanosystem.69 Specifically, HA and CHI synergistically reduced hydrogel cytotoxicity, and GA enhanced tissue adhesion, particularly for hepatic and muscular interfaces. Dopamine conjugated HA served as both a dispersion matrix preventing MnO2NPs aggregation and a mild microenvironment facilitating subsequent exosome-mediated miR-582-5p delivery. This combined of green synthesis for chemical safety and surface modification for metal safety establishes a robust framework for safer nanomedicine applications.

Figure 5 Synergistic mechanisms of natural product–metal nanomaterial systems against liver fibrosis. (A) Enhanced biosafety and biocompatibility. (B) Overcoming bioavailability limitations. (C) Multi-target synergistic counteraction of pathological cascade reactions.

Overcoming Bioavailability Limitations of Natural Produce

MNMs enhance the bioavailability of natural products through optimized physicochemical properties, representing their fundamental synergistic mechanism (Figure 5B).

First, MNMs enhance the solubility and bioavailability of natural compounds. For example, AuNPs increased silymarin’s water solubility from 0.5 g/L to stable dispersion and improve bioavailability from 20% to 96%.119 Curcumin exhibits extremely low bioavailability (< 1%) when administered alone. However, AgNPs conjugation significantly enhances its pharmacokinetic properties, demonstrating a 3-fold increase in hepatic accumulation and a 10-fold solubility improvement.134 Similarly, β-sitosterol achieves 50% greater intestinal absorption efficiency and 60% liver targeting rate when conjugated with AgNPs.161Rhizophora apiculata extract-conjugated AgNPs reduce the required dose by 50% while maintaining therapeutic efficacy and minimizing tannin-induced hepatotoxicity.163 Curcumin exhibits susceptibility to oxidation and degradation under physiological conditions. PtNPs mitigate this instability by coordinating with curcumin’s hydroxyl groups (-OH) through Pt0 binding, thereby reducing systemic degradation, extending blood circulation half-life, and prolonging therapeutic activity.135

Second, targeted enrichment mechanisms. Passive targeting is achieved through nanoparticles that accumulate in the liver via the enhanced permeability and retention (EPR) effect, increasing hepatic drug enrichment 3–10 folds.131,132,162 Chitosan-modified AgNPs bind to sialic acid receptors, extending drug retention time to 16 hours.134 AgNPs modified with plant extracts of Cymbopogon citratus binds to fat-soluble receptors on hepatocytes, directing lemongrass-modified nanoparticles (LG-NPs) to damaged liver tissue.164 Flavonoids in Rhizophora apiculata leaf extract interact with glycoproteins on Toxoplasma, improving anti-parasite targeting of AgNPs.162 This dual strategy (“passive nano-size targeting + active phytochemical guidance”) overcomes two major limitations of natural products: (1) systemic dispersion and (2) low local concentrations. Additionally, it extends circulation time (eg, AgNPs’ half-life increases from 2 to 8 hours) and reduces metabolic loss.165

Mechanism Synergy: Covering Multiple Pathological Links of Liver Fibrosis

The pathological progression of liver fibrosis follows a cascading sequence of oxidative stress -inflammatory response - HSCs activation - ECM deposition.174 This complex cascade cannot be effectively interrupted by either natural products or nanomaterials alone. However, their combined application enables multi-target modulation through synergistic mechanisms.

Nanomaterials and natural products synergistically regulate oxidative stress and inflammation (Figure 5C). AuNPs exhibit intrinsic antioxidant and anti-inflammatory properties, while also potentiating the biological efficacy of conjugated natural products. For example, silymarin exhibits classical hepatoprotective properties, including antioxidant activity, anti-inflammatory effects, and inhibition of HSCs activation. Through conjugation with AuNPs, SGNPs not only enhance bioavailability but also leverage the inherent antioxidant and anti-inflammatory properties of AuNPs for synergistic antifibrotic effects. Mechanistically, SGNPs exhibit a multi-target action involving (1) oxidative stress reduction, (2) inflammation suppression, and (3) modulation of fibrogenic pathways. Notably, SGNPs demonstrate 11.3–33.5% superior therapeutic efficacy compared to SIL or GNPs alone in preclinical models (33.5% enhancement in MDA inhibition, 11.3% increase in TGF-β1 suppression, 33.3% elevation in protective miRNA expression).119 Similarly, Kabir et al used AuNPs to enhance the solubility and targeting of silymarin. Silymarin inhibits oxidative stress, HSCs activation and KCs infiltration, and synergistically blocks the fibrotic cascade reaction.131 ScAuNPs reduce the transformation of HSCs into myofibroblasts and the expression of α-SMA by inhibiting the expression of TGF-β1, or blocking the activation of its receptors (TGFβR1 and TGFβR2). The area of liver fibrosis decreased from 3.5% to 0%. Cornus sanguinea L. polyphenol-coated AuNPs (NPCS) attenuated TGF-β-mediated fibrogenesis by synergistic suppressing oxidative stress, inflammation (NF-κB signaling pathway), and α-SMA expression.133 Besides, AgNPs can activate the Nrf2 pathway, chelate cadmium ions, and potentiate antioxidant enzyme activity (SOD and CAT). When combined with modified H. lippii extract, the treatment demonstrated synergistic hepatoprotective, antioxidant (activate Nrf2 pathway), and anti-inflammatory (Inhibit NF-κB pathway) anti-fibrotic effects (ALT decreased by 70.4%, MDA reduced by 42.0%, Complete resolution of portal fibrosis) in cadmium-exposed models. Specifically, biochemical analyses revealed.165 The same synergistic mechanism was observed when AgNPs were combined with β-sitosterol and lemongrass (Cymbopogon citratus) extract. Both compounds demonstrated complementary antioxidant effects in the treatment of liver fibrosis, and inhibited the expression of TGF-β1, blocking its downstream Smad2/3 phosphorylation, and reversed the activation of HSCs.161,164 In addition, curcumin suppresses collagen expression by modulating fibrosis-associated signaling pathways, notably the HIF-1α/ERK pathway.26 Concurrently, PtNPs demonstrate dual antifibrotic mechanisms: (1) ROS scavenging via enzyme-mimetic activity, and (2) HSCs activation inhibition. These complementary actions yield synergistic attenuation of fibrotic progression.135

The combination exerted synergistic antifibrotic effects were also found in MnO2NP@PLGA/Ssb1 nanosystem, in which MnO2NPs catalyzed H2O2 decomposition to alleviate hypoxia and oxidative stress, while Ssb1 suppressed HSCs activation markers (α-SMA and COL-I) by inhibiting the TGF-β1/Smad3 pathway.166 Further synergy was observed when MnO2 was combined with platycodonin D.145Platycodonin D upregulated Kruppel-like factor 2 (KLF2) and phosphorylated endothelial nitric oxide synthase (p-eNOS), reversing liver sinusoidal endothelial cell (LSE