Advances in ultrasound-mediated nanozyme systems in therapeutic applications

Abstract

Ultrasound-mediated nanozyme systems are emerging therapeutic platforms that couple externally controllable acoustic fields with enzyme-mimetic catalysis. Ultrasound can improve tissue delivery, penetration and local activation through cavitation, sonoporation, phase transition and piezoelectric effects, whereas nanozymes remodel pathological microenvironments by generating or scavenging reactive oxygen species (ROS), modulating oxygen availability, perturbing metabolism and restoring redox homeostasis. A central challenge is to distinguish true nanozyme catalysis from related ultrasound-responsive processes, including sonodynamic activation, piezocatalysis and ultrasound-triggered delivery. This review summarizes recent progress across tumors, drug-resistant bacterial and biofilm-associated infections, atherosclerosis, neurodegenerative diseases, osteochondral disorders, ischemia–reperfusion (I/R) injury, and tendon repair and tendinopathy. We highlight how ultrasound and nanozymes cooperate to overcome biological barriers, regulate oxidative stress and inflammation, remodel disease microenvironments and support tissue repair or functional recovery. We further discuss mechanistic heterogeneity, incomplete attribution of therapeutic effects, insufficient standardization of ultrasound parameters, material complexity, biodistribution and clearance, biosafety, and manufacturing challenges. The review argues that progress should be judged by catalytic attribution, disease-matched acoustic design and material exit routes, rather than by adding functions to already complex platforms. Overall, ultrasound-mediated nanozyme therapy should be viewed as a mechanism-dependent and disease-context-specific strategy whose translation will require clearer catalytic validation, acoustic reporting and safety evaluation.

1. Introduction

Nanomaterials with intrinsic enzyme-like catalytic activities, commonly referred to as nanozymes, have become an important class of artificial enzymes for biomedical applications [1]. Unlike natural enzymes, whose catalytic functions are encoded by defined protein structures, nanozymes derive their activity from engineerable physicochemical features, including composition, particle size, morphology, exposed crystal facets, defect structures, valence states and surface chemistry [2], [3]. These features regulate substrate adsorption, interfacial electron transfer and intermediate conversion at surface-accessible or atomically defined active sites, enabling diverse enzyme-mimicking activities such as peroxidase-, oxidase-, catalase-, superoxide dismutase- and glutathione peroxidase-like catalysis [4], [5]. This structural tunability provides a versatile basis for designing nanozymes that either amplify pathological oxidative stress or restore redox homeostasis in disease microenvironments.

Despite these advantages, the therapeutic translation of nanozymes remains constrained by the difficulty of controlling where, when and how catalysis occurs in vivo. Catalytic activities measured in simplified buffer systems can be altered substantially by protein corona formation, particle aggregation, substrate availability, pH, oxygen tension, reducing equivalents and immune clearance in pathological tissues [6], [7], [8]. As a result, high apparent activity in vitro does not necessarily produce effective or selective catalysis at diseased sites. The central bottleneck is therefore no longer whether nanozymes can catalyze model reactions, but whether catalysis can be delivered, acoustically controlled and mechanistically verified in diseased tissue.

Ultrasound offers a clinically established external modality for addressing this control problem because it combines non-invasiveness, centimeter-scale tissue penetration and adjustable spatiotemporal dosing [9]. In ultrasound-mediated nanozyme systems, acoustic energy can contribute to therapy through several coupled but mechanistically distinct routes. First, acoustic cavitation, radiation force, microstreaming and sonoporation can improve local transport, vascular or tissue permeability and cellular uptake, thereby increasing nanozyme access to pathological sites [10], [11], [12], [13]. Second, ultrasound can activate sonosensitizers or ultrasound-responsive components to generate reactive species, remodel oxygen or redox conditions, or trigger drug release [14], [15], [16]. Third, in piezoelectric or defect-engineered materials, ultrasonic stimulation may promote charge separation and interfacial electron transfer, thereby modulating nanozyme-like catalytic reactions [17], [18], [19]. This multiplicity is useful only when the dominant therapeutic driver is defined, because delivery enhancement, sonodynamic activation, piezocatalysis and genuine enzyme-mimetic catalysis can otherwise be conflated.

Accordingly, this review uses the term ultrasound-mediated nanozyme systems to refer to therapeutic platforms in which enzyme-mimetic catalysis is either directly regulated by ultrasound or functionally integrated with ultrasound-responsive processes. This scope covers four related platform types. These include nanozymes whose catalytic activity is enhanced by acoustic or mass-transfer effects; nanozyme-containing sonodynamic systems in which catalytic reactions reshape oxygen or redox substrates; piezoelectric nanozyme hybrids in which ultrasound-induced polarization modulates interfacial electron transfer or active-site redox cycling; and ultrasound-triggered delivery systems in which acoustic activation improves nanozyme accumulation, penetration or release. By contrast, ultrasound-responsive nanomedicines without demonstrable enzyme-like catalytic activity are discussed only as related comparators rather than as core nanozyme systems. This distinction is essential for evaluating whether therapeutic efficacy arises primarily from nanozyme catalysis, sonosensitizer activation, piezocatalysis, delivery enhancement or deliberately coupled mechanisms.

Several recent reviews have provided useful perspectives on nanozyme and ultrasound-responsive therapeutic systems, including the in vivo design of nanozymes, ultrasound-activated theranostic materials, smart metal–organic framework nanozymes, and colloidal nanozymes for cancer diagnosis and therapy [9], [20], [21], [22]. These works have highlighted key issues such as catalytic activity, biointerface regulation, biodistribution, clearance, biosafety, acoustic activation mechanisms, dual catalysis, theranostics, surface chemistry and precision medicine. However, they have generally treated nanozyme design, ultrasound-responsive theranostics, MOF nanozymes and colloidal nanozymes as partly separate literatures. What remains less resolved is how to judge therapeutic benefit when ultrasound-enhanced delivery, sonosensitizer activation, piezocatalysis and genuine enzyme-mimetic turnover coexist in the same platform. In this context, the present review treats ultrasound not only as an activation source, but also as an experimental and translational variable that must be matched to catalytic mechanism, disease microenvironment, acoustic reporting and safety constraints. This perspective clarifies when a system should be interpreted as nanozyme-mediated therapy, when it is better viewed as Sonodynamic therapy (SDT) or ultrasound-triggered delivery, and how studies can be compared despite non-standardized ultrasound parameters.

Guided by this framework, we first summarize the catalytic basis of nanozymes, with emphasis on material classes, structure–activity relationships, catalytic kinetics, biointerface-dependent activity and redox-regulatory mechanisms. We then discuss the major ultrasound-mediated processes that can be coupled with nanozyme systems, including targeted delivery, sonodynamic activation and piezoelectric modulation, while highlighting the need to separate genuine enzyme-mimetic catalysis from related ultrasound-responsive effects. On this basis, we compare therapeutic applications across tumors, drug-resistant bacterial and biofilm-associated infections, atherosclerosis, neurodegenerative diseases, osteochondral disorders, I/R injury, and tendon repair and tendinopathy, focusing on the disease barriers that determine the desired catalytic output. Finally, we discuss current challenges and future design principles, focusing on disease-matched catalytic outputs, ultrasound parameter reporting, biodistribution and clearance, biosafety, manufacturing reproducibility and clinical translation. Together, these sections move from definition and catalytic attribution to acoustic control, disease-matched application and translational evaluation, so that comparisons are made by mechanism and therapeutic task rather than by material novelty alone.

2. Catalytic mechanisms of nanozymes

Nanozymes are catalytic nanomaterials that emulate enzymatic functions through structurally defined or surface-accessible active sites. The field is commonly traced to the discovery of the intrinsic peroxidase-like activity of Fe₃O₄ nanoparticles [4], [23]. Since then, nanozymes have expanded from iron oxide nanoparticles to a broad range of material platforms, including metals, metal oxides, carbon-based materials, metal–organic frameworks, covalent organic frameworks, Prussian blue analogues and single-atom catalysts. Across these systems, catalytic behavior is governed by active-site composition, surface electronic structure, substrate adsorption, electron transfer and reaction-environment coupling.

Nanozymes and natural enzymes should be viewed as complementary catalytic systems rather than as direct substitutes. Nanozymes offer physicochemical stability, convenient storage, engineerable composition and the ability to integrate catalytic, optical, magnetic or electrical functions within a single platform [2], [24], [25]. Natural enzymes, in contrast, possess evolutionarily optimized active sites that usually provide higher substrate specificity, biological selectivity and single-site catalytic efficiency. The heterogeneous surfaces of nanozymes can support broad substrate conversion and catalytic cascades, but they may also cause nonspecific reactions and complicate active-site normalization. Therefore, kinetic comparisons between nanozymes and natural enzymes are meaningful only when substrates, reaction conditions and active-site definitions are matched [26].

These differences make nanozyme function highly context-dependent. A mechanistic understanding of nanozymes therefore requires more than assigning enzyme-like activity labels. It requires consideration of material composition, active-site structure, catalytic kinetics and the reaction environment in which catalysis occurs [3], [20], [27], [28]. This section first outlines the major material classes of nanozymes and their catalytic-function spectrum, then examines how structural and kinetic features determine catalytic performance. It further discusses biointerface-dependent activity, ROS- and RNS-related redox catalysis, the balance between pro-oxidant and antioxidant pathways, and emerging catalytic mechanisms beyond direct ROS/RNS regulation. This organization emphasizes the evidence needed to attribute therapeutic effects to nanozyme catalysis rather than to nonspecific material or assay effects.

2.1. Material platforms and catalytic-function spectrum of nanozymes

Nanozymes are broadly classified by material composition and by the structural nature of their catalytic sites. Metal-based nanozymes, such as Pt- and Au-containing systems, often rely on surface metal atoms and interfacial electron transfer to mediate oxidoreductase-like reactions. Metal oxide nanozymes, represented by Fe3O4 and CeO2, illustrate how mixed valence states, oxygen vacancies, exposed facets and surface defects can regulate catalytic behavior [6], [25]. Other transition-metal compounds, including metal sulfides [29], nitrides [30], carbides/MXenes [31] and phosphides [32], have further expanded the chemical space of nanozyme design. Across these inorganic platforms, catalytic performance is less determined by composition alone than by the accessibility, electronic structure and reaction environment of active sites.

Beyond these inorganic platforms, carbon-based, framework-based and biomimetic/hybrid nanozymes provide a broader and more precisely tunable chemical-modification space for active-site engineering and biological integration. Carbon materials (e.g., graphene derivatives and carbon dots) can serve as either intrinsic catalytic scaffolds [33], [34] or supports for additional active components [32]. Metal–organic frameworks (MOFs), covalent organic frameworks (COFs) and coordination polymers offer ordered environments for embedding catalytic sites within metal nodes, organic linkers or confined cavities [35], [36], [37], [38]. More active-site-defined systems include Prussian blue analogues and single-atom nanozymes. Prussian blue analogues represent a distinct cyanometallate coordination-network family with multienzyme-like redox activity [39], [40]. In contrast, single-atom nanozymes feature isolated metal centers stabilized by metal–ligand, M–N–C, protein or amyloid-like coordination environments [41], [42], [43], [44]. Biomimetic and hybrid nanozymes further integrate catalytic nanomaterials with cell membranes, proteins or cascade modules, shifting nanozyme design from bare particles toward systems better adapted to complex biological settings [32], [38], [44], [45].

Across these material platforms, nanozymes can also be classified by catalytic function. Most therapeutic nanozymes exhibit oxidoreductase-like activities, including peroxidase-like, oxidase-like, catalase-like, superoxide dismutase-like and glutathione peroxidase-like reactions. Less common but increasingly important functions include haloperoxidase-like, nitrite reductase-like, hydrolase-like, protease-like and NADH oxidase-like catalysis. Material composition and active-site architecture therefore define the structural basis of nanozyme classification, whereas enzyme-like activity defines the catalytic task. The structure–activity and kinetic principles underlying these functions are discussed in Section 2.2, redox mechanisms in Section 2.3 and catalytic mechanisms beyond direct ROS/RNS regulation in Section 2.4.

2.2. Structure–activity relationships and catalytic kinetics of nanozymes

Nanozyme performance is governed by the coupling between structural features, catalytic kinetics and the biological reaction environment. Structural parameters define the accessibility and electronic properties of active sites, kinetic parameters describe their operation under defined assay conditions, and biointerfacial processes can reshape apparent activity in biological media. This section therefore treats nanozyme activity as an integrated structure–kinetics–biointerface relationship rather than as a fixed material property.

2.2.1. Structural determinants of nanozyme activity

The catalytic behavior of nanozymes is governed by structural determinants operating at multiple levels, from particle geometry to atomic coordination and surface interfaces. These descriptors are often coupled, because changes in size, defect density or surface chemistry can also alter metal valence, substrate adsorption and electron transfer. Structure–activity relationships should therefore be established using matched controls that isolate the variable under investigation.

At the particle level, size, morphology and exposed facets determine the number, accessibility and chemical identity of surface catalytic sites. In a size-controlled series of iron oxide nanozymes, peroxidase-like activity varied non-monotonically and reached its maximum at 7.82 nm rather than at the smallest size of 3.17 nm [46]. Although smaller particles provided a larger specific surface area, they contained less surface iron and a lower Fe²⁺/Fe³⁺ ratio, reducing the contribution of catalytically active Fe²⁺ sites. Facet effects were more directly isolated in cysteine-modified PdPt₃ hollow nanocages with matched size, composition, wall thickness and surface area [47]. Nanocages exposing {100} facets showed 2.2-fold higher peroxidase-like activity and stronger enantiomer recognition than those exposing {111} facets. These examples show that surface area alone is insufficient to explain nanozyme activity.

Defect chemistry and metal valence states further modulate catalytic performance by reconfiguring charge distribution and the energetics of adsorbed intermediates. Loading Cu and Pt onto CeO₂ increased the Ce³⁺/Ce⁴⁺ ratio and oxygen-vacancy concentration, resulting in stronger SOD-like and CAT-like activities than those of the pristine oxide [48]. This enhancement was attributed to metal–support interactions that facilitated electron transfer and lowered the energy barriers for vacancy formation and subsequent catalytic steps. Oxygen vacancies play a related but mechanistically distinct role in ε-MnO₂. In this system, they stabilize high-spin Mn(III)–O species and induce Jahn–Teller distortion. These electronic changes promote oxygen activation and preserve OXD-like activity from −20 to 45 °C [49]. In Fe₃O₄, surface-specific analyses showed that H₂O₂ activation involves electron donation from subsurface Fe²⁺ centers, whereas interfacial water stabilizes key intermediates [50]. Progressive Fe²⁺ oxidation correlated with catalytic deactivation, consistent with the size-dependent behavior observed in iron oxide systems [46]. Local coordination environments can further determine whether catalytically active valence states are preserved. For example, bovine serum albumin established a confined coordination microenvironment that shielded isolated Cu(I) sites from rapid oxidative degradation [44]. Collectively, these observations indicate that defects, redox couples and the local medium operate in concert rather than as independent descriptors of catalytic activity.

Atomic coordination provides a more precise route to tune substrate binding and reaction barriers. Cu–N₃ single-atom sites showed substantially higher peroxidase-like activity than Cu–N₄ sites because the lower-coordinate geometry favored H₂O₂ adsorption and dissociation [51]. A related trend was observed for Fe–N₃ antioxidant nanozymes [52]. However, undercoordination should not be treated as a universal design rule, because the preferred geometry depends on both the metal center and the catalytic reaction. Regulation can also be achieved through the second coordination sphere or by directly engineering the primary coordination environment. Adjacent Mn atoms shifted the d-band center of Fe sites in Fe–Mn dual-atomic nanozymes, strengthening substrate adsorption and increasing catalytic efficiency relative to isolated Fe sites [53]. Axial chlorine coordination directly modified Fe–N₄ centers and improved both peroxidase-like and glutathione oxidase-like activities relative to conventional Fe–N₄ sites [54]. Together, these examples show that atomic coordination regulates nanozyme catalysis through coordination number, neighboring-site effects and axial ligand modulation.

Surface ligands provide an additional level of catalytic control by modulating charge transfer, substrate access and intermediate binding. Among similarly sized polymer-coated Ru nanozymes, polystyrene sulfonate produced substantially higher POD-like activity than polyacrylic acid or polyvinylpyrrolidone [55]. This difference did not correlate with zeta potential, but was instead attributed to charge transfer from Ru to the ligand, which weakened excessive binding of hydroxyl-radical intermediates. A distinct bioinspired strategy was used in gold clusterzymes, where Fe–TCPP was attached to glutathione-stabilized gold nanoclusters as a heme-like catalytic motif [56]. Fe–TCPP increased H₂O₂ affinity and facilitated electron transfer from the gold core, resulting in enhanced peroxidase-like catalysis. Thus, ligand engineering can either tune an existing metallic surface or introduce an additional catalytic center and electron-transfer pathway.

Overall, nanozyme activity emerges from the coupling of geometric structure, electronic state, atomic coordination and interfacial chemistry. Structure–activity comparisons are most informative when materials share a common scaffold or differ by a well-controlled design variable. Even then, structural optimization cannot be judged from activity labels or end-point signals alone. Quantitative kinetic analysis is therefore needed to show whether a structural change improves catalytic turnover, substrate affinity or reaction efficiency under defined conditions.

2.2.2. Catalytic kinetics and reaction-environment effects

Catalytic performance should be quantified from initial reaction rates rather than end-point signals, because end-point readouts may reflect substrate depletion, particle instability or assay interference as well as catalytic turnover. Nanozyme kinetics are commonly assessed by varying one substrate concentration while holding the others constant, followed by nonlinear fitting of initial-rate data to the Michaelis–Menten equation, v₀ = V_max[S]/(K_m + [S]) [25]. The tested substrate range should straddle K_m; a range from 0.5 × K_m to 5 × K_m has been recommended for fitting POD-like nanozyme kinetics [57]. In this framework, K_m is an apparent saturation parameter rather than a substrate dissociation constant, whereas V_max denotes the maximum initial rate under defined assay conditions. The catalytic constant, k_cat = V_max/[E_t], depends on how the concentration of catalytically accessible sites is defined, and k_cat / K_m describes catalytic efficiency in the low-substrate regime.

Kinetic parameters are most informative when both the material comparison and the substrate conditions are well controlled. In a matched single-atom nanozyme system, atom-pair engineering of Zn-SA/CNCl increased V_max, k_cat and k_cat / K_m for TMB oxidation by 346-, 1,496- and 133-fold, respectively, compared with Zn-SA/CN lacking the Zn–N₄Cl₁ regulatory sites [58]. This example shows that kinetic analysis can quantify the effect of a defined coordination regulator when the active-site framework and assay conditions are internally matched. Substrate identity adds a second level of control. Au@Pt nanozymes showed a K_m of 0.21 ± 0.05 mM for TMB, comparable to horseradish peroxidase, but a K_m of 540 ± 70 mM for H₂O₂, approximately three orders of magnitude higher than that of horseradish peroxidase [59]. Thus, substrate conditions suitable for enzyme-based assays can be inadequate for nanozymes. In an Au@Pt-based immunoassay, non-optimized TMB and H₂O₂ concentrations increased the detection limit by up to 30-fold compared with kinetically favorable conditions [53]. These findings indicate that nanozyme assays should be optimized according to measured kinetic parameters rather than inherited from natural-enzyme protocols.

Reaction conditions provide a further layer of control over nanozyme kinetics. Even when the material system and substrates are defined, the measured reaction rate depends on the local chemical environment around the active site. In PCN-222-Fe, confining poly(acrylic acid) within MOF channels increased the local proton concentration and shifted the highest peroxidase-like activity to pH 7.4, yielding an approximately fourfold activity increase over the unmodified nanozyme at the same bulk pH [60]. Medium composition also affects measured catalytic output in a material-dependent manner. In a systematic TMB-based assay across four common nanozymes, alternative buffers increased signal intensity by 41–68% for several materials, whereas sodium acetate remained suitable for LaNiO₃ [61]. These findings indicate that pH, buffer identity, ionic strength, solvent and substrate concentrations should be treated as part of the kinetic context rather than as interchangeable assay background.

Particle state is another major source of apparent kinetic error. Electrolyte concentration can alter both colloidal stability and substrate accessibility. In Prussian blue nanozymes, increasing ionic strength compressed the electric double layer and improved guaiacol access to the particle surface, thereby accelerating HRP-like reactions; Cs⁺ produced a particularly strong activity increase [62]. However, the effect of ionic strength is interface-dependent. In a profiling study of 14 oxidoreductase-mimicking nanozymes, bare Au@Pt nanoparticles aggregated at ionic strengths of 60 mM or higher, leading to reduced peroxidase-like activity [63]. The same study further showed that apparent kinetic parameters can be distorted not only by aggregation, but also by multiple catalytic activities and chemical transformation. Neglecting multiple activities caused a sixfold underestimation of Mn₂O₃ oxidative activity. TMB oxidation by Au@Ag originated from released Ag⁺ rather than intrinsic nanoparticle catalysis, whereas storage-dependent transformation of Au@Prussian blue produced a more than fourfold overestimation of peroxidase-like activity after 24 h of storage [57]. These findings indicate that reliable kinetic analysis requires verification of colloidal stability, chemical stability and reaction specificity under the actual assay conditions.

Accordingly, K_m, V_max, k_cat and k_cat/K_m should be treated as condition-dependent descriptors rather than immutable material constants. Reliable comparisons require initial-rate measurements, explicit active-site or mass normalization, matched substrates, controlled temperature and medium composition, and verification of chemical and colloidal stability. These requirements are also the minimum basis for deciding whether ultrasound changes catalysis itself or only changes substrate availability, particle state or assay chemistry.

2.2.3. Biointerface-dependent catalytic behavior

Upon contact with biological fluids, nanozymes acquire a biomolecular interface that can differ markedly from the surface characterized in protein-free buffers. This interface can regulate apparent catalysis by altering active-site accessibility, substrate transport and colloidal state. In AuNR@Pt@PDDAC nanorods, plasma proteins and individual corona components suppressed multiple enzyme-like activities [59]. The extent of inhibition depended on protein architecture. Fibrous proteins formed denser protein networks with smaller mesh pores than spherical proteins. These denser networks restricted substrate permeation more strongly and therefore produced greater catalytic inhibition [64].

Protein-corona effects can also change during intracellular trafficking. For 5-nm Pt nanozymes, BSA and serum-derived coronas reduced multiple enzyme-like activities in extracellular-mimicking media by shielding the catalytically active Pt surface [60]. In artificial lysosomal fluid, proteolytic corona degradation progressively restored catalytic activity. This process produced an extracellular-to-lysosomal catalytic bioswitch, with an activity increase exceeding tenfold [65]. Biointerfaces can also enhance nanozyme activity in specific material systems. In a CeO₂–Bioglass hybrid, protein coating increased CAT-like activity, probably because improved colloidal stability outweighed active-site shielding [66]. This observation indicates that biological interfaces do not simply block catalytic sites; rather, they reshape nanozyme catalysis through the competing effects of active-site accessibility, substrate transport and particle stability.

Together, these findings indicate that protein coronas can suppress, restore or occasionally enhance nanozyme activity depending on protein organization, surface coverage, colloidal state and intracellular processing. Nanozyme performance measured in simple buffers should therefore not be extrapolated directly to biological environments. Catalytic evaluation should combine corona characterization with measurements of colloidal stability, substrate accessibility and activity in relevant extracellular and intracellular media. These measurements define the biological baseline against which ultrasound-mediated catalytic activation should be judged.

2.3. Redox catalytic mechanisms of nanozymes

ROS comprise radical species, such as superoxide (O₂^•−), hydroxyl radical (•OH), and peroxyl radical (ROO·), as well as non-radical species including hydrogen peroxide (H₂O₂), singlet oxygen (¹O₂), hypochlorous acid (HOCl), and peroxynitrite (ONOO⁻) [67], [68]. Their biological effects are strongly concentration- and context-dependent. At controlled levels, ROS participate in redox signaling and regulate cellular proliferation, differentiation and immune responses [69], [70]. By contrast, excessive ROS accumulation overwhelms endogenous antioxidant defenses and causes oxidative damage to lipids, proteins and nucleic acids, thereby contributing to the development and progression of numerous pathological conditions [67], [68], [71]. This duality provides the mechanistic basis for nanozyme-mediated redox regulation: pro-oxidant nanozymes amplify oxidative stress by generating ROS, whereas antioxidant nanozymes eliminate excessive ROS and restore redox homeostasis. Nanozymes can additionally regulate reactive nitrogen species (RNS), including nitric oxide and peroxynitrite, further extending their redox functions. Accordingly, this section discusses ROS-generating catalysis, ROS-scavenging catalysis, RNS-related regulation and the context-dependent balance between pro-oxidant and antioxidant catalytic pathways (Fig. 1).

Fig. 1.

Schematic overview of multifunctional redox catalytic behaviors of nanozymes. Nanozymes regulate biological redox homeostasis through coordinated pro-oxidant, antioxidant, RNS-related, and non-redox catalytic activities. The red panel summarizes pro-oxidant pathways for ROS generation, the blue panel depicts antioxidant mechanisms for ROS scavenging, the purple panel illustrates RNS-related redox regulation, the green panel presents non-redox catalytic functions beyond ROS/RNS modulation, and the central panel highlights nanozyme-mediated redox balance and pathological outcomes associated with abnormal ROS/RNS levels.

2.3.1. ROS-generating oxidative mechanisms

ROS-generating nanozymes enhance oxidative stress mainly through H₂O₂-dependent POD-like reactions, O₂-dependent OXD-like reactions and substrate-supplying catalytic cascades. In a broad operational sense, Fenton and Fenton-like reactions can be regarded as important chemical routes underlying H₂O₂-dependent POD-like nanozyme activity, because they also convert H₂O₂ into highly oxidizing ROS, particularly •OH. However, they should not be considered strictly equivalent to POD-like nanozyme catalysis. Classical Fenton chemistry refers to Fe²⁺-mediated H₂O₂ decomposition to produce •OH, whereas Fenton-like reactions usually involve other redox-active metal ions or alternative iron valence cycles that activate H₂O₂ to generate •OH or related reactive intermediates. By contrast, POD-like nanozymes are typically evaluated as enzyme mimics using peroxidase substrates and Michaelis-Menten-type kinetic parameters, such as K_m and V_max, reflecting substrate-dependent catalytic behavior. Therefore, Fenton/Fenton-like reactions may contribute to, or mechanistically explain, POD-like activity, but POD-like catalysis does not necessarily proceed exclusively through a classical Fenton pathway involving freely diffusible •OH. In radical-mediated systems, redox-active metal centers activate H₂O₂ to generate •OH or other reactive intermediates. For example, Cu(I)-BSA single-atom nanozymes catalyzed H₂O₂ to generate •OH, O₂^•− and ¹O₂ [44]. In MgCaFe-LDH@Au nanozymes, GSH-mediated reduction of Fe³⁺ to Fe²⁺ enhanced POD-like activity, which was attributed to Fe²⁺-mediated Fenton conversion of H₂O₂ into •OH [72]. Sulfur coordination in a hemin–cysteine–Fe nanozyme likewise promoted selective POD-like activity under mildly acidic conditions, with •OH identified as the principal product [73].

By contrast, POD-like substrate oxidation may also proceed through an electron-transfer (ET) pathway. Au nanoparticles confined within quantum-dot coacervate droplets showed no detectable •OH formation in the presence of H₂O₂. Electrochemical measurements instead indicated that the Au-containing droplets facilitated electron transfer from the reducing substrate to H₂O₂. The authors therefore proposed a sequential pathway in which electrons are transferred from the substrate to the confined Au nanoparticles and subsequently to H₂O₂, reducing H₂O₂ to water while oxidizing the substrate [74]. Prussian blue nanozymes exhibited a more complex mechanism. Electron flow occurred from a reducing substrate, such as TMB or ABTS, to H₂O₂ through either a conduction-band-mediated or valence-band-mediated pathway, depending on the oxidation state and surface structure of the nanozyme. Fe–OH and Fe=O intermediates were also detected, indicating that this process should not be interpreted as a simple radical-free direct ET mechanism [39]. Chromogenic-substrate oxidation alone therefore cannot distinguish freely diffusible radical formation, surface-bound intermediates and ET-mediated catalysis.

OXD-like catalysis provides a complementary route that uses O₂ as the initial substrate. Au³⁺-functionalized UiO-67 nanoparticles generated O₂^•− under aerobic conditions and •OH in the presence of H₂O₂, demonstrating coexisting OXD-like and POD-like activities [75]. PtAu nanoparticles similarly exhibited OXD-like and POD-like activities associated with O₂^•− and •OH production, respectively [76]. ROS generation can be amplified further through glucose oxidase-like (GOx-like) cascades, in which glucose oxidation supplies H₂O₂ and increases local acidity for downstream POD-like or Fenton-like reactions [72], [76], [77]. HPO-like nanozymes operate through a distinct H₂O₂-dependent pathway involving halide oxidation. V₂O₅ nanowires catalyzed bromide oxidation to HOBr, which subsequently reacted with H₂O₂ to form ¹O₂ in the absence of an organic acceptor [78]. Electronically engineered vanadium-based artificial enzymes also generated HClO in the presence of H₂O₂ and Cl⁻, alongside •OH and O₂^•− [79]. These studies indicate that HPO-like platforms may generate mixtures of reactive oxygen and halogen species rather than a single oxidant. Consequently, ROS-generating mechanisms should be assigned using substrate-dependent controls and direct product identification, such as radical-quenching assays, electron paramagnetic resonance or species-selective probes, rather than chromogenic-substrate oxidation alone.

2.3.2. ROS-scavenging antioxidative mechanisms

Antioxidative nanozymes restore redox homeostasis mainly through SOD-like, CAT-like and GPx-like reactions. SOD-like nanozymes catalyze the dismutation of O₂^•− into H₂O₂ and O₂, although the underlying mechanism varies among material classes. In carbon-dot nanozymes, surface-modification experiments suggest that carboxyl, hydroxyl and amino groups anchor O₂^•− through hydrogen bonding. One O₂^•− transfers an electron to the electron-deficient carbon framework and is oxidized to O₂, whereas the reduced carbon-dot intermediate subsequently transfers the electron to another O₂^•− to produce H₂O₂ and regenerate the original structure [80]. By contrast, metal-containing nanozymes mediate redox reactions through coordination-defined metal centers and accessible valence states. Prussian blue nanozymes, for example, contain a mixed-valence Fe coordination network and exhibit both SOD-like and CAT-like activities [40]. Nevertheless, structural characterization alone is insufficient to establish the elementary mechanism of O₂^•− dismutation.

Because H₂O₂ remains an oxidizing intermediate, SOD-like activity alone does not complete ROS detoxification. CAT-like nanozymes decompose H₂O₂ into H₂O and O₂, whereas GPx-like nanozymes couple H₂O₂ reduction to GSH oxidation. The spatially organized MOC-R system illustrates a sequential SOD-like/CAT-like cascade: its external TEMPO-containing ligands contributed predominantly to O₂^•− removal, whereas its internal copper clusters mediated H₂O₂ decomposition and O₂ production [81]. GPx-like activity was demonstrated in a selenophospholipid liposozyme that required GSH as the reducing substrate and showed little activity toward O₂^•− or •OH [82]. Mn₃O₄-containing protein nanoassemblies similarly promoted the GSH-dependent conversion of H₂O₂ into H₂O, accompanied by GSSG formation [83]. Coupling SOD-like activity with CAT-like or GPx-like reactions can therefore facilitate sequential ROS elimination while limiting secondary H₂O₂ accumulation.

Some antioxidant platforms have also been reported to decrease •OH or other radical signals. For example, Pt@LF reduced O₂^•− and •OH signals in electron spin resonance and competitive-reaction assays [84]. Such signal reductions should not, however, be interpreted automatically as evidence of regenerative catalysis. As with ROS-generating mechanisms, the assignment of ROS-scavenging pathways requires substrate-dependent controls, direct product identification and evidence of repeated catalytic turnover. Probe-signal suppression may instead arise from radical interception, reversible surface binding, probe adsorption, optical interference or stoichiometric material oxidation. Reversible binding can be examined by separating or washing the material and determining whether the bound species are released or the signal is restored. By contrast, sustained substrate-to-product conversion, retained activity across repeated cycles and stable post-reaction composition provide stronger evidence for catalytic ROS elimination [20]. These complementary measurements are essential for distinguishing genuine antioxidant catalysis from transient radical sequestration, sacrificial material consumption and assay-dependent signal suppression.

Taken together, ROS-generating and ROS-scavenging activities should not always be considered independently, because multifunctional nanozymes may integrate competing and sequential catalytic pathways. SOD-like catalysis generates H₂O₂ from O₂^•−, whereas CAT-like and GPx-like reactions remove the resulting H₂O₂ through O₂-producing decomposition or GSH-dependent reduction. Conversely, POD-like catalysis can use the same H₂O₂ pool to drive oxidative substrate conversion or generate reactive intermediates. The net redox outcome therefore depends on the relative rates of H₂O₂ production and consumption rather than on the presence of any single enzyme-like activity. This balance can be regulated intrinsically: valence engineering of MoO_3−x shifted H₂O₂ activation toward predominantly CAT-like or POD-like pathways [85]. Nanozyme redox behavior should therefore be evaluated under matched, physiologically relevant conditions and reported as a balance of competing catalytic fluxes, not as a single activity label.

2.3.3. RNS-related redox regulation

RNS, including nitric oxide (NO), nitrogen dioxide radical (NO₂^•) and peroxynitrite (ONOO⁻), exert concentration- and context-dependent biological effects. At physiological levels, NO regulates vascular tone and platelet function and contributes to angiogenesis, neurotransmission and wound repair. Conversely, excessive RNS accumulation can induce protein nitration, lipid peroxidation and DNA damage. ROS–RNS crosstalk may further amplify these effects because NO reacts rapidly with superoxide (O₂^•−) to form ONOO⁻ [86], [87], [88], [89]. In the cerebral microvasculature, modeling further indicates that the spatial action of NO is shaped by its site of production, vascular geometry and local degradation or scavenging, favoring localized signaling [90].

Nanozyme-mediated RNS generation is primarily exemplified by nitrite reductase (NiR)-like conversion of nitrite into NO. CuNIR-inspired nanohybrids comprising ultrasmall CuS clusters confined on nanofibrillar lysozyme assemblies generated NO from nitrite. Their temporally regulated NO release was associated with initial antibacterial activity, followed by enhanced angiogenesis and infected-wound repair [91]. MoS₂ nanosheets also displayed nitrite-reductase-like behavior, generating NO from nitrite, as confirmed by chemiluminescence and electron paramagnetic resonance, and inducing NO-dependent antibacterial effects [92]. Together, these studies indicate that catalytic RNS generation is not restricted to copper-based systems.

Conversely, RNS-scavenging nanozymes are designed to attenuate nitrosative stress by removing excessive NO and ONOO⁻. A hollow cobalt single-atom nanozyme scavenged NO and ONOO⁻ together with multiple ROS. Its combination with sustained minocycline delivery reduced microglial inflammation and promoted neural repair and functional recovery in a rat model of spinal cord injury [93]. More recently, 6-nm iron oxide nanoparticles with SOD-like and CAT-like activities were shown to scavenge NO through the formation of stable nitrosyl–metal complexes [94]. These studies illustrate two complementary modes of RNS regulation: direct removal of NO or ONOO⁻ and simultaneous ROS scavenging that may limit secondary ONOO⁻ formation. Nevertheless, metal-mediated NO sequestration should be distinguished from catalytic RNS decomposition unless active-site regeneration and sustained catalytic turnover are demonstrated.

• Other catalytic mechanisms

Beyond ROS/RNS-centered redox regulation, nanozymes can also act on defined metabolites, cofactors and biomacromolecular substrates, thereby extending their biological functions beyond oxidative-stress modulation. For example, a metal–ligand dual-site single-atom nanozyme, Ni-DAB, mimicked urate oxidase by selectively oxidizing uric acid to allantoin while reducing O₂ to H₂O₂, and this substrate-specific activity was further used in a urine-powered biofuel-cell system [41]. A photoactivated metal-free AIE polymersome nanozyme similarly catalyzed O₂-independent NADH-to-NAD⁺ conversion under hypoxia, disrupting the NAD⁺/NADH balance and impairing tumor energy metabolism [95]. In addition to these metabolite- and cofactor-converting reactions, hydrolytic nanozymes have expanded the substrate scope of nanozyme catalysis. MOF-808-Al exhibited acetylcholinesterase-like activity by hydrolyzing acetylcholine and also degraded an organophosphate simulant, thereby providing neuroprotection against organophosphate-induced injury [96]. Ce-doped lignin nanozymes hydrolyzed phosphate esters and protein amide bonds and disrupted bacterial biofilms by degrading proteinaceous and nucleotide components of extracellular polymeric substances [97]. Protease-like catalysis has also been demonstrated using a ferritin–Pd nanozyme that preferentially cleaved pro-caspase-3 and induced caspase-dependent apoptosis after targeted vesicular delivery into tumor cells [98]. Together, these examples indicate that nanozyme catalysis can be extended from ROS/RNS regulation to substrate-specific metabolic conversion, hydrolysis and protease-like biomolecular intervention, although such assignments require direct evidence of substrate consumption, product formation and catalytic specificity.

Despite their catalytic versatility, nanozymes often show reduced activity in complex biological environments, limited lesion accumulation and insufficient spatiotemporal control in vivo [99], [100]. These limitations create a need for external regulatory strategies that can activate or enhance nanozyme functions at defined sites. Ultrasound provides such a modality because it is non-invasive, deeply tissue-penetrating and spatially focusable. In addition to improving delivery and penetration through cavitation, acoustic streaming and sonoporation, ultrasound can also modulate catalytic reactions through mechanical energy input, sonochemical effects and piezoelectric polarization in responsive materials. The following section therefore defines the ultrasound effects that are most relevant to nanozyme delivery, activation and mechanistic attribution.

3. Ultrasound effects for biomedical applications

Ultrasound (US) refers to acoustic waves with frequencies above the upper limit of human hearing, generally 20 kHz [101]. These waves propagate through biological media as alternating cycles of compression and rarefaction, thereby delivering mechanical energy in a spatially and temporally controllable manner. Owing to its non-invasive nature, real-time operability, portability and favorable tissue penetration, US has long been used for diagnostic imaging and is increasingly being developed as a therapeutic energy source [102]. In biomedical settings, the effects of US are commonly divided into thermal effects and non-thermal mechanical effects, including cavitation, acoustic radiation force, acoustic streaming and sonochemical reactions. The dominant effect depends strongly on acoustic parameters, such as frequency, acoustic pressure, acoustic intensity, duty cycle, pulse duration and total exposure time [103], [104].

Thermal effects arise from the absorption of acoustic energy and its conversion into heat. When acoustic energy is focused within a defined tissue volume, local temperature elevation can be used for ablation or other thermally mediated interventions. For example, magnetic resonance-guided focused ultrasound (MRgFUS) can raise the focal temperature above 60 °C, enabling local tissue ablation and pain relief in patients with bone metastases [105]. By contrast, many ultrasound-mediated nanotherapeutic strategies are designed to operate under low-heating or non-thermal conditions, where cavitation, acoustic forces and sonochemical reactions are expected to dominate. Distinguishing these regimes is essential, because uncontrolled heating may confound mechanisms otherwise attributed to sonodynamic reactions, cavitation-enhanced delivery or ultrasound-regulated catalytic activity.

Acoustic cavitation is one of the most important non-thermal bioeffects of US. It refers to the nucleation, oscillation and possible collapse of gas bubbles in a liquid under an acoustic field. Cavitation is generally classified as non-inertial, or stable, cavitation and inertial, or transient, cavitation [106], [107]. At relatively low acoustic pressures, microbubbles undergo sustained oscillations around an equilibrium radius without violent collapse. These oscillations generate local microstreaming and shear stress, which can enhance mass transport, increase membrane permeability and facilitate the movement of therapeutic agents across biological barriers [108], [109]. At higher acoustic pressures, bubbles expand rapidly during rarefaction and collapse violently during compression. This inertial cavitation concentrates acoustic energy into a small volume over a short time scale, producing shock waves, microjets and intense local shear stress. These events can disrupt nearby structures and may also promote the formation of reactive species through cavitation-associated sonochemical reactions [107], [109].

US can also exert mechanical forces through momentum transfer and fluid motion. Acoustic radiation force arises from the interaction between the acoustic field and suspended particles, microbubbles, cells or tissues, producing a net force that can displace these objects within the insonated region [110], [111]. Acoustic streaming refers to steady fluid flow generated by the absorption and dissipation of acoustic energy in a viscous medium. This process produces local microcurrents and shear stress, particularly near boundaries, and can improve fluid mixing, interstitial transport and tissue penetration [112]. In biological settings, these flow-related effects may further contribute to improved local distribution, cellular uptake and barrier transport when coupled with cavitation or microbubble-mediated delivery.

These intrinsic bioeffects provide the mechanistic foundation for ultrasound-mediated therapeutic systems. Depending on acoustic conditions and material design, US can act as a thermal ablation source, a cavitation trigger, a mechanical force field, a sonochemical stimulus or a delivery enhancer. This distinction is particularly important for ultrasound-mediated nanozyme systems, because therapeutic efficacy may arise from different levels of coupling between the acoustic field and the material, including enhanced lesion accumulation, improved mass transfer around catalytic sites, increased membrane permeability, sonochemical reactive-species generation and modulation of catalytic activity. The following sections use this distinction to separate delivery enhancement, sonodynamic activation and piezoelectric coupling before considering how each process can be experimentally linked to nanozyme catalysis.

Collectively, targeted delivery, sonodynamic activation and piezoelectric regulation represent the three principal mechanisms through which ultrasound enhances nanozyme-based therapeutic systems. Rather than acting independently, these mechanisms often function synergistically to improve nanozyme accumulation, catalytic efficiency and therapeutic efficacy. Their mechanistic relationships are illustrated in Fig. 2, which provides an overview of how the major ultrasound bioeffects are integrated with nanozyme catalysis before each mechanism is discussed in detail.

Fig. 2.

Schematic illustration of ultrasound-triggered synergistic nanozyme therapy. Ultrasound generates multiple physical bioeffects, including cavitation, acoustic radiation force, acoustic streaming and sonochemical reactions, which collectively enable three major ultrasound-mediated therapeutic mechanisms. A) Targeted delivery enhances nanozyme accumulation by transiently overcoming biological transport barriers. B) Sonodynamic therapy (SDT) activates sonosensitizers and synergizes with nanozyme catalysis to promote reactive oxygen species (ROS) generation and catalytic amplification. C) Piezoelectric effect converts ultrasound-induced mechanical deformation into polarization charges that enhance catalytic activity and bioelectrical regulation, ultimately leading to downstream cellular responses.

3.1. Targeted delivery effect

Targeted delivery remains a major constraint on nanozyme therapy in vivo. Even when nanozymes show robust catalytic activity in vitro, their effects in tissues are often limited by poor lesion accumulation, insufficient penetration across dense extracellular matrices or endothelial barriers, and suboptimal cellular uptake. Ultrasound is therefore important not only as an external trigger, but also as a means of transiently lowering transport barriers and creating the conditions under which nanozyme catalysis can occur at the relevant site and depth [109], [113], [114].

A central mechanism is sonoporation. Early scanning electron microscopy revealed localized membrane disruptions after ultrasound exposure [115], and later single-cell patch-clamp studies showed rapid increases in transmembrane current during microbubble-assisted insonation, consistent with transient pore formation and ion flux across the plasma membrane [116]. Uptake route depends strongly on bubble dynamics. Under milder acoustic conditions, stable cavitation and membrane deformation tend to favor endocytic uptake, whereas stronger bubble-cell interactions promote pore formation and more direct cytoplasmic entry [109], [117]. Recent work has further shown that ultrasound-driven microbubbles can generate stable cyclic microjets that puncture the membrane and facilitate delivery even at relatively mild pressures, providing a more explicit physical basis for microbubble-mediated transport [118].

These effects underpin ultrasound-targeted microbubble destruction and related delivery strategies. By amplifying vascular permeability, extravasation and local transport, microbubbles can improve target-site payload accumulation while confining the effect to the insonated region. Focused ultrasound combined with microbubbles has been particularly important in the central nervous system, where it enables transient and localized blood-brain barrier (BBB) opening for targeted brain delivery [119]. This concept has now progressed to early clinical studies, including recent work in Alzheimer's disease showing transient BBB opening with real-time cavitation monitoring and acceptable short-term safety [120]. A complementary strategy is to eliminate the gas core altogether. Acoustically activatable liposomes, for example, enable low-intensity ultrasound-triggered drug uncaging with reduced thermal burden and improved translational plausibility [14].

For ultrasound-mediated nanozyme systems, however, improved delivery is enabling rather than sufficient. Transient barrier opening, deeper penetration or enhanced uptake become mechanistically meaningful only when they increase nanozyme access to pathological substrates, intracellular compartments or lesion-resident reaction environments. This distinction matters because many ultrasound-responsive carriers improve localization without contributing catalytic function themselves. By contrast, in delivery-enabled nanozyme systems, ultrasound-assisted transport directly expands the opportunity for enzyme-mimetic catalysis in vivo. The key benchmark is therefore not delivery alone, but whether delivery increases catalytic engagement at the disease site and improves the same therapeutic endpoint.

• Sonodynamic therapy (SDT) effect

Unlike photodynamic therapy (PDT), which is limited by shallow light penetration, SDT uses ultrasound as a deeply penetrating (>10 cm) and externally controllable energy source to generate cytotoxic ROS in situ [15]. The prevailing mechanism of SDT involves acoustic cavitation, in which bubble oscillation and collapse create transient high-energy microenvironments and, in some systems, sonoluminescence that activates nearby sonosensitizers [16]. Recent studies further suggest that SDT efficacy depends not only on sensitizer design, but also on how efficiently cavitation is induced and how effectively cavitation-derived energy is converted into chemical reactivity [121].

However, conventional SDT remains strongly constrained by the oxygen and redox status of diseased tissues. Most sonodynamic processes require molecular oxygen to generate cytotoxic ROS, but solid tumors, infected wounds and chronic inflammatory lesions often contain hypoxic regions and abnormal H₂O₂/GSH levels [122], [123], [124]. These microenvironmental features restrict oxygen-dependent ROS production and weaken the therapeutic efficacy of sonosensitizer-based SDT. In this context, nanozymes provide a route to overcome the microenvironmental limitations of conventional SDT by reshaping local oxygen and redox homeostasis. CAT-like nanozymes can decompose endogenous H₂O₂ to generate O₂, thereby relieving hypoxia and sustaining oxygen-dependent ROS production during SDT. In parallel, POD-like or Fenton/Fenton-like activities can convert H₂O₂ into highly cytotoxic •OH, while OXD-like activity can promote O₂-derived ROS generation. These enzyme-mimetic reactions convert pathological substrates into therapeutic chemical species and enable cascade ROS amplification beyond sonosensitizer activation alone.

The key question, therefore, is not whether ultrasound increases ROS, but whether it produces a nanozyme-dependent catalytic advantage beyond conventional SDT. Some recent ultrasound-triggered therapeutic nanoplatforms, including acoustic-sensitive cuproptosis systems and mechanophore-based ultrasound therapies, show strong biological activity but are conceptually closer to ultrasound-responsive therapeutic materials than to bona fide nanozyme-driven catalytic systems, because enzyme-like turnover is not the dominant validated mechanism [125], [126]. Within the broader landscape of ultrasound-activated theranostic materials, SDT-coupled nanozyme systems are best viewed as a distinct mechanistic subclass in which sonodynamic and enzyme-mimetic processes act in concert rather than merely in parallel [9].

3.2. Piezoelectric effect

Unlike traditional sonodynamic systems that mainly rely on the cavitation effect of ultrasound, piezoelectric materials can convert ultrasound-–induced mechanical deformation into surface polarized charges and local electric potentials, which promote charge separation and directional migration of charge carriers [127], [128]. The transferred electrons then react with O2, H2O or H2O2 to form ROS such as H2O2, •OH and O^2•− [129], [130], [129], [130]. This piezoelectric-–enabled ROS generation circumvents the nonlinear threshold inherent to cavitation, thereby enabling precise tuning of ROS output via ultrasound parameters.

Materials design can further couple piezoelectric charge generation to intrinsic nanozyme catalysis. In oxygen-vacancy-rich (BiFe)₀.₉(BaTi)₀.₁O_3−x (BFBT) nanoreactors, oxygen vacancies not only enhance piezoelectric charge separation but also reshape the electronic structure of catalytic Fe sites [19]. Electron-rich vacancies increase the Fe(II) fraction and accelerate Fe(III)/Fe(II) cycling, thereby strengthening peroxidase-like activity and sustaining Fenton-like ROS production without externally added H₂O₂. This defect-engineering strategy differs from ultrasound power switching, which mainly biases enzyme-like pathway preference, because it improves the intrinsic catalytic competence of the nanozyme. Consistently, ultrasound intensity can regulate nanozyme behavior by modulating piezoelectric charge-carrier flux: low power (≤ 0.5 W cm^-2) favors catalase-like activity, whereas higher power (> 0.5 W cm^-2) enhances peroxidase-like and sonodynamic activities [131].

In addition to catalytic redox regulation, piezoelectric nanozymes can function as physical bioelectrical interfaces. Ultrasound-induced polarization charges reshape local electric fields, surface potentials and cell–material interactions, thereby eliciting cellular responses that are not reducible to ROS chemistry. The biological outcome is governed by the magnitude, polarity and temporal profile of the surface potential [3]. Depending on these parameters, polarization charges may favor interfacial electron transfer for ROS generation [17], induce transient ionic gradients and membrane depolarization [128], or perturb protein adsorption and receptor-mediated signaling at the cell–material interface [132]. Two recent studies have begun to define how these piezoelectric bioelectrical cues are transduced in nanozyme-containing systems. One study showed that piezoelectric stimulation promotes Ca²⁺ influx through voltage-gated and mechanosensitive channels, followed by Ca²⁺/calmodulin-dependent regenerative signaling in infected bone [133]. Another study identified an integrin β1-mediated route, in which the FAK–PI3K–AKT/ERK axis promotes annulus fibrosus repair [134]. These findings indicate that piezoelectric nanozymes integrate catalytic redox modulation with a distinct bioelectrical regulatory axis.

Therefore, piezoelectric coupling should be discussed as a specific mechanistic interface rather than as a broad synonym for ultrasound responsiveness. Pure piezocatalytic systems rely mainly on ultrasound-induced polarization and ROS generation, whereas piezoelectric bioelectrical systems regulate ion flux and cell signaling. Piezoelectric–nanozyme hybrids, in turn, require evidence that polarization charges modulate enzyme-like catalysis, active-site redox cycling or bioelectrical transduction.

4. Therapeutic applications of ultrasound-mediated nanozyme systems across disease contexts

The preceding sections outlined the catalytic properties of nanozymes and the major ways in which ultrasound can regulate delivery, penetration, catalytic activation and piezoelectric responses. These mechanisms provide the basis for disease-oriented therapeutic design, but their value depends strongly on pathological context. In tumors and infections, ultrasound-mediated nanozyme systems are often used to amplify oxidative damage and disrupt protective microenvironments. In contrast, in inflammatory, degenerative and ischemic diseases, they are more commonly designed to buffer oxidative stress, regulate inflammation and support tissue repair. Thus, the desired catalytic output must be redefined for each disease context before ultrasound responsiveness can be considered therapeutic.

Collectively, the major therapeutic mechanisms of ultrasound-mediated nanozyme systems across seven representative disease categories covered in this review are schematically illustrated in Fig. 3. The following section therefore uses disease context as the organizing principle, emphasizing how ultrasound-responsive mechanisms and nanozyme catalytic functions are matched to disease-specific barriers, therapeutic objectives and translational constraints.

Fig. 3.

Schematic illustration of ultrasound-mediated nanozyme therapeutic applications across multiple disease models.

4.1. Tumors

Conventional cancer therapies, including chemotherapy, radiotherapy and surgical resection, remain limited by drug resistance, systemic toxicity, and the high risk of tumor metastasis and recurrence [135]. Nanozyme-based therapeutic systems have emerged as a promising strategy for tumor treatment because they can exploit the biochemical abnormalities of the tumor microenvironment (TME) to drive catalytic reactions. However, their efficacy is constrained by the complex architecture of solid tumors. Tumor tissues undergo extensive extracellular matrix (ECM) remodeling, which creates a major barrier to nanozyme delivery. Collagen deposition, matrix crosslinking and altered ECM organization increase tumor stiffness, compress blood and lymphatic vessels, elevate interstitial fluid pressure and reduce convective transport [136], [137]. As a result, nanomedicines often accumulate near tumor vessels but penetrate poorly into deeper tumor regions. Beyond this physical barrier, the ECM also regulates tumor-cell invasion, pre-metastatic niche formation, immune-cell infiltration and responses to immunotherapy [138]. Therefore, effective ultrasound-mediated nanozyme therapy requires not only tumor accumulation, but also sufficient access to ECM-restricted regions where catalytic activation can exert therapeutic effects.

Although the ECM imposes a formidable transport barrier, ultrasound can partially improve nanoparticle access to solid tumors. In head and neck squamous cell carcinoma xenografts, non-destructive pulsed focused ultrasound reduced interstitial fluid pressure in the tumor core and increased nanoparticle delivery and penetration into central tumor regions [139]. However, most ultrasound-mediated nanozyme systems reported so far do not rely solely on direct ultrasound-induced ECM remodeling. Instead, they more often use ultrasound-responsive or tumor-responsive nanoplatform design to improve nanozyme access.

One strategy is to use ultrasound as a trigger for structural disassembly. Lin et al. developed ultrasound- and glutathione-responsive vesicles assembled from Janus Au-MnO nanoparticles (Fig. 4) [140]. Ultrasound converted the approximately 100 nm vesicles into Janus Au-MnO nanoparticles of about 20 nm, thereby improving tumor penetration; subsequent glutathione-mediated MnO degradation released smaller Au nanoparticles of about 10 nm as cavitation nucleation sites and Mn²⁺ for chemodynamic therapy. A complementary strategy is to engineer nanozymes that shrink in response to the TME. In the acid-dissociable ZIF@GOx/GQDs nanozyme generator, glucose oxidase-mediated acidification promoted ZIF-8 framework disassembly in the mildly acidic TME, converting an approximately 150 nm parental nanoplatform into ultrasmall peroxidase-like graphene quantum dots of about 3 nm [141].

Fig. 4.

Mechanisms of ultrasound-mediated nanozyme therapy in tumor. a) The mechanism of US-responsive disassembly and synergistic SDT and CDT. b) The SDT ability of JNP Ve under US irradiation with and without GSH treatment. Reproduced with permission from ref. [140], Copyright © 2020 Wiley-VCH. c) Schematic illustration of the activation mechanism of P-Por and P-Por-Os under US irradiation. d) •OH and •O₂^– trapped by DMPO in different substrates under US irradiation. Reproduced with permission from ref. [148], Copyright © 2024 American Chemical Society. e) Se²⁻ from Mn/Se-NE@FCSinduces apoptosis, pyroptosis, and necroptosis in MB49 tumor cells, leading to PANoptosis and ICD; f) Mn²⁺ from Mn/Se-NE@FCS promotes the binding of PANoptosis-released dsDNA to cGAS, enhancing STING pathway activation. Reproduced with permission from ref. [160], Copyright © 2026 Wiley-VCH GmbH.

For tumors protected by specialized anatomical barriers, ultrasound can instead act as an upstream delivery-enabling step. This strategy is particularly relevant to glioblastoma, where the BBB and blood-brain tumor barrier (BBTB) restrict systemic access of nanomedicines. In a glioblastoma-targeting nanozyme study, SonoVue microbubbles were injected through the tail vein, followed by ultrasound irradiation at 1 MHz and 1 W cm⁻² with a 50% duty cycle for 3 min [142]. Evans blue imaging confirmed transient BBB opening, and NIR-II fluorescence imaging further showed enhanced nanozyme accumulation at the glioblastoma site. Thus, in anatomically shielded tumors, ultrasound may improve nanozyme delivery not by reducing ECM resistance directly, but by transiently opening vascular barriers before catalytic therapy. Beyond size transformation and ultrasound-assisted barrier opening, surface engineering with targeting ligands, such as folic acid [143], RGD peptides [144] or hyaluronic acid [142], can further enhance tumor-cell recognition and local retention of nanozyme-containing systems.

After tumor penetration, therapeutic efficacy depends on whether nanozymes can generate sufficient ROS within the TME. Solid tumors provide a favorable redox niche for catalytic therapy, including weak acidity, elevated H₂O₂, abundant GSH and regional hypoxia. Nanozymes can exploit these cues through peroxidase-like, Fenton-like, oxidase-like or cascade catalytic reactions to generate cytotoxic •OH, O₂^•− and ¹O₂, thereby overwhelming tumor antioxidant defenses [145]. Ultrasound can further intensify this process when it directly enhances nanozyme catalysis. In CaF₂ nanozymes, ultrasound strengthened peroxidase-like, as shown by increased TMB oxidation at 652 nm and stronger DMPO-•OH ESR signals in a duration- and power-density-dependent manner [10].

Hypoxia and antioxidant buffering are two major barriers to ROS-based tumor therapy. Hypoxic tumor regions restrict oxygen-dependent ROS generation [146], whereas elevated intracellular GSH buffers oxidative damage [147]; together, these features weaken oxidative-stress-based treatment. Ultrasound-mediated nanozyme systems can counter these barriers by supplying O₂, depleting GSH or coupling both processes with catalytic ROS amplification. Porphyrin-based platforms illustrate how these functions can be integrated within a single ultrasound-responsive architecture. In P-Por-Os, the osmium-coordinated sites serve as artificial enzyme centers (Fig. 4) [148]. These atomically dispersed Os sites confer POD-like activity, enabling H₂O₂ conversion into •OH and O₂^•− under acidic TME-like conditions, while also supporting H₂O₂ decomposition to generate O₂ for hypoxia relief. The conjugated porphyrin framework provides the sonosensitizing component: porphyrin units mediate ultrasound-responsive ¹O₂ generation, and Os coordination narrows the bandgap of the framework, facilitating ultrasound-induced charge excitation and further enhancing •OH and O₂^•− production.

This design principle is further extended in PEG-CuP-COF@ΔSt, where the porphyrin-containing COF framework provides sonodynamic activity, while the Cu-coordinated network introduces a multienzyme-like redox cascade [149]. Its SOD-like activity converts O₂^•− into H₂O₂, thereby supplying substrate for POD-like •OH generation. In parallel, GPx-like activity consumes GSH and converts it into GSSG, with nearly 80% GSH depletion reported in vitro. Thus, beyond porphyrin-mediated sonodynamic ROS generation, PEG-CuP-COF@ΔSt weakens tumor antioxidant buffering through a SOD/POD/GPx-like cascade.

Excessive ROS generation provides the mechanistic link between catalytic redox disruption and regulated tumor-cell death. Once ROS production exceeds the buffering capacity of GSH and related antioxidant systems, tumor cells undergo lipid peroxidation, mitochondrial injury and ion-homeostasis collapse, which together initiate multiple death programs [150], [151], [152]. Ferroptosis provides a representative route by which ultrasound-amplified catalytic stress is converted into lipid-peroxide-driven tumor-cell death. This process is defined by iron-dependent lipid peroxidation and failure of the GSH/GPX4 antioxidant axis, rather than by nonspecific ROS accumulation alone. Consistent with this mechanism, ultrasound-mediated cascade catalysis has been shown to reduce GPX4 expression, enhance lipid peroxidation and induce ferroptosis-associated mitochondrial shrinkage and cristae loss in CT26 tumor cells [153]. Ferroptotic stress can further intersect with immunogenic cell-death signaling. In ZFPG nanozymes, this ferroptotic state was accompanied by calreticulin exposure, HMGB1 release, ATP release and HSP90 upregulation, thereby enhancing dendritic-cell maturation and CD8⁺ T-cell infiltration [154]. Thus, ferroptosis-oriented sononanozyme design should be evaluated not only by lipid peroxidation and GPX4 inhibition, but also by whether ferroptotic damage can be converted into productive antitumor immune signaling.

Pyroptosis is an inflammatory form of regulated cell death that links ultrasound-amplified oxidative stress to antitumor immunity. Its execution depends on gasdermin cleavage and membrane-pore formation, but different sononanozyme systems can engage distinct gasdermin axes. Both systems were reported to induce pyroptosis, but through different execution routes: LFO@GOx [155] activates the canonical inflammasome-caspase-1-GSDMD axis, whereas PEG-CuP-COF@ΔSt [149] mainly converts caspase-3 signaling into GSDME-dependent pyroptotic membrane disruption. In LFO@GOx nanoreactors, ultrasound-enhanced catalytic ROS generation activated TXNIP, promoted NLRP3 inflammasome signaling, and increased caspase-1 cleavage and GSDMD processing in 4T1 breast cancer cells. By contrast, PEG-CuP-COF@ΔSt increased cleaved caspase-3 and generated the pore-forming N-terminal fragment of GSDME (N-GSDME) in RM-1 prostate tumors, whereas N-GSDMD was not obviously detected. This distinction matters because the GSDMD axis reflects inflammasome-driven pyroptosis, whereas the GSDME axis can convert apoptosis-associated caspase-3 signaling into inflammatory membrane rupture. Therefore, pyroptosis-oriented sononanozyme systems should be evaluated by the specific gasdermin pathway they activate, rather than by general inflammatory cell-death markers alone.

Cuproptosis introduces a metal-metabolic mode of tumor-cell death into ultrasound-mediated nanozyme therapy. In biodegradable Cu₃P-Cel sonozymes, the exogenous Cu component serves as a tumor-responsive metal reservoir. After degradation in the acidic and reducing TME, released Cu ions accumulate intracellularly. Excess Cu can bind to lipoylated components of the tricarboxylic-acid cycle, promoting protein aggregation, destabilizing mitochondrial metabolic enzymes and inducing proteotoxic stress. In parallel, ultrasound amplifies ROS generation, which further weakens redox buffering and aggravates mitochondrial dysfunction [156]. Thus, Cu-containing sononanozymes provide a route to couple catalytic oxidative stress with copper-driven metabolic collapse, extending ultrasound-mediated nanozyme therapy beyond ROS-dominant ferroptosis or gasdermin-dependent pyroptosis.

Tumor-cell death can initiate antitumor immunity by releasing tumor antigens and DAMPs that promote dendritic-cell activation and T-cell priming. However, these signals are often restrained by the immunosuppressive TME, where inefficient antigen presentation, immune-checkpoint signaling, Treg and MDSC accumulation, and M2-like tumor-associated macrophages limit durable immune control [157]. Thus, local ROS-mediated killing does not necessarily translate into systemic antitumor immunity. Ultrasound-mediated nanozyme therapy is therefore moving beyond catalytic tumor ablation toward immune-microenvironment remodeling. By integrating ultrasound-enhanced ROS generation, regulated tumor-cell death and catalytic TME modulation, these systems can amplify immunogenic signaling, reshape immune-cell composition and improve responsiveness to immunotherapy. In 4T1 tumor models, PdSA/Ti_3−xC₂Ty plus ultrasound induced ICD signals, including calreticulin exposure, HMGB1 release and ATP secretion, promoting dendritic-cell maturation [158]. It also shifted tumor-associated macrophages from an M2-like to an M1-like phenotype and increased CD4+ and CD8+ T-cell responses, contributing to suppression of untreated distant tumors. This immune-remodeling strategy was further refined in γ-MnO_2−x/CD sonozymes, which released Mn ions in response to the TME to activate the cGAS-STING pathway [159]. This innate immune pathway enhanced type I interferon signaling, dendritic-cell maturation and downstream T-cell priming.

Compared with γ-MnO_2−x/CD, Mn/Se-NE@FCS further links Mn-based STING activation with PANoptosis-associated tumor-cell death. Although the study detected concurrent changes in apoptosis-, pyroptosis- and necroptosis-related markers, the upstream coordination of these death pathways was not fully resolved. Nevertheless, in bladder cancer models, Mn/Se-NE@FCS-mediated immune activation improved the response to PD-1 blockade, suggesting a potential direction for designing nanozyme therapies that combine ultrasound-amplified oxidative stress, innate immune sensing and checkpoint sensitization (Fig. 4) [160].

Antigen-centered immune remodeling represents a distinct direction from redox- or STING-centered strategies. In CPIP@EV-CM, the melanoma cell membrane provides tumor-associated antigens, whereas S. aureus-derived extracellular vesicles supply pathogen-associated adjuvant signals that promote dendritic-cell uptake and maturation [161]. This heterogenic EV-CM membrane therefore functions as a vaccine-like interface, co-delivering antigenic and costimulatory cues to strengthen dendritic-cell-mediated antigen presentation and cytotoxic T-cell priming. More broadly, antigen-centered immune engineering can also be achieved by amplifying tumor-cell recognition rather than by delivering tumor antigens to dendritic cells. In PATCH, ultrasound-activated porphyrinic coordination network (PCN) nanozymes catalyze proximity labelling on tumor-cell surfaces to generate high-density synthetic antigen clusters. These clusters improve tumor-cell recognizability and drive BiTE-mediated T-cell receptor clustering, offering another route to strengthen antitumor immune responses [162].

Ultrasound-mediated nanozyme therapy may also be extended to metabolic intervention by loading glucose oxidase onto nanozyme platforms to consume intratumoral glucose and implement starvation therapy [163]. Another emerging direction is to remodel cellular redox metabolism through the NADH/NAD+ redox couple, thereby influencing mitochondrial electron transport and intracellular oxygen consumption [164].

Successful ultrasound-mediated nanozyme cancer therapy depends less on the number of incorporated functions than on whether catalytic design is matched to specific tumor barriers. Effective systems generally need to improve access to poorly penetrable tumor regions, respond selectively to tumor-microenvironmental cues and convert local redox damage into durable biological outcomes, including regulated tumor-cell death, immune activation or sensitization to immune-checkpoint blockade. Recent original studies published in 2025–2026 suggest that the field is moving in this direction, from broadly multifunctional formulations toward acoustically programmable and tumor-context-matched designs. This shift is reflected first in piezoelectric or mechanically charged nanozymes, which use ultrasound-induced polarization or charge regulation to activate catalytic centers [165], [166]. A second direction involves defect-engineered, heterojunction and high-entropy sonozymes that link improved charge separation to multienzyme-like ROS amplification [167], [168], [169]. A third direction is the development of biomimetic, biodegradable or clinically inspired platforms that enhance tumor targeting, hypoxia/GSH remodeling, STING activation or checkpoint sensitization [154], [156], [160], [170].

At present, no single catalytic mechanism can be considered universally superior, because efficacy depends on tumor model, material dose, ultrasound parameters and therapeutic endpoints. Cascade catalytic systems are generally more versatile than single-function nanozymes because they can integrate substrate generation, oxygen supply, O₂-dependent oxidase-like reactions, POD/Fenton-like •OH production and antioxidant depletion within tumor-specific reaction networks. In this framework, GOx- or SOD-like activities supply H₂O₂, CAT-like activity relieves hypoxia and supports oxygen-dependent ROS generation, OXD-like and POD/Fenton-like reactions amplify oxidative stress, and GPx-like or GSHOx-like activities weaken antioxidant buffering. Piezoelectric and sonozyme systems add another layer of control by exploiting ultrasound-driven charge separation, although their performance remains highly dependent on acoustic conditions. Future evaluation should therefore move beyond ROS yield alone and test whether each catalytic module explains defined outcomes, including regulated tumor-cell death, immune activation and checkpoint sensitization, while reporting acoustic parameters, catalytic mechanisms, material degradation, metal-ion release and immune-safety windows more systematically. Representative platforms, their nanozyme activities, ultrasound parameters and therapeutic mechanisms for tumor are summarized in Table 1, Table 2, respectively.

Table 1.

Comparative summary of representative ultrasound-mediated nanozyme systems for therapeutic applications.

Disease category	Material name	Disease model	Material platform/type	Nanozyme activity	Main mechanism	Therapeutic outcome	Key limitation	Ref.
Tumor	Janus Au–MnO NP vesicles	97H orthotopic liver tumor-bearing mice	Ultrasound-responsive delivery system; Mn-based nanozyme; Au-based sonosensitizer	Mn2+-mediated Fenton-like	US/GSH-triggered JNP Ves disassembly enhances tumor penetration and couples SDT with Mn2+-mediated CDT	Inhibited orthotopic liver tumor growth with enhanced tumor penetration and MR/PA imaging	Mixed delivery/SDT/CDT contribution; incomplete in vivo acoustic reporting	[140]
Tumor	CaF2 nanozyme	4 T1 breast tumor- and H22 liver tumor-bearing mice	Ultrasound-amplified nanozyme; inorganic nanozyme; hydrogel depot system	POD-like	US-amplified POD-like catalysis converts H2O2 into ·OH and induces Ca2+-overload-associated tumor killing.	Achieved 69.2% and 68.8% tumor-growth inhibition in 4 T1 and H22 models, respectively	Intratumoral hydrogel administration; acoustic pressure NR	[10]
Tumor	BTO/MoS2@CA	CT26 colon tumor-bearing mice	Piezoelectric material; MoS2-based nanozyme	POD-like	US-induced BTO piezoelectric polarization enhances MoS2 POD-like catalysis, while CA supplies H2O2 to drive ferroptotic oxidative stress.	Suppressed CT26 tumor growth with ferroptosis-associated tumor damage	Key acoustic parameters NR; piezoelectric contribution needs clearer in vivo validation	[153]
Tumor	P-Por-Os	Hepa1-6 subcutaneous and orthotopic HCC-bearing mice	Porphyrin-based sonosensitizer; Os-based artificial enzyme	POD- and CAT-like	Os-mediated catalysis supplies O2 and amplifies ROS, while porphyrin-mediated SDT disrupts redox homeostasis.	Reduced subcutaneous HCC tumor growth and orthotopic liver tumor nodules	Os-catalysis and porphyrin-SDT contributions require clearer separation	[148]
Tumor	PEG-CuP-COF@ΔSt	RM-1 prostate tumor-bearing mice with metastasis	Porphyrin-based sonosensitizer; Cu-based multienzyme nanozyme	SOD-, POD-, and GPx-like	Bacterial tumor targeting delivers Cu-porphyrin COF for SDT-coupled SOD/POD/GPx-like redox cascades and GSH depletion	Suppressed primary, distant and bone-metastatic RM-1 tumors with enhanced antitumor immunity	Bacteria-mediated delivery adds biosafety and regulatory complexity	[149]
Tumor	PdSA/Ti3−xC2Ty	Bilateral 4 T1 breast tumor-bearing mice	Ti3−xC2Ty MXene-based sonosensitizer; Pd single-atom nanozyme	POD- and CAT-like	Pd single atoms and US-excited MXene coordinate catalytic ROS generation, SDT and immune microenvironment remodelling	Eradicated primary 4 T1 tumors, suppressed distant tumors and prolonged survival	Mixed NCT/SDT contribution; duty cycle and acoustic pressure NR	[158]
Tumor	Cu3P-Cel sonozymes	Bilateral 4 T1 breast tumor-bearing mice	Cu3P-based sonosensitizing nanozyme	Fenton-like	US-responsive Cu3P-Cel promotes ROS generation, cuproptosis and cGAS–STING activation to sensitize immune-checkpoint blockade	Enhanced primary and distant 4 T1 tumor control and sensitized tumors to αPD-L1 therapy	US exposure duration NR	[156]
Tumor	Mn/Se-NE@FCS	Subcutaneous and bilateral MB49 bladder tumor-bearing mice	Ultrasound-responsive Mn/Se nanozyme; folic acid–chitosan-coated immunomodulatory platform	POD- and GPx-like	US-amplified Mn/Se nanozyme catalysis links oxidative stress to STING activation and PANoptosis-associated antitumor immunity	Improved local and distant MB49 tumor control and enhanced anti-PD-1 responsiveness	PANoptosis mechanism remains incompletely resolved; intratumoral administration	[160]
Antibacterial	Pip-loaded catalytic MBs (MB-Pip)	PAO1 biofilm-induced chronic lung infection in mice	Ultrasound-responsive microbubble delivery system; Fe3O4 nanozyme	POD-like	US-induced microbubble cavitation disrupts PAO1 biofilms and enhances Fe3O4/Pip penetration, while Fe3O4 POD-like catalysis generates ·OH for EPS degradation and bacterial killing	Efficiently treated PAO1 biofilm-induced chronic lung infection, improving bacterial clearance, lung inflammation and survival	Limited bacterial targeting of MB-Pip; mixed physical disruption, POD-like catalysis, antibiotic release and macrophage activation complicate mechanism attribution	[176]
Antibacterial	ACPCAH	MRSA-infected diabetic wound in rats	P-C3N5-based piezoelectric nanozyme system; Au/Cu1.6O/P-C3N5 multienzyme-like nanozyme	SOD-, CAT-, GOx-, POD- and NOS-like	US-induced piezoelectric polarization and sonothermal effects amplify the SOD-CAT-GOx-POD/NOS cascade for ROS/RNS-mediated antibacterial activity and microenvironment repair	Accelerated MRSA-infected diabetic wound healing by reducing inflammation, relieving hypoxia, lowering local glucose and promoting angiogenesis	Complex pentaenzyme-like cascade with concurrent piezoelectric and sonothermal enhancement; individual catalytic and ultrasound-mediated contributions remain difficult to isolate in vivo	[186].
Antibacterial	Ti-MnO2-CPA@Ce6 coating	S. aureus-infected diabetic implant-associated infection in rats	Ce6 sonodynamic implant coating; MnO2 cascade nanozyme	GOx- and POD-like	US-activated Ce6 SDT kills S. aureus and releases antigens, while MnO2 GOx/POD-like catalysis and Mn2+ release promote adaptive immune activation.	Suppressed S. aureus diabetic implant infection, activated adaptive antibacterial immunity and promoted bone-implant osseointegration	Mixed SDT, cascade nanozyme catalysis and adaptive immune activation; implant-coating format may limit applicability to surgically accessible implant infections	[192]
Atherosclerosis	SMC-HA	ApoE−/− carotid and aortic atherosclerotic plaque-bearing mice	HA-modified Mn-N4 single-atom sonozyme	CAT- and POD-like	Mn-N4 sites mediate SCT under US, while CAT-like activity relieves plaque hypoxia and POD-like activity promotes ROS generation to induce mitochondrial apoptosis of inflammatory macrophages	Reduced plaque burden in carotid and aortic AS models, decreased plaque macrophages and increased collagen content	Long-term recurrence/chronic toxicity and large-animal validation needed; US parameter optimization required	[199]
Atherosclerosis	LyP-1Lip@HS	Atherosclerotic aortic plaque-bearing mice	LyP-1-functionalized LIFU-responsive lipid-gated Prussian blue nanozyme/H2S-donor delivery system	CAT-, SOD- and POD-like	LIFU-induced cavitation disrupts the lipid membrane, exposing HMPB to restore nanozyme activity and releasing SAC for endogenous H2S generation	Reduced aortic plaque burden, necrotic core formation, oxidative stress and inflammatory cytokines; improved ABCA1/CD206 expression	Frequency, duty cycle and exposure duration NR in main text; gas delivery/release control remains a translation issue	[203]
Atherosclerosis	Nb2C-Pt@HA-PEG	High-cholesterol diet-fed ApoE−/− aortic and aortic-root atherosclerotic plaque-bearing mice	HA/PEG-modified Pt-decorated piezoelectric-like Nb2C MXene nanozyme	CAT- and SOD-like	Pt decoration induces piezoelectric-like Nb2C-Pt interfacial polarization under US, promoting charge separation/electron transfer and enhancing ROS-scavenging catalysis	Reduced whole-aorta and aortic-root plaque area, decreased intraplaque ROS and macrophage infiltration, increased collagen/fibrous-cap stability and reduced necrotic core	Frequency, intensity and duty cycle NR in main text; POD-like activity still lower than natural enzymes or SAzymes	[204]
Neurodegenerative disease (PD)	Q@CeBG	MPTP-induced PD mice; SH-SY5Y/BV2 cell models	FUS-assisted BBB delivery system; GSH-modified BSA nanoreactor containing CeO2 nanozyme and quercetin	SOD- and CAT-like	FUS/MB-mediated BBB opening enhances brain delivery, while CeO2-mediated ROS scavenging and quercetin-associated immunomodulation reduce neuronal oxidative stress and promote M1-to-M2 microglial polarization	Protected dopaminergic neurons, reduced oxidative and inflammatory injury, and improved behavioural deficits in PD mice	FUS mainly serves as a delivery enhancer rather than catalytic trigger; CeO2, quercetin and FUS-delivery contributions are coupled	[221]
Neurodegenerative disease (AD)	USI-MHOF@KD8	3 × Tg-AD mice; Aβ42 aggregation and HT22 cell models	Ultrasound-responsive HOF sonosensitizer; KD8-modified Aβ-targeting artificial enzyme platform	CAT- and SOD-like	US activates HOF to generate 1O2 for Aβ oxidation and aggregation inhibition, while CAT/SOD-like activities decompose excess ROS and support microenvironmental recovery	Reduced cerebral Aβ plaque burden, alleviated neuronal injury and improved cognitive deficits in 3 × Tg-AD mice	ROS-generating Aβ oxidation and antioxidant microenvironmental effects are coupled; long-term clearance and acoustic safety need further validation	[229]
Osteochondral diseases (OA)	Mn-nanobowls	ACLT-induced OA in SD rats	Ultrasound-propelled polymer nanobowls decorated with MnO2 nanozyme	CAT-like	US propulsion enhances deep cartilage penetration, while MnO2 decomposes H2O2, scavenges ROS and generates O2 to protect chondrocytes and cartilage matrix	Penetrated up to 307 μm into cartilage after US treatment; reduced cartilage degeneration and improved histological and micro-CT outcomes in OA rats	Mainly local intra-articular delivery; US contribution is primarily transport/penetration rather than direct catalytic activation	[242]
Osteochondral diseases (OA)	MOF(Hf)-Pt	ACLT-induced OA in SD rats	US-responsive piezoelectric MOF(Hf) framework loaded with Pt nanozyme	SOD- and CAT-like	US-induced piezoelectric response promotes charge transfer and enhances SOD/CAT-like catalytic activity for ROS scavenging and cartilage protection	Reduced OARSI scores by 76.90% and 82.66% after 4 and 8 weeks; inhibited inflammation and promoted cartilage regeneration	Frequency and detailed acoustic reporting are incomplete; piezoelectric catalysis and biological regeneration effects remain partly coupled	[243]
Osteochondral diseases (RA)	Rh/SPX-HSA	CIA mouse model of RA	Sparfloxacin sonosensitizer-loaded concave-cubic Rh nanozyme with HSA-mediated inflammatory-joint targeting	POD- and CAT-like	US activates SPX-mediated SDT to generate 1O2, while Rh nanozyme supplies O2, produces •OH and relieves hypoxia, inducing mitochondrial dysfunction and apoptosis in pathogenic FLSs	Suppressed paw swelling and arthritis score, reduced inflammatory cytokines, alleviated bone/cartilage destruction and angiogenesis in CIA mice	Pro-oxidative strategy requires precise lesion targeting and acoustic dosing to avoid collateral oxidative injury	[246]
Osteochondral diseases (RA)	BEVs@Ru-HFO	CIA mouse model of RA	Bone marrow stem cell-derived EVs encapsulating Ru-cluster-anchored hydroxylated Fe2O3 nanozyme	SOD- and CAT-like	BEVs improve inflammatory-joint targeting and local retention; Ru-HFO scavenges ROS, relieves hypoxia, repolarizes macrophages and suppresses osteoclastogenesis; US enhances accumulation/permeability	Reduced synovitis, inflammatory cytokines and bone erosion; promoted M2-like macrophage responses and inhibited osteoclast formation in CIA mice	US parameters are not fully reported in the main text; EV-based manufacturing and batch consistency may complicate translation	[247]
Ischemia-reperfusion injury	CeO2/MnOx@UK@PFP	Rat inferior vena cava thrombosis model	LIPUS-triggered ultrasmall nanobubble; Ce/Mn oxide nanozyme; urokinase/PFP-loaded thrombolytic platform	CAT-, SOD- and ·OH-scavenging mimetic activities	LIPUS induces PFP liquid-to-gas phase transition, enhancing US/CEUS visibility and causing nanobubble burst for localized UK and CeO2/MnOx release; CeO2/MnOx scavenges ROS and improves the thrombotic microenvironment	Enabled real-time US-guided thrombus visualization, enhanced thrombolysis and reduced ROS/inflammatory injury; in vivo thrombolysis rate reached 87 ± 5.01% with LIPUS	Thrombosis model rather than cerebral/myocardial I/R model; frequency and duty cycle were not reported in the main text; cavitation, UK release and nanozyme antioxidation contributions are difficult to isolate	[257]
Ischemia-reperfusion injury	BaTiO3/CeO2/GQDs	MCAO/R rat model and OGD/R-stimulated SH-SY5Y cells	Injectable thermosensitive hydrogel; piezoelectric heterojunction enzyme system	SOD-mimetic activity; H2O2-decomposing/O2-generating nanozyme activity	US activates BaTiO3/CeO2/GQD piezoelectric heterojunction to provide local electrical cues, restore mitochondrial function and shift metabolism from anaerobic glycolysis toward aerobic respiration	Reduced cerebral infarct volume, improved neuronal survival, restored ATP/MMP and alleviated Ca2+ overload and metabolic dysfunction	Intracerebral in situ injection is invasive; systemic delivery and long-term clearance remain less clinically straightforward	[258]
Ischemia-reperfusion injury	Ce-CLMs-BMN	Rat acute myocardial infarction model; H2O2-injured H9C2 cells	ROS/US dual-responsive bilayer microneedle; CeO2 nanozyme and US-responsive NO donor micro-nano reactors	SOD- and CAT-like	ROS-responsive upper layer releases CeNPs for acute redox-inflammatory buffering, followed by US-triggered NO generation from lower CLMs to promote angiogenesis and anti-fibrotic repair	Improved cardiac function, reduced infarct size/fibrosis, promoted angiogenesis and attenuated inflammatory remodeling	Local microneedle implantation is procedure-dependent; broad multicomponent design complicates attribution of individual effects	[261]
Tendon injury	ENEVs	Rat Achilles tendon defect model	Ultrasound-augmented EV-cloaked enzymatic nanohybrid	CAT-like	US-enhanced cellular uptake delivers EV-cloaked Ru-doped Zn nanocubes for H2O2 decomposition, ROS scavenging, sustained Zn2+ release and macrophage-polarization regulation	Promoted functional recovery and matrix reconstruction, restored tendon morphology, suppressed intratendinous scarring and reduced peritendinous adhesion	Direct US-nanozyme evidence remains limited to one platform; EV cloaking and catalytic/ionic contributions are difficult to fully separate	[275]

Note: US, ultrasound; SDT, sonodynamic therapy; NCT, nanozyme catalytic therapy; CDT, chemodynamic therapy; POD, peroxidase; CAT, catalase; SOD, superoxide dismutase; GPx, glutathione peroxidase; GOx, glucose oxidase; NOS, nitric oxide synthase; ROS, reactive oxygen species; RNS, reactive nitrogen species; GSH, glutathione; BBB, blood–brain barrier; FUS, focused ultrasound; OA, osteoarthritis; RA, rheumatoid arthritis; PD, Parkinson’s disease; AD, Alzheimer’s disease; I/R, ischemia–reperfusion; NR, not reported. Mechanistic descriptions were assigned conservatively to distinguish nanozyme catalysis from sonodynamic effects, piezoelectric responses, cavitation-enhanced delivery, and drug or ion release.

Table 2.

Reported ultrasound parameters and translational features of representative ultrasound-mediated nanozyme systems.

Material name	Frequency	Intensity/power density	Duty cycle	Exposure duration	Biodistribution/clearance	Toxicity/safety window	Ref.
Janus Au–MnO NP vesicles	1.0 MHz	2.0 W cm−2	NR	10 min for disassembly/SDT assays; 30 min for penetration assay; in vivo duration NR	Time-dependent clearance; main-organ accumulation decreased to low levels by 7 days	Negligible body-weight fluctuation; no evident systemic toxicity in reported assays	[140]
CaF2 nanozyme	1.0 MHz	1.2 W cm−2	50%	15 min in vivo; 1–2 min common in vitro; 10 s for Ca2+ imaging	NR for biodistribution/clearance; local hydrogel retention used	No significant body-weight change; negligible inflammatory lesions in major organs	[10]
BTO/MoS2@CA	NR	NR	NR	5 min; applied 10 min and 3 days after injection	NR	No significant body-weight change; no major-organ H&E injury; blood tests normal	[153]
P-Por-Os	1.0 MHz	2.0 W cm−2 in vivo; 0.5 W in vitro optimization	30% in vivo; 50% in vitro optimization	5 min in vivo; 120 s in vitro optimized condition	Lung/spleen accumulation at high dose decreased by > 50% after 72 h	No significant liver-function, blood-routine or biochemical changes; no vital-organ H&E damage at 10 mg kg−1	[148]
PEG-CuP-COF@ΔSt	1.0 MHz	1.0 W cm−2	50%	5 min; in vivo on days 1, 3 and 5	Tumor accumulation peaked at 24 h; ΔSt showed time-dependent organ clearance	Dose-dependent bacterial tolerability; H&E and blood biochemistry showed no obvious organ damage	[149]
PdSA/Ti3−xC2Ty	50 kHz	2.0 W cm−2	NR	5 min in mechanistic assays; in vivo duration NR	Tumor accumulation peaked at 24 h; blood t1/2 = 6.1 ± 1.4 h; Ti signal absent in major organs by 14 days	No obvious body-weight fluctuation; major-organ H&E and blood tests normal by day 14	[158]
Cu3P-Cel sonozymes	50 kHz	1.0 W cm−2	50%	10 min in vitro; in vivo duration NR	Tumor fluorescence peaked at 24 h; ex vivo imaging showed predominant liver accumulation at 12 h and tumor accumulation at 24 h. The liver was a major RES/metabolic organ, and Cu signals were nearly undetectable after 14 days, indicating hepatic and renal clearance.	No weight loss; blood markers normal; major-organ H&E showed negligible abnormalities	[156]
Mn/Se-NE@FCS	1.0 MHz	2.0 W cm−2	50%	5 min	NR	Body weight, serum biochemistry and major-organ H&E indicated no systemic toxicity	[160]
Pip-loaded catalytic MBs (MB-Pip)	variable 400–900 kHz	1.0 W cm−2	NR	100 cycles (5 s on, 1 s off per cycle)	Liver/lung/kidney fluorescence; infected-lung signal peaked at 8 h and increased ∼ 1.6-fold with US. Fe levels normalized by day 28, indicating gradual clearance	Low hemolysis and negligible L-O2 cytotoxicity up to 1 mg mL−1 Fe3O4 NPs; no obvious changes in body weight, blood biochemistry, routine blood indices or major-organ H&E after i.v. injection	[176]
ACPCAH	1.0 MHz	1.0 W cm−2	NR	10 min	ICP-MS showed Cu was detectable only in wound tissue, not in major organs, indicating local retention and no evident systemic accumulation	No significant body-weight change during treatment; H&E staining showed no obvious abnormalities, inflammation or damage in major organs; Cu was detected only in wound skin, not in vital organs	[186].
Ti-MnO2-CPA@Ce6 coating	1.0 MHz	1.5 W cm−2	50%	2 min for 5 cycles in vivo; 10 min in vitro	As a localized implant coating rather than a systemic delivery platform, systemic biodistribution and clearance were not systematically reported. The study mainly assessed local Mn2+ release, hydrogel degradation and tissue safety	Good osteoblast viability with or without US; the initial slight decrease under US recovered after extended culture. MnO2 NFs showed no significant cytotoxicity at 50–200 μg mL−1, and long-term histology showed no obvious pathological damage around implants	[192]
SMC-HA	1.0 MHz	1.0 W cm−2 in vivo; 0.5–1.0 W cm−2 in vitro	50%	10 min in vivo; 10 min for cellular experiments in vitro; 2–15 min for sonocatalytic assays in vitro	Plaque fluorescence peaked at 24 h; mainly accumulated in liver and kidney, with minimal cardiac deposition	No significant body-weight change, hemolysis, blood toxicity or major-organ H&E injury; no hemolysis up to 200 μg mL−1	[199]
LyP-1Lip@HS	NR	3 W cm−2 in vivo; 3 W cm−2 in vitro; 4 W cm−2 caused reduced cell viability	NR	NR	Aortic fluorescence reached stable peak at 4 h; accumulated mainly in liver and spleen via RES; enhanced plaque/macrophage localization	3 W cm−2 disrupted lipid membrane while preserving ∼ 95% cell viability; 4 W cm−2 reduced viability to 73%; no obvious organ histology, blood, liver/kidney-function or hemolysis abnormalities	[203]
Nb2C-Pt@HA-PEG	NR	NR	NR	10 min in vivo	Blood t1/2 ≈1.52 h; plaque/aorta accumulation at 4 h = 0.13% ID vs 0.03% ID for non-HA control; abundant liver/spleen accumulation, limited kidney accumulation; liver Nb decreased by ∼ 50% at 4 weeks	No significant body-weight, hematological, liver/kidney-function or major-organ H&E toxicity; RBC change in US group remained within normal range	[204]
Q@CeBG	1.0 MHz	1.5 W cm−2	50%	2 min in vivo after 50 μL SonoVue MB injection; 30 s in vitro BBB/cell assays; 5 treatments every 2 days over 7 days	FUS enhanced Cy5-Q@CeBG brain accumulation; brain fluorescence assessed at 2, 4 and 8 h; blood/brain Que and Ce monitored up to 12 h; complete clearance NR	Good cytocompatibility and low hemolysis reported; selected FUS parameters caused no obvious brain H&E injury in normal mice; long-term systemic safety limited	[221]
USI-MHOF@KD8	1.0 MHz	1.5 W cm−2	50%	20 min daily over a 15-day in vivo treatment period; 1 h for Aβ oxidation/aggregation assays in vitro	Ex vivo brain/organ fluorescence assessed at 24 h and 14 days after i.p. injection; brain delivery was evaluated, but detailed clearance kinetics NR	>90% HT22 cell viability at 150 μg mL−1; major-organ H&E showed no discernible damage after treatment; body-weight and blood-biochemistry safety window NR	[229]
Mn-nanobowls	1.0 MHz in vivo; 1–3 MHz tested for propulsion	1.0 W cm−2 in vivo; 20 Vp-p amplitude used for optimized in vitro propulsion	60%	5 min per treatment; repeated every 48 h for penetration assay or weekly during the 6-week therapy study	Local IA delivery; cartilage penetration increased to 307 μm with US; systemic biodistribution/clearance NR	Knee skin temperature remained within a safe range after 5 min US; major-organ H&E showed no obvious lesions	[242]
MOF(Hf)-Pt	NR	0.35 W cm−2 for standard US treatment; 0–1.25 W cm−2 tested for catalytic regulation	NR	60 s standard treatment; 30, 60 and 90 s tested for catalytic regulation	IA retention tracked by fluorescence; knee-joint signal nearly disappeared after 120 h; liver/kidney signal indicated main clearance routes	Low hemolysis; major-organ H&E showed no obvious abnormalities; US adjuvant therapy caused no evident organ damage	[243]
Rh/SPX-HSA	1.0 MHz	1.0 W cm−2	50%	1 min in vitro; 3 min every 3 days for 5 treatments in vivo	ZnPc-labeled Rh/SPX and Rh/SPX-HSA were intravenously injected into normal or CIA mice and imaged at 0.5, 2, 4, 6, 12, and 24 h. Rh/SPX-HSA showed selective accumulation in inflamed joints, slower fluorescence decay than Rh/SPX, and ∼ 2.8-fold higher paw fluorescence in CIA mice than in normal mice at 24 h.	Major-organ H&E staining showed negligible damage in heart, liver, spleen, lung, and kidney; hematological indices were within normal ranges, indicating good biosafety and blood compatibility.	[246]
BEVs@Ru-HFO	NR	NR	NR	NR	IVIS showed effective accumulation in inflamed joints at 6 h after intravenous administration. BEVs encapsulation and US enhanced inflamed-joint targeting and prolonged joint retention; ex vivo imaging showed Ru-HFO tended to retain in the liver, while BEVs@Ru-HFO targeted inflamed joints. Hepatic fluorescence peaked at 12 h and decreased by 24 h, suggesting metabolic clearance	No significant body-weight fluctuation was observed; H&E staining and blood biochemical analysis indicated good biocompatibility.	[247]
CeO2/MnOx@UK@PFP	NR	0.8, 1.6 and 2.4 W cm−2 tested for LIPUS-triggered imaging; 1.6 W cm−2 selected for subsequent treatment	NR	Echo signal peaked at 3 min under 1.6 W cm−2 and weakened after 7 min; in vitro thrombolysis assessed after 2 and 4 h; in vivo imaging monitored at 1, 5 and 15 min, and thrombolysis assessed after 7 h	Tail-vein injection; real-time US monitoring showed thrombus-site signal enhancement after LIPUS, but systemic biodistribution/clearance was not systematically reported in the main text	No significant RAW264.7/HUVEC cytotoxicity; no significant hemolysis up to 150 μg mL−1; major-organ H&E and serum biochemistry showed no obvious damage or significant abnormalities	[257]
BaTiO3/CeO2/GQDs	1.0 MHz in vitro; typical transcranial focused US: 1–3 MHz in vivo	1.5 W cm−2 in vitro; typical 0.5–2 W cm−2 in vivo	50%	3 min in vitro; typical 5–10 min in vivo	Intracerebral in situ injection produced lesion-localized brain signal at 24 h with little off-target organ distribution; intravenous BCGTs showed weaker brain accumulation and kidney-associated clearance	Major-organ H&E after 7 days showed no obvious histological toxicity; local intracerebral depot reduced systemic exposure compared with intravenous administration	[258]
Ce-CLMs-BMN	NR	2.0 W cm−2	NR	10 min for in vitro NO release/cellular assays; 5 min every other day for seven in vivo sessions	Local microneedle implantation at infarct area; systemic biodistribution/clearance not systematically reported	Good hemocompatibility/cytocompatibility in vitro; blood routine, coagulation tests and major-organ H&E indicated no obvious in vivo toxicity	[261]
ENEVs	1.0 MHz	0.5 W cm−2	NR	1 min in cellular uptake assay; applied 30 min after ENEV addition; in vivo duration NR	Local injection; EV cloaking prolonged local retention. US increased day-1 local retention and was associated with enhanced tissue penetration, cellular uptake and subsequent renal clearance	ENs < 30 μg mL−1 showed no evident cytotoxicity in vitro; ENEVs showed no pathological major-organ changes by H&E and hemolysis ratio < 5%	[275]

Note: US, ultrasound; LIFU, low-intensity focused ultrasound; LIPUS, low-intensity pulsed ultrasound; FUS, focused ultrasound; CEUS, contrast-enhanced ultrasound; IA, intra-articular; i.v., intravenous; RES, reticuloendothelial system; ICP-MS, inductively coupled plasma mass spectrometry; H&E, hematoxylin and eosin staining; NR, not reported. Acoustic parameters are listed as reported in the cited studies; missing frequency, intensity, duty cycle, or exposure duration values were retained as NR to highlight reporting gaps and avoid over-interpretation.

4.2. Drug-resistant bacteria and biofilm-associated infections

Biofilm-associated infections caused by multidrug-resistant bacteria represent a persistent challenge in anti-infective therapy. The extracellular polymeric substance (EPS)-rich biofilm matrix forms a physical and biochemical barrier that limits antibiotic penetration, reduces bacterial metabolic activity, and protects pathogens from host immune clearance [171], [172]. The pathological microenvironment of infected tissues further complicates treatment. Local oxygen limitation, pH dysregulation, inflammatory signaling, and nutrient shifts can reprogram bacterial metabolism and stress responses, promoting antimicrobial tolerance and weakening host defense [173], [174], [175]. Therefore, effective antibacterial treatment needs to address not only planktonic bacterial killing, but also biofilm matrix disruption, penetration into protected niches, modulation of the infection microenvironment, and prevention of recurrence. These layered barriers make antibacterial therapy an instructive test case for ultrasound-mediated nanozyme design. Success requires both barrier opening and stage-dependent catalysis, because ultrasound can mechanically perturb biofilms and improve local penetration, whereas nanozymes can convert endogenous or infection-specific cues into bactericidal, matrix-degrading, oxygen-regulating or immunomodulatory effects.

A key advantage of ultrasound-mediated antibacterial systems is their ability to overcome the physical barrier imposed by the biofilm matrix before catalytic killing occurs. Ultrasound-responsive microbubbles provide a representative example of this barrier-opening function. In a chronic lung infection model caused by P. aeruginosa biofilms, ultrasound-responsive catalytic microbubbles loaded with Fe₃O₄ nanoparticles and piperacillin (MB-Pip) were used to improve biofilm penetration [176]. Upon ultrasound irradiation, microbubbles underwent cavitation to physically disrupt the PAO1 biofilm architecture, generating holes and channels, reducing biofilm thickness, and enhancing the penetration of both Fe₃O₄ nanoparticles and piperacillin into the protected biofilm interior. However, physical disruption alone may not fully eliminate the intact biomacromolecular scaffold of EPS. Because extracellular DNA, exopolysaccharides, and proteins are major structural components of bacterial EPS, nanozyme-based antibiofilm strategies can weaken the matrix either by loading or immobilizing natural EPS-degrading enzymes, such as DNase [177], α-amylase [178], or proteases [179], or by directly constructing DNase-mimetic or hydrolase-like nanozymes [180]. In this framework, ultrasound-mediated microbubble cavitation mainly opens and loosens the biofilm barrier at the structural level, whereas nanozyme-based or nano-enabled matrix-disrupting strategies further weaken EPS at the molecular level, thereby facilitating deeper catalytic reactions and antibacterial action.

HPO-like nanozymes provide an antibacterial mechanism that is distinct from conventional POD- or OXD-like ROS generation. By catalyzing the H₂O₂-dependent oxidation of halide ions, they produce hypohalous species such as HOCl and HOBr. These reactive halogen species can damage bacterial envelopes and biofilm-associated biomolecules, and may also interfere with bacterial adhesion, colonization and quorum-sensing-regulated biofilm formation [181]. This mechanism is particularly relevant to antibacterial therapy, where catalytic activity must act on extracellular biofilm architecture as well as bacterial redox homeostasis. For example, a GOx/HPO cascade nanozyme integrating Au nanoparticles with V₂O₅ nanowires generated HOBr from glucose, O₂ and Br⁻, enabling antibacterial and antibiofilm effects against nontypeable Haemophilus influenzae [182]. P,S-codoped Fe-ZIF-8-derived artificial enzymes further showed that HPO-like activity can be combined with POD-like ROS generation and glutathione depletion to improve oral biofilm eradication under low-H₂O₂ conditions [183]. Although these systems have not yet been integrated with ultrasound activation, they define an antibacterial-specific catalytic route that could complement ultrasound-mediated biofilm opening and substrate transport.

Because biofilm-associated infections arise in distinct tissue contexts, ultrasound-mediated nanozyme strategies should be designed according to the specific pathological demands of each infection type rather than as a uniform antibacterial approach. Diabetic wounds are commonly characterized by local hypoxia, chronic inflammation, impaired angiogenesis and delayed ECM remodeling, all of which restrict granulation tissue formation, collagen deposition and re-epithelialization [184]. Therefore, in diabetic wound infection, ultrasound-mediated nanozyme systems should not be designed solely to eradicate bacteria and disrupt biofilms, but should also address the metabolic and regenerative deficits of the diabetic wound microenvironment by regulating glucose metabolism, oxygen availability, inflammation and vascular regeneration. Previous studies have shown that CAT-like nanozyme-mediated oxygen generation has been combined with nanosonosensitizers to improve antibacterial SDT in diabetic infected wounds [185]. In this design, nanozyme-catalyzed O₂ supply alleviates local hypoxia and provides substrate for ultrasound-triggered ¹O₂ production, thereby enhancing oxidative bacterial killing.

Apart from oxygen-assisted sonodynamic killing, more sophisticated ultrasound-responsive nanozyme systems further couple catalytic antibacterial activity with pH-regulated repair functions in infected diabetic wounds. ACPCAH, a hyaluronic-acid-encapsulated hydrogel spray loaded with L-arginine and Au/Cu_1.6O nanoparticles co-deposited on phosphorus-doped C₃N₅ nanosheets, represents an integrated design that combines ultrasound-amplified catalysis with pH-responsive microenvironment regulation in infected diabetic wounds (Fig. 5) [186]. Under ultrasound stimulation, the Au/Cu_1.6O/P-C₃N₅ heterostructure generates piezoelectric polarization and mild sonothermal effects. Piezoelectric polarization enhances charge separation and interfacial electron transfer, whereas the sonothermal effect accelerates enzyme-like reaction kinetics, together amplifying the SOD/CAT/GOx/POD/NOS cascade. In the acidic bacterial infection microenvironment, this cascade is dominated by GOx-like substrate supply and POD/NOS-like ROS/RNS generation, which enhances bacterial killing and membrane disruption. In contrast, during the neutral-to-alkaline repair phase, sustained NOS-mediated NO release cooperates with CAT-like oxygen generation to promote VEGFA upregulation and functional angiogenesis. Bi₂WO₆@Cu₂O-GOx demonstrates another metabolic strategy by inhibiting bacterial glucose uptake and glycolysis, which restricts energy supply and reshapes the diabetic wound microenvironment [187].

Fig. 5.

Mechanisms of ultrasound-mediated nanozyme therapy in drug-resistant bacteria and biofilm-associated infections. a) Structure of Pip-loaded MBs (MB-Pip) and their capabilities to disrupt biofilm, degrade EPS, and activate macrophages under US stimulation. Reproduced with permission from ref. [13], Copyright © 2023 The Authors. Distributed under a Creative Commons Attribution NonCommercial License 4.0 (CC BY-NC). b) Five-enzyme cascade catalysis of ACPCAH. Reproduced with permission from ref. [186], Copyright © 2023 American Chemical Society. c) Schematic illustration of PtPdRuRhIr-driven suppression of M1 activation and promotion of an M2-like, pro-healing phenotype under inflammatory challenge. Reproduced with permission from ref. [191], Copyright © 2025 Wiley-VCH GmbH. d) The mechanism of the piezoelectric effect enhanced the sonodynamic performance and nanozyme activity as well as the mechanism of synergistically killing bacteria and boosting osteoblast differentiation of the hydrogel. Reproduced with permission from ref. [131], Copyright © 2025 The Author(s). Published by the Royal Society of Chemistr.

Oral and endodontic infections represent anatomically confined biofilm infections, in which bacteria persist within dental plaque, periodontal pockets, root canal systems, lateral canals and dentinal tubules [188]. In this setting, ultrasound-induced acoustic cavitation, microstreaming, and mechanical disturbance can improve the transport of irrigants and nanomaterials into confined anatomical structures while weakening biofilm barriers [189]. Furthermore, ultrasound can activate catalytic materials in situ. However, examples of ultrasound-mediated nanozyme systems for oral and endodontic infections remain limited. Current studies mainly suggest that ultrasound can activate catalytic antibacterial processes in situ, with both SDT and piezoelectric catalysis enhancing ROS generation to promote local biofilm eradication in confined dental microenvironments [190]. Meanwhile, nanozyme-based oral infection therapy is evolving beyond simple ROS-mediated antibacterial activity. Recent pH-switchable multienzyme systems further promote tissue repair, including alveolar bone regeneration, by scavenging oxidative stress and driving macrophage polarization toward the reparative M2 phenotype (Fig. 5) [191].

Distinct from macrophage-centered M2 polarization, studies of implant-associated infection show that ultrasound-mediated nanozyme systems can also activate antigen-presenting immunity. In a diabetic orthopedic implant infection model, a Ti-MnO₂-CPA@Ce6 coating combined MnO₂-based cascade nanozyme catalysis with Ce6-mediated SDT to combat Staphylococcus aureus infection [192]. Mn²⁺ released from the MnO₂ component acted as an immunostimulatory signal, promoting dendritic-cell maturation and antigen presentation. Transcriptomic analysis further indicated activation of both innate immune pathways and adaptive immune responses, which may help establish systemic immune memory to prevent infection recurrence and distal dissemination. Recent studies published in 2025-2026 further point to additional emerging directions, including metal-ion-metabolism-related cuproptosis-like [193] and ferroptosis [194] bacterial killing, as well as integration with photothermal therapy (PTT) [195].

Unlike tumor therapy, where ROS generation is mainly used to induce cancer-cell death, antibacterial nanozyme therapy must balance bacterial eradication with host-tissue preservation and repair. This requirement varies across infection sites, where inflammation resolution, redox homeostasis, vascular regeneration, bone repair or implant integration may become equally important therapeutic goals. Future ultrasound-mediated nanozyme systems should therefore move toward controllable catalytic switching, enabling strong ROS generation within infected biofilms followed by ROS scavenging, immunomodulation and tissue-specific regeneration after bacterial clearance. pH-responsive systems provide a useful starting point by linking acidic biofilms to POD- or Fenton-like catalysis and near-physiological conditions to antioxidase-like repair. However, pH alone may be insufficient because infected biofilms are heterogeneous and passive pH responsiveness provides limited control over the timing and dose of ROS production. Ultrasound offers an externally controllable temporal switch for redox regulation in a power-dependent manner– (Fig. 5) [131]. In the future, more advanced systems could also involve the integration of endogenous infection-associated cues (e.g., elevated H₂O₂, glucose, lactate, or bacterial enzyme activity) with externally adjustable ultrasound parameters. Such dual-responsive designs would allow treatment to shift from biofilm destruction to host-tissue repair as the infection microenvironment evolves. Future studies should therefore distinguish antibacterial clearance from regenerative microenvironment remodeling when defining therapeutic success. Representative platforms, their nanozyme activities, ultrasound parameters and mechanisms underlying antibacterial therapy are summarized in Table 1, Table 2.

4.3. Atherosclerosis

Atherosclerosis (AS) is initiated by lipid retention and endothelial activation, but its progression is largely orchestrated by lesional macrophages. Under pro-inflammatory stimulation, endothelial cells express adhesion molecules and chemokines that recruit circulating monocytes into the arterial intima, where they differentiate into macrophages and take up retained lipoproteins to form lipid-laden foam cells [196]. This macrophage-centered pathology provides a cellular rationale for applying SDT to AS, as ultrasound-activated sonosensitizers or sono-responsive nanoplatforms can generate ROS in plaques to induce apoptosis or phenotypic modulation of inflammatory macrophages and foam cells [197], [198]. Conventional sonosensitizers are often limited in atherosclerotic plaques by local hypoxia, which restricts oxygen-dependent ROS generation. SMC-HA addresses this limitation through sonocatalytic therapy (SCT) (Fig. 6) [199]. This platform consists of tetranitrogen-coordinated single-atom manganese sites embedded in a nitrogen-doped carbon framework and is further modified with hyaluronic acid for macrophage targeting. The Mn-N₄ sites serve as the catalytic core of SCT: under ultrasound irradiation, they promote ROS generation, while their CAT-like activity decomposes endogenous H₂O₂ into O₂ to relieve plaque hypoxia and support oxygen-dependent sonocatalysis. In parallel, their POD-like activity further contributes to ROS production within the inflammatory plaque microenvironment.

Fig. 6.

Mechanisms of ultrasound-mediated nanozyme therapy in atherosclerosis. a) Schematic illustration of the synthesis procedure for SMC-HA and its application in SCT targeting plaque treatment. b) Schematic illustration of the mechanism of SMC-mediated SCT procedure. c) CAT-like activity of CNF, MnO₂, and SMC. d) POD-like activity of CNF, MnO₂, and SMC. Reproduced with permission from ref. [199], Copyright © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH. e) Schematic illustration of a versatile nanoplatform integrating endogenous H₂S gas therapy with a multienzyme-like nanozyme for the treatment of AS. ^LyP−1Lip@HS actively targets atherosclerotic lesions and subsequently “unlocks” the enzyme-like activity of HMPB and facilitates the release of the endogenous H₂S donor by LIFU irradiation. The H₂S is generated endogenously, benefiting for preventing oxidative stress, inflammation and lipid metabolism disorders. In addition, ^LyP−1Lip@HS serves as PA/MR imaging contrast agent, providing visual monitoring for AS treatment. (f) Controlled release of SAC from ^LyP−1Lip@HS triggered by LIFU irradiation. g) TEM of ^LyP−1Lip@HS after LIFU irradiation (scale bar: 100 nm). Reproduced with permission from ref. [203], Copyright © 2024 Elsevier Ltd. h) Schematic illustration of the Pt-adsorption-induced local symmetry breaking of Nb₂C, leading to the emergent US-responsive properties. i) Calculated d-band center of Nb₂C and Nb₂C-Pt. j) ELF featuring the interfaces between Nb₂C and adsorbed Pt. k) CDD diagram of interfaces within Nb₂C-Pt, the yellow and cyan areas correspond to electron accumulation and depletion, respectively. Reproduced with permission from ref. [204], Copyright © 2025 The Author(s). Advanced Materials published by Wiley-VCH GmbH.

With AS progresses, macrophage-derived foam cells become embedded in plaques marked by unresolved inflammation and oxidative stress. Excess ROS sustains this injurious state by driving cytokine release, endothelial activation and lipid peroxidation, while ongoing macrophage death can exceed efferocytic capacity and contribute to necrotic-core formation [200], [201]. In advanced lesions, sustained redox injury compromises fibrous-cap integrity, so further amplifying ROS becomes increasingly difficult to justify. The therapeutic goal therefore shifts from macrophage ablation to plaque stabilization. ROS-scavenging nanozymes fit this requirement because they function as catalytic redox regulators rather than stoichiometric antioxidants. Through SOD-, CAT- and GPx-like activities, they can convert superoxide, hydrogen peroxide and lipid peroxides into less reactive products, reducing the oxidative pressure that sustains inflammation and plaque vulnerability. [202]. When integrated with plaque-targeting designs or ultrasound-gated activation, these catalytic reactions can reshape the plaque redox microenvironment with spatial and temporal control, restraining inflammation-lipid dysregulation and supporting fibrous-cap preservation.

LyP-1Lip@HS further shows how ultrasound can gate catalytic activity within atherosclerotic plaques (Fig. 6) [203]. After plaque-targeted accumulation, local LIFU irradiation at 3 W cm⁻² induces cavitation-mediated lipid membrane disruption, exposing HMPB to restore its nanozyme activity while releasing the H₂S donor SAC within the lesion. This condition preserved approximately 95% cell viability, whereas 4 W cm⁻² reduced viability to 73%, underscoring the need for a defined acoustic safety window. Ultrasound-mediated control can also operate at the level of catalytic efficiency. In Nb₂C-Pt@HA-PEG, Pt nanoparticle decoration reshaped the local symmetry and electronic structure of the Nb₂C MXene scaffold (Fig. 6) [204]. The Pt-Nb₂C interface shifted the d-band center to a higher energy level, increasing the affinity of the catalytic surface for oxygen-containing ROS intermediates. Under ultrasound irradiation, piezoelectric polarization promoted charge separation and interfacial electron transfer, which accelerated ROS-scavenging reactions on the catalytic surface. Consistent with this mechanism, CAT-like H₂O₂ scavenging increased from 54.8% without ultrasound to 80.2% after ultrasound irradiation. Thus, acoustic stimulation in this system does more than trigger material activation; it directly enhances nanozyme catalysis.

As noted above, lipid metabolic dysfunction is a core driver of AS progression; accordingly, nanozymes have been developed to reduce ox-LDL uptake by suppressing ROS/NF-κB-associated inflammatory signaling in macrophages [205], [206]. This oxidative-inflammatory axis also extends beyond macrophage lipid handling to the endothelial compartment, where persistent ROS exposure promotes endothelial senescence, barrier dysfunction and leukocyte adhesion. CZALO extends this strategy by coupling Zr⁴⁺-doped ceria nanozyme-mediated ROS scavenging with NO gas therapy [207]. In this system, co-delivered L-arginine is converted by macrophage-overexpressed nitric oxide synthase (NOS) into NO, which diffuses into neighboring endothelial cells and modulates senescence-associated pathways, particularly SASP signaling and the p53/p21/p16 axis. In this way, nanozyme-mediated redox buffering and NO signaling cooperate to delay endothelial senescence and preserve endothelial homeostasis.

Plaque heterogeneity adds another layer of complexity to anti-senescent therapy in AS, as dysfunctional cells are unevenly distributed within fibrous, lipid-rich and inflammatory plaque regions, while nanotherapeutics must overcome plaque-specific delivery barriers before reaching these targets [208], [209]. This challenge has motivated the development of nanozyme-based strategies that combine plaque accumulation with functional regulation of senescent cells. Ding et al. addressed this issue using D-PtPd₂CuFe, a chirality/phase dual-engineered high-entropy alloy senozyme with concentration-dependent senolytic and senomorphic activities [210]. At higher local exposure, D-PtPd₂CuFe selectively eliminated senescent cells by disrupting mitochondrial Fe-S cluster homeostasis and activating cuproptosis- and ferroptosis-associated cell death. When local availability was limited, it acted instead as a senomorphic agent, suppressing SASP-associated inflammatory signaling without requiring overt cell killing. Such dose-adaptive behavior is well matched to heterogeneous plaques, allowing the same nanozyme to either clear senescent vascular cells or restrain their pro-inflammatory secretory phenotype according to local material exposure.

Ultimately, the therapeutic value of nanozyme-based AS intervention depends on whether cellular and metabolic improvements are translated into structural plaque stabilization. This endpoint is not defined simply by reduced inflammatory or lipid burden, but by preservation of the architectural features that protect plaques from rupture. Recent studies published in 2025-2026 have therefore increasingly evaluated plaque stability using histological and molecular indices, including MMP-9 downregulation, increased collagen deposition, fibrous-cap preservation and reduced necrotic-core area [211], [212]. These readouts are essential because they link macrophage regulation, endothelial protection and lipid remodeling to a less vulnerable plaque phenotype. They also point to the need for imaging-guided nanozyme therapy, as plaque localization, activation timing and structural treatment response cannot be reliably inferred from dosing alone. Accordingly, recent nanozyme-based AS platforms have incorporated imaging functions, including fluorescence, photothermal, MRI and PA imaging, to support lesion localization, guide therapeutic timing and monitor treatment response [203], [213], [214], [215]. This theranostic capability is particularly relevant to ultrasound-mediated nanozyme therapy, where catalytic activation must be spatially matched to plaque distribution and maintained within a safe acoustic window.

Taken together, ultrasound-integrated nanozyme therapy for AS is best understood as a lesion-adaptive acoustic-catalytic strategy. Unlike tumor or antibacterial applications, where therapeutic efficacy often depends on intensifying cytotoxic stress, AS requires calibrated intervention within a chronic, heterogeneous and structurally fragile vascular niche. This distinction is particularly evident in the divergent use of ROS. In macrophage-rich lesions, ultrasound-activated nanozymes or sonocatalytic systems can generate ROS to promote foam-cell apoptosis or inflammatory macrophage ablation, provided that cell death remains spatially confined and can be resolved by efferocytosis. In advanced or vulnerable plaques, however, excessive ROS may aggravate lipid peroxidation, endothelial dysfunction, smooth-muscle-cell injury, necrotic-core expansion and fibrous-cap weakening. In this setting, ROS-scavenging nanozymes offer a more plaque-stabilizing strategy by buffering oxidative stress rather than amplifying it.

Ultrasound adds spatial control, penetration enhancement and energy-gated activation, whereas nanozymes provide catalytic routes to generate, scavenge or rebalance ROS according to plaque context. As a result, the therapeutic focus is shifting from macrophage ablation alone toward coordinated regulation of foam-cell lipid handling, inflammatory signaling, endothelial integrity, cellular senescence and fibrous-cap stability. The translational standard is therefore not whether ultrasound or nanozymes can modulate plaques, but whether their effects can be matched to plaque stage, cellular composition, local redox tone and acoustic exposure. Future studies should define plaque-specific redox windows, standardize ultrasound parameters, incorporate imaging-guided activation and prioritize durable structural endpoints such as fibrous-cap preservation and necrotic-core reduction. These advances will be needed to move ultrasound-mediated nanozyme therapy from experimental plaque modulation toward clinically meaningful stabilization of vulnerable atherosclerotic lesions. Representative platforms, their nanozyme activities, ultrasound parameters and atheroscierosis-therapeutic mechanisms are summarized in Table 1.

4.4. Neurodegenerative diseases

Neurodegenerative diseases are characterized by an interconnected pathological network involving mitochondrial dysfunction, oxidative stress, aberrant protein misfolding and aggregation, and chronic neuroinflammation [216], [217], [218], [219]. These processes reinforce one another and create a sustained injurious microenvironment in vulnerable brain regions. In this context, antioxidant nanozymes provide a rational therapeutic direction because they can catalytically regulate excessive ROS and help restore redox homeostasis [220]. However, catalytic activity alone is insufficient for effective intervention in the central nervous system. Nanozyme-based therapeutics must cross biological barriers, accumulate in diseased neural sites and remain active within microenvironments where oxidative injury, proteotoxic stress and glial activation are closely coupled. The key challenge is therefore to connect redox-regulating catalysis with brain-region access, pathological target engagement and functional recovery, rather than treating ROS scavenging alone as sufficient evidence of disease modification.

In the Q@CeBG study, FUS was used mainly as a BBB-opening tool rather than as a direct catalytic trigger (Fig. 7) [221]. After intravenous administration of sulphur hexafluoride MBs, low-intensity FUS was applied at 1.0 MHz and 1.5 W cm⁻² with a 50% duty cycle for 2 min, inducing acoustic cavitation and transiently increasing BBB permeability. This enabled greater Q@CeBG entry into the brain without making ultrasound the primary catalytic stimulus. After FUS-assisted delivery, Q@CeBG reduced neuronal oxidative stress through CeO₂-mediated ROS scavenging and promoted microglial polarization from the M1 phenotype toward the M2 phenotype.

Fig. 7.

Mechanisms of ultrasound-mediated nanozyme therapy in neurodegenerative disease. a) Schematic illustrating the preparation process of Q@CeBG nanoreactor and its therapeutic mechanisms for Parkinson's disease based on neuroprotection and modulation of the brain microenvironment. b) Schematic demonstration of the transwell assay for the coculture of BBB model to assess the penetration of Q@CeBG with or without FUS. c) Fluorescence intensities of SH-SY5Y cells detected by flow cytometry after different treatments. Reproduced with permission from ref. [221], Copyright © 2024 Elsevier B.V. d) Illustration for the mechanism of Aβ aggregation inhibition and ROS elimination in 3 × Tg-AD model mice. Under US irradiation, USI-MHOF@KD8 produce singlet oxygen (¹O₂) for Aβ oxygenation and further attenuation of β-Amyloidopathy in vivo. Furthermore, USI-MHOF@KD8 exhibits CAT- and SOD-like activities to scavenge ROS for brain microenvironment reprogramming. e) O₂^•− scavenging activities of USI-MHOF. f) Time-dependent O₂ generation of USI-MHOF (50 µg mL⁻¹) in the presence of H₂O₂ (5 mm) by a portable dissolved oxygen meter. Data are shown as mean ± SD (n = 3). Reproduced with permission from ref. [229], Copyright © 2024 Wiley-VCH GmbH.

Other Parkinson’s disease (PD)-oriented nanozyme studies, although not combined with ultrasound, also mainly target the interaction between oxidative stress and microglia-mediated neuroinflammation. QPMP, a mannitol-modified PLGA nanoreactor co-loading dihydroquercetin and Pt nanozymes, regulated ROS-associated neuronal injury and microglial activation, as indicated by reduced CD86 and IL-6 levels and increased CD206 and IL-10 expression [222]. Prussian blue nanozyme addressed a more specific inflammatory cell-death pathway. In MPTP-induced PD models, PBzyme scavenged ROS and suppressed the microglial NLRP3-caspase-1-GSDMD pathway, thereby reducing pyroptosis and dopaminergic neurodegeneration [223]. Together, these studies suggest that PD nanozyme therapy is moving from broad antioxidant protection toward more defined control of microglial phenotype and inflammasome-mediated injury. At present, however, Q@CeBG remains the clearest example showing that FUS-enabled brain delivery can be coupled with redox and immune regulation in PD.

In Alzheimer’s disease (AD), nanozyme therapy should be organized according to the pathological process being targeted. One major direction is to reduce oxidative and inflammatory stress within the AD microenvironment. Mn₃O₄ nanozymes illustrate this immunomodulatory strategy. By inhibiting the TLR4/NOX2-related ROS pathway, they shifted microglia toward an M2-like phenotype. In 5×FAD mice, this early inflammatory remodeling was followed by reduced fibrillar Aβ burden and improved cognitive performance after longer treatment [224]. Cu_2−xSe-TPP nanozymes address a different source of redox injury. Through SOD- and CAT-like activities, they reduced mitochondrial oxidative stress in microglia, thereby targeting an intracellular source of inflammatory ROS [225]. NM@PB-Ce further links redox regulation to inflammasome injury. This cerium-doped Prussian blue nanozyme suppressed NLRP3 activation and pyroptosis, which are closely associated with Aβ oligomer-induced neuronal damage [226]. Tau-related pathology has also begun to be explored. Ultrasmall Prussian blue nanozymes inhibited tau fibril formation in an okadaic acid-induced tauopathy model, suggesting a possible route to connect nanozyme catalysis with tau-centered protein aggregation [227].

A second strategy uses locally generated ROS to remodel Aβ aggregates, rather than simply removing excessive ROS. Ru3+-NMOF represents this mechanism through POD-like catalysis. By converting H₂O₂ into •OH, Ru3+-NMOF oxidized Aβ, increased its hydrophilicity and weakened hydrophobic interactions within amyloid fibrils [228]. USI-MHOF@KD8 extends this concept into an ultrasound-responsive format. Under US irradiation, the HOF sonosensitizer generated ¹O₂, oxidized Aβ and reduced its aggregation propensity (Fig. 7) [229]. In parallel, its CAT- and SOD-like activities helped decompose excessive ROS and mitigate neuronal oxidative stress, providing microenvironmental support for Aβ-targeted intervention. Compared with the H₂O₂-dependent Ru3+-NMOF system, USI-MHOF@KD8 provides external ultrasound control and was validated in 3×Tg-AD mice, where it reduced cerebral Aβ plaques and improved cognitive deficits. Together, these studies show that AD-oriented nanozyme therapy is moving beyond broad antioxidant protection toward more defined regulation of microglial activation, mitochondrial stress, inflammasome injury, tau aggregation and Aβ remodeling.

Building on these redox-centered disease-modifying strategies, recent studies have begun to connect nanozyme catalysis with broader regulatory nodes in neurodegenerative pathology. Recent designs now extend nanozyme catalysis to endogenous antioxidant transcription, copper-ion homeostasis and metabolism-oriented regulation of NAD+/NADH balance, mitochondrial function and microglial bioenergetics [230], [231], [232], [233]. In PD, nanozyme systems have also been used to regulate ferroptosis and mitophagy, linking redox catalysis to iron handling and mitochondrial quality control [234]. These advances also highlight an important mechanistic distinction within redox-based therapy. Some nanozymes act by removing excessive pathological ROS to relieve oxidative injury, inflammation or mitochondrial stress, whereas others use locally generated ROS as controlled effectors to oxidatively remodel Aβ aggregates [228], [229]. These directions point to a central design principle: successful nanozyme therapy must match catalytic function to the dominant disease-amplifying loop, rather than simply adding more enzyme-like activities. They also show that therapeutic efficacy should be judged by whether early redox, immune or metabolic correction translates into protein-pathology reduction, neuronal preservation and behavioral recovery. This creates a higher bar for multifunctional platforms, because each module should have a defined role in delivery, catalytic regulation or pathological readout. Such mechanisms will be therapeutically meaningful only if catalytic activity can reach the correct brain region and operate within an appropriate pathological window. Regional BBB opening and acoustic activation by ultrasound may help confine these interventions to disease-relevant brain areas, but their value will depend on whether delivery gains are linked to catalytic activity and functional recovery. Without this linkage, BBB opening remains a delivery achievement rather than evidence of nanozyme-mediated neuroprotection. Representative platforms, their nanozyme activities, ultrasound parameters and therapeutic mechanisms for neurodegenerative diseases are summarized in Table 1, Table 2, respectively.

4.5. Osteochondral diseases

Joint diseases involving articular cartilage, including osteoarthritis (OA), rheumatoid arthritis (RA) and gouty arthritis (GA), differ in their etiology but share several therapeutic barriers that limit effective local intervention. Articular cartilage is avascular, sparsely cellular and embedded in a dense extracellular matrix (ECM), which restricts passive diffusion of therapeutic agents to chondrocytes and matrix targets and contributes to its limited intrinsic repair capacity after injury [235], [236]. Although intra-articular administration can increase local exposure, rapid clearance from the joint cavity and insufficient retention within cartilage further reduce therapeutic bioavailability [237]. In parallel, persistent inflammation and excessive ROS promote matrix degradation, chondrocyte dysfunction and progressive impairment of joint function [235], [238], [239], [240]. These coupled physical and biochemical barriers provide the rationale for ultrasound-mediated nanozyme systems, which may combine enhanced local transport with catalytic regulation of oxidative and inflammatory stress in the joint microenvironment.

A first strategy is to use ultrasound to overcome the transport barrier imposed by dense, avascular cartilage. Previous studies have highlighted that ultrasound-induced cavitation and acoustic streaming can facilitate molecular transport not only in the superficial zone but also significantly enhance the penetration of large molecules into the deep layers of articular cartilage, with concentrations increasing by more than twofold under optimized parameters [241]. Building on this principle, Gong et al. developed MnO₂-decorated antioxidant nanobowls that could be directionally propelled under an external ultrasound field (Fig. 8) [242]. Upon ultrasound irradiation, these nanobowls reached a propulsion speed of 40.35 μm s⁻¹ in vitro and penetrated cartilage to a depth of up to 307 μm in vivo, markedly exceeding the non-ultrasound group. This study demonstrates that ultrasound can actively improve cartilage penetration and local nanozyme access, thereby addressing a key delivery limitation in cartilage-targeted therapy.

Fig. 8.

Mechanisms of ultrasound-mediated nanozyme therapy in osreochondral diseases. a) Mn-nanobowls motion images captured at different ultrasound amplitudes (5, 10, 15, and 20 V_p-p) in 5 s. b) Analysis of the average MSD vs time interval based on the tracking trajectory. c) The average velocity of Mn-nanobowls was analyzed from the tracking trajectory at different ultrasound amplitudes (n = 3). Reproduced with permission from ref. [242], Copyright 2025 Wiley-VCH GmbH. d) SOD viability of MOF (Hf)-Pt with different US power regulations. Reproduced with permission from ref. [243], Copyright © 2025 Elsevier B.V. e) Schematic illustration for the superior remodeling properties of the redox in RA. Reproduced with permission from ref.[247], Copyright © 2026 Wiley-VCH GmbH. f) Schematic illustration for the preparation of Rh-SPX/HSA and its related mechanisms for the treatment of Rheumatoid Arthritis. Reproduced with permission from ref. [248]Copyright © 2021 Elsevier Ltd.

Another therapeutic role of ultrasound is to turn piezoelectric nanozymes into externally tunable catalytic systems. Yao et al. developed an ultrasound-triggered MOF(Hf)-Pt piezocatalytic nanozyme for OA treatment, in which MOF(Hf) served as the piezoelectric framework and Pt acted as the catalytic center (Fig. 8) [243]. Under ultrasound stimulation, the piezoelectric response promoted charge transfer and enhanced enzyme-mimicking activity. This catalytic output could be adjusted by ultrasound parameters: for example, the SOD-like superoxide-scavenging readout increased from 15.61% without ultrasound to approximately 60% after 90 s of ultrasound exposure at 0.35 W cm⁻², and further increased at 1.25 W cm⁻² under the same exposure time. This finding shows that ultrasound can function as a controllable external switch for regulating piezoelectric nanozyme activity.

This strategy can also be extended to other ultrasound-triggered therapeutic outputs. In MoS₂@Bi₂S₃∈HA, ultrasound-evoked piezoelectricity promoted in situ H₂ generation from water while strengthening antioxidant enzyme-like activity, thereby coupling catalytic redox regulation with H₂-mediated anti-inflammatory therapy [244]. In CeMOF@BTO-functionalized biphasic scaffolds, ultrasound-activated piezoelectric nanozymes may guide osteochondral repair through bioelectrical regulation of Ca²⁺-cAMP/PKA-SOX9-associated signaling and subsequent stem-cell chondrogenesis [245].

After ultrasound-enhanced cartilage delivery and piezoelectric therapeutic activation, the focus can shift from cartilage-centered intervention to the inflamed synovial microenvironment. This shift is especially relevant to RA, in which persistent immune-mediated synovitis drives synovial hyperplasia, hypoxia, macrophage activation and osteoclastogenesis. Fibroblast-like synoviocytes (FLSs) are key effector cells in RA synovitis, as their abnormal proliferation contributes to pannus formation, inflammatory mediator release and cartilage invasion [239]. Rh/SPX-HSA illustrates a pro-oxidative ultrasound-driven strategy for this cellular target [246]. Ultrasound activated the sparfloxacin sonosensitizer to generate ¹O₂, whereas the concave-cubic Rh nanozyme provided POD- and CAT-like activities that generated •OH, supplied O₂ and relieved hypoxia. By amplifying ROS production within the lesion, this mutually reinforcing sonodynamic platform induced mitochondrial dysfunction and suppressed activated FLSs. This strategy is attractive for controlling synovial hyperplasia, but it requires accurate joint targeting and acoustic dosing because excessive ROS may injure surrounding cartilage or normal synovial cells.

An alternative strategy is to reduce oxidative and immune stress rather than exploit it for cytotoxicity. Extracellular vesicles (EVs), which can transfer bioactive cargo and modulate recipient-cell behavior, provide a biological delivery platform for integrating joint targeting with nanozyme-mediated redox and immune regulation [246]. This concept is exemplified by BEVs@Ru-HFO, a biocatalytic nanoarchitecture composed of Ru-cluster-anchored hydroxylated Fe₂O₃ encapsulated within bone marrow stem cell-derived EVs (Fig. 8) [247]. The design separates delivery and catalysis into two coordinated modules: BEVs enhance inflammatory joint targeting and local retention, while Ru-HFO restores redox balance through multi-enzyme-like ROS scavenging. Ultrasound stimulation further increased BEVs@Ru-HFO accumulation in inflamed joints, consistent with an ultrasound-mediated permeability enhancement effect. Once enriched at the lesion site, BEVs@Ru-HFO alleviated hypoxia, promoted anti-inflammatory macrophage responses and suppressed osteoclastogenesis. This antioxidant and immunoregulatory strategy is more aligned with chronic inflammatory remodeling and tissue protection, although it may be less direct in eliminating established aggressive FLS populations.

Considering the studies discussed above, ultrasound-mediated nanozyme therapy for articular-cartilage-related diseases should move beyond simple ROS elimination toward context-dependent redox control. In cartilage and stem-cell niches, antioxidant nanozymes are preferable because they protect matrix integrity, preserve regenerative cells and support tissue repair. By contrast, in pathological synovial compartments, especially hyperproliferative FLSs and pannus tissue, selectively pro-oxidative strategies may be useful for removing pathogenic synovial cells, as illustrated by the ultrasound-triggered sonodynamic nanozyme Rh/SPX-HSA (Fig. 8) [248] and the light-free ¹O₂-generating nanozyme CP@M [249]. The challenge is to match redox output to the affected compartment: excessive ROS generation may damage cartilage and normal synovium, whereas indiscriminate ROS scavenging may be insufficient to eliminate aggressive synovial cell populations. Future cascade and biomimetic nanozymes could therefore be designed to integrate oxygen supply, ROS regulation, immune modulation and regenerative signaling within one platform, provided that pro-oxidative and antioxidant phases are spatially and temporally separated. In parallel, image-guided and ultrasound-enhanced delivery represents an important translational direction, as shown by photoacoustic-capable CuAP systems combined with ultrasound-targeted microbubble destruction [250].

Gouty arthritis remains an underexplored area for ultrasound-mediated nanozyme therapy. Although standalone nanozyme platforms, including FALNZs, HMPB-Pt@MM and HTO-MnO, have been investigated for reducing local urate/MSU burden, scavenging inflammatory ROS and limiting flare recurrence [251], [252], [253], direct studies combining ultrasound with nanozyme-based GA therapy are still scarce. Future designs could integrate urate-degrading or MSU-responsive nanozyme activity with ultrasound-enhanced joint delivery or ultrasound-triggered local release. At present, this direction should be framed as a mechanism-driven opportunity that requires direct ultrasound-nanozyme validation rather than as an established application. Representative platforms, their nanozyme activities, ultrasound parameters and therapeutic mechanisms for Osteochondral diseases are summarized in Table 1, Table 2.

4.6. Ischemia–reperfusion injury

Ischemia–reperfusion (I/R) injury arises when blood flow returns to previously ischemic tissue, yet reperfusion also initiates a secondary wave of damage. This injury is commonly driven by abrupt ROS production, mitochondrial dysfunction, Ca²⁺ overload, inflammatory amplification and microvascular impairment [254]. These processes reinforce one another. Mitochondrial damage accelerates oxidative stress and energetic failure, whereas endothelial dysfunction and leukocyte recruitment aggravate local hypoxia, edema and perfusion defects. For ultrasound-mediated nanozyme therapy, this creates one of the clearest examples of disease-inverted catalytic logic. In I/R injury, the goal is not to amplify ROS for cell killing, but to buffer oxidative stress, preserve mitochondrial and vascular function, and support repair within a narrow therapeutic window [255].

Rapid vascular recanalization is therefore an upstream requirement for reducing the severity of subsequent I/R injury. In this context, ultrasound-responsive cerium nanozyme systems have been used to accelerate thrombolysis by strengthening acoustic cavitation. Ce-UiO-CM enhances cavitation through a gas-generating mechanism [256]. Its porous Ce-MOF structure provides nucleation sites, while CAT-like H₂O₂-to-O₂ conversion supplies oxygen to amplify ultrasound-induced cavitation. CeO₂/MnOx@UK@PFP uses a different strategy. Under LIPUS, perfluoropentane undergoes liquid-to-gas phase transition, causing nanobubble expansion, urokinase release and real-time ultrasound visualization (Fig. 9) [257].

Fig. 9.

Mechanisms of ultrasound-mediated nanozyme therapy in I/R injury. a) The scheme of CeO₂/MnO_X@UK@PFP burst after US irradiation. b) Ultrasound images of B-mode and CEUS after different treatment and LIPUS irradiation. Reproduced with permission from ref. [257], Copyright © 2025, Wiley-VCH GmbH. c) Schematic diagram of aerobic metabolism and lactate regulation induced by US + BCGTs. d) Representative images of the TTC staining of brain from MCAO/R rats treated with different administrations (n = 5). Reproduced with permission from ref. [258], Copyright © 2026, Elsevier B.V. e, f) The mechanisms of AMI in-situ sequential therapy using Ce-CLMs-BMN assisted by US irradiation. M0, M1 and M2 indicate different types of macrophages. Reproduced with permission from ref.[261], Copyright © 2024ElsevierB.V.

These thrombolysis-oriented designs show how ultrasound-mediated nanozymes may reduce ischemic burden at the recanalization stage. However, restoring vessel patency does not fully prevent reperfusion-associated secondary injury. Direct protection of reperfused brain tissue requires correction of mitochondrial metabolic failure, not merely ROS removal. US+BCGTs@TSG was designed for this purpose by embedding a BaTiO₃/CeO₂/graphene quantum dot piezoelectric heterojunction into an injectable thermosensitive hydrogel (Fig. 9) [258]. After local gelation in the infarct region, ultrasound activates the heterojunction and provides wireless electrical stimulation to injured neural cells. This stimulation suppresses LDHA-associated lactate production, restores PDH activity, and redirects pyruvate metabolism toward the TCA cycle and oxidative phosphorylation. As a result, US+BCGTs@TSG improves mitochondrial membrane potential and ATP production. In this system, CeO₂ and graphene quantum dots mainly stabilize the piezoelectric process by improving charge transfer and limiting material-derived ROS, allowing ultrasound-triggered electrical cues to drive metabolic repair.

In myocardial I/R injury, cardiomyocyte loss is coupled to microvascular obstruction, ferroptosis-associated lipid peroxidation and fibrotic remodeling [259], [260]. These intertwined processes make ultrasound useful less as a destructive energy source than as a trigger for timed nanozyme action and payload release. Ce-CLMs-BMN exemplifies a local, sequential repair strategy for post-infarction injury (Fig. 9) [261]. Its ROS-responsive upper layer releases CeNPs during the acute phase, limiting oxidative stress and inflammatory activation. After this initial buffering, ultrasound activates Cur/L-Arg micro-nano reactors in the lower layer to generate NO, thereby supporting angiogenesis and restraining fibrotic remodeling. The strength of this design is temporal separation: CeNPs address the early redox-inflammatory burst, whereas ultrasound-induced NO release supports later repair. Its limitation is equally important. Local microneedle implantation is procedure-dependent, and the broad multi-component design makes individual therapeutic contributions difficult to separate.

MP@T NPs adopt a more reperfusion-focused timing strategy [262]. The TA-Ce shell first provides ROS-scavenging and anti-inflammatory protection, aided by collagen affinity for injured myocardium. At 24 h after reperfusion, ultrasound induces PFP vaporization and releases ML351 to suppress ferroptosis. This later effect is linked to ACSL4 downregulation, GPX4 restoration and reduced lipid peroxidation. Compared with Ce-CLMs-BMN, MP@T NPs offer a sharper mechanistic match to MI/RI because ultrasound release is aligned with the ferroptotic phase. The trade-off is a narrow temporal requirement and greater translational complexity from the drug-loaded phase-change platform.

Despite these advances, several issues remain unresolved. Future systems should better integrate vascular recanalization with downstream tissue protection, because restored blood flow can also intensify oxidative stress, inflammatory activation and microvascular dysfunction. A central challenge is to define the redox direction of ultrasound activation in I/R therapy. Unlike tumor or antibacterial treatment, where ROS amplification can be therapeutically useful, I/R injury requires ROS buffering as the dominant strategy. In piezoelectric platforms, ultrasound-induced ROS should therefore be regarded as a controllable side effect rather than a therapeutic output. The useful electrical stimulation needs to be preserved through efficient charge separation and interfacial transfer, while ROS-forming reactions are restrained by nanozyme components such as CeO₂ [258]. Ultrasound parameters also need to be matched to specific therapeutic purposes, including cavitation-enhanced thrombolysis, phase-transition release and piezoelectric activation. In hepatic and renal I/R injury, ultrasound has been explored mainly for molecular imaging, severity assessment and responsive nanotheranostics rather than direct therapeutic activation [263], [264], [265], [266]. These studies suggest that imaging-guided or biomarker-responsive activation may improve treatment timing, dosing and organ-specific precision. Finally, multifunctional platforms require controls that quantify how much of the benefit comes from ultrasound, nanozyme catalysis and loaded therapeutic agents, rather than treating the combined outcome as mechanistic proof. Where feasible, ultrasound-only, nanozyme-only, payload-only and catalytically inactive material controls should be included. Representative platforms, their nanozyme activities, ultrasound parameters and therapeutic mechanisms for I/R injury are summarized in Table 1, Table 2.

4.7. Tendon repair and tendinopathy

Tendon injury remains difficult to treat because tendons are hypovascular and metabolically quiescent tissues with limited intrinsic regenerative capacity; consequently, healing is slow and often reparative rather than regenerative, producing scar-prone tissue with inferior structural organization and mechanical competence [267]. After injury, the inflammatory, proliferative and remodeling phases should coordinate immune resolution, tenocyte activity, collagen deposition and matrix maturation, but this sequence is frequently disrupted by unresolved inflammation and maladaptive macrophage activation [268]. Persistent ROS further impair tenocyte function and matrix homeostasis, favoring adhesion, impaired gliding and a disorganized collagen profile enriched in type III collagen rather than mature type I collagen architecture [269]. Thus, tendon repair failure is not simply a structural defect, but a regenerative microenvironment disorder involving redox imbalance, immune dysregulation, collagen-type mismatch and inadequate mechanical matrix reconstruction.

Therapeutic ultrasound alone has not shown consistent clinical benefit as a stand-alone treatment for Achilles tendinopathy [270]. However, preclinical studies suggest that low-intensity pulsed ultrasound can influence collagen organization, tissue mechanics and inflammatory remodeling during tendon healing [271], [272]. This distinction reframes ultrasound less as an independent regenerative treatment than as a non-invasive physical trigger. Its penetration, focusability and parameter tunability allow acoustic energy to be delivered to dense and poorly vascularized tendon tissue with spatial and temporal control. These features make ultrasound particularly suitable for activating responsive biomaterials that convert external energy into local biochemical, mechanical or electrical cues.

Recent advances in ultrasound-responsive piezoelectric biomaterials suggest a strategy to reconstruct the bioelectrical component of the tendon repair niche. By converting external acoustic energy into local electrical cues, these systems can promote tendon stem/progenitor cell proliferation and tenogenic differentiation, supporting the concept that bioelectrical signals can instruct reparative cell behavior [273]. Mechanistically, related piezo-bioelectric tendon studies link these effects to mechanosensitive ion channels and Ca²⁺-dependent signaling, which may influence tendon-associated transcriptional programs and matrix-forming responses [274]. Thus, ultrasound-responsive piezoelectric materials provide a physical route to restore bioelectrical signaling in damaged tendons, but they do not fully address the redox and inflammatory abnormalities that define the injured tendon microenvironment.

Nanozyme-based systems offer a complementary route by targeting redox abnormalities that are not fully addressed by ultrasound-responsive piezoelectric materials. Rong et al. provided a direct example by developing enzymatic nanohybrids for ultrasound-augmented tendon matrix reconstruction (Fig. 10) [275]. The Ru-doped Zn nanocube core exhibited CAT-like activity, enabling H₂O₂ decomposition and ROS scavenging in the injured tendon microenvironment. Sustained Zn²⁺ release further supported tenocyte proliferation and shifted matrix production toward type I collagen synthesis while reducing type III collagen deposition. This study connects ultrasound-assisted delivery with catalytic and ionic regulation of tendon matrix repair.

Fig. 10.

Mechanisms of ultrasound-mediated nanozyme therapy in tendon repair. a) Construction of ENEVs, treatment of rat Achilles tendon injury, and the therapeutic mechanisms involved in the regulation of tenocyte metabolism and redirection of macrophage polarization. b) Representative images of Masson staining, and c) polarized light microscopy (stained with Sirius Red) of tendons from different groups. d) Representative immunofluorescent images of COL I (green) and COL III (red). Reproduced with permission from ref. [275], Copyright © 2023 American Chemical Society.

Redox imbalance also affects tendon repair through mitochondrial dysfunction. Mitochondria support reparative tendon-cell function by sustaining energy production and redox balance. CeO₂-loaded anisotropic nanofiber scaffolds provide a strong mechanistic example: ceria nanozymes reduced ROS, preserved mitochondrial membrane potential and partially restored ATP production in tendon-derived stem cells under oxidative stress [276]. In a rat Achilles tendon defect model, the same scaffold also improved mitochondrial ultrastructure, linking catalytic ROS buffering to mitochondrial preservation during repair. However, this synthetic aligned scaffold provides only partial biochemical mimicry of native tendon ECM. MnO₂-modified decellularized tendon membranes approach this limitation from the opposite direction. By growing MnO₂ nanozymes in situ on collagen fibers, this design combines CAT/POD-like ROS-scavenging activity with the structural and biochemical cues of a tendon-derived matrix [277]. Its mitochondrial evidence is narrower, mainly involving preserved membrane potential and reduced apoptosis, but it offers a more tendon-mimetic carrier. Together, these two systems suggest that future ultrasound-nanozyme platforms should combine energy-preserving catalytic cores with acoustically responsive, tendon-like structural matrices.

The same redox-centered logic can be extended to regulated cell death and inflammatory signaling. RuO₂ nanozymes shifted the focus toward ferroptosis, a cell-death program driven by iron-dependent lipid peroxidation. In collagenase-induced tendinopathy, hollow RuO₂ nanozymes limited lipid peroxidation and Fe²⁺ accumulation in tenocytes, while reducing COX2 and increasing GPX4 and FTH1 expression [278]. Redox regulation can also reshape macrophage-associated inflammatory amplification. PBzymes linked ROS scavenging to reduced MAPK activation in M1 macrophages, lower IL-1β/TNF-α production and enhanced M2 reparative polarization [279]. These findings indicate that tendon nanozyme design is moving from general antioxidant supplementation toward pathway-specific microenvironment remodeling, including matrix reconstruction, mitochondrial preservation, ferroptosis control and inflammatory resolution. Although direct ultrasound-combined evidence remains limited, these mechanisms provide a rationale for designing ultrasound-responsive nanozyme systems with spatial, temporal and pathological specificity.

Future ultrasound-nanozyme strategies for tendon repair may be most effective when designed as integrated microenvironment-regulating systems. In this framework, ultrasound would serve as an external regulator rather than a stand-alone therapy. Its penetration, focusability and parameter tunability could enhance local delivery, activate responsive materials and control treatment timing in dense, poorly vascularized tendon tissue. Piezoelectric components could convert acoustic input into local bioelectrical or electromechanical cues that support tenogenic cell behavior. Nanozyme components, in turn, could provide catalytic regulation of the injured tendon microenvironment, including redox balance, inflammatory resolution and matrix remodeling. Together, these modules could address the physical, biochemical and structural deficits that jointly limit tendon regeneration.

A practical design could combine a tendon-mimetic scaffold, an ultrasound-responsive piezoelectric module and a nanozyme-based catalytic module. The scaffold would provide aligned mechanical guidance, the piezoelectric component would restore ultrasound-triggered signaling, and the nanozyme component would correct pathological microenvironmental stress. This concept may also extend to more complex tendon-associated interfaces. For example, nanozyme-based treatment in osteoporotic rotator cuff tear has been used to coordinate inflammatory control with repair across tendon, bone and muscle tissues [280]. Ultrasound may further improve nanoparticle penetration, cellular uptake, local retention and on-demand release, making it especially useful for hypovascular tendon tissue. Thus, the key future direction is not simply to combine ultrasound with nanozymes, but to define which module provides mechanical guidance, bioelectrical signaling, catalytic repair and sustained functional recovery. This module-level accountability would make tendon systems easier to compare with scaffold-only or ultrasound-only regenerative approaches. Representative platforms, their nanozyme activities, ultrasound parameters and therapeutic mechanisms for tendon repair are summarized in Table 1, Table 2.

5. Discussion and perspective

Rather than restating individual studies, this section uses the evidence summarized in Table 1, Table 2 to identify design rules, acoustic determinants and translational constraints for ultrasound-mediated nanozyme therapy. The discussion therefore moves from disease-specific catalytic logic to ultrasound standardization, in vivo fate, safety windows and future design principl.

5.1. Matching catalytic output to disease-specific redox needs

As summarized in Table 1, ultrasound-mediated nanozyme design is better understood through lesion-specific redox tasks than through disease labels or material complexity alone. Tumor therapy usually requires pro-oxidative catalysis, in which ultrasound amplifies ROS/RNS production, Fenton-like reactions, oxygen supply or GSH depletion. Infection-related applications are more stage dependent: oxidative catalysis is needed to kill bacteria and disrupt biofilms, whereas antioxidant activity may be required later to limit inflammation and support tissue repair. In atherosclerosis, neurodegenerative disorders and osteochondral diseases, the same distinction also applies, as ultrasound-enhanced ROS can be therapeutic when directed at pathological targets but harmful when it aggravates bystander oxidative injury. In I/R injury and tendon repair, ROS generated by piezoelectric or sonodynamic processes should therefore be controlled rather than assumed to be beneficial. Future systems should therefore begin with the required redox task and then choose catalytic modules and acoustic inputs that support that task at the appropriate disease stage.

5.2. Ultrasound parameters as determinants of therapeutic mechanism

As summarized in Table 2, the acoustic conditions used in ultrasound-mediated nanozyme studies are concentrated within a relatively narrow but incompletely reported window. Frequencies generally range from approximately 50 kHz to 3 MHz, with 1 MHz being the most frequently used; acoustic intensity or power density is commonly distributed between 0.5 and 2.0 W cm⁻²; reported duty cycles are usually around 30-60%; and exposure durations range from seconds to approximately 20 min per session. These parameters should be regarded as mechanistic variables rather than procedural details, because they jointly determine tissue penetration, energy deposition, cavitation probability, sonodynamic ROS generation and nanozyme activation efficiency [281], [282].

Frequency and intensity are among the most frequently reported parameters, but their biological meaning depends on the therapeutic mechanism. Frequency affects acoustic penetration, focal geometry and bubble dynamics: lower frequencies generally favor cavitation and deeper propagation, whereas higher frequencies may improve spatial confinement but can reduce penetration depth. Intensity determines the acoustic energy delivered to the tissue or reaction system and can strengthen cavitation-related mechanical effects, radical formation and thermal deposition. In ultrasound-mediated nanozyme systems, these effects may translate into enhanced delivery, catalytic activation or sonodynamic ROS generation, but they also increase the need to balance therapeutic activation with tissue safety [108].

Duty cycle and exposure duration further define the cumulative acoustic dose. A higher duty cycle or longer exposure can increase cavitation persistence, energy deposition and catalytic output, but it can also increase heat accumulation and nonspecific tissue injury. Pulsed ultrasound may reduce thermal loading while preserving mechanical or cavitation-mediated effects, which is particularly relevant for delivery-enhanced nanozyme systems and piezoelectric nanozyme platforms. Therefore, duty cycle and exposure time should be reported together with intensity and frequency, rather than treated as secondary parameters [283].

Acoustic pressure and cavitation conditions remain insufficiently described in most nanozyme studies, although they are central to ultrasound-mediated therapy. Stable cavitation can promote microstreaming, convective transport and transient membrane permeabilization, thereby improving nanozyme penetration, carrier release and cellular uptake. In contrast, inertial cavitation can produce stronger mechanical stress, local heating, sonoluminescence and radical formation, which may enhance sonodynamic ROS production but also increase the risk of tissue injury if uncontrolled [108]. When cavitation is proposed to mediate nanozyme activation or delivery, future studies should report acoustic pressure, mechanical index, pulse mode and cavitation-related evidence.

A further limitation revealed by Table 2 is that the ultrasound mode is clearly defined in only a limited subset of studies. low-intensity focused ultrasound (LIFU)-induced cavitation-mediated lipid membrane disruption [203], FUS/MB-mediated BBB opening [221], low-intensity pulsed ultrasound (LIPUS)-triggered phase transition [257] and LIPUS-assisted nanozyme propulsion [242] are explicitly described in selected examples, whereas many other studies report the stimulus only as “US” or “ultrasound”. This ambiguity is not merely semantic. Delivery-enhanced nanozymes may require permeability enhancement, cavitation or phase transition, sonodynamic nanozymes depend on ROS generation, and piezoelectric nanozymes require acoustic conditions that drive polarization and charge separation. Thus, the ultrasound mode should be matched to the intended nanozyme function.

Thermal and non-thermal effects should also be separated more explicitly. Only a small number of ultrasound-mediated nanozyme studies have examined whether heat contributes to catalytic activation. For example, an ultrasound-augmented multienzyme-like nanozyme hydrogel spray showed enhanced ROS generation at elevated temperature after sonication, suggesting that sonothermal effects can participate in nanozyme activation [186]. However, the enhanced catalytic performance in that study was attributed to coupled sonothermal and piezoelectric effects rather than to heat alone. More broadly, mild hyperthermia can increase blood flow, vascular permeability and thermoresponsive release, whereas excessive heating may denature proteins, damage tissue and confound the interpretation of nanozyme-mediated therapy [283]. Some LIFU settings can produce biological effects with minimal temperature elevation, indicating that cavitation, acoustic radiation force and mechanical stress may dominate under selected conditions [282].

A related translational barrier is the mismatch between how ultrasound is reported in nanozyme studies and how it is delivered by clinical systems. Preclinical studies usually describe frequency, nominal intensity and exposure time, whereas clinical platforms are defined by probe geometry, focal depth, pulse sequence, mechanical index, thermal index and tissue-dependent acoustic fields. These parameters determine local pressure, heating and cavitation, but they are rarely calibrated or reported together in current nanozyme studies. Consequently, the same nominal ultrasound setting may produce different levels of nanozyme activation, ROS generation, tissue penetration and therapeutic efficacy across instruments and anatomical sites. Future standardization should therefore include calibrated acoustic pressure, duty cycle, pulse repetition parameters, focal configuration, coupling conditions, temperature change and cavitation evidence. Standardized catalytic evaluation under ultrasound should also include matched no-ultrasound, ultrasound-only, inert-particle, sonosensitizer-only and nanozyme-control groups, together with temperature monitoring and reaction-volume calibration. Only by linking these parameters and controls to catalytic output, biodistribution and safety can ultrasound-mediated nanozyme therapy become reproducible and compatible with clinical ultrasound platforms.

5.3. Biodistribution, clearance and safety windows

For ultrasound-mediated nanozyme therapy, efficacy should be interpreted together with biodistribution, clearance and long-term safety. As summarized in Table 2, most studies report encouraging short-term biosafety, including body-weight monitoring, blood biochemistry, hemolysis assays and major-organ histology. However, pharmacokinetics, organ retention, degradation products and complete clearance pathways remain inconsistently reported. This gap is important because ultrasound is not only an activation source, but also a physical input that can reshape tissue permeability, local retention and cellular uptake. Safety should therefore be evaluated as a material-acoustic property, with the treated tissue and ultrasound exposure considered together.

The biodistribution patterns in Table 2 indicate that administration route and material architecture strongly shape in vivo fate. Systemically administered platforms often achieve lesion accumulation within several hours to 24 h, but many also show liver, spleen or kidney distribution, reflecting uptake by clearance-related organs. Only a subset of studies report time-dependent elimination over 7-28 days, such as reduced organ signals, disappearance of elemental markers or normalization of metal levels. Local hydrogels, coatings, microneedles, intra-articular depots and intracerebral implants can reduce systemic exposure, but they still require evidence for local degradation, ion release, chronic inflammation and eventual elimination.

Ultrasound further complicates this safety landscape. As summarized in Table 2, acoustic exposure can enhance BBB opening, lesion accumulation, cartilage penetration, local retention and cellular uptake. These effects may improve therapeutic efficiency by increasing the amount of active nanozyme that reaches the disease microenvironment. However, the same mechanisms may also increase exposure in adjacent normal tissues, particularly when cavitation, vascular permeability or high acoustic pressure is involved. Safety assessment should therefore compare biodistribution, clearance and tissue toxicity with and without ultrasound under the same acoustic parameters used for therapy.

Long-term safety cannot be inferred from acute cytotoxicity or major-organ histology alone. Many ultrasound-mediated nanozyme systems still contain metal-based, high-Z, piezoelectric, vesicular or polymer-coated components that may persist in tissues, release ions, alter redox balance or interact with the immune system. This is particularly relevant for platforms using liposomes, extracellular-vesicle-like carriers, biomimetic coatings, PEGylated surfaces or repeated ultrasound treatment. Evidence from extracellular vesicles, lipid nanoparticles and PEGylated nanomedicines shows that biomimetic or stealth modification does not necessarily confer immune inertness; immune recognition can be shaped by source materials, surface antigens, administration route, repeat dosing, anti-PEG antibodies and biomolecular corona formation [284], [285], [286]. Because corona proteins, complement factors and opsonins can influence phagocytic uptake, clearance and cytokine responses, future safety assessment should include complement activation, inflammatory cytokines, macrophage responses and immune-cell uptake, in addition to routine histology and serum biochemistry [287], [288], [289], [290].

More tissue-friendly design may require a shift from maximal functionality to functionality with biological exit routes. Polysaccharide-derived materials provide one useful direction because they combine biodegradability, biocompatibility, mechanical flexibility and tunable surface chemistry [291], [292]. Safer design, however, should not be limited to polysaccharides. Natural piezoelectric biomaterials, including collagen, silk fibroin, amino acids and peptides, and clinically familiar inorganic nanozymes such as Prussian blue and iron oxide may also expand the design space when degradation, ion release and immune compatibility are controlled [293], [294], [295]. These directions point toward ultrasound-responsive nanozyme systems designed around catalytic activity, acoustic responsiveness, tissue compliance and post-treatment clearance rather than maximal functional loading.

Finally, safety and manufacturability should be linked rather than considered separately. Ultrasound-mediated nanozyme systems often combine catalytic centers, sonosensitizers, piezoelectric components, targeting ligands, vesicles, hydrogels or coatings. Small batch-to-batch changes in particle size, surface chemistry, oxidation state, crystallinity, catalytic activity or acoustic responsiveness may alter biodistribution and therapeutic output. Translation will require scalable synthesis, sterilization-compatible processing, storage stability and quality-control assays that connect material attributes with biological function. Regulatory evaluation is also likely to be complex because many systems combine a nanomaterial therapeutic with an external ultrasound device and, in some cases, biologic-like or implantable components. This will require product-specific characterization, risk-based biocompatibility assessment and clear combination-product classification.

5.4. Translational challenges and future design principles

The next stage of ultrasound-mediated nanozyme therapy should be defined by mechanism-resolved design rather than multifunctional assembly. Many reported systems integrate catalytic activity, sonodynamic effects, piezoelectric responses, targeted delivery, immune modulation or imaging within a single platform. Such integration can improve therapeutic performance, but it can also obscure the dominant driver of efficacy. Future studies should therefore define the primary therapeutic logic before interpreting efficacy: catalytic redox regulation, sonodynamic amplification, piezoelectric charge separation, delivery enhancement or a deliberately coupled mechanism. This distinction is also needed to define where ultrasound-mediated nanozymes offer advantages over delivery-only ultrasound systems, conventional SDT without nanozyme catalysis, photodynamic or photothermal therapy, piezocatalytic materials and passive catalytic nanomedicine. It is particularly important because pro-oxidative tumor and antibacterial strategies require different catalytic outputs, acoustic inputs and safety controls from redox-restorative applications in neurodegeneration, OA or I/R injury. Claims involving cavitation-enhanced catalysis, piezoelectric charge biology, regulated cell death pathways and immune modulation should therefore be separated from experimentally validated mechanisms and supported by pathway-specific evidence, rather than inferred only from therapeutic outcomes or downstream markers.

A second priority is to make platforms comparable across studies. At present, efficacy is often reported without enough information to determine whether the improvement arises from nanozyme catalysis, acoustic activation, enhanced delivery, disease-model sensitivity or increased energy input. The field would benefit from a reporting framework that links catalytic activity, acoustic exposure, disease microenvironment, biodistribution and safety to the same therapeutic endpoint. At minimum, this framework should specify the material category and active-site design, validated enzyme-like activity, ultrasound mode, frequency, intensity or power density, acoustic pressure or mechanical index when available, duty cycle, exposure time, disease model, primary endpoint, safety readout and control group that separates catalysis from delivery or energy input. Such a framework would make it easier to identify transferable design variables, disease-specific constraints, genuinely effective ultrasound-responsive mechanisms and the controls needed to justify mechanism-based claims.

Preclinical validation should also better reflect clinical constraints. Many target diseases involve chronic inflammation, repeated oxidative stress, impaired perfusion, dense extracellular matrices, biofilms, calcified plaques or neural barriers. These factors can alter nanozyme delivery, acoustic propagation, catalytic substrate availability and immune responses. Therefore, proof-of-concept efficacy in short-term small-animal models should be complemented by models that capture chronicity, repeated treatment, clinically relevant tissue depth and tissue heterogeneity. This shift would help distinguish platforms that are experimentally effective from those with a plausible translational path, especially for diseases in which treatment depth, repeat dosing and chronic tissue remodeling determine clinical feasibility.

Ultimately, the most translatable ultrasound-mediated nanozyme systems may not be those with the largest number of functions, but those in which mechanism, acoustic activation, safety and manufacturability are aligned from the outset. This requires simpler and better-defined architectures, reproducible catalytic and acoustic performance, disease-matched treatment logic, and material designs with credible biological exit routes. In this sense, the field should prioritize platforms whose therapeutic mechanism, acoustic control and safety profile can be understood, compared and reproduced before clinical translation is proposed.

Ultrasound-mediated nanozyme therapy provides a versatile strategy for coupling catalytic regulation with spatially controlled acoustic activation. Across current studies, its therapeutic value arises not from ultrasound or nanozyme activity alone, but from their coordinated effects on redox modulation, ROS/RNS generation, tissue penetration, catalytic activation and local delivery. Future progress will depend on moving from function-stacked platforms toward systems in which catalytic task, acoustic input, disease barrier and safety endpoint are aligned. This requires clearer acoustic parameter reporting, disease-matched catalytic design, reliable biodistribution and clearance assessment, and long-term safety evaluation. Such advances will determine which ultrasound-mediated nanozyme systems remain promising experimental constructs and which can become clinically meaningful therapeutic platforms

CRediT authorship contribution statement

Yuzhen Shu: Writing – review & editing, Writing – original draft. Xingheng Wang: Writing – review & editing. Shuwei Zhang: Writing – review & editing. Sujiao Cao: Writing – review & editing. Ruiqian Guo: Writing – review & editing. Li Qiu: Writing – review & editing, Supervision, Methodology, Conceptualization. Yuanjiao Tang: Writing – review & editing, Supervision, Methodology, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.