Visual theranostic oral nanozymes for IBD from principles to preclinical design Authors: Shangyong Zuo, Kefan Li, Liang Feng, Xingyan Yu, Yifan Wang, Chengwei Wu, Wenjian Weng, Yaoting He, Xiujun Cai, Kui Cheng Published: 2026 ABSTRACT Inflammatory bowel disease (IBD), which includes ulcerative colitis (UC) and Crohn's disease (CD), is a chronic, relapsing inflammatory disorder of the intestine driven by disruption of mucosal immune balance. In this context, mucosal inflammation, oxidative stress, and damage to the epithelial barrier reinforce one another and form a self-perpetuating cycle. Oral nanozymes that combine catalytic removal of reactive oxygen species with in vivo imaging readouts provide a potential platform for site-focused intervention and imaging-based evaluation of IBD. This review focuses on how oral nanozymes are designed to couple imaging and therapy and summarizes recent progress in two overarching directions: single-modality imaging-guided strategies and multimodal imaging-guided strategies. Within the single-modality group, current studies are discussed mainly as computed tomography (CT)-guided strategies and optical/photoacoustic (PA)-guided strategies according to their dominant signal-generation mechanisms. Multimodal imaging-guided strategies, by contrast, integrate structural and functional information on a single platform to support more informative preclinical assessment of IBD. This paper also discusses current limitations in oral delivery, characterization of enzyme-like activity, imaging assessment, and long-term biosafety, thereby outlining key design principles for the next generation of oral nanozymes for visual theranostics in IBD. Introduction Inflammatory bowel disease (IBD) is a chronic inflammatory disorder of the gastrointestinal tract [1]. It is characterized by heterogeneous mucosal lesions with ulceration and structural remodeling, alternating phases of activity and remission, and a high rate of relapse [2]. IBD mainly comprises two forms (Scheme 1). Ulcerative colitis (UC) typically starts in the rectum and extends proximally in a continuous pattern, with diffuse inflammation that is mainly confined to the mucosa and submucosa [3]. In contrast, Crohn's disease (CD) shows patchy, segmental, and transmural inflammation along the bowel and tends to form deep ulcers and granulomas [4]. Despite these distinct distribution patterns, UC and CD share similar pathogenic mechanisms characterized by a positive feedback loop between mucosal inflammation-driven oxidative stress and disruption of the intestinal epithelial barrier. Barrier disruption facilitates microbial and antigen translocation across the leaky mucosa, aggravates inflammation, and ultimately creates a self-perpetuating cycle [5]. These features make IBD not only a chronic inflammatory disease, but also a spatially heterogeneous and dynamically evolving target, for which lesion-localized intervention and longitudinal in vivo readout are particularly desirable [6].Scheme 1Multiscale pathological features of healthy intestine, UC, and CD, with the shared pathogenic cascade in IBD. Multiscale pathological features of healthy intestine, UC, and CD, with the shared pathogenic cascade in IBD. Reflecting this spatial and temporal complexity, clinical assessment of IBD relies on symptoms, endoscopy, and cross-sectional imaging to evaluate disease extent, inflammatory burden, and treatment response [7,8]. Endoscopy remains central for the direct evaluation of mucosal lesions and biopsy acquisition, particularly in UC, where disease activity is mainly confined to the mucosa and submucosa [9]. In contrast, cross-sectional and transmural imaging techniques, including computed tomography enterography (CTE), magnetic resonance enterography (MRE), and intestinal ultrasound (IUS), are especially important in CD for assessing bowel-wall abnormalities, transmural involvement, and extraintestinal complications [8,[10], [11], [12]]. The major roles, advantages, and limitations of these clinical assessment approaches are summarized in Table 1. However, these approaches are primarily intended to evaluate disease status at the tissue or organ level, rather than to determine whether a formulation has reached the inflamed segment, how long it remains there, or how its local presence relates to subsequent therapeutic effects [13].Table 1Common clinical assessment approaches for IBD mentioned in this review.Table 1ApproachMain roleAdvantageLimitationRef.SymptomsClinical presentationSimple; readily availableLimited lesion specificity[7]EndoscopyMucosal lesions; biopsyDirect visualization; tissue samplingInvasive; mucosa-focused[9]CTEDisease extent; wall lesions; complicationsRapid; cross-sectional viewRadiation; limited treatment-linked readout[10]MREWall lesions; transmural involvement; complicationsNo radiation; soft-tissue contrastCost and time burden; limited formulation-level information[12]IUSWall lesions; inflammatory activityDynamic; repeatableOperator dependence; limited deep-segment assessment[8] Table 1Common clinical assessment approaches for IBD mentioned in this review.Table 1ApproachMain roleAdvantageLimitationRef.SymptomsClinical presentationSimple; readily availableLimited lesion specificity[7]EndoscopyMucosal lesions; biopsyDirect visualization; tissue samplingInvasive; mucosa-focused[9]CTEDisease extent; wall lesions; complicationsRapid; cross-sectional viewRadiation; limited treatment-linked readout[10]MREWall lesions; transmural involvement; complicationsNo radiation; soft-tissue contrastCost and time burden; limited formulation-level information[12]IUSWall lesions; inflammatory activityDynamic; repeatableOperator dependence; limited deep-segment assessment[8] Common clinical assessment approaches for IBD mentioned in this review. In clinical management, conventional systemic administration often fails to balance efficacy and safety and shows marked interindividual variability [14]. Consequently, long-term disease control requires delivery routes that are easy to administer, capable of sustaining stable therapeutic effects, and associated with minimal additional systemic burden [15,16]. In this context, oral administration is particularly attractive because it offers practical advantages for repeated dosing in chronic treatment [17]. However, the highly complex intestinal environment introduces substantial uncertainty into oral delivery. As formulations transit through the gastrointestinal tract, they encounter along-tract variations in pH, enzymes, and the microbiota [18]. They are also continuously influenced by the mucus layer and peristalsis [19]. As a result, lesion entry efficiency and residence time are difficult to predict. Oral nanozymes therefore need to reconcile several partly competing requirements, including protection against gastric degradation, sufficient mucus penetration or mucosal adhesion, prolonged retention at inflamed segments, and minimal unintended systemic absorption through the damaged epithelial barrier [20]. For oral formulations, the key challenge is therefore not simply disease assessment, but whether their delivery behavior within the diseased intestine can be visualized in vivo [21]. Platform-linked imaging can provide longitudinal information on formulation localization and retention, while also helping track changes in inflammatory burden during treatment [[22], [23], [24], [25]]. When imaging is co-integrated with therapy in the same oral carrier, the readout can be made aligned with delivery [26], such that CT attenuation can be rendered approximately quantitative under standardized imaging and analysis conditions, reflecting the in situ coverage and retention of the active nanoplatform [27]. This linkage enables response attribution, distinguishing insufficient lesion accumulation and residence from limited pharmacodynamic efficacy after delivery [28]. In a chronic, relapsing disease such as IBD, quantitative imaging at prespecified time points can inform treat-to-target adjustment of both dose and dosing frequency [29], instead of relying on infrequent terminal measurements [30]. Importantly, imaging signals do not directly equate to therapeutic efficacy. Depending on the platform design, they may be used mainly for formulation localization and retention, lesion visualization and inflammation burden assessment, treatment response monitoring, or a combination thereof [31]. These signals should therefore be interpreted within the biological context of the readout. Among the many nanoformulations developed for IBD, most systems still function mainly as carriers for loaded drugs or biologics [32,33]. They rely on loaded small molecules or biologics to provide anti-inflammatory effects and have limited ability to offer sustained, amplifiable catalytic control over inflammation-driven reactive oxygen species (ROS) imbalance [34,35]. In contrast, nanozymes can exhibit multiple enzyme-like activities, including superoxide dismutase (SOD)-like, catalase (CAT)-like, and glutathione peroxidase (GPx)-like functions [36]. These catalytic activities allow them to directly regulate oxidative-stress-related processes without markedly increasing the number of therapeutic components [37]. Thus, compared with conventional drug-carrier nanoplatforms, nanozymes provide an active catalytic basis for regulating inflammation-driven ROS imbalance, going beyond the passive delivery role of conventional carriers. Building on this catalytic basis, oral nanozyme platforms can be further coupled with imaging readouts when appropriate imaging moieties are introduced or intrinsic contrast properties are used [38]. In this context, their value in imaging-guided IBD theranostics lies in the integration of ROS regulation with trackable lesion delivery and retention [39,40]. Beyond formulation tracking, some nanozyme designs may, in principle, couple catalytic-site states or substrate conversion to imaging signals [41], enabling catalytic reactions, microenvironment modulation, and imaging readouts to form a functional closed loop—an integration that is difficult to achieve with most inert imaging nanoprobes or conventional drug-loaded nanoparticles [42,43]. To date, however, orally administered nanozyme platforms that integrate imaging and therapy in IBD remain predominantly preclinical, and clinical evidence is still limited [38]. When nanozymes are interfaced with readouts such as computed tomography (CT), near-infrared (NIR) optical imaging, or photoacoustic (PA) imaging, they may support longitudinal in vivo monitoring and provide a basis for optimizing dose and dosing schedule in preclinical studies [38], help relate delivery-associated imaging information to downstream pharmacodynamic outcomes [44], and better address the management needs arising from lesion heterogeneity and fluctuating disease activity in IBD [31]. Current studies at the interface between nanozymes and in vivo imaging for IBD can broadly be divided into two overarching directions [45]. One direction involves single-modality imaging-guided systems, in which imaging is achieved primarily through one dominant readout mode [38]. Within this category, current studies are mainly represented by CT-guided strategies and optical/PA-guided strategies [46]. CT-guided platforms are typically constructed around high-Z (high-atomic-number) components to provide stable and relatively quantitative CT contrast, and can be further integrated with nanozyme catalytic domains to achieve simultaneous therapeutic intervention and imaging-based monitoring. Optical/PA-guided strategies, by contrast, rely on fluorescence or light-to-acoustic conversion to follow oral formulations in vivo [47]. In these designs, ratiometric fluorescence improves robustness, whereas PA imaging typically provides deeper penetration and higher spatial resolution than conventional fluorescence imaging in scattering tissues. These signals are mainly employed to evaluate arrival and retention at inflamed intestinal segments and to relate these parameters to treatment-associated changes [48]. The second overarching direction involves multimodal integration, in which two or more imaging modalities are combined within one platform to provide more complete in vivo assessment by integrating structural and functional information [49]. Although nanozyme–imaging integration has been increasingly explored in IBD models, the available evidence remains difficult to compare across studies [50]. Imaging signals are used for different purposes, including lesion accumulation, intestinal retention, inflammation visualization, and treatment-associated monitoring, but these roles are not always clearly distinguished [51]. Moreover, catalytic activity, imaging performance, oral-delivery behavior, and therapeutic outcomes are often evaluated using different endpoints and models [52,53]. A focused comparison is therefore needed to clarify what each imaging readout demonstrates and how it relates to catalytic function, oral delivery, and in vivo therapeutic evidence [54,55]. Herein, this review focuses primarily on orally administered nanozyme platforms with built-in imaging readouts for IBD and discusses their roles in imaging and therapy for visual management (Scheme 2), which conceptually relates the major compositional families discussed in this review to the principal functions they are intended to serve, including lesion mapping and segmental distribution, inflammatory activity grading, mucosal restoration and remodeling, microenvironment modulation, and longitudinal monitoring. The representative studies included here were selected according to four practical criteria: oral administration or clear relevance to oral platform design, integration of a nanozyme or nanozyme-like therapeutic component with an imaging readout, in vivo evaluation in IBD-related models, and sufficient mechanistic or functional information to support cross-study comparison. A small number of boundary cases that are mechanistically or conceptually informative are also discussed explicitly as boundary cases, not as core examples within the main framework. Accordingly, this review is intended as a structured narrative overview.Scheme 2Schematic diagram of the major design routes of oral nanozymes for IBD and their corresponding imaging roles and therapeutic goals. Schematic diagram of the major design routes of oral nanozymes for IBD and their corresponding imaging roles and therapeutic goals. As illustrated in Scheme 3, the main body is first organized according to whether imaging is achieved through a single modality or through multimodal integration. Within the single-modality section, representative studies are further discussed as CT-guided strategies and optical/PA-guided strategies according to their dominant signal-generation mechanisms. Across these sections, representative studies from recent years are compared using the same core dimensions, including imaging function, catalytic activity, oral delivery constraints, and therapeutic readout. Through this framework, the review aims to compare representative design routes, clarify the distinct information provided by different imaging modes, and summarize transferable principles for the future development of safer, more interpretable, and more clinically relevant oral nanozyme systems for IBD.Scheme 3Review framework: single-modality systems, including CT-guided and optical/PA-guided strategies, and multimodal systems. Review framework: single-modality systems, including CT-guided and optical/PA-guided strategies, and multimodal systems. Imaging readout mechanisms For the single-modality systems emphasized in this review, nanoplatforms with imaging readouts, including oral nanozymes, can be broadly divided into two routes: CT-guided structural imaging and optical/PA-guided functional imaging. This section aims to clarify the signal-generation pathway of these two modalities. It discusses how nanoplatforms modulate X-ray or light propagation and energy conversion in the intestinal environment, and how these changes are converted into quantitative image intensities through signal detection and image reconstruction. It also summarizes key materials design and engineering strategies proposed in recent years to enhance imaging performance (Table 2), thereby providing a basis for the subsequent imaging-based evaluation of visual theranostic oral nanozymes. To clarify the underlying imaging principles, some examples cited in this section extend beyond the core oral nanozyme scope of this review and are included only as mechanistic references.Table 2Summary of representative imaging readout strategies for oral nanoplatforms in IBD.Table 2ModalitySystemSize (nm)Key variableWhat it improvesRef.CTPEG-Au5-75Lesional entry amount, local volume fraction, spatial distributionHigher CT conspicuity and greater enhancement yield in inflamed intestinal segments[56]CTBi-GLUEn/rMucus-interface retention and anchoring, spreading uniformityBroader coverage and more stable, reproducible real-time CT readouts[57]CTEr-DOTAn/rEnergy-dependent attenuation near a K-edge for discriminationHigher specificity and interpretability, reduced luminal background[58]Optical (NIR-II)Hemicyaninen/rInflammation-triggered activation, high fecal clearanceBetter background suppression and more reliable oral readout[59]PAGeTen/rStable absorption profile in the gut, sustained non-radiative dissipationMore robust and repeatable PA readout[60]n/r - not reported. Table 2Summary of representative imaging readout strategies for oral nanoplatforms in IBD.Table 2ModalitySystemSize (nm)Key variableWhat it improvesRef.CTPEG-Au5-75Lesional entry amount, local volume fraction, spatial distributionHigher CT conspicuity and greater enhancement yield in inflamed intestinal segments[56]CTBi-GLUEn/rMucus-interface retention and anchoring, spreading uniformityBroader coverage and more stable, reproducible real-time CT readouts[57]CTEr-DOTAn/rEnergy-dependent attenuation near a K-edge for discriminationHigher specificity and interpretability, reduced luminal background[58]Optical (NIR-II)Hemicyaninen/rInflammation-triggered activation, high fecal clearanceBetter background suppression and more reliable oral readout[59]PAGeTen/rStable absorption profile in the gut, sustained non-radiative dissipationMore robust and repeatable PA readout[60]n/r - not reported. Summary of representative imaging readout strategies for oral nanoplatforms in IBD. n/r - not reported. CT readout of nanoplatforms As shown in Scheme 4A, when an X-ray beam passes through tissue and nanoparticles, the transmitted intensity I recorded by the detector can be expressed as:(1)I=I0exp(−∑iμixi)=I0exp[−(μtissuextissue+μparticlexparticle)]Here, I0 is the incident X-ray intensity; μi and xi denote the linear attenuation coefficient and the effective path length of medium i, respectively; μtissue and xtissue refer to tissue, and μparticle and xparticle refer to nanoparticles [61,62]. High-Z nanoplatforms usually have μ values higher than those of soft tissue. At the microscopic level, this arises from the higher nuclear charge and more strongly bound inner-shell electrons of high-Z elements. X-rays therefore undergo photoelectric interactions more readily, followed by relaxation processes such as characteristic X-ray emission and Auger electron release. These processes increase local energy deposition and the overall attenuation [63]. After image reconstruction, the transmitted intensity is converted into a voxel-wise map of attenuation coefficients. These are commonly expressed as CT values in Hounsfield units (HU) by normalizing the voxel μ to that of water:(2)HU=1000×μ−μwaterμwater−μair≈1000×μ−μwaterμwaterwhere μwater and μair are the linear attenuation coefficients of water and air, respectively (μair ≈ 0) [64]. Notably, HU values depend on scan and reconstruction settings. Therefore, cross-study comparisons require analysis normalized to a common imaging protocol.Scheme 4Schematic illustration of the signal generation mechanisms of CT-guided (A) and optical/PA-guided (B) strategies based on oral nanozymes. Schematic illustration of the signal generation mechanisms of CT-guided (A) and optical/PA-guided (B) strategies based on oral nanozymes. Researchers have recognized that the CT conspicuity of nanoparticles is not determined by atomic number alone, but is jointly regulated by their effective entry into inflamed intestinal segments [65], their local volume fraction, and their spatial distribution [66]. Using a PEGylated gold nanoparticle platform as a model, one study systematically compared particle sizes from 5 to 75 nm and evaluated their impact on CT enhancement in colitis [56]. The results showed only a weak correlation between in vivo CT contrast and particle size. In the context of oral administration, this suggests that increasing the effective distribution and coverage of the nanoplatform in the target region is more important than simply pursuing higher atomic numbers or more extreme particle sizes. Building on this understanding, a major recent design trend is to regard the material-mucus interface as a key factor in increasing the product μparticle xparticle. One approach is to enhance retention and anchoring at inflamed mucosa to increase the effective path length xparticle [67]. Another is to improve spreading and coverage along the mucosal surface, which raises the effective volume fraction and makes the spatial distribution more uniform [68]. For example, the Bi-GLUE microgel network is deformable and strongly mucoadhesive, so it can rapidly spread and anchor to mucus in the intestinal environment. This leads to broader gastrointestinal coverage and more stable real-time CT readouts, making CT signals under oral administration more stable and reproducible [57]. In parallel, emerging spectral CT adds a stronger functional dimension to structural imaging. When the contrast element has a high K-edge, the energy dependence of μ becomes more pronounced, and spectral CT can exploit attenuation differences across energy ranges to achieve material discrimination [69]. Er-DOTA shows marked energy-dependent attenuation between 40 and 160 keV and yields higher CT values and brighter signals in inflamed colonic segments in dextran sulfate sodium (DSS)-induced colitis models [58]. These findings indicate that adding an energy-resolved dimension can improve detection sensitivity in oral intestinal imaging. It also helps distinguish true accumulation in the target region from nonspecific signals arising from luminal contents. As a result, the reliability and interpretability of CT readouts are increased. Accordingly, the CT-guided nanozyme systems discussed below derive image contrast mainly from X-ray attenuation by high-Z components. Optical/PA readout of nanoplatforms As shown in Scheme 4B, when incident light propagates through intestinal tissue and a nanoplatform, it undergoes both absorption and scattering, and the local light fluence Φ decays with depth. For strongly scattering media, the diffusion approximation provides a first-order description:(3)ϕ=ϕ0exp(−μeffz)(4)μeff=3μa(μa+μs')Here, Φ0 is the incident fluence; μa and μ′s are the absorption coefficient and the reduced scattering coefficient, respectively; μeff​ is the effective attenuation coefficient; and z is the depth along the direction of light propagation [70]. Local absorption is generally contributed by both tissue and nanoparticles:(5)μa=μa,tissue+μa,particlewhere μa,tissue and μa,particle are the absorption coefficients of tissue and nanoparticles, respectively [71]. For dispersed nanoparticles, their absorption contribution can be written as:(6)μa,particle=ln(10)εCwhere ε is the molar extinction coefficient and C is the local effective concentration at the inflamed intestinal segment [72]. Taken together, Eqs. (3)–(6) indicate that the core contribution of nanoplatforms to optical/PA imaging is to increase the local μa, thereby reshaping light energy deposition and ultimately generating detectable image contrast after signal detection and reconstruction. As illustrated by the energy-level diagram in Scheme 4B, nanoplatforms absorb excitation light and undergo electronic transitions from the ground state S0 to excited states, followed by rapid non-radiative relaxation to the lowest excited state S1. On this basis, the key difference between the two imaging strategies lies in the relaxation pathways from S1 back to S0 [73]. In optical imaging, relaxation occurs mainly via radiative decay: particles return from S1 to S0 and emit photons, and the detector collects this emission to generate the optical image signal. In this case, the primary role of the nanoplatform is to provide a stable and detectable exogenous emission channel [74]. In PA imaging, relaxation is dominated by non-radiative decay: the energy released during the S1→S0 transition is dissipated as heat through vibrational relaxation, the resulting heat deposition induces transient thermoelastic expansion, and ultrasound waves are generated for imaging. Here, the nanoplatform enhances local light absorption and photothermal (PT) conversion efficiency, which improves PA contrast and resolvability [75]. Building on this mechanistic framework, recent studies aiming to improve the optical/PA interpretability of oral nanoplatforms in the intestinal setting have gradually converged on two clear technical routes. One route is emission-based readout, in which the optical signal arises mainly from the intrinsic emission of the nanoplatform, covering the first near-infrared window (NIR-I, ∼700-900 nm) and the second near-infrared window (NIR-II, ∼1000-1700 nm) [76]. The other route is absorption-thermalization-based readout, in which broadband absorbers with strong NIR absorption and efficient non-radiative energy dissipation convert deposited light energy into PA signals. Mechanistically representative advances along these two routes are outlined below. For the former, namely emission-based systems, environment-responsive or activatable designs are widely used. Probes are engineered to maintain low signal output under non-lesion conditions and to switch on or markedly increase their emission in response to inflammation-related stimuli. This approach improves specificity and interpretability in the complex intestinal background and makes signal changes more closely reflect inflammation-related molecular activity [77,78]. For example, an activatable NIR-II hemicyanine system with high fecal clearance efficiency has shown good performance in IBD assessment. This research suggests that, in oral applications, background suppression, target specificity, clearance behavior, and readout reproducibility should be optimized within a unified design framework, instead of improving a single parameter in isolation [59]. For the latter, broadband absorption-thermalization platforms, the robustness of the PA readout depends on more than simply increasing absorption. It is equally important that the absorption spectrum, and thus the local μa, remain relatively stable in the gastrointestinal environment (aggregation, mucus interactions, protein corona formation, etc.) [79], and that a high fraction of non-radiative dissipation can be maintained over time [80], so that the acoustic source term becomes more reproducible. GeTe nanosheets developed for multimodal theranostics in IBD demonstrate a feasible route to achieve PA and PT readouts from broadband absorbers under oral conditions, providing an intuitive mechanistic illustration for absorption-thermalization platforms [60]. Accordingly, the optical/PA platforms discussed below can be understood mainly as emission-based systems or absorption-thermalization-based systems, depending on how the readout is generated. CT-guided visual theranostic nanozyme strategies In CT-guided approaches, orally administered nanozymes are often organized around a radiopaque component that also serves as the structural center of the platform. Although the specific materials vary, many designs face similar questions: how to integrate imaging capability with catalytic function without compromising either, how to maintain stability in the gastrointestinal tract, and how to configure composite structures so that each component performs reliably under oral dosing conditions. The subsections that follow outline representative ways researchers have addressed these considerations within CT-guided formulations. High-Z nanozyme-based CT imaging Cerium oxide platforms The K-edge of cerium at 40.4 keV, together with its relatively high density [81], endows CeO2 with high and quantifiable X-ray attenuation, resulting in good CT contrast [82]. Consequently, many studies have adopted cerium oxide nanozymes as the central platform [83], making use of the reversible Ce3+/Ce4+ redox cycle to provide multi-enzyme-like ROS scavenging (Fig. 1A) and to couple imaging and therapy within a single material system [11,84]. In the following discussion, cerium oxide or ceria is therefore used as the family-level term, CeO2 denotes stoichiometric cerium oxide formulations, CeO2-x denotes oxygen-deficient variants, and CeOx is reserved for formulations explicitly designated in the original study as nonstoichiometric cerium oxide.Fig. 1(A) Schematic illustration of the catalytic reaction of cerium oxide nanozymes for hydroxyl radical scavenging. (B) Intestinal CT images of colitis and healthy mice after oral administration of Dex-CeO2 nanozymes. Reproduced with permission [11]. Copyright 2020, American Chemical Society. (C) Schematic illustration of the anti-inflammatory and antifibrotic mechanisms of Dex-CeO2-x@CS/SA-gel nanozymes in IBD therapy. Reproduced with permission [38]. Copyright 2023, BioMed Central (Springer Nature). (D) Schematic illustration of the anti-inflammatory mechanism of Cur@PC-HA/CeO2 nanozymes in IBD therapy. Reproduced with permission [21]. Copyright 2024, BioMed Central (Springer Nature). (E) Gastrointestinal CT images of colitis and healthy mice after oral administration of Inu-CeO2 nanozymes or ioversol. Reproduced with permission [26]. Copyright 2025, Dove Medical Press. (A) Schematic illustration of the catalytic reaction of cerium oxide nanozymes for hydroxyl radical scavenging. (B) Intestinal CT images of colitis and healthy mice after oral administration of Dex-CeO2 nanozymes. Reproduced with permission [11]. Copyright 2020, American Chemical Society. (C) Schematic illustration of the anti-inflammatory and antifibrotic mechanisms of Dex-CeO2-x@CS/SA-gel nanozymes in IBD therapy. Reproduced with permission [38]. Copyright 2023, BioMed Central (Springer Nature). (D) Schematic illustration of the anti-inflammatory mechanism of Cur@PC-HA/CeO2 nanozymes in IBD therapy. Reproduced with permission [21]. Copyright 2024, BioMed Central (Springer Nature). (E) Gastrointestinal CT images of colitis and healthy mice after oral administration of Inu-CeO2 nanozymes or ioversol. Reproduced with permission [26]. Copyright 2025, Dove Medical Press. In one study, dextran-coated cerium oxide (Dex-CeO2) nanoparticles with a diameter of ∼17.5 nm were synthesized by a precipitation method, in which a solution of cerium nitrate was mixed with dextran and then added dropwise into ammonia solution, followed by reaction at 90 °C for 1 h [11]. Dextran was used to exploit its tendency to accumulate at inflamed segments, allowing cerium oxide to be preferentially enriched in inflamed intestinal tissue. This preferential enrichment partially overcomes the poor specificity of conventional iodine-based and barium sulfate-based contrast agents for IBD lesions, which hampers radiological identification of intestinal inflammation. In vivo CT imaging showed that, 24 h after oral administration of Dex-CeO2, no obvious enhancement was observed in the colon of healthy mice, whereas DSS colitis mice still exhibited marked colonic enhancement (Fig. 1B), indicating selective visualization of inflamed intestinal segments. In terms of attenuation performance, Dex-CeO2 displayed slightly higher mass attenuation than the iodine contrast agent iopamidol at all tested tube voltages, with an attenuation coefficient of 44.0 HU mL (mg Ce)−1 at a clinical tube voltage of 80 kV. This result suggests that, in addition to depicting bowel morphology, the CT signal may provide information on lesion-associated localization of the orally delivered therapeutic component. Building on this work, a subsequent study refined both the synthesis and delivery strategy (Fig. 1C). By optimizing the dextran-coating procedure, the authors replaced the previous fast reaction (90 °C, 1 h) with a slower reaction (25 °C, 24 h) and obtained Dex-CeO2-x nanoparticles with a hydrodynamic diameter of ∼30 nm [38]. The surface Ce3+ fraction was markedly increased compared with bare CeO2 (41% vs 28%). To enhance acid resistance after oral administration, Dex-CeO2-x was further embedded in a chitosan/sodium alginate hydrogel (Dex-CeO2-x@CS/SA-gel), enabling gastric stability and intestinal release. The main contribution of this study was to provide the previously missing in vivo evidence of anti-inflammatory efficacy. It also demonstrated suppression of fibrosis-related signaling, manifested as ROS scavenging, downregulation of TGF-β1 and α-SMA/collagen expression, and inhibition of myofibroblast activation. In chronic 2,4,6-trinitrobenzene sulfonic acid (TNBS)-induced and DSS-induced colitis models, oral Dex-CeO2-x (30 mg/kg) reduced the levels of inflammatory cytokines such as IL-1β and TNF-α and downregulated fibrotic markers including α-SMA and collagen. Histological evaluation and colonoscopic examination both indicated a reduction in colonic inflammation. Compared with relying solely on nanozyme-mediated catalytic ROS scavenging, coadministration with small-molecule anti-inflammatory drugs can provide more controllable and synergistic anti-inflammatory effects [85], owing to their well-established mechanisms of action and pharmacokinetic profiles [86]. Consistent with this concept, cerium oxide nanozymes were integrated with a conventional anti-inflammatory drug (Fig. 1D) [21]. CeO2 nanozymes were synthesized on a hyaluronic acid (HA) template to obtain HA/CeO2. Curcumin-loaded poly(lactic-co-glycolic acid) (Cur@PLGA) nanoparticles were coated with chitosan to form chitosan-coated Cur@PLGA (Cur@PC). The two components were then assembled via DMTMM-mediated amide coupling into core-shell Cur@PC-HA/CeO2 nanoparticles with a hydrodynamic diameter of ∼270 nm, and a reversal of the ζ potential (from +12.4 mV to −47.0 mV). At the same cerium content (normalized by Ce mass), Cur@PC-HA/CeO2 exhibited somewhat lower CT attenuation (∼24.2 HU mL (mg Ce)−1) than the Dex-CeO2 system, and comparable CAT-like activity required a higher cerium concentration (29 vs 20 μg Ce·mL−1), as shown in Table 3. This trade-off likely reflects its larger particle size and multilayer core-shell structure, in which CeO2 is confined to the outer layer and diluted by a substantial number of low-Z polymers, thereby reducing the effective electron density and accessible catalytic sites per unit cerium content. In DSS-induced colitis mice, however, oral Cur@PC-HA/CeO2 produced persistent CT enhancement in inflamed colonic segments, whereas the colonic signal in healthy mice rapidly declined. Together with ICP-MS evidence of cerium accumulation in inflamed tissue, these findings indicate that integration of a curcumin-loaded core with a CeO2 nanozyme shell affords a formulation that maintains preferential localization to inflamed colonic segments, enables selective visualization of inflamed intestinal segments, and achieves combined drug-mediated and nanozyme-mediated anti-inflammatory effects.Table 3Summary of nanozymes and nanozyme-like materials for integrated imaging and therapy in IBD.Table 3StrategySystemImaging coreSynthesis methodMean size (nm)Basis for preferential localizationPrimary imaging roleImaging performanceReported lesion retention window (h)CAT-like activity/O2 generationSOD-like activityAnti-inflammatory animal modelRef.CT-Guided StrategiesDex-CeO2CeO2Precipitation method∼17.5n/rFLR, LVB∼44.0 HU mL (mg Ce)−1 (80 kV, clinical scanner)24n/rn/rn/r[11]Dex-CeO2-x@CS/SA-gelCeO2Precipitation method∼30pH responsive releaseFLR, LVBQualitatively confirmedn/r∼0.014a mg·L−1·s−1 (material 200 μg mL−1; [Ce] 20 μg mL−1)∼0.9 U mL−1 at 100 μg mL−1DSS-induced acute colitis and TNBS-induced chronic colitis[38]Cur@PC-HA/CeO2CeO2Precipitation method; emulsion solvent evaporation method; electrostatic adsorption coating; amide-bond crosslinking self-assembly∼270CD44 receptor mediated uptakeFLR, LVB, TRM∼24.2 HU mL (mg Ce)−1 (80 kV, clinical scanner)24∼0.015a mg·L−1·s−1 (material 200 μg mL−1; [Ce] 29 μg mL−1)∼0.2 U mL−1 at 100 μg mL−1DSS-induced acute colitis[21]Inu-CeO2CeO2Precipitation method∼8.3n/rFLR, LVB∼47.7 HU mL (mg Ce)−1 (80 kV, clinical scanner)12∼0.041a mg·L−1·s−1 (material 200 μg mL−1; [Ce] 106 μg mL−1)Qualitatively confirmedDSS-induced acute colitis[26]SrDy2O4SrDy2O4High-temperature solid-state methodn/rn/rFLR, LVBQualitatively confirmedn/r∼327 U mg−1∼102 U mg−1DSS-induced acute colitis[87]Bi-HuABi (III)chelation-driven self-assembly method∼172n/rFLR, LVBQualitatively confirmed96n/rn/rDSS-induced acute and chronic colitis[51]BPCDBaSO4Precipitation method∼250Electrostatic adsorptionFLR, LVBQualitatively confirmed∼48∼0.010a mg·L−1·s−1 (material 100 μg mL−1; [Ce] 7 μg mL−1)∼0.6 U mL−1 at 100 μg mL−1DSS-induced acute colitis[88]Optical/PA-Guided StrategiesGB-CDsGB-CDsSolvothermal method∼2.4n/rFLR, LVBQualitatively confirmedn/rn/r3299 U mg−1bDSS-induced acute colitis[46]CR-CDsCR-CDsHigh-temperature carbonization∼2.2n/rFLR, LVBQualitatively confirmedn/rn/rQualitatively confirmedDSS-induced acute colitis[89]Tris/PEG-Mn@QDsMn@Ag2Te QDsOrganic-phase hot-injection method∼5.2Electrostatic adsorptionLVBQualitatively confirmedn/rQualitatively confirmedQualitatively confirmedDSS-induced acute colitis[90]AU-LIP-CMcAu clustersSingle-atom substitution; thin-film hydration followed by sonication and extrusion; biomimetic cell-membrane coating∼230Cell membrane enabled homingFLR, LVBQualitatively confirmedn/aQualitatively confirmedQualitatively confirmedDSS-induced acute colitis[91]PB@N3 and LR@DBCOPB@N3Co-precipitation; bioorthogonal click chemistry∼200n/rFLRQualitatively confirmedn/rQualitatively confirmedn/rDSS-induced acute colitis[47]MNPs-Lf-EVsMNPsChemical oxidative polymerization; sonication-reassembly130-150n/rFLR, LVBQualitatively confirmedn/rn/rn/rDSS-induced acute colitis[92]Multimodal Imaging-Guided StrategiesCeNP-PEGcIodinated contrast agent (for CTE) [18F] DPA-714 (for PET)Non-hydrolytic sol-gel reaction; phase-transfer PEGylation∼10n/rLVB, TRMQualitatively confirmedn/aQualitatively confirmedQualitatively confirmedDSS-induced acute colitis[49]MPN@CeOxCeOx (for CT)EGCG-Fe MPN (for MRI)Hydrothermal method; chelation-driven self-assembly method∼250ROS triggered activationFLR, LVBQualitatively confirmedn/rn/rQualitatively confirmedDSS-induced acute colitis[93]n/r - not reported; n/a - not applicable.FLR - formulation localization and retention; LVB - lesion visualization and inflammation burden assessment; TRM - treatment response monitoring; Note - Primary imaging role was assigned according to the main function supported by the original study.aThe O2 evolution rate over 0-300 s (d[O2]·dt−1, 0-300 s) was used as the quantitative index for CAT-like activity.bApparent SOD-like activity determined with a WST-based commercial SOD kit. The value is highly assay-dependent and should not be directly compared with U·mg−1 values measured by other methods.cnon-oral reference system. Table 3Summary of nanozymes and nanozyme-like materials for integrated imaging and therapy in IBD.Table 3StrategySystemImaging coreSynthesis methodMean size (nm)Basis for preferential localizationPrimary imaging roleImaging performanceReported lesion retention window (h)CAT-like activity/O2 generationSOD-like activityAnti-inflammatory animal modelRef.CT-Guided StrategiesDex-CeO2CeO2Precipitation method∼17.5n/rFLR, LVB∼44.0 HU mL (mg Ce)−1 (80 kV, clinical scanner)24n/rn/rn/r[11]Dex-CeO2-x@CS/SA-gelCeO2Precipitation method∼30pH responsive releaseFLR, LVBQualitatively confirmedn/r∼0.014a mg·L−1·s−1 (material 200 μg mL−1; [Ce] 20 μg mL−1)∼0.9 U mL−1 at 100 μg mL−1DSS-induced acute colitis and TNBS-induced chronic colitis[38]Cur@PC-HA/CeO2CeO2Precipitation method; emulsion solvent evaporation method; electrostatic adsorption coating; amide-bond crosslinking self-assembly∼270CD44 receptor mediated uptakeFLR, LVB, TRM∼24.2 HU mL (mg Ce)−1 (80 kV, clinical scanner)24∼0.015a mg·L−1·s−1 (material 200 μg mL−1; [Ce] 29 μg mL−1)∼0.2 U mL−1 at 100 μg mL−1DSS-induced acute colitis[21]Inu-CeO2CeO2Precipitation method∼8.3n/rFLR, LVB∼47.7 HU mL (mg Ce)−1 (80 kV, clinical scanner)12∼0.041a mg·L−1·s−1 (material 200 μg mL−1; [Ce] 106 μg mL−1)Qualitatively confirmedDSS-induced acute colitis[26]SrDy2O4SrDy2O4High-temperature solid-state methodn/rn/rFLR, LVBQualitatively confirmedn/r∼327 U mg−1∼102 U mg−1DSS-induced acute colitis[87]Bi-HuABi (III)chelation-driven self-assembly method∼172n/rFLR, LVBQualitatively confirmed96n/rn/rDSS-induced acute and chronic colitis[51]BPCDBaSO4Precipitation method∼250Electrostatic adsorptionFLR, LVBQualitatively confirmed∼48∼0.010a mg·L−1·s−1 (material 100 μg mL−1; [Ce] 7 μg mL−1)∼0.6 U mL−1 at 100 μg mL−1DSS-induced acute colitis[88]Optical/PA-Guided StrategiesGB-CDsGB-CDsSolvothermal method∼2.4n/rFLR, LVBQualitatively confirmedn/rn/r3299 U mg−1bDSS-induced acute colitis[46]CR-CDsCR-CDsHigh-temperature carbonization∼2.2n/rFLR, LVBQualitatively confirmedn/rn/rQualitatively confirmedDSS-induced acute colitis[89]Tris/PEG-Mn@QDsMn@Ag2Te QDsOrganic-phase hot-injection method∼5.2Electrostatic adsorptionLVBQualitatively confirmedn/rQualitatively confirmedQualitatively confirmedDSS-induced acute colitis[90]AU-LIP-CMcAu clustersSingle-atom substitution; thin-film hydration followed by sonication and extrusion; biomimetic cell-membrane coating∼230Cell membrane enabled homingFLR, LVBQualitatively confirmedn/aQualitatively confirmedQualitatively confirmedDSS-induced acute colitis[91]PB@N3 and LR@DBCOPB@N3Co-precipitation; bioorthogonal click chemistry∼200n/rFLRQualitatively confirmedn/rQualitatively confirmedn/rDSS-induced acute colitis[47]MNPs-Lf-EVsMNPsChemical oxidative polymerization; sonication-reassembly130-150n/rFLR, LVBQualitatively confirmedn/rn/rn/rDSS-induced acute colitis[92]Multimodal Imaging-Guided StrategiesCeNP-PEGcIodinated contrast agent (for CTE) [18F] DPA-714 (for PET)Non-hydrolytic sol-gel reaction; phase-transfer PEGylation∼10n/rLVB, TRMQualitatively confirmedn/aQualitatively confirmedQualitatively confirmedDSS-induced acute colitis[49]MPN@CeOxCeOx (for CT)EGCG-Fe MPN (for MRI)Hydrothermal method; chelation-driven self-assembly method∼250ROS triggered activationFLR, LVBQualitatively confirmedn/rn/rQualitatively confirmedDSS-induced acute colitis[93]n/r - not reported; n/a - not applicable.FLR - formulation localization and retention; LVB - lesion visualization and inflammation burden assessment; TRM - treatment response monitoring; Note - Primary imaging role was assigned according to the main function supported by the original study.aThe O2 evolution rate over 0-300 s (d[O2]·dt−1, 0-300 s) was used as the quantitative index for CAT-like activity.bApparent SOD-like activity determined with a WST-based commercial SOD kit. The value is highly assay-dependent and should not be directly compared with U·mg−1 values measured by other methods.cnon-oral reference system. Summary of nanozymes and nanozyme-like materials for integrated imaging and therapy in IBD. n/r - not reported; n/a - not applicable. FLR - formulation localization and retention; LVB - lesion visualization and inflammation burden assessment; TRM - treatment response monitoring; Note - Primary imaging role was assigned according to the main function supported by the original study. The O2 evolution rate over 0-300 s (d[O2]·dt−1, 0-300 s) was used as the quantitative index for CAT-like activity. Apparent SOD-like activity determined with a WST-based commercial SOD kit. The value is highly assay-dependent and should not be directly compared with U·mg−1 values measured by other methods. non-oral reference system. Simplifying the synthetic sequence helps reduce system complexity, lower the risk of batch-to-batch variation [94], and improve the reproducibility and scalability of the preparation process [95,96]. Accordingly, one study adopted a one-pot alkaline precipitation route: inulin and cerium nitrate were mixed under alkaline conditions and reacted at 37 °C for 4 h to obtain Inu-CeO2 nanoparticles with a hydrodynamic diameter of ∼8.3 nm (Fig. 1E) [26]. Inulin endowed the particles with a certain degree of acid resistance and colonic preference, thereby mitigating the tendency of bare CeO2 to aggregate in the gastrointestinal environment. These nanoparticles also exhibited favorable in vivo CT performance. At a clinical tube voltage of 80 kV, the attenuation coefficient reached ∼47.7 HU mL (mg Ce)−1. Moreover, 12 h after oral administration of Inu-CeO2, DSS colitis mice still showed pronounced enhancement in the colon, whereas healthy mice showed almost no enhancement. These findings indicate selective visualization of inflamed intestinal segments and suggest utility mainly for indicating lesion-associated entry and short-term retention of the oral therapeutic carrier. Dy-based AB2X4 platforms Starting from the requirements of oral therapy for IBD, a machine-learning (ML)-assisted high-throughput screening strategy for AB2X4 nanozymes was introduced [87]. Quantifiable descriptors such as antioxidant capacity, acid stability, and ζ potential were digitized and fed into ML models combined with SISSO regression to establish a semiquantitative structure-property relationship for the AB2X4 family. To ensure adequate CT contrast, high-Z rare-earth oxides were prioritized. In particular, Dy-containing compositions with relatively large X-ray attenuation cross-sections over the clinical CT tube-voltage range were selected to combine ROS scavenging with CT visibility. Fig. 2A illustrates the ML-assisted screening of a large AB2X4 candidate library, which identified SrDy2O4 as a nanozyme that combines SOD-like and CAT-like activities with a moderately negative ζ potential. Well-crystallized bulk particles were then obtained by high-temperature solid-state synthesis. The material maintained its crystal structure and enzyme-like activity in simulated gastric fluid, indicating suitable physicochemical robustness for oral administration. Under clinical CT conditions, SrDy2O4 displayed favorable mass attenuation. In DSS-induced acute colitis models, oral administration of SrDy2O4 led to sustained enhancement of inflamed colonic segments, with contrast retention markedly longer than that of iodine-based agents. Therapeutically, SrDy2O4 significantly improved disease phenotypes, reduced multiple pro-inflammatory cytokines, and suppressed mitochondria-associated apoptosis. In parallel, 16S rRNA sequencing showed partial restoration of intestinal microbiota diversity and enrichment of beneficial taxa. Together, this workflow illustrates one ML-assisted route for designing CT-guided oral nanozyme platforms, particularly for extending the imaging window during which residence of the orally delivered material at inflamed sites can be followed.Fig. 2(A) ML-guided design workflow of SrDy2O4 nanozymes and their gastrointestinal CT imaging evaluation. Reproduced with permission [87]. Copyright 2025, Wiley-VCH. (B) Schematic illustration of the synthesis of Bi-HuA nanoparticles and their anti-inflammatory mechanism in IBD therapy. Reproduced with permission [51]. Copyright 2025, Elsevier. (C) Gastrointestinal CT imaging evaluation and anti-inflammatory mechanism of BPCD nanozymes for IBD therapy. Reproduced with permission [88]. Copyright 2025, American Chemical Society. (A) ML-guided design workflow of SrDy2O4 nanozymes and their gastrointestinal CT imaging evaluation. Reproduced with permission [87]. Copyright 2025, Wiley-VCH. (B) Schematic illustration of the synthesis of Bi-HuA nanoparticles and their anti-inflammatory mechanism in IBD therapy. Reproduced with permission [51]. Copyright 2025, Elsevier. (C) Gastrointestinal CT imaging evaluation and anti-inflammatory mechanism of BPCD nanozymes for IBD therapy. Reproduced with permission [88]. Copyright 2025, American Chemical Society. Bi-based platforms Previous studies have shown that Bi-based platforms can provide excellent in vivo imaging performance. For example, Bi2S3@SiO2 achieves higher linear CT contrast than BaSO4 and iodine-based agents at lower doses, and 28-day follow-up indicates minimal toxicity and limited disturbance of the intestinal microbiota, with a possible tendency to increase microbial diversity [97]. Dextran-coated Bi2O3 (Dex-Bi2O3) likewise offers higher linear contrast than iodine-based contrast agents and low toxicity, while also showing preferential accumulation at inflamed sites [98]. However, these systems did not incorporate therapeutic modules, and the therapeutic potential of bismuth-based platforms in IBD has not yet been clearly defined. As shown in Fig. 2B, a chelation-driven metal-phenolic network (MPN) self-assembly strategy was employed to construct Bi-humic acid (Bi-HuA, humic acid is abbreviated as HuA in this work) nanoparticles [51]. Bi(NO3)3·5H2O was first dissolved in dilute HNO3 and then added dropwise into an alkaline HuA solution, followed by the addition of a small amount of FeCl3 to promote MPN formation. After reaction at room temperature for 12 h, Bi-HuA nanoparticles with a hydrodynamic diameter of ∼172 nm were obtained. In this system, HuA serves both as an organic scaffold providing multiple coordination sites and as a therapeutic component with anti-inflammatory, antioxidant, and microbiota-modulating functions. Coordination with Bi effectively enhances its in vivo stability. In Bi-HuA, Bi(III) acts as the high-Z attenuating element, giving a mass attenuation coefficient markedly higher than that of ioversol. In DSS-induced colitis, the nanoparticles selectively accumulated at inflamed colonic sites, and the resulting CT enhancement persisted for up to 96 h. In addition, Bi-HuA was well tolerated in vivo and improved disease-related readouts by protecting mitochondria, suppressing the NF-κB pathway, lowering pro-inflammatory cytokine levels, and improving multiple inflammation-related indices. In parallel, the composition and diversity of the intestinal microbiota showed a trend toward restoration. Although ROS-scavenging capacity has been demonstrated, direct evidence for enzyme-like catalysis such as CAT-like activity (for example, O2 evolution, increases in dissolved oxygen, or Michaelis-Menten kinetics) is still lacking, so whether Bi-HuA qualifies as a nanozyme in the strict sense remains to be clarified. In addition, the precise molecular mechanism underlying its preferential accumulation in inflamed intestinal segments has not yet been elucidated. Contrast core-based CT imaging The above studies indicate that ceria-based and Bi-based platforms intrinsically possess theranostic potential. However, using them directly as contrast agents requires relatively high oral doses [99,100], and such intake levels may raise concerns about biosafety [101]. To reduce the amount of these metals with potential safety risks, a more rational strategy is to decouple imaging from anti-inflammatory catalysis: employ clinically established contrast agents for visualization, and let nanozymes focus on catalysis so that their dosage can be minimized. Based on this concept, the BaSO4@PDA@CeO2/DSP (BPCD) nanozyme was designed (Fig. 2C) [88]. BaSO4 particles were first coated with a polydopamine (PDA) shell via base-triggered oxidative polymerization, then dispersed in ethylene glycol containing Ce(NO3)3 to grow CeO2 nanosheets in situ on the PDA shell while loading dexamethasone sodium phosphate (DSP), yielding particles with a hydrodynamic diameter of ∼250 nm. The negatively charged PDA shell promotes electrostatic binding to positively charged proteins in the inflammatory microenvironment, facilitating phagocytic uptake and enhancing local retention. CT imaging is provided by the BaSO4 cores, which afford discernible contrast at lesions for approximately 48 h after oral administration, suggesting its value as a readout of formulation coverage and local retention under oral dosing. Benefiting from the dispersed distribution of CeO2 on PDA and the markedly higher surface Ce3+ fraction compared with bare CeO2 (42% vs 28%), BPCD exhibits clear CAT-like behavior and elevated dissolved oxygen levels in vitro. Based on the dissolved oxygen evolution rate, BPCD at 7 μg Ce·mL−1 achieves an O2 generation rate of ∼0.010 mg L−1 s−1, indicating efficient utilization of Ce and an enzymatic profile closer to “high efficiency at low loading”. In a DSS-induced acute colitis model, BPCD exerted therapeutic effects by scavenging ROS and remodeling the immune microenvironment. It promoted macrophage polarization toward the M2 phenotype, suppressed JAK/STAT and NF-κB activation, and reduced pro-inflammatory cytokines and myeloperoxidase (MPO) activity, with concomitant improvements in disease activity index (DAI), colon length, and epithelial barrier function. Systemically, complete blood counts, serum biochemistry and histopathology remained within physiological limits after oral administration, indicating better tolerability than bare CeO2 nanozymes. Optical/PA-guided visual theranostic nanozyme strategies Optical/PA strategies rely on platforms whose photophysical behaviors can be adjusted through composition, band structure, and surface chemistry. Unlike CT-guided systems, these approaches typically emphasize tunability and responsiveness, allowing catalytic activity, signal generation, and intestinal interactions to influence one another. As a result, the studies in this section are organized by the type of optical module, separating intrinsically fluorescent nanozyme platforms from broadband absorber-based platforms. The following subsections summarize these general routes and the design considerations that guide them. Nanozyme-intrinsic fluorescence-based optical imaging Carbon dot platforms Carbon dots (C-dots) are a class of fluorescent nanomaterials composed of a carbon core and surface functional groups [102]. In recent years, they have been shown to display enzyme-like activities such as ROS scavenging and have gradually emerged as a typical carbon-based nanozyme system [103]. Owing to their high photostability, tunable emission, and good biocompatibility, C-dot nanozymes are promising candidates to serve as both antioxidant catalytic centers and optical imaging probes within a single platform [104]. Based on this concept, the researchers constructed an oral nanozyme system in which red-emissive C-dots acted as a combined imaging-therapeutic core for visual intervention in ulcerative colitis. Specifically, GSH/biotin-derived C-dots (GB-CDs) with an average diameter of ∼2.4 nm were synthesized by a solvothermal method using formamide as the solvent and glutathione (GSH) and D-biotin as precursors, followed by reaction at 160 °C for 8 h (Fig. 3A) [46]. Compared with blue or green emission, red fluorescence experiences less scattering and lower autofluorescence background in biological tissues, which improves penetration depth and signal-to-noise ratio for in vivo intestinal imaging [105]. Red-emissive GB-CDs are thus more suitable for whole-intestine imaging and assessment of inflammation distribution. Intestinal imaging performance was quantified by ex vivo fluorescence analysis of colonic regions of interest (ROIs) after in vivo imaging. Oral GB-CD nanozymes produced stronger fluorescence in inflamed than in non-inflamed colonic tissue, suggesting inflammation-associated accumulation and macroscopic lesion imaging capability. In this context, the optical signal is valuable mainly because it makes the lesion-associated distribution of the oral therapeutic platform more visible. However, this strategy mainly relies on passive enrichment driven by the colitis microenvironment (such as pH and charged proteins), and direct evidence for molecular-level specificity of imaging within lesions is still lacking. Benefiting from their relatively high SOD-like activity, GB-CD nanozymes efficiently scavenged ROS in cell models of inflammation and oxidative stress, attenuated inflammatory responses, and preserved cell viability. The reported SOD-like activity, calculated using a WST-based commercial assay, reached approximately 3.3 × 103 U mg−1. As WST-based readouts are sensitive to assay conditions and data processing [106], such values are best interpreted together with complementary catalytic evaluations and used with caution for cross-comparison between different nanozyme systems [107]. In DSS-induced acute colitis mice, oral administration of GB-CDs ameliorated the overall disease phenotype and reduced colonic levels of pro-inflammatory cytokines.Fig. 3(A) Schematic illustration of the synthesis of GB-CD nanozymes, their gastrointestinal optical imaging, and anti-inflammatory mechanism. Reproduced with permission [46]. Copyright 2024, Elsevier. (B) Schematic illustration of the in vivo macrophage-associated localization of CR-CD nanozymes and their anti-inflammatory mechanism in IBD therapy. Reproduced with permission [89]. Copyright 2025, American Chemical Society. (C) Schematic illustration of Tris/PEG-Mn@QDs nanozymes for NIR-IIb imaging in IBD models and the positive correlation between NIR-IIb signal intensity and the degree of inflammatory injury. Reproduced with permission [90]. Copyright 2025, American Chemical Society. (D) Schematic illustration of the inflammation-associated mechanism underlying preferential accumulation of AU-LIP-CM nanozymes at inflamed intestinal segments. Reproduced with permission [91]. Copyright 2025, Wiley-VCH. (A) Schematic illustration of the synthesis of GB-CD nanozymes, their gastrointestinal optical imaging, and anti-inflammatory mechanism. Reproduced with permission [46]. Copyright 2024, Elsevier. (B) Schematic illustration of the in vivo macrophage-associated localization of CR-CD nanozymes and their anti-inflammatory mechanism in IBD therapy. Reproduced with permission [89]. Copyright 2025, American Chemical Society. (C) Schematic illustration of Tris/PEG-Mn@QDs nanozymes for NIR-IIb imaging in IBD models and the positive correlation between NIR-IIb signal intensity and the degree of inflammatory injury. Reproduced with permission [90]. Copyright 2025, American Chemical Society. (D) Schematic illustration of the inflammation-associated mechanism underlying preferential accumulation of AU-LIP-CM nanozymes at inflamed intestinal segments. Reproduced with permission [91]. Copyright 2025, Wiley-VCH. In another study, a C-dot nanozyme with blue-green emission was reported (Fig. 3B) [89]. This type of C-dot nanozyme showed high fluorescence brightness and was highly compatible with standard confocal and immunofluorescence channels [108]. It was therefore suitable for high-resolution localization and colocalization analysis of inflamed segments and specific cell subsets (such as macrophages) in tissue sections and superficial tissues. Cimicifugae rhizome-derived C-dots (CR-CDs) with an average diameter of ∼2.2 nm were obtained by a combined carbonization-water extraction procedure, using Cimicifugae rhizome as the precursor, followed by carbonization at 350 °C for 1 h and subsequent water extraction [89]. Compared with the aforementioned red-emissive C-dots strategy that mainly enhanced whole-colon imaging, this work placed greater emphasis on the distribution characteristics at inflammatory foci and at the macrophage level. After oral administration, CR-CDs preferentially accumulated in inflamed colonic segments in DSS-induced colitis mice. They colocalized with F4/80+ macrophages, suggesting macrophage-associated localization in the inflammatory microenvironment. Owing to their multi-enzyme-like antioxidant activity, CR-CDs scavenged excess ROS, activated the colonic Nrf2/HO-1 antioxidant pathway, downregulated pro-inflammatory cytokines, and reshaped UC-associated intestinal dysbiosis. However, because this system originates from a complex carbonized herbal product, its specific structural motifs and active sites have not yet been clearly resolved. Inorganic semiconductor quantum-dot platforms Inorganic semiconductor quantum-dot (QD) nanozymes are nanomaterial systems that use semiconductor QDs as an inorganic scaffold and acquire enzyme-like catalytic activity through elemental doping or surface ligand regulation [109]. They offer precisely tunable emission wavelengths, high fluorescence quantum yields and photostability, and relatively well-defined crystal structures and chemical compositions [110]. Their bandgaps and surface chemistries can be engineered to extend emission into the NIR-II region, enabling high signal-to-noise deep-tissue imaging [111]. Mn@Ag2Te QDs were synthesized by an organic-phase hot-injection method at 180 °C [90]. The QDs were then coated with branched oleylamine-poly(acrylic acid) (OPA) to introduce surface carboxyl groups and subsequently coupled with 8-arm PEG-NH2 and tris(hydroxymethyl)aminomethane (Tris), yielding Tris/PEG-modified Mn@Ag2Te QDs (Tris/PEG-Mn@QDs) with pH-responsive surface charge reversal and an average diameter of ∼5.2 nm. Owing to the enhancement of radiative recombination by Mn doping, these QDs showed high fluorescence quantum yield and a high signal-to-noise ratio in the NIR-IIb region (∼1500-1700 nm). The NIR-IIb readout in this platform therefore follows an emission-based mechanism, with radiative recombination within the QDs serving as the signal source. Compared with visible or NIR-I emission, NIR-IIb imaging experiences weaker tissue scattering and lower tissue autofluorescence, which markedly improves the effective penetration depth and spatial resolution for intestinal imaging. By combining efficient NIR-IIb imaging with inflammation-driven mucosal accumulation mediated by surface charge reversal, Tris/PEG-Mn@QDs enabled integrated imaging readout and therapy in DSS-induced acute colitis. After oral administration, NIR-IIb signal intensity in the colon correlated positively with the degree of inflammation (Fig. 3C). Its sensitivity to early, mild lesions surpassed that of conventional indices such as DAI, colon length, and tight junction protein expression, indicating that the optical readout of this system is highly sensitive to inflammatory changes. In parallel, the SOD-like and CAT-like activities of Tris/PEG-Mn@QDs alleviated oxidative stress and epithelial apoptosis in the inflamed mucosa, mitigated colon shortening and histological inflammatory infiltration, and reduced levels of pro-inflammatory cytokines. Thus, a single inorganic QD platform achieved in vivo grading of ulcerative colitis lesions together with ROS-targeted intervention. Noble metal cluster platforms Similar to the C-dot and inorganic QD systems described above, noble metal clusters (NMCs) can integrate optical imaging and enzyme-like catalytic functions within a single nanostructure [112]. Au cluster nanozymes consist of cores containing several to a few dozen gold atoms [113]. Under quantum confinement, they exhibit tunable intrinsic fluorescence and abundant surface-active sites [114]. These features make them suitable NIR emitters and allow them to mimic antioxidant enzymes such as SOD and CAT via electron-transfer processes [115]. A nanoplatform composed of Au cluster nanozymes and a dual-membrane biomimetic liposome was designed for ROS-scavenging-driven IBD therapy [91]. Au cluster nanozymes with SOD-like and CAT-like activities were first obtained by a single-atom doping strategy and then encapsulated in liposomes to form AU-LIP. The liposomes were subsequently camouflaged with a hybrid membrane derived from fused neutrophil and red blood cell membranes, yielding a dual-membrane biomimetic construct termed AU-LIP-CM. Under 780 nm excitation, AU-LIP-CM emitted NIR-II fluorescence with wavelengths above 1200 nm and thus served as an imaging core. At the same time, the membrane shell endowed the platform with improved inflammatory-site accumulation and immune-evasive properties, supporting preferential accumulation in inflamed intestinal segments in DSS-induced colitis mice (Fig. 3D). The system showed prolonged NIR-II imaging and clear anti-inflammatory activity in a DSS-induced acute colitis model. Because this platform has so far been evaluated only by intravenous administration, it is discussed here mainly as a conceptual reference. Broadband absorber-based optical/PA imaging Prussian Blue platforms Prussian Blue (PB) nanozymes are inorganic enzyme-like nanomaterials built on a ferrocyanide coordination framework [116]. With appropriate surface modification, they can achieve good dispersibility and robust enzyme-like activity [117,118]. A cooperative therapeutic system combining PB nanozymes and probiotics was developed (Fig. 4A) [47]. PB nanoparticles were synthesized by a polyethyleneimine (PEI)-stabilized co-precipitation method and subsequently functionalized with azido (N3) groups to yield PB@N3 nanozymes. Lactobacillus reuteri was modified to bear DBCO groups, generating LR@DBCO probiotics. At the inflamed colonic mucosa, PB@N3 and LR@DBCO underwent an in situ N3-DBCO click reaction to form a stable complex. This design enabled a spatiotemporal cooperative effect under oral administration: PB@N3 first scavenged excess ROS and improved the local microenvironment, and probiotic colonization was subsequently promoted on the PB@N3 surface. In DSS-induced acute colitis, this cooperative system produced greater improvement in clinical and histological indices, as well as reduced colonic pro-inflammatory cytokines and increased anti-inflammatory cytokines, compared with PB or probiotics alone. It is noteworthy that this N3-DBCO-based in vivo click-guided colonization strategy builds on earlier concepts of bioorthogonal click chemistry for nanoparticle and cell delivery [119,120]. Here, it was extended to an oral nanozyme-probiotic combination system and combined with small-animal PA imaging to compare PB and PB@N3. PB@N3 showed stronger accumulation in inflamed colon and superior PA performance. However, the main source of the enhanced PA signal remains unclear: it is not yet resolved whether the increase arises primarily from higher local colonization of PB@N3 in inflamed tissue, or from improved per-dose PA output and catalytic amplification of the nanozyme. This question will require clarification through quantitative biodistribution studies and normalized imaging analyses. For now, the PA signal is more informative for interpreting local enrichment and microenvironment-responsive behavior than for directly judging efficacy.Fig. 4(A) Schematic illustration of in vivo PA imaging of PB@N3 nanozymes and their anti-inflammatory mechanism in IBD models. Reproduced with permission [47]. Copyright 2024, Elsevier. (B) Schematic illustration of in vivo PA and fluorescence imaging of MNPs-Lf-EVs nanozymes and their anti-inflammatory mechanism in IBD models. Reproduced with permission [92]. Copyright 2025, Elsevier. (A) Schematic illustration of in vivo PA imaging of PB@N3 nanozymes and their anti-inflammatory mechanism in IBD models. Reproduced with permission [47]. Copyright 2024, Elsevier. (B) Schematic illustration of in vivo PA and fluorescence imaging of MNPs-Lf-EVs nanozymes and their anti-inflammatory mechanism in IBD models. Reproduced with permission [92]. Copyright 2025, Elsevier. Melanin-based platforms Melanin nanomaterials (MNPs) possess broad-spectrum free-radical scavenging capability and good biocompatibility [121], and can mimic the functions of endogenous antioxidant enzymes [122]. In this context, bacterial extracellular vesicles derived from Lactobacillus fermentum were employed as carriers for MNPs (Fig. 4B). Negatively charged MNPs were first synthesized by chemical oxidative polymerization and then loaded into the vesicle lumen through a sonication-reassembly process, yielding MNPs-Lf-EVs with a hydrodynamic diameter of 130-150 nm [92]. This system took advantage of the probiotic tropism retained by extracellular-vesicle membrane proteins, enabling enrichment and prolonged retention at inflamed colonic mucosa in DSS colitis mice after oral administration. MNPs-Lf-EVs thereby acted as broadband absorbers for PA and near-infrared imaging, allowing in vivo visualization of inflammation extent and nanoplatform distribution. By reducing intracellular and tissue ROS levels and promoting epithelial barrier repair as well as intestinal-microbiota modulation, MNPs-Lf-EVs achieved greater improvements in colitis than free MNPs. The antioxidant behavior of this material was mainly supported by endpoint assays (radical scavenging, reduced ROS levels, and mitochondrial protection), without canonical enzyme-kinetics validation such as dissolved oxygen validation and Michaelis-Menten analysis. Under a strict definition, it therefore resembles a potent antioxidant nanoplatform with nanozyme-like features, although the redox-active sites in melanin still leave open the possibility of genuine nanozyme catalytic activity. Multimodal imaging-guided visual theranostic nanozyme strategies Multimodal imaging can integrate the complementary strengths of different modalities. In an organ as structurally complex and highly motile as the intestine, it can concurrently support precise lesion localization, assessment of inflammation severity, and monitoring of drug distribution [123]. Polyethylene glycol modified cerium oxide (CeNP-PEG) nanozymes with a hydrodynamic diameter of ∼10 nm were prepared [49]. In an acute colitis mouse model, intravenous administration reduced disease severity and histological inflammation. Because this formulation lacked an intrinsic imaging module, treatment responses were evaluated using CTE and TSPO targeted [18F] DPA-714 positron emission tomography/computed tomography (PET/CT) (Fig. 5A). These tools provided structural and inflammatory readouts of treatment effect. Because this study involved neither oral administration nor an intrinsic imaging module, it is better regarded here as an example of imaging-assisted evaluation of nanozyme therapy, instead of being presented as a representative oral imaging-guided nanozyme platform.Fig. 5(A) CT enterography (CTE) and [18F] DPA-714 PET/CT images of CeNP-PEG nanozyme-treated IBD models. Reproduced with permission [49]. Copyright 2022, BioMed Central (Springer Nature). (B) Schematic illustration of the in vivo anti-inflammatory mechanism of MPN@CeOx nanozymes and their CT and MRI imaging mechanisms. Reproduced with permission [93]. Copyright 2023, Elsevier. (A) CT enterography (CTE) and [18F] DPA-714 PET/CT images of CeNP-PEG nanozyme-treated IBD models. Reproduced with permission [49]. Copyright 2022, BioMed Central (Springer Nature). (B) Schematic illustration of the in vivo anti-inflammatory mechanism of MPN@CeOx nanozymes and their CT and MRI imaging mechanisms. Reproduced with permission [93]. Copyright 2023, Elsevier. Combining magnetic resonance imaging (MRI) and CT in IBD allows MRI to provide soft-tissue contrast and functional information [124]. CT offers high spatial resolution and quantitative attenuation signals; together with positron emission tomography (PET), it establishes a more reliable framework for imaging-guided evaluation of orally administered nanomedicines [125]. Nonstoichiometric cerium oxide (CeOx) nanozymes were synthesized by a hydrothermal method and encapsulated within an EGCG-Fe MPN shell to form core-shell MPN@CeOx particles with a hydrodynamic diameter of ∼250 nm. An enteric coating was further introduced to enable colon-specific release at colonic pH after oral administration (Fig. 5B) [93]. In this system, the inner CeOx layer, as a high-Z component, served as the CT contrast core and showed stronger X-ray attenuation in vitro than a clinical iodine contrast agent. The outer Fe-containing MPN responded to the H2O2-enriched inflammatory microenvironment and released metal ions capable of T1-weighted enhancement. This design enabled MRI-selective imaging of ROS-rich inflamed segments and provided complementary CT/MRI readouts on a single nanoplatform. Owing to its negatively charged surface and polyphenolic groups, this material also showed adhesion to and retention on inflamed colonic mucosa, which helped increase local exposure to the formulation. Functionally, the nanozyme system exhibited typical enzyme-like antioxidant activity and ROS-responsive behavior, alleviating oxidative-stress-induced cell damage and suppressing inflammatory responses in vitro. In a DSS-induced acute colitis model, oral administration of enteric-coated MPN@CeOx was tracked by multimodal imaging, confirming its accumulation in inflamed colon and the coordinated enhancement of CT/MRI signals at lesion sites. It also improved overall disease phenotype under both prophylactic and therapeutic dosing regimens, accompanied by global downregulation of inflammation-related transcripts. To the best of our knowledge, among orally administered nanozyme (or nanozyme-like) platforms that integrate CT and MRI readouts within the same carrier and report therapeutic outcomes in IBD models, only a limited number of examples have been reported to date. Within a single system, it combines imaging guidance, anti-inflammatory and antioxidant effects, and remodeling of the intestinal microenvironment, illustrating one possible route for the future development of nanozyme-based IBD therapies with built-in visualization and evaluability. Discussion and perspectives The studies reviewed above indicate that oral nanozymes for IBD are increasingly judged not only by anti-inflammatory efficacy, but also by whether their performance can be evaluated and compared in a consistent, clinically relevant manner. In practice, platform design should reconcile several recurring tensions, including the integration of imaging and catalytic functions under oral dosing constraints, and the pursuit of higher activity with a relatively lower material burden. Accordingly, the following discussion is organized as a practical framework that links imaging purpose, platform architecture, catalytic design, safety control, and validation strategy. This section summarizes the shared design considerations and suggests practical ways to compare different systems more fairly. Delivery barriers and design of oral nanozymes Gastrointestinal delivery directly affects both local therapeutic exposure and the interpretation of imaging readouts in oral nanozyme systems [126]. Variations in pH, digestive components, mucus properties, microbiota composition, and motility can alter platform integrity, lesion entry, local residence, and signal generation [55,127]. Protection against gastric degradation and premature release is the first requirement. In the systems reviewed here, this is mainly achieved by hydrogel embedding, enteric coating, or polymer-assisted stabilization. Dex-CeO2-x@CS/SA-gel improves gastric stability and supports intestinal release [38], whereas enteric-coated MPN@CeOx enables colonic release after enteric protection [93]. Inu-CeO2 represents a simpler stabilization strategy in which inulin contributes to oral stability while preserving CT visibility after administration [26]. Passage through the mucus layer is the second requirement. Smaller size, hydrophilic interfaces, and charge regulation may favor mucus entry [128], whereas stronger interfacial interaction may support persistence after lesion entry [129]. This point is particularly important in optical/PA systems, because fluorescence or PA signals are difficult to interpret if they mainly reflect luminal residence instead of lesion-associated localization. Tris/PEG-Mn@QDs illustrates one route through pH-responsive surface charge reversal under inflammatory conditions [90]. Mucoadhesion and prolonged local retention form the third requirement, and they also determine how lesion enrichment should be interpreted in oral systems. Because these platforms first interact with the intestinal lumen, mucus layer, and inflamed mucosal surface, increased lesion-associated signals should not be automatically interpreted as molecular targeting. In many cases, they more appropriately reflect microenvironment-mediated preferential localization or retention, arising from mucus anchoring [57,129], electrostatic interactions with proteins enriched in inflamed mucosa [88], pH- or charge-responsive behavior [90], or macrophage-associated uptake and colocalization within inflammatory foci [89]. In CT-guided systems, attenuation depends not only on the high-Z component itself but also on local volume fraction, spatial distribution, and retained mass [65,66]. This general principle has already been illustrated by Bi-GLUE through improved mucus coverage and anchoring [57]. In the oral nanozyme systems emphasized here, a similar retention-oriented design is seen in BPCD, where the PDA-containing structure may support lesion-associated residence [88]. Local retention in such systems can influence both therapeutic exposure and the persistence of the CT readout. Microbiota-associated responsiveness represents a fourth, system-specific design route. This route appears mainly in probiotic-assisted or vesicle-assisted systems and is particularly relevant to colonic delivery. PB@N3 combined with LR@DBCO promotes local probiotic colonization after oral administration, linking ROS regulation, local enrichment, and PA visibility within one cooperative framework [47]. MNPs-Lf-EVs use probiotic membrane-derived extracellular vesicles to increase lesion-associated enrichment and retention at inflamed colonic mucosa while also supporting PA and near-infrared readout [92]. These examples indicate that microbiota-associated design can contribute not only to localization and retention, but also to local microenvironment modulation. Gastric protection, mucus traversal, lesion retention, and microbiota-associated responsiveness do not contribute equally across imaging strategies. CT-guided systems depend more strongly on coverage and residence, whereas optical/PA-guided systems place greater weight on signal interpretability and background control. Multimodal systems additionally require coordinated release and temporal consistency between different readouts. Future evaluation should therefore extend beyond simple buffer-based assays to include simulated gastric and intestinal environments, not only for delivery-related testing but also for enzyme-like kinetic measurements [130]. Within the same framework, future studies should also assess lesion retention time, systemic leakage across the damaged epithelial barrier, and the relationship between readout intensity and pathological severity [131]. Approaches to enhance imaging and diagnostic performance Research in oral nanozymes is shifting from purely therapeutic use toward imaging-guided therapy and gradually toward the use of quantitative imaging to evaluate both efficacy and safety [21,38,130]. For platforms aiming to move to clinical translation, simply providing stronger radical scavenging is unlikely to be sufficient. It will be increasingly important to prioritize three core aspects: lesion-preferential delivery, microenvironment responsiveness, and imaging-based monitoring that reflects disease burden and therapeutic response as reliably as possible [55,132]. Among the oral nanozyme strategies reviewed in this article, most imaging schemes can be broadly divided into CT-guided high-spatial-resolution strategies and optical/PA-guided high-sensitivity strategies, with a smaller number of multimodal platforms integrating structural and functional information. CT-guided platforms are currently the closest to routine diagnostic workflows and may therefore be more readily aligned with translational development [133]. In these designs, high-Z elements such as Ce, Dy, and Bi are sometimes incorporated into a single nanozyme core that provides both CT contrast and catalytic activity. In other cases, imaging and catalysis are decoupled by using a clinically established radiopaque core (e.g., BaSO4 or iodine-based contrast agents) together with an outer layer that carries nanozymes and drugs. In the latter case, imaging and catalysis are handled by separate components, allowing independent optimization and reducing the dependence of imaging performance on the dose of catalytic metals. Compared with CT, optical/PA imaging strategies place greater emphasis on sensitivity and temporal resolution [134,135]. C-dots, QDs, and NMCs are frequently used as the main imaging components and often serve simultaneously as antioxidant centers through intrinsic or ligand-installed functional groups [136,137]. Red and NIR/NIR-II emission can reduce tissue autofluorescence and scattering, enabling higher signal-to-noise ratios and deeper intestinal imaging [138]. NIR-II emissive QDs and NMCs are particularly attractive because their band structures and surface ligands can be finely tuned [139,140]. Doping and bandgap engineering make it possible to match excitation and emission wavelengths precisely, and to integrate fluorescence, PA, and PT readouts within a single nanostructure [141,142]. When combined with enzyme-like ROS scavenging, such platforms can simultaneously report on inflammation severity and deliver active intervention. Nevertheless, the penetration depth of optical imaging remains inherently limited, and most NIR-II studies so far have been performed in mice or other small animals [143,144]. In human IBD, a thicker abdominal wall and deeper visceral tissues lead to marked signal attenuation even in the NIR-II range, and the clinical availability of NIR-II equipment is still limited [145]. As a result, most NIR-II studies related to colorectal disease currently focus on focal imaging and local navigation, rather than true whole-intestine imaging through the body surface [146]. Multimodal imaging offers a way to reconcile these complementary strengths. CT and MRI provide high-resolution structural information on bowel wall thickening, luminal narrowing, fistulas, abscesses, and perienteric fat stranding [133], and are already recommended in combination for disease staging, activity assessment, and preoperative planning [147]. In practical terms, imaging strategy design should be guided by the intended clinical question. When the core objective is lesion localization and clinically interpretable structural assessment, CT-based imaging routes are better aligned with current imaging practice for IBD [133]. For scenarios that prioritize sensitive readout of local inflammatory status or in situ formulation retention, optical/PA imaging approaches may be more informative [148]. When lesion localization and inflammatory activity need to be interpreted together, multimodal imaging integration becomes more justified. More broadly, imaging-oriented oral nanozymes will need to describe IBD in three dimensions. In time, readouts should follow how inflammation changes during treatment; in space, they should indicate where lesions are located and how far they extend; in terms of severity, signal intensity should be meaningfully related to accepted measures of disease activity and response. As these ideas are translated into material design, the focus will likely shift from simply generating brighter signals to building imaging-disease relationships that are stable, interpretable, and useful for clinical decision-making. Approaches to enhance catalytic performance For IBD, oxidative stress and inflammatory amplification are central features of disease progression. From a catalytic standpoint, nanozymes that can provide robust SOD-like and CAT-like activities under gastrointestinal conditions are ideal, because they directly target this core pathological loop without increasing the number of drugs. To translate this conceptual advantage into practical benefit, future platforms need to increase catalytic output per metal atom introduced, maintain activity in complex gastric and colonic environments, and match reaction kinetics to the timescale of inflammatory injury. Ceria-based nanozymes are a typical example in which catalytic performance and imaging capacity are both intimately linked to structure. Several studies in related fields have shown that reducing particle size [149], tuning exposed facets [150], and engineering oxygen-vacancy content can markedly enhance SOD-like and CAT-like activities [151]. Porous carbon frameworks and core-shell architectures such as Pd@CeO2 further illustrate how cooperative interactions between metal oxide and surrounding matrix can amplify apparent catalytic activity per unit Ce [[152], [153], [154]]. Transferring these concepts to oral IBD nanozymes would make it possible to enhance ROS regulation without increasing the Ce dose, thereby improving the safety profile. In addition to structural regulation, an emerging strategy is to introduce external-field responsiveness to enable on-demand activation of nanozyme catalysis. In particular, ultrasound stimulation has been employed to enhance catalytic reactions via sonodynamic effects, and can further reinforce enzyme-like activity through piezocatalytic pathways [155]. Representative examples have been reported in other disease contexts. For instance, hollow Co-CeO2@PEG nanospheres can amplify ROS generation under ultrasound through cavitation and band-structure modulation, while their multi-enzyme-like behavior participates in cascade reactions for synergistic tumor therapy [156]. In anti-inflammatory and tissue-repair settings, an ultrasound-triggered piezocatalytic MOF(Hf)-Pt nanozyme enables ultrasound-tunable enhancement of SOD-mimicking and CAT-mimicking activities, thereby achieving efficient ROS scavenging for osteoarthritis treatment [157]. A recent study in acute liver injury further supports this external-field-gated strategy, in which an ultrasound-responsive piezoelectric nanozyme heterojunction regulated ferroptosis-related oxidative damage through mechano-chemical coupling and strengthened catalytic output only under externally applied stimulation [158]. Other external stimuli, including NIR irradiation and alternating magnetic fields, have also been used to increase peroxidase-like or catalase-like outputs through localized photothermal or magnetothermal heating [159]. For example, Au nanorod@CeO2 hybrid nanozymes can potentiate peroxidase-associated antibacterial effects under NIR photothermal conditions [160]. Notably, although external-field-augmented nanozyme concepts have been extensively explored in tumors, infections, fibrosis, osteoarthritis, and related settings, they have rarely been systematically incorporated into nanozyme designs for IBD theranostics to date. Future development may leverage time-programmable stimulation, such as pulsed activation, to better align nanozyme therapy with fluctuations in lesion activity, and couple external activation with CT and NIR-II imaging readouts so that stimulation parameters can be dynamically adjusted according to lesion location and inflammatory burden. Beyond ceria-based systems, a number of nanozyme systems with strong antioxidant and immunoregulatory effects in inflammatory or intestinal models have emerged but have not yet been systematically explored for integrated imaging and therapy in IBD. H2S-responsive Cu5.4O nanozymes, for example, can scavenge both ROS and H2S and possess PT and PA properties, suggesting potential evolution into a multimodal therapeutic platform [161]. Nanozymes such as NiCo2O4@PVP [162], GSH-Se [163], and related compositions have also shown significant anti-inflammatory activity and barrier protection while avoiding rare-earth accumulation at the compositional level [164,165]. Taken together, these considerations indicate that future nanozyme design for IBD should develop along two practical routes. One is structure-based optimization, in which particle size, defect chemistry, and interfacial synergy are fine-tuned to maximize enzyme-like activity per metal site while limiting overall metal loading. The other is demand-responsive regulation, in which external-field or microenvironment-triggered strategies are used to better match catalytic output to fluctuating lesion activity. Together, these routes provide a more practical basis for improving therapeutic efficacy while facilitating subsequent optimization of safety and imaging performance. Approaches to improve metabolism, clearance, and long-term biosafety Safety considerations are crucial for oral platforms intended for repeated administration in a chronic disease [166]. The current design based on BaSO4 cores coated with ceria nanozymes already illustrates one way to decouple imaging and catalysis. This architecture improves the utilization efficiency of Ce-based catalytic centers and enables CT imaging windows that are not strictly tied to the catalytic metal dose. However, most systems still rely on rare-earth elements and their oxides as major catalytic components [167,168]. Long-term or high-dose oral use may lead to intestinal accumulation and associated safety risks [169]. Several studies have pointed out that rare-earth-based nanoparticles tend to show more complex issues of long-term retention [170], which calls for more conservative choices in dose design and size control [171]. From the viewpoint of balancing efficacy with safety, two directions are particularly important. First, as discussed above, catalytic efficiency per metal atom should be increased. Second, the catalytic core should be built from components with a long record of clinical or preclinical use. PB-based nanozymes, for example, have already shown reproducible multi-enzyme-like protective effects in UC and CD models [172,173]. Se-based nanozymes and related systems also demonstrate potent antioxidant and anti-inflammatory properties [[163], [164], [165]]. Integrating such comparatively safer catalytic systems with BaSO4 or conventional iodine-based contrast agents may thus provide CT-guided platforms for integrated diagnosis and therapy without introducing substantial new compositional risks. Future oral nanozyme design should also consider post-functional fate. For systems that require lesion-associated retention during the active phase, retention should not be the final objective [174]. A more desirable design is one that remains stable during gastric transit and local action, but subsequently undergoes degradation, deactivation, or conversion into more readily eliminated forms once the imaging and therapeutic window has passed [175]. Adjacent nanozyme literature supports this principle. PB nanoparticles have been reported to degrade under physiological conditions [176], and recent reviews of biodegradable enzyme-mimicking nanocatalysts have likewise emphasized degradation pathway and in vivo metabolic fate as integral design parameters [177]. Safety challenges are equally prominent in optical and NIR-II emissive systems. Many high-brightness NIR-II fluorophores contain heavy-metal components or exhibit prolonged in vivo retention [178], and their long-term pharmacokinetic and toxicological profiles remain incompletely defined [73]. This issue makes them difficult to apply to repeated oral follow-up in the near term. To mitigate these risks, an optimization direction is to move toward NIR-II platforms built primarily from light elements and structurally engineered emitters. Systems based on organic aggregation-induced emission (AIE) probes, C-dots, or polymer dots use light elements such as C, H, N, and O as the emissive backbone. By extending conjugated chains [179,180], designing donor-acceptor units [181,182], and tuning molecular packing [183], their emission can be shifted into the NIR-II region. This approach substantially reduces dependence on heavy metals while preserving, as far as possible, the advantages in imaging depth and signal-to-noise ratio. Beyond compositional risk reduction, the long-term safety of oral nanozymes and nanozyme-like imaging platforms should be established by systematically characterizing their in vivo fate after gastrointestinal administration. Key aspects include structural persistence, partial degradation, fecal elimination, hepatobiliary or renal clearance, uptake by the mononuclear phagocytic system, and unintended redistribution across the damaged intestinal barrier [11,184,185]. Among the core studies discussed in this review, Dex-CeO2 is one of the few examples reporting explicit clearance information, with 97.6% of the oral dose removed within 24 h [11]. By contrast, many other oral nanozyme or nanozyme-like systems report short-term tolerability, lesion retention, or organ histology, but do not quantify fecal versus urinary elimination or cumulative intestinal burden under repeated dosing [37,131]. They also rarely assess blood exposure after barrier damage or its potential hematological, hemolytic, inflammatory, or immune consequences. This gap is especially relevant to metal-containing systems, because prolonged lesion-associated residence, although favorable for imaging and local therapy, can also complicate interpretation of delayed clearance and residual metal burden [184,186]. Related ceria literature further supports this concern: nanoceria has shown size-dependent accumulation and prolonged tissue persistence under non-oral exposure conditions [167]. These observations do not directly resolve the oral IBD setting, but they indicate that long-term colonic exposure to metal-containing nanozymes cannot be inferred safely from short-term tolerance alone. A further limitation lies in model duration. Although a few studies have extended oral nanozyme evaluation to chronic TNBS- or DSS-based settings, the current evidence base is still dominated by short-duration murine models. This remains insufficient for a chronic, relapsing-remitting disease such as human IBD. Future studies should therefore incorporate longer follow-up, repeated-dose evaluation, and, where feasible, chronic or relapsing models, with eventual validation in more human-relevant systems such as large-animal models or patient-derived intestinal platforms [187,188]. Challenges and prospects for CT-NIR-II multimodal nanozymes Scheme 5 brings together the complementary strengths and major bottlenecks of current CT-guided and optical/PA-guided strategies, and links these challenges to corresponding optimization directions, including high-Z core-within-shell designs, reduced rare-earth-metal loading, safer catalytic centers, a shift toward light-element NIR-II emitters, and eventual integration with CT. Against this background, the central design task is to combine clinically reliable structural imaging, sensitive functional readouts of inflammation, and efficient localized catalytic therapy within a single oral platform. At the same time, metal loading and structural complexity need to be kept under control.Scheme 5Schematic summary of the current strategies, challenges, and possible solutions of oral nanozymes for visual theranostics. Schematic summary of the current strategies, challenges, and possible solutions of oral nanozymes for visual theranostics. At the imaging level, several multimodal examples, such as combining CeNP-PEG with CT enterography and [18F] DPA-714 PET/CT, or using oral MPN@CeOx to obtain CT/MRI signals, have shown that organ-scale structural information and molecular-scale functional information can be acquired within one system. However, integrated oral platforms that combine whole-intestine CT readouts with NIR-II functional signals on the same carrier remain rare, and representative demonstrations are still lacking under comparable dosing and imaging conditions [189]. Within oral theranostic design for IBD, CT-NIR-II imaging can offer a complementary division of labor that is not always matched by CT-MRI combinations. CT is well suited to providing anatomical information on bowel wall thickening, luminal narrowing, and extraintestinal complications, and can be used for whole-intestine staging. NIR-II imaging, within its accessible field of view, has the potential to sensitively detect local inflammatory activity, oxidative stress, or microvascular changes. In principle, an oral nanoplatform could adopt a high-Z imaging core with a good safety record (e.g., BaSO4 or iodine-based contrast agents) and incorporate light-element NIR-II emitters (e.g., organic AIE probes, C-dots, or polymer dots) in a structurally simple configuration, such as a single-layer core-shell architecture. Within such a construct, a highly efficient nanozyme module with minimized metal loading could be integrated, enabling structural imaging, functional readouts, and antioxidant intervention on a single carrier. To realize such CT-NIR-II multimodal nanozymes, several interrelated challenges must still be addressed. At the compositional level, dependence on heavy metals for both imaging and catalysis needs to be reduced as far as possible without sacrificing performance. At the structural level, multiple functions should be integrated into architectures that remain sufficiently simple to allow scalable synthesis and predictable in vivo behavior. At the translational level, regulatory and manufacturing constraints should be taken into account, favoring platforms built from components with established safety data and relatively well-defined pharmacokinetics. Against this backdrop, Scheme 5 outlines a pragmatic evolutionary route: starting from a decoupled CT core that already meets clinical imaging standards, progressively introducing structurally optimized nanozyme modules with lower metal loading. Once safety and pharmacokinetic profiles are better defined, light-element NIR-II emitters can then be integrated into this framework. In this way, the route balances ambition with feasibility and provides a materials-based framework for CT-NIR-II multimodal nanozymes that could support treat-to-target-oriented preclinical evaluation of IBD. Methodological considerations for imaging and enzyme-like activity From a methodological perspective, the comparability of imaging and enzyme-like activity measurements will directly determine whether different strategies can be meaningfully optimized against one another [190,191]. In imaging characterization, most studies on oral nanozymes or nanoprobes still focus on presenting representative in vivo fluorescence images or comparing average signal intensities between inflamed and control groups at a single time point. Only a few CT studies have systematically reported HU concentration calibration curves for the metal component and time-resolved pharmacokinetics of contrast agent retention. For optical, PA, and MRI modalities, more standardized quantitative metrics are needed. These may include normalization using reference phantoms and the use of ratiometric or time-resolved signals to reduce tissue background. Such quantitative readouts should then be correlated with clinical endpoints, including endoscopic scores, radiologic grading, histologic findings, and serum inflammatory markers. Recent IBD studies suggest that the immediate value of artificial-intelligence-assisted image analysis lies mainly in automated segmentation and standardized quantification, particularly in endoscopy, intestinal ultrasound, and radiomics-based cross-sectional imaging [192,193]. For oral nanozyme studies, this may be most useful for reducing reader dependence when quantifying inflamed segments, retained-mass-defined CT regions, or serial signal changes across time points [192]. Even with such tools, the biological meaning should not be assumed to be identical across modalities [21]. As discussed above, CT-derived measures are often more directly influenced by local retained mass and spatial coverage, whereas optical or PA signals may be more sensitive to lesion-associated activation and microenvironmental change [39,194]. Intestinal residence time should likewise be interpreted with caution, because it is shaped not only by mucosal injury but also by formulation properties such as surface chemistry and mucoadhesion [37,195]. When imaging data and clinically relevant endpoints are reported together under the same disease model and comparable dosing regimens, the incremental value of different imaging strategies for disease staging, response assessment, and complication prediction can be evaluated more reliably [196]. In the characterization of enzyme-like activity, current nanozyme studies still suffer from several issues, including catalyst concentrations that fall outside the linear initial-rate regime, insufficient sampling time points, and inconsistent conversion to standard enzyme units. Systematic studies have shown that nanozyme structure and dispersibility have a strong impact on apparent activity [197,198], leading to several-fold differences in calculated values [149,199,200]. Future kinetic studies should restrict catalyst and substrate concentrations to the initial-rate linear regime and use multiple sampling time points to obtain reliable slopes. The results should then be converted to standard units (e.g., U·mg−1) under harmonized assay conditions. This approach will facilitate comparison between related materials. It will also support quantitative analysis, from a structural-design perspective, of how metal usage per unit mass relates to catalytic efficiency. For multi-enzyme-like nanozymes, it will also be necessary to clarify the relative contributions of different pathways to the overall therapeutic effect to avoid situations in which a single assay index appears high but the net treatment benefit remains limited. In general, quantitative, standardized imaging readouts and rigorously designed enzyme-like activity assays are needed, so that data from different platforms can be compared on the same scale and used to guide rational optimization. Conclusions In summary, studies have shown that orally administered nanozymes for IBD can combine local ROS scavenging with in vivo imaging on a single platform, offering a promising tool for lesion-focused intervention and evaluation in preclinical models. Current studies can be summarized into single-modality imaging-guided strategies and multimodal imaging-guided strategies. Within single-modality strategies, CT-guided systems and optical/PA-guided systems represent the two main routes according to their dominant signal-generation mechanisms. For CT-guided strategies, two representative design routes have emerged. One uses integrated high-Z nanozymes based on oxides of Ce, Dy, or Bi, in which the same component serves as both CT contrast core and catalytic center. The other uses BaSO4 as a clinically established CT core and places catalytic shells such as CeO2 on the outside, thereby decoupling imaging and catalysis. The former route demonstrates feasibility but still requires relatively high doses of rare-earth elements and carries a risk of long-term accumulation. The latter route highlights the advantage of maintaining CT contrast while reducing the catalytic metal dose and increasing its utilization efficiency. For optical/PA strategies, systems based on C-dots, QDs, and NMCs use intrinsic fluorescence or NIR/NIR-II emission to achieve optical imaging, whereas strong absorbers such as PB and melanin-like materials provide deeper readouts through PA imaging. These platforms offer higher sensitivity for tracking changes in the inflammatory microenvironment over time, but they still need to be combined with CT or MRI for whole-intestine staging and quantitative assessment. Multimodal imaging-guided strategies, by contrast, integrate CT or MRI with NIR, PET, or PA signals on the same oral nanozyme platform, allowing structural indices such as wall thickening and luminal narrowing to be interpreted together with functional readouts of oxidative stress and inflammatory activity. This combined framework is more conceptually aligned with imaging endpoints relevant to treat-to-target assessment. Looking ahead, the design of oral nanozymes for visual theranostics in IBD should focus on reducing rare-earth and heavy-metal exposure at both the compositional and structural levels and on increasing catalytic efficiency per metal atom. In practical terms, this may involve moving toward catalytic systems with more favorable safety profiles, such as PB-based nanozymes, Se-based nanozymes and related selenium-containing systems, while in optical modules greater emphasis may be placed on light-element emitters such as C-dots, organic AIE probes, and polymer dots. It should also prioritize nanozyme systems that provide multi-enzyme-like activities within a favorable safety window. At the same time, more standardized protocols are needed for enzyme-like activity assays and quantitative imaging evaluation. Imaging readouts should be actively correlated with histological healing and clinical scores, rather than being reported in isolation. These efforts will help move oral nanozyme platforms from concept-level demonstrations toward more interpretable preclinical platforms for IBD. Ethics approval and consent to participate This is a review article. No animal experiments or clinical studies were performed. Ethical approval and patient consent were not required. CRediT authorship contribution statement Shangyong Zuo: Conceptualization, Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review & editing. Kefan Li: Data curation, Investigation, Methodology, Writing – original draft, Writing – review & editing. Liang Feng: Validation, Writing – review & editing. Xingyan Yu: Conceptualization, Funding acquisition. Yifan Wang: Supervision, Writing – review & editing. Chengwei Wu: Funding acquisition, Writing – review & editing. Wenjian Weng: Funding acquisition, Supervision. Yaoting He: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. Xiujun Cai: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. Kui Cheng: Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing. 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.