Biomimetic Compartmentalization of Enzymes for Sustainable Coenzyme Recycling in Oxidative Photocatalysis

Abstract

Coenzyme regeneration is essential for oxidoreductase‐based biocatalysis, yet integrating enzymatic and photocatalytic systems remains challenging due to ROS‐mediated enzyme inactivation. Here, we developed a metabolic‐inspired compartmentalization strategy that enables sustainable NAD⁺/NADH cycling by mimicking the spatial organization of cellular redox metabolism. Glucose dehydrogenase (GDH) was encapsulated within silica nanoparticles, exploiting the intrinsic instability of hydroxyl radicals and their limited diffusion distance to shield the enzyme from TiO₂‐generated ROS while preserving substrate and cofactor accessibility. The encapsulated GDH maintained catalytic competence and achieved five consecutive coenzyme regeneration cycles, whereas the free enzyme lost activity after a single photocatalytic exposure. Beyond cofactor recycling, NADH generated by this system functioned as an endogenous antioxidant, protecting cells from photocatalytic oxidative stress. This work demonstrates that principles from cellular biochemistry, including compartmentalization and redox homeostasis, can be translated into artificial systems for sustainable biocatalysis and cell protection.

This study presents a novel strategy inspired by the spatial compartmentalization of cellular metabolic pathways to address the challenge of coenzyme regeneration in enzyme–photocatalyst hybrid systems. Glucose dehydrogenase was encapsulated within silica nanoparticles, exploiting the short half‐life and limited diffusion distance of hydroxyl radicals to protect the enzymes against photocatalytic oxidation stress. The encapsulated enzyme maintained catalytic activity and achieved continuous NAD⁺/NADH regeneration cycles.

1. Introduction

Oxidoreductases represent a vital class of enzymes with extensive applications in industrial biocatalysis, pharmaceutical synthesis, and biomedical research [1, 2, 3]. These enzymes catalyze a wide range of redox reactions with exceptional selectivity and efficiency under mild conditions. However, most oxidoreductases require nicotinamide cofactors such as NAD(P)⁺ and NAD(P)H as essential electron carriers to mediate hydride transfer during catalysis [4]. The high cost of these cofactors renders their stoichiometric consumption economically prohibitive for large‐scale applications. Consequently, efficient cofactor regeneration systems are indispensable for the practical implementation of oxidoreductase‐based biocatalysis [5].

Several strategies have been developed for cofactor regeneration, including chemical, electrochemical, photocatalytic, and enzymatic approaches [6, 7, 8, 9, 10, 11, 12, 13, 14, 15]. Among these, photocatalytic cofactor regeneration has emerged as a sustainable strategy for driving enzymatic redox reactions using light energy. Various photocatalytic materials have been developed for NAD(P)H regeneration, including TiO₂, graphitic carbon nitride, CdS quantum dots, metal‐organic frameworks, covalent organic frameworks, and carbon dots [16, 17, 18]. To achieve selective regeneration of enzymatically active 1,4‐NADH, electron mediators such as rhodium complexes are commonly employed to convert single‐electron photochemical processes into selective hydride transfer [19]. Recent advances have demonstrated efficient coupling of photocatalytic NADH regeneration with enzymatic reactions for CO₂ reduction to formic acid, asymmetric ketone reduction, and amino acid synthesis [20, 21]. Strategies to improve photocatalytic efficiency have focused on enhancing visible light absorption, promoting charge separation through heterojunction architectures, and optimizing surface reactions [16]. However, a critical limitation in these photocatalytic‐enzymatic coupling systems is the generation of reactive oxygen species (ROS) during photocatalysis, which causes rapid enzyme deactivation and undermines long‐term system stability.

To address this enzyme‐photocatalyst incompatibility, several protective strategies have been explored. One approach involves encapsulating the photocatalyst within protective shells to delay ROS release into the bulk solution [22]. However, this strategy often compromises cofactor regeneration efficiency due to impeded mass transfer across the encapsulation barrier. Alternatively, co‐encapsulation of target enzymes with antioxidant enzymes such as catalase or superoxide dismutase has been investigated [23]. While this approach provides some degree of protection, it introduces additional system complexity and increases costs, and the antioxidant enzymes themselves remain susceptible to ROS‐mediated inactivation over prolonged operation. More recently, thylakoid membrane‐inspired capsules have been developed to physically separate photocatalytic and enzymatic centers using porous TiO₂ membranes [24], and self‐assembly approaches integrating electron mediators onto photocatalyst surfaces have shown promise in suppressing enzyme deactivation [25]. Core‐shell architectures employing mesoporous silica shells have also been designed to trap and deactivate ROS within the shell layer before they reach the enzymatic material [26]. Despite these advances, most strategies require complex fabrication processes or additional components, and achieving sustained cofactor cycling while preserving enzyme integrity remains a significant challenge.

Interestingly, living cells face a similar challenge. ROS are continuously generated as byproducts of aerobic metabolism, particularly within mitochondria during oxidative phosphorylation. To protect ROS‐sensitive enzymes and maintain cellular redox homeostasis, cells employ spatial compartmentalization strategies, sequestering ROS‐generating processes within membrane‐bound organelles while localizing vulnerable enzymes in separate compartments. For instance, peroxisomes contain both ROS‐generating oxidases and protective catalase within the same organelle, whereas the mitochondrial inner membrane physically separates the electron transport chain from cytosolic enzymes. Inspired by these natural compartmentalization strategies, we hypothesized that confining the enzyme within a protective matrix could mimic the protective role of organelle membranes in biological systems.

A key insight enabling our approach lies in the intrinsic physicochemical properties of ROS. Hydroxyl radicals possess an extremely short half‐life of ≈10⁻⁹ seconds in aqueous solution and can diffuse only a few nanometers before reacting with surrounding molecules. Superoxide anions, though slightly longer‐lived, also exhibit limited diffusion distances at physiological pH. Based on these characteristics, we reasoned that encapsulating only the target enzyme within a protective matrix of appropriate thickness would be sufficient to shield it from photocatalyst‐generated ROS, without requiring additional antioxidant components or compromising photocatalytic efficiency. Unlike previous approaches that encapsulate photocatalysts to limit ROS diffusion or employ auxiliary antioxidant enzymes for protection, our strategy exploits the inherent instability of ROS as the primary protective mechanism.

Herein, we present a compartmentalized enzyme‐photocatalyst system for sustainable coenzyme regeneration (Figure 1). Glucose dehydrogenase (GDH) was encapsulated within silica nanoparticles (SiNPs) via a sol–gel process, creating a physical barrier that attenuates ROS diffusion while permitting free passage of substrates and cofactors (Figure 1c). This GDH@SiNPs system was coupled with TiO₂ nanoparticles to establish a complete NAD⁺/NADH recycling platform driven by UV irradiation. We demonstrate that this biomimetic compartmentalization strategy, inspired by the spatial organization of metabolic pathways in living cells, enables repeated cofactor cycling without enzyme inactivation. Furthermore, we explore the potential of this system for cytoprotective applications, wherein the continuous generation of NADH serves as an exogenous ROS scavenger to protect living cells from photocatalytic oxidative stress.

FIGURE 1.

The compartmentalized enzyme‐photocatalyst system for sustainable coenzyme regeneration. (a) The glucose dehydrogenase‐loaded silica nanoparticles (GDH@SiNPs) and TiO₂ nanoparticle system complete the NAD⁺/NADH cycle, enabling continuous coenzyme regeneration while preventing enzyme inactivation by reactive oxygen species generated during photocatalysis. (b) Bare GDH is easily deactivated by the reaction of TiO₂ nanoparticles. (c) The silica matrix quenched ROS before reaching the enzyme core due to their short half‐life in an aqueous environment, effectively shielding GDH from oxidative inactivation.

2. Results and Discussion

2.1. Design of Key Idea in This Study

To develop a sustainable NAD⁺/NADH recycling system, we encapsulated GDH into silica nanoparticles, creating a protective barrier that exploits the short diffusion distance and high reactivity of ROS such as hydroxyl radical (·OH) and superoxide (·O₂ ⁻). As ROS travel through the silica matrix, they are quenched before reaching the enzyme core due to their short half‐life in an aqueous environment, effectively shielding GDH from oxidative inactivation while still allowing the diffusion of substrates and cofactors. This compartmentalization strategy enables the simultaneous operation of both enzymatic and photocatalytic reactions in a single system, achieving continuous coenzyme cycling while preserving enzyme activity.

2.2. Synthesis and Characterization of GDH@SiNPs

GDH‐encapsulated silica nanoparticles were synthesized via a sol–gel process using tetraethyl orthosilicate (TEOS) and (3‐aminopropyl)‐trimethoxysilane (APTMS) as precursors, following our previously established protocol [27]. During the condensation process, the carboxyl groups of GDH formed covalent amide bonds with the amine groups of APTMS, resulting in the stable immobilization of the enzyme within the silica matrix. The resulting nanoparticles exhibited an amorphous and non‐porous structure, which limits the diffusion of molecules through the silica matrix, a key feature for attenuating ROS penetration toward the encapsulated enzyme.

Transmission electron microscopy (TEM) revealed that the synthesized nanoparticles possessed a spherical morphology with diameters ranging from 50 to 120 nm (Figure 2b). Dynamic light scattering (DLS) analysis indicated a hydrodynamic diameter of 139 nm (peak intensity) or 165 nm (z‐average), with a polydispersity index (PDI) of 0.142, confirming a narrow and uniform size distribution (Figure 2c). The zeta potential was measured at −11.2 ± 2.5 mV, indicating a slightly negative surface charge. This moderate negative charge provides sufficient electrostatic repulsion to prevent particle aggregation in aqueous suspension, thereby ensuring colloidal stability. Furthermore, the nanoscale dimensions facilitate adequate dispersion in the reaction medium while providing sufficient silica thickness to serve as a protective barrier against external ROS. BET analysis of the silica nanoparticles revealed a surface area of ≈ 60 m²/g, indicating the presence of heterogeneously distributed channels within the silica matrix. These channels enable the diffusion of small molecules such as substrates (glucose, ∼180 Da) and cofactors (NAD⁺/NADH, ∼663 Da) while retaining the encapsulated enzyme (GDH, ∼50 kDa).

FIGURE 2.

Preparation and characterization of GDH@SiNPs. (a) Fabrication scheme of GDH@SiNPs. (b) TEM micrograph of GDH@SiNPs. (c) DLS measurement for size distribution of GDH@SiNPs. (d) Michaelis–Menten kinetic of GDH@SiNPs and bare GDH (n = 3).

The encapsulation efficiency of GDH within silica nanoparticles was determined by dissolving the nanoparticles in ammonium bifluoride and quantifying the released protein using the bicinchoninic acid (BCA) assay. The encapsulation efficiency was calculated to be 86 ± 11%, indicating that the sol–gel process effectively incorporated the majority of the enzyme into the silica matrix. This high encapsulation efficiency ensures sufficient catalytic capacity within the nanoparticles while minimizing enzyme loss during the synthesis process.

To determine the catalytic properties and to distinguish intrinsic enzyme damage from mass‐transfer limitations imposed by the silica shell, we compared the Michaelis–Menten kinetics of encapsulated GDH with native, free GDH under identical buffer conditions. Native GDH exhibited Km = 25.5 ± 1.7 mM, kcat = 209 ± 16 s⁻¹, and kcat/Km = 8.2 ± 0.8 mM⁻¹ s⁻¹, whereas encapsulated GDH showed Km = 20.1 ± 2.5 mM, kcat = 52 ± 1.2 s⁻¹, and kcat/Km = 2.6 ± 0.3 mM⁻¹ s⁻¹, retaining ≈ 31.7% of the native catalytic efficiency (Figure 2d). The lower Km value indicates the enhanced substrate affinity upon encapsulation, while the reduced kcat reflects the decreased turnover capacity. This kinetic profile is consistent with previous observations of enzymes in confined nanoscale environments [27], where spatial restriction enhances substrate binding but limits product release due to diffusional constraints within the matrix. The catalytic efficiency (kcat/Km) provides an overall measure of enzyme performance, and the obtained value confirms that the encapsulated GDH maintains functional biocatalytic activity suitable for cofactor regeneration applications. Notably, while some reduction in kinetic parameters compared to free enzyme occurs due to mass transfer limitations within the silica shell, the encapsulated system offers the critical advantage of enzyme protection in harsh oxidative environments. These results demonstrate that our encapsulation strategy successfully preserves the catalytic function of GDH while enabling its integration into ROS‐generating photocatalytic systems.

2.3. ROS‐Mediated Enzyme Inactivation and Protective Effect of Silica Encapsulation

Having demonstrated the protective effect of silica matrix, we first examined the stability of GDH in the presence of photoactivated TiO₂ nanoparticles (Figure 3a). Basic characteristics of TiO₂ nanoparticles are described in Figure S1 and Table S1. Upon UV irradiation (365 nm), TiO₂ generates ROS, primarily hydroxyl radicals (·OH) and superoxide anions (·O₂ ⁻), which are responsible for oxidizing NADH to NAD⁺. However, these ROS species are also highly damaging to proteins, attacking amino acid residues and disrupting enzyme structure. When bare GDH was incubated with TiO₂ nanoparticles under UV illumination, the enzyme lost nearly all catalytic activity within 10 min, retaining less than 10% of its initial activity (Figure 3b). In contrast, GDH encapsulated within silica nanoparticles (GDH@SiNPs) maintained over 50% of its activity even after 20 min of UV exposure under identical conditions.

FIGURE 3.

Oxidative stress resistance, and recycling of NAD⁺ and NADH by alternating addition of glucose in the dark and irradiation with UV light. (a) NAD⁺ and NADH conversion in this study. (b) Monitoring of GDH activity in the presence of ROS (generated by TiO₂ nanoparticles under UV light irradiation) (n = 3). (c) Recycling of NAD⁺ and NADH using GDH@SiNPs and TiO₂ nanoparticles (n = 3). (d) Recycling of NAD⁺ and NADH using native GDH and TiO₂ nanoparticles. For c and d, an increase in the y‐axis value indicates NADH accumulation by GDH‐catalyzed reduction, while a decrease reflects NAD⁺ regeneration by TiO₂‐mediated oxidation (n = 3).

This dramatic difference in enzyme stability can be attributed to the intrinsic properties of ROS and the protective architecture of the silica shell. Hydroxyl radicals possess an extremely short half‐life of ≈10⁻⁹ seconds (1 ns) in aqueous solution and react with organic molecules at near diffusion‐controlled rates (10⁹ – 10¹⁰ M⁻¹ s⁻¹), limiting their effective diffusion distance to only a few nanometers, typically less than 50 molecular diameters from the site of formation [28, 29, 30]. Superoxide anions, while somewhat longer‐lived with half‐lives ranging from microseconds to minutes depending on pH and solution conditions, also undergo rapid spontaneous dismutation at neutral pH (∼10⁵M⁻¹s⁻¹ at pH 7) [31]. Given that the silica nanoparticles exhibit a shell thickness of ≈25–60 nm (based on TEM observations of 50–120 nm diameter particles with the enzyme at the core), ROS generated at the TiO₂ surface must traverse this substantial distance to reach the encapsulated enzyme. This shell thickness far exceeds the typical diffusion range of hydroxyl radicals (2–6 nm), providing ample opportunity for ROS to be quenched before reaching the enzyme core. The amorphous, non‐porous nature of the silica matrix further impedes ROS diffusion, providing ample opportunity for these reactive species to be quenched through reactions with the silica network or surrounding water molecules before reaching the enzyme core.

To further evaluate whether enzyme leaching occurs during photocatalytic reactions, the supernatant of GDH@SiNPs and TiO₂ mixture after 1 h of UV irradiation was collected, concentrated, and analyzed for protein content using both Bradford and BCA assays, respectively. No detectable enzyme was found in the supernatant, confirming that GDH remains stably encapsulated within the silica matrix throughout the photocatalytic process.

2.4. Continuous NAD⁺/NADH Recycling Through Cyclic Operation

Building upon the demonstrated protective effect, we next evaluated the feasibility of continuous coenzyme recycling using the GDH@SiNPs and TiO₂ nanoparticle system. Upon UV irradiation, TiO₂ generates reactive oxygen species through well‐established photocatalytic mechanisms, primarily converting water molecules to hydroxyl radicals and dissolved oxygen to superoxide, with minor production of H₂O₂ and singlet oxygen [32]. These ROS, particularly hydroxyl radicals with their high oxidation potential (2.8 V vs NHE), are responsible for the oxidation of NADH back to NAD⁺. The recycling experiment was designed as a cyclic process consisting of two phases, which are an enzymatic reduction phase and a photocatalytic oxidation phase. In each cycle, GDH@SiNPs and TiO₂ nanoparticles were introduced into a reaction solution containing NAD⁺, followed by the addition of glucose to initiate the enzymatic conversion of NAD⁺ to NADH over 50 min. Subsequently, UV light was applied for 10 min to drive the TiO₂‐mediated oxidation of NADH back to NAD⁺, completing one full cycle.

The encapsulated GDH system successfully sustained coenzyme recycling over five consecutive cycles, demonstrating remarkable operational stability (Figure 3c). To provide a quantitative assessment as suggested, we present the cycle‐by‐cycle data. In the first cycle, 0.241 mM NADH was generated by GDH@SiNPs, and 0.165 mM NAD⁺ was regenerated by TiO₂. In the second, third, and fourth cycles, NADH production was 0.281, 0.314, and 0.307 mM, respectively, while NAD⁺ regeneration was 0.241, 0.237, and 0.251 mM, respectively. In contrast, the bare GDH system failed to proceed beyond the first cycle, as the enzyme was irreversibly inactivated during the initial UV exposure phase (Figure 3d). This result unequivocally demonstrates that silica encapsulation enables the repeated use of GDH in an otherwise incompatible oxidative environment. The ability to perform multiple cycles has profound implications for practical biocatalytic applications, as it directly translates to reduced cofactor consumption and improved process economics. Furthermore, this cyclic operation mimics the continuous cofactor turnover observed in cellular metabolic pathways, where NAD⁺ and NADH are perpetually interconverted to sustain redox reactions. By achieving five complete cycles without significant loss of enzymatic function, our encapsulation system represents a significant step toward establishing metabolic‐inspired, self‐regenerating cofactor systems for sustainable biocatalysis. We additionally measured the morphology of GDH@SiNPs after five cycling reactions. No apparent change was observed on TEM image (Figure S2).

2.5. Cytoprotective Application of the GDH@SiNPs/TiO₂ Systems

To demonstrate the potential applicability of our coenzyme regeneration system beyond in vitro biochemical reactions, we evaluated its capacity to protect living cells from TiO₂‐induced oxidative damage. TiO₂ nanoparticles are widely employed in photocatalytic applications, including antimicrobial treatments and environmental remediation; however, the ROS generated upon UV irradiation pose significant cytotoxic risks to mammalian cells. We hypothesized that GDH@SiNPs could mitigate this oxidative stress by continuously supplying NADH, which serves as a direct ROS scavenger (Figure 4a). Upon reaction with ROS, NADH is oxidized back to NAD⁺, which can then reenter the GDH‐catalyzed reaction cycle, thereby establishing a self‐sustaining antioxidant system.

FIGURE 4.

Cell protection of GDH@SiNPs against TiO2‐triggered ROS stress. (a) GDH@SiNPs can mitigate oxidative stress by continuously supplying NADH, which serves as a direct ROS scavenger. (b) Cell viability measured by CCK‐8 assay of HeLa cells in the presence of GDH@SiNPs and TiO2 nanoparticles under UV light irradiation. Effective cell protection of the GDH@SiNPs was achieved by enzymatic NADH production (n = 4). Data are presented as mean ± SD (n = 4). Statistical significance was determined by one‐way ANOVA followed by Tukey's post‐hoc test. Significant differences are indicated as **(p < 0.01), and ***(p < 0.001).

Cells were co‐incubated with GDH@SiNPs (20 μg/mL), TiO₂ nanoparticles(10 μg/mL), and NADH (10 mM) in glucose‐supplemented medium and subsequently exposed to UV irradiation (365 nm) for 20 min. The concentration of GDH@SiNPs used was confirmed to be nontoxic to HeLa cells (Figure S3). Cell viability assays revealed that the GDH@SiNPs/TiO₂ system afforded substantial cytoprotection, with ≈87% of cells remaining viable after UV exposure (Figure 4b). In contrast, cells treated with TiO₂ nanoparticles alone exhibited severe oxidative damage, with viability dropping to ≈58%. Notably, UV irradiation at 365 nm in the absence of TiO₂ did not cause significant direct cellular damage, consistent with the understanding that UVA primarily induces cytotoxicity through ROS‐mediated mechanisms rather than direct DNA photodamage. These results demonstrate that GDH@SiNPs effectively intercepts TiO₂‐generated ROS before they can inflict lethal oxidative injury to cells.

To exclude nonspecific protective effects from the silica material itself, such as ROS adsorption, light path obstruction, or UV scattering, control experiments were performed using empty SiNPs without an encapsulated enzyme. Empty SiNPs failed to provide cytoprotection under identical TiO₂ and UV irradiation conditions, confirming that the protective effect is dependent on GDH‐mediated NADH generation rather than physical or chemical shielding by the nanoparticle matrix (Figure S4).

The cytoprotective efficacy of this system holds considerable promise for several applications. In photodynamic therapy and photocatalytic cancer treatment, TiO₂‐based systems can selectively destroy tumor cells through localized ROS generation. However, collateral damage to healthy tissues remains a critical concern. The integration of compartmentalized GDH could enable spatial protection of non‐target cells while maintaining therapeutic ROS levels at malignant sites. Similarly, in whole‐cell biotransformation processes where photocatalytic cofactor regeneration is desirable [33], our system provides a solution by decoupling beneficial oxidative chemistry from its cytotoxic consequences. The self‐regenerating nature of the NADH/NAD⁺ cycle ensures sustained protection without continuous external antioxidant supplementation, offering a biocatalytically autonomous strategy for managing many types of oxidative stresses in complex biological environments [34, 35, 36, 37].

The protective mechanism demonstrated in this study bears a resemblance to compartmentalization strategies employed by living cells. In eukaryotic cells, ROS generated at the mitochondrial electron transport chain are confined within membrane‐bound organelles, protecting cytosolic enzymes from oxidative damage. The mitochondrial inner membrane serves as a physical barrier that limits ROS diffusion while permitting metabolite transport through specific carriers.

Our silica encapsulation strategy functionally mimics this biological architecture. The silica shell (25–60 nm thickness) serves an analogous role to organelle membranes, exploiting the short half‐life and limited diffusion distance of hydroxyl radicals to attenuate ROS before they reach the encapsulated enzyme. Meanwhile, the porous nature of the shell permits free passage of substrates and cofactors, analogous to the selective permeability of biological membranes. Furthermore, the continuous NADH generation parallels the role of reduced cofactors in cellular antioxidant defense, where NADPH maintains the pool of reduced antioxidants. This functional analogy between our artificial system and natural compartmentalization validates the biomimetic design principle and suggests potential for further development toward more sophisticated, adaptive protection mechanisms.

3. Conclusion

In this study, we established a metabolic‐inspired coenzyme recycling system that mimics the continuous NAD⁺/NADH turnover observed in cellular redox metabolism. By encapsulating glucose dehydrogenase within silica nanoparticles, we created an artificial compartmentalization strategy analogous to organelle‐based spatial segregation in living cells, effectively shielding the enzyme from ROS‐mediated oxidative inactivation. This biomimetic architecture exploits the inherent instability of reactive oxygen species, particularly the nanosecond‐scale half‐life and nanometer‐range diffusion distance of hydroxyl radicals, to achieve selective protection while preserving enzyme‐substrate accessibility. The encapsulated GDH maintained its catalytic competence for coenzyme reduction and sustained multiple regeneration cycles, demonstrating that cofactor recycling can be achieved in oxidatively challenging environments without compromising enzyme integrity. Moreover, the NADH generated by this system served as an endogenous ROS scavenger, providing cell protection against photocatalytic oxidative stress and highlighting the dual functionality of reduced cofactors in both biocatalysis and antioxidant defense. This work demonstrates that principles derived from cellular biochemistry, including compartmentalization, cofactor cycling, and redox homeostasis, can be successfully translated into artificial systems, offering a biologically informed strategy for sustainable biocatalysis and oxidative stress management. Future studies should address several aspects to further advance this system. Direct measurement of ROS diffusion and attenuation within the silica matrix would provide a deeper mechanistic understanding of the protective effect. Systematic optimization of silica shell thickness is needed to establish the relationship between shell dimensions and protective efficacy. Additionally, extending this compartmentalization strategy to other oxidoreductases would demonstrate the generalizability of this approach for broader biocatalytic applications.

4. Materials and Methods

4.1. Materials

NAD⁺‐dependent glucose dehydrogenase (GDH) from Pseudomonas sp. (EC 1.1.1.47; 550 unit/mg), D‐glucose, N‐hydroxysuccinimide (NHS), tetraethyl orthosilicate (TEOS), and (3‐Aminopropyl)trimethoxysilane (APTMS) were purchased from Merck (Darmstadt, Germany). Titanium dioxide (TiO2) nanoparticles were purchased from EVONIK (Essen, Germany). Nicotinamide adenine dinucleotide reduced form (NADH), nicotinamide adenine dinucleotide oxidized form (NAD⁺), N‐(3‐dimethylaminopropyl)‐N′‐ethylcarbodiimide hydrochloride (EDC), and n‐hexanol were obtained from Carl Roth (Karlsruhe, Germany). A UV hand lamp (6W) was purchased from VILBER LOURMAT (Collegien, France). Cyclohexane, Triton X‐100, and n‐hexanol were purchased from Daejung Chemicals (Korea, Republic of).

4.2. Preparation of GDH@SiNPs

The enzyme‐loaded silica nanoparticles were prepared following our previous report [18]. The mixture of cyclohexane (2 mL), 1‐hexanol (0.5 mL), Triton X‐100 (0.45 mL), TEOS (58.5 μL) was prepared as an organic phase. The mixture of NHS (2 mg) and EDC (2 mg), APTMS (7.5 μL), and 1.6 mg of GDH in 0.3 mL of sodium phosphate buffer (3 mM, pH 7.4) was prepared as an aqueous phase. To prepare a water‐in‐oil (w/o) microemulsion, aqueous phase (0.3 mL) was added to the organic phase (≈3 mL) under magnetic stirring at 500 rpm and room temperature. The coupling reaction between GDH and APTMS was allowed to proceed for an hour. Then, 60 μL of a 25% ammonia solution was added to the resultant emulsions. After reaction for 24 h under stirring, the resulting reactant was precipitated by the addition of 10 mL of ethanol and centrifuged at 11,000 rpm for 15 min. The pellet was redispersed in 1 mL of deionized water.

4.3. Enzymatic Assay for GDH

Each stock solution was prepared containing NAD⁺ at 60 mM and glucose at 2 M. To determine the Michaelis‐Menten kinetics of GDH, reactions were carried out in a total volume of 100 μL. Each reaction contained 90 μL of phosphate buffer (50 mM, pH 7.4) and 0.8 μL of NAD⁺. The enzyme was introduced as either free GDH (0.16 mg, equivalent to 0.009 unit) or silica nanoparticle‐encapsulated GDH (GDH@SiNPs, 16 mg) in a volume of 2 μL. Reactions were started by adding 10 μL of glucose at concentrations ranging from 0 to 400 mM. Absorbance changes at 340 nm were recorded every 15 s using a TECAN Infinite M Plex plate reader (Männedorf, Switzerland).

4.4. ROS‐Induced Oxidative Stress Resistance Test for GDH

Each stock solution was prepared containing NAD⁺ (60 mM), glucose (120 mM), phosphate buffer (50 mM, pH 7.4), TiO₂ nanoparticles (0.2 mg/mL), GDH@SiNPs (16 mg/mL), and native GDH (1.6 mg/mL), respectively. In a 1 mL glass vial, TiO₂ nanoparticles (10 μL) and GDH samples (40 μL) were combined with phosphate buffer (50 mM, pH 7.4) to reach a final volume of 1000 μL. The GDH samples used were either native GDH or GDH@SiNPs. The mixture was exposed to UV light, and 98 μL aliquots were collected at 30‐minute intervals. Each aliquot was then mixed with NAD⁺ (2 μL) and glucose (2 μL) to start the GDH reaction, and absorbance changes at 340 nm were recorded. The illumination setup had an irradiated area of ≈6 cm² and a distance of 4 cm from the light source.

4.5. NAD⁺/NADH Recycling

Each stock solution was prepared containing NAD⁺ (60 mM), glucose (120 mM), phosphate buffer (50 mM, pH 7.4), TiO₂ nanoparticles (0.2 mg/mL), GDH@SiNPs (16 mg/mL), and Lutensol AT50, respectively. In a 1 mL glass vial, NAD⁺ (2.6 μL), TiO₂ nanoparticles (5 μL), GDH@SiNPs (50 μL), Lutensol AT‐50 (15 μL), and phosphate buffer (527 μL) were mixed together. The GDH reaction was started by adding glucose (1.3 μL), and the mixture was incubated in the dark. The total reaction volume was 600 μL, with NAD⁺ and glucose each at a final concentration of 0.25 mM. After 50 min, absorbance at 340 nm was measured. The vial was then exposed to UV light for 10 min to activate the TiO₂ nanoparticles, followed by another absorbance measurement to complete one cycle. This process was repeated to obtain five consecutive cycles in total.

4.6. Cell Protection Assay

HeLa cells (Research Resource Identifier number: CVCL_0030) were obtained from Korea Cell Line Bank (KCLB No: 10 002, Seoul, Korea) and cultured in DMEM medium (Welgene, Gyeongsan, Korea) supplemented with 10% fetal bovine serum (FBS; qualified grade, Gibco, Thermo Fisher Scientific, Waltham, MA, USA). Cells were maintained at 37°C in a humidified atmosphere containing 5% CO₂ with penicillin and streptomycin. Cells were passaged at ≈70–90% confluency using 0.25x trypsin‐EDTA and used for experiments within 20 passages.

For the cell protection assay, cells were seeded in a 96‐well plate and incubated overnight. The cells were washed twice with DPBS and then treated with serum‐ and glucose‐free medium (100 μL) containing glucose (50 mM), NADH (10 mM), GDH@SiNPs (20 μg/mL), and/or TiO₂ nanoparticles (10 μg/mL). The cells were incubated under UV irradiation for 20 min, followed by three washes with DPBS. Cell viability was evaluated using the Cell Counting Kit‐8 assay.

Funding

This study was supported by the Pusan National University (10.13039/501100002543)

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplementary Material

Jung Myung Kyung, Park So‐Hee, You Dong‐Hyun, Lee Ji‐Soo, Ryu Da‐Hye, Hwang Dae Youn, An Beum‐Soo, Yang Seung Yun, Seo Sungbaek, Choi Youngwoo, Jo Seong‐Min, Biomimetic Compartmentalization of Enzymes for Sustainable Coenzyme Recycling in Oxidative Photocatalysis, ChemBioChem 2026, 27, e70381. 10.1002/cbic.70381

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.