Oppositely Charged Single Enzyme Nanogels Form Versatile Coacervates for Efficient Enzyme Cascade Catalysis

ABSTRACT

Enzymatic cascade systems enhance the efficiency of biocatalysis by mimicking natural metabolic pathways. However, conventional platforms face challenges such as ineffective enzyme distribution, uncontrolled stoichiometry, limited substrate accessibility, and a microenvironment that is suboptimal for enzyme performance. Inspired by cellular biomolecular condensates, we here present a design based on oppositely charged single enzyme nanogels (SENs) that phase separate into denser microdroplet coacervates, which thereby serve both as an immobilization scaffold and functional unit. These coacervates, the first fabricated exclusively from oppositely charged nanogels, form under broad environmental conditions and enable specific microenvironments by adjusting SEN ratios to match the requirements of coupled enzymes. Selective substrate enrichment and efficient diffusion of intermediates enhance cascade productivity. The applicability of the platform is demonstrated with pharmaceutically relevant enzymes, highlighting its potential for biosynthetic applications. This work establishes a new paradigm that integrates protein‐polymer engineering with phase separation chemistry to overcome key limitations in cascade enzyme catalysis.

Oppositely charged single enzyme nanogels (SENs) phase‐separate into bi‐enzymatic coacervate microdroplets, acting as both scaffold and functional units. By tuning SEN ratios, these coacervates create specific microenvironments that enable selective small‐molecule enrichment and efficient intermediate diffusion. Demonstrated with pharmaceutically relevant enzymes, this first nanogel‐only coacervate platform establishes a design principle for integrating protein–polymer engineering with phase separation‐driven catalysis.

1. Introduction

Inspired by natural metabolic pathways, enzyme cascades offer a powerful strategy to enhance the efficiency and sustainability of complex synthesis routes [1]. Key factors such as enzyme spatial co‐localization, optimized enzyme stoichiometry, and rapid diffusion of intermediates critically influence catalytic performance [2]. Beyond enzyme organization, the selective substrate recruitment and local enrichment within the catalytic microenvironment are equally important and yet often overlooked [3]. Inspired by cellular compartmentalization, diverse multi‐enzyme platforms have been engineered to achieve high local enzyme concentrations, including polymeric particles [4, 5], metal‐organic frameworks (MOFs) [6], protein‐inorganic nanoflowers [7], combined cross‐linked enzyme aggregates (combi‐CLEAs) [8, 9], and DNA and protein scaffolds [10, 11]. However, due to their solid state, these particles often face significant challenges, including steric hindrance limiting substrate access and altering protein conformation, restricted tunability of the microenvironment, and stochastic enzyme distribution [12]. For instance, MOF pores are often too small for large enzymes or impede the diffusion of bulky substrates and intermediates [13], while combi‐CLEAs suffer from poor spatial control and activity loss during enzyme aggregation and cross‐linking [14, 15]. DNA scaffolds, despite enabling precise enzyme positioning, exhibit limited mechanical stability, complex assembly, and scalability issues [16]. Consequently, these limitations indicate the urgent need for advanced multi‐enzyme platforms that overcome these constraints to enable more efficient biocatalysis.

Among the different cellular compartments, membrane‐less organelles formed via liquid–liquid phase separation (LLPS), commonly referred as biomolecular condensates, have attracted growing interest [17, 18]. LLPS often arises through complex coacervation, wherein oppositely charged macromolecules such as proteins and nucleic acids undergo associative phase separation to form dynamic, droplet‐like assemblies [19]. These condensates facilitate enzymatic reactions through multiple mechanisms: (i) spatial organization that increases local concentrations of specific enzymes and substrates (e.g., purinosomes) [20], (ii) insulation of metabolic pathways to ensure reaction specificity (e.g., G‐bodies) [21], and (iii) regulation of enzyme activity via allosteric interactions or post‐translational modifications (e.g., stress granules and P‐bodies) [22, 23]. Synthetic analogs, termed coacervates, have been engineered through scaffold design to compartmentalize both enzymes and substrates—typically driven by electrostatic interactions—thereby enhancing local substrate availability [24]. For example, the SUMOylation reaction was accelerated by up to 36‐fold within coacervates compared to the bulk phase due to mass action effects and altered substrate affinity [25]. Despite these advantages, the implementation of coacervates in enzyme cascade catalysis remains constrained by challenges in targeted recruitment of multiple enzymes, substrates, and cofactors, as well as limited operational stability [26].

The chemistry of complex coacervates covers a wide range of biological and synthetic components, including synthetic polymers [27], derivatized polysaccharides [28], polypeptides [29], and intrinsically disordered proteins [30]. Notably, certain pairs of globular proteins with opposing net charges can undergo phase separation under low ionic strength (<20 mM NaCl) and within pH ranges between their respective isoelectric points [31]. Proteins with patchy charge distribution form stronger electrostatic interactions, promoting heteroprotein complex coacervation and enabling protein co‐localization and up‐concentration at very specific molar ratios [32]. However, only a limited number of globular proteins—such as lactoglobulin and lactoferrin—have demonstrated this capability [32], and the distinct environmental requirements for coacervation versus cascade catalysis further complicate their practical integration [33]. Moreover, these narrow phase separation conditions hinder control over protein ratios and internal coacervate microenvironment, which is critical for efficient substrate recruitment and enhanced catalytic performance.

To improve coacervation compatibility of globular proteins, strategies such as protein engineering and chemical modification have been explored [34, 35, 36, 37, 38]. In this context, single enzyme nanogels (SENs) represent a promising alternative. These nanometric hybrid biocatalysts consist of individual enzymes encapsulated within a thin hydrogel shell formed via protein‐templated in situ free radical polymerization [39, 40]. The porous polymeric shell not only stabilizes the enzyme under harsh conditions but also allows for precise modulation of surface charge and hydrogen bonding capacity, thereby tuning interparticle interactions [41]. Successful SEN synthesis relies on the confinement of polymerization monomers at the enzyme surface, driven primarily by weak interactions [42]. Reported polymerizations generally occur under neutral to alkaline conditions, where most proteins carry a net negative charge due to their isoelectric points below this range. Consequently, negatively charged monomers are electrostatically repelled from the enzyme surface, making the fabrication of negatively charged SENs particularly challenging.

Building on the concept of heteroprotein complex coacervation, this study explores the self‐assembly of oppositely charged SENs that function in linear enzyme cascades, such as Glucose Oxidase (GOx) and Horse Radish Peroxidase (HRP). To promote coacervation, nanogels bearing complementary chemical functionalities were synthesized on the enzyme surfaces (Scheme 1i). Specifically, quaternary amines were strategically introduced onto GOx, while carboxylic acid groups were incorporated onto HRP. The modified physicochemical surface anisotropies of these enzymes with uniform, highly charged polymeric mantels broadened the conditions for hybrid heteroprotein coacervate formation (Scheme 1ii). This multi‐enzyme compartmentalization design enabled fine control over the phase separation behavior and enzyme stoichiometry, and coacervate microenvironment, leading to optimized cascade productivity. Finally, the versatility of the platform was demonstrated by the coacervation of SENs of pharmaceutical relevance involved in parallel enzyme cascades.

SCHEME 1.

Overview of single enzyme nanogel (SEN) synthesis and heteroprotein coacervation. (i) Fabrication of positively or negatively charged SENs using pure enzymes as templates, resulting in core‐shell architectures with hydrogel shells. (ii) Mixing oppositely charged SENs across a broad pH range under moderate salt conditions triggers liquid–liquid phase separation, forming microdroplets that co‐localize multiple enzymes at high concentrations with tunable stoichiometry. The resulting densely packed coacervates can recruit small molecules (A and C) and favor reaction intermediate confinement (B), enhancing product formation rate (D).

2. Results and Discussion

2.1. Synthesis, Characterization, and Activity of SENs

Using the single enzyme nanogel technology, individual proteins were encapsulated within flexible polyacrylamide‐based hydrogels bearing tailored functionalities, thereby generating a unique and customizable protein interface [43]. The synthesis protocol for SENs was systematically optimized to maximize enzyme encapsulation efficiency by varying monomer to protein concentrations and adjusting polymerization pH (Table S1). Specifically, GOx was encapsulated within a hydrogel synthesized from (3‐acrylamidopropyl)trimethylammonium chloride (APTAC) and N‐hydroxyethyl acrylamide (HEAA), introducing quaternary amine and hydroxyl functionalities, respectively, resulting in the positively charged GOx@SEN+. Similarly, HRP was encapsulated using 2‐carboxyethyl acrylate (CEAA) and HEAA to introduce carboxylic acid and hydroxyl groups, yielding negatively charged HRP@SEN‐. HRP@SEN‐ polymerization was conducted at pH 5.0, between the pKa of CEAA (4.5) and the pI of HRP (6.3), thereby maximizing ionic interactions between the enzyme surface and the monomer (Table S2). In both systems, N,N’‐methylene bisacrylamide (BIS) was used as the crosslinker.

Dynamic light scattering (DLS) analysis revealed that the hydrodynamic diameters of GOx@SEN+ and HRP@SEN‐ were significantly larger than those of the corresponding free enzymes, with average values of 28 nm and 15 nm, respectively (free GOx: 7 nm; free HRP: 5 nm; Figure 1A). The normalized volume distributions indicated well‐defined populations, further supported by correlograms exhibiting smooth single‐exponential decays (Figure 1B). Enzyme encapsulation efficiency and the corresponding increase in molecular weight were confirmed by SDS‐PAGE (Figure S1). HRP@SEN‐ was further characterized using asymmetric flow field‐flow fractionation (AF4) coupled with UV detection, multi‐angle light scattering (MALS), and DLS. GOx@SEN+ could not be characterized by AF4 because positively charged SENs adhered to the AF4 membrane. UV profiles revealed a significant size increase compared to native HRP, with retention times of 51 and 16 min, respectively, and with only 10% of the sample corresponding to free enzyme. Simultaneous MALS and DLS analysis provided morphological insights via the shape factor (Rg Rh⁻¹). HRP@SEN‐ exhibited values between 1.5 and 2.0, which may be related to asphericity, intraparticle density variations, or other factors (Figure 1C) [44].

FIGURE 1.

Physicochemical and functional characterization of GOx@SEN+ and HRP@SEN‐ at pH 6.0. (A) Volume size distribution profiles obtained by DLS, showing increased hydrodynamic diameters upon encapsulation. (B) Correlograms from DLS measurements. (C) AF4 fractogram of HRP and HRP@SEN‐ (red). Ratio between the radius of gyration (R _g) and hydrodynamic radius (R _h) to determine the shape factor (black). (D) SAXS profiles recorded from the enzymes, GOx and HRP (light markers), as well as the nanogels, GOx@SEN+ and HRP@SEN‐ (dark markers), along with their respective fits (solid lines: sphere model for free enzymes and gel model for SENs). (E) CryoTEM micrographs of GOx@SEN+, scale bar: 500 nm. (F) CryoTEM micrographs of HRP@SEN‐, scale bar: 500 nm. (G) SEN size distribution determined from cryoTEM images. (H) Zeta‐potential measurements comparing free enzymes and their corresponding SENs. (I) Turnover number (k _cat) comparison between free enzymes and SENs.

The SAXS profiles of the free enzymes and SENs exhibited pronounced differences in the low‐q scattering regime (Figure 1D). A q‐independent plateau was observed in the SAXS profiles of both GOx and HRP due to their small size (R < 1/q _min). This plateau was absent in the SAXS profiles of SENs due to their larger size (R > 1/q _min). Guinier analysis in the low‐q regime (𝑞𝑅_𝑔 < 1.3) yielded radii of gyration of R _g = 34.0 ± 0.1 Å for GOx and 22.6 ± 0.1 Å for HRP (Table S3). The good agreement for HRP with the predicted R _g and molecular weight (M _w) indicated that HRP was predominantly monomeric (Figure S2 red). In contrast, GOx exhibited an R_g and M_w approximately 30% larger than predicted, consistent with dimerization and in agreement with its established quaternary structure (Figure S2 blue) [45]. The experimental SAXS profiles of both enzymes were well described by a simple spherical form factor model (Table S3). In contrast, SEN profiles were analyzed using a gel model, which provided key structural parameters: overall particle size (R), correlation length (ξ), and fractal dimension (D) (Table S4, see methods S4.2). Interestingly, GOx@SEN+ were larger, and exhibited a less connected network (R = 210.0 ± 1.0 Å, ξ = 80.3 ± 0.3 Å) compared to HRP@SEN‐ (R = 107.3 ± 0.2 Å and ξ = 27.8 ± 0.1 Å), while their fractal dimensions were similar (D = 2.7‐2.8). The higher ξ value for GOx@SEN+ likely reflected lower crosslinking density (crosslinker to total monomer ratio: 0.29 vs. 0.40 for HRP@SEN‐, Table S1). These structural differences between SENs arise due to variations in polymerization conditions, including pH, monomer chemistry, and enzyme surface anisotropy [42].

CryoTEM micrographs of GOx@SEN+ (Figure 1E; Figure S3A) and HRP@SEN‐ (Figure 1F; Figure S3B) revealed differences in morphology and size distribution (Figure 1G). GOx@SEN+ exhibited larger and more heterogeneous particles, with diameters ranging from approximately 10 to 45 nm and occasional multi‐lobed morphology. In contrast, HRP@SEN‐ formed smaller, predominantly uniform particles within the 6–17 nm range, with only a minor fraction between 20–35 nm. From these dimensions, and using crystallographic estimates of the underlying enzyme sizes, we calculated average nanogel shell thicknesses of 8.5 nm for GOx@SEN+ and 4.3 nm for HRP@SEN‐. Both systems displayed excellent solubility, with no detectable aggregates observed in the images. These morphological trends align with hydrodynamic dimensions determined by DLS and AF4, and structural parameters derived from SAXS analysis.

Zeta potential measurements (Figure 1H) demonstrated pronounced shifts in surface charge upon encapsulation. GOx@SEN+ exhibited a reversal in charge, shifting from approximately −8 mV (free GOx) to +10 mV, while HRP transitioned from near neutral to −19 mV. These changes in electrostatic properties indicate that the encapsulation process significantly alters the surface charge of the nanoparticles.

The catalytic performance of SENs was also evaluated (Figure 1I; Figures S4 and S5). Both SENs retained substantial activity, achieving turnover numbers (k _cat) corresponding to approximately 50%–60% of those measured for free enzymes. Although the nanogel shells impose a minor diffusion limitation (Figure S6) [42], the retained enzymatic activity remains impressive, with values reaching 5.3 × 10³ min⁻¹ for GOx@SEN+ and 3.7 × 10³ min⁻¹ for HRP@SEN‐.

In summary, the detailed characterization of SENs confirmed the successful encapsulation of single active enzymes within thin hydrogel shells bearing opposite surface charges.

2.2. Oppositely Charged Single Enzyme Nanogel (SEN) Pairs Enable Hybrid Heteroprotein Complex Coacervation

Heteroprotein complex coacervation has so far been limited to a narrow set of pairs of globular proteins and demands highly specific conditions (narrow pH ranges, precise protein stoichiometries, and low ionic strengths) which restricts its practical applicability [32, 46]. Reported examples, including protein combinations and environmental parameters required for successful phase separation, are summarized in Table S5. For instance, GOx and HRP, commonly used as model enzymes in cascade reactions, failed to undergo phase separation at pH 6.0 (between their isoelectric points) across varying molar ratios and in the absence of salt (Figure S7). With our oppositely charged SENs in hand we expected to be able to induce robust coacervation under a broad range of conditions, including variations in pH and ionic strength, to construct catalytically active microdroplets.

Brightfield and confocal laser scanning microscopy (CLSM) imaging of 1.5 µM GOx@SEN+ and 2.5 µM HRP@SEN‐ confirmed the formation of discrete spherical microdroplets (3.3 ± 0.8 µm) (Figure 2A), indicative of thermodynamically stable phase separation with uniform interfacial properties [47]. Moreover, they exhibited excellent phase stability, persisting as functional condensed phase for at least 20 h under physiologically relevant conditions (Figure S8). Dual‐color fluorescence verified co‐localization of both enzymes, key for ensuring high local protein concentrations and, thus, shorter distances for reaction intermediate transfer [48]. The labelling of SENs did not alter their physicochemical properties (Table S6). Three‐dimensional reconstruction of CLSM z‐stacks further demonstrated single‐phase spherical droplets with a homogeneous internal environment. Satisfyingly, liquid‐liquid phase separation occurred at physiological ionic strength (150 mM), a condition that traditionally prevents coacervation in all previously reported heteroprotein systems, demonstrating the exceptional multivalent electrostatic interactions and dynamic network flexibility of our oppositely charged SENs. Consistent with this, CLSM imaging revealed solid precipitates in the absence of NaCl (Figure S9), indicative of strong electrostatic interactions driving phase separation. In contrast, at 500 mM NaCl, coacervate formation was abolished (Figure S9), as confirmed by DLS (Figure S10).

FIGURE 2.

Characterization of coacervates formed from 1.5 µM GOx@SEN+ and 2.5 µM HRP@SEN‐ at pH 6.0 and 150 mM NaCl. (A) Representative bright‐field (BF) and confocal laser scanning microscopy (CLSM) images of SEN coacervates: (i) GOx@SEN+ labelled with sulfoCy3 (λ_ex = 561 nm); (ii) HRP@SEN‐ labelled with sulfoCy5 (λ_ex = 638 nm), and (iii) merged image. Scale bar: 10 µm. Inset: 3D reconstruction of z‐stacks of coacervates. The distance between scale bar ticks in the 3D reconstruction is 5 µm. (B) Phase diagram showing the phase separation fraction (1 red, 0 blue) across varying SEN concentrations in sodium phosphate buffer (pH 6.0 and 150 mM NaCl). (C) Estimated number of GOx@SEN+ (blue) and HRP@SEN‐ (red) incorporated into coacervates. Enzyme molar ratio (n_GOx: n_HRP) within the dense phase represented by black rhombuses. For coacervate assembly HRP@SEN‐ concentration was fixed at 2.5 µM, while GOx@SEN+ was varied from 1.0 to 2.5 µM. (D) Partitioning of low‐molecular‐weight dyes into coacervates. (E) SAXS profiles of coacervates (purple) compared to individual SENs (red and blue). (F) FRAP analysis of micrometer‐size coacervates to assess dynamics (purple) and diffusion of 5(6)‐carboxyfluorescein (black).

Using the high‐throughput PhaseXplorer microfluidic system [49], we mapped the phase behavior of GOx@SEN+ and HRP@SEN‐ across a range of SEN concentrations ranging from 0.5 to 2.5 µM at pH 6.0 in 150 mM NaCl (Figure 2B). This analysis revealed a broad phase‐separated regime, attributed to the high and balanced charge density and uniform charge distribution of SENs. Their structural flexibility may promote polymer entanglement and multivalent interactions, strengthening ionic attractions between oppositely charged SENs. Furthermore, other interactions, including hydrogen bonding mediated by hydroxyl groups present in both SENs, probably expanded the phase separation window [50]. The observed asymmetry in the phase boundary likely originates from differences in SEN surface charge density (Figure 1H) and from unequal SEN sizes [51]. Next, we investigated the phase separation of oppositely charged SENs across a broad pH range (4.0–9.0) to assess their versatility under diverse environmental conditions (Figure S11). Remarkably, coacervation persisted throughout all tested pH values and was compatible with every buffer system except citrate, whose multivalent counterions strongly disrupted complex formation. This broad phase stability significantly expands the applicability of SEN‐based coacervates and enables their integration into most enzyme cascades.

To assess the tunability of enzyme stoichiometry within coacervates, protein incorporation into the dense phase was quantified for formulations prepared with an excess of HRP@SEN‐ (2.5 µM). Increasing the concentration of GOx@SEN+ led to a modest increase in its immobilization yield (from 35% to 50%, Figure S12). In contrast, HRP@SEN‐ exhibited consistently lower incorporation (approximately 25%–30%), probably reflecting the combined effects of its small size, higher crosslinking density, and greater surface charge relative to GOx@SEN+. Protein immobilization values were utilized to estimate the number of enzymes incorporated into the dense phase (Figure 2C), revealing equimolar enzyme incorporation at formulation 1.5 µM GOx@SEN+ and 2.5 µM HRP@SEN‐. Below 1.5 µM GOx@SEN+, HRP@SEN‐ incorporation exceeded that of GOx@SEN+, whereas above this concentration, the trend reversed, demonstrating tunable control over GOx:HRP molar ratios from 0.5 to 1.5 within the dense phase. Varying the SEN ratio also affected coacervate size (Figure S13).

The volume fraction of the dense coacervate phase (1.5 µM GOx@SEN+ and 2.5 µM HRP@SEN‐, pH 6.0, 150 mM NaCl) was determined to be 0.3% v/v, indicating that the concentrations of both GOx and HRP within the coacervate reached approximately 0.2 mM, corresponding to 33 mg mL⁻¹ for GOx and 9 mg mL⁻¹ for HRP. This results in a total protein concentration of 0.4 mM within the coacervate phase. Notably, this represents a 233‐fold enrichment for GOx and a 96‐fold enrichment for HRP relative to the dilute phase. Such up‐concentration levels approach the protein concentration from the cytosol of living cells [52], underscoring the ability of SEN‐based coacervates to mimic biologically relevant environments. Furthermore, the high density of 1.5 g mL⁻¹, determined from the calculated coacervate mass and volume, reflects the tightly packed internal architecture.

The local physicochemical environment within coacervates critically influences enzyme activity by altering substrate accessibility or shifting reaction equilibria [53]. To characterize the internal milieu of SEN coacervates (1.5 µM GOx@SEN+ and 2.5 µM HRP@SEN‐, pH 6.0, 150 mM NaCl), we examined the uptake of small fluorescent molecules (Figure 2D). A panel of low‐molecular‐weight compounds (MW < 1000 g mol⁻¹) with varying net charges was added to the coacervate solution and analyzed by CLSM. Negatively charged pyranine and 5(6)‐carboxyfluorescein exhibited stronger partitioning into micrometric coacervates compared to positively charged rhodamine 6G and crystal violet, although overall partitioning remained moderate. These observations suggest that the coacervate interior was slightly positively charged, likely due to the larger size and higher occupancy of GOx@SEN+ relative to HRP@SEN‐, thereby favoring electrostatic interactions with anionic dyes. Beyond electrostatic interactions, hydrogen bonding may also contribute: 5(6)‐carboxyfluorescein, despite being less negatively charged than pyranine, exhibits a greater capacity for hydrogen bond formation, which could explain its stronger partitioning. A similar effect is evident when comparing rhodamine 6G and crystal violet, where differences in hydrogen bonding rather than hydrophobicity account for their partitioning behavior.

Zeta‐potential measurements of coacervates containing equimolar enzyme amounts revealed a negative surface potential (−34 mV), indicating enrichment of HRP@SEN‐ particles at the coacervate surface, likely due to their excess in the mixture. Comparison of SAXS profiles for SENs and coacervates revealed a distinct change in morphology and structure upon mixing the SENs. For the coacervates, the scattering pattern lacks a q‐independent plateau, indicating that such structures exceed the accessible length scale of the SAXS instrument (R > π/q _min ∼45 nm). Consequently, the Guinier approximation could not be performed. Furthermore, comparing the q‐dependent decay, the intensity showed a stronger power‐law relation, α ≈ 2.7 for the coacervates, indicating the rough surface of these coacervates (Figure 2E). These findings suggest that the interior of equimolar coacervates contains a slightly higher ratio of positively charged species, while the interface is enriched with smaller HRP@SEN‐.

Fluorescence recovery after photobleaching (FRAP) was employed to evaluate the dynamics of the coacervates. Photobleaching of fluorescently labelled GOx@SEN+ (purple) yielded an immobile fraction of 0.55 ± 0.08, and a half‐time recovery (t_1/2) of 155 ± 25 s (Figure 2F; Figure S14). These recovery kinetics indicate that the coacervate interior is partially dynamic and exhibits viscoelastic behavior, combining fluid‐like and solid‐like characteristics [54]. Notably, as demonstrated above, these properties can be tuned by adjusting ionic strength. In contrast, the small fluorescent dye 5(6)‐carboxyfluorescein (black) showed no immobile fraction and a t_1/2 of 2.1 s, two orders of magnitude faster than SENs, which demonstrates the high diffusion dynamics of small molecules within coacervates.

These results demonstrate that oppositely charged SENs act as building blocks for coacervation under broad formulation conditions, enabling the co‐localization of enzymes at exceptionally high concentrations. This platform offers precise control over coacervate composition through tunable SEN ratios, while ionic strength provides additional handles to modulate the nature of phase separation. Equimolar protein formulations in the dense phase preferentially enrich negatively charged small molecules within the coacervates due to a slightly positive charge environment, while these small molecules remain highly diffusive.

2.3. SEN Coacervates Boost Cascade Performance

The cascade activity of the previously characterized equimolar coacervates was evaluated by monitoring HRP‐mediated oxidation of negatively charged ABTS into its chromogenic product, using hydrogen peroxide generated in situ by GOx during glucose oxidation (Figure 3A). Based on dye recruitment experiments, we anticipated glucose and ABTS partitioning into coacervates via hydrogen bonding and ionic interactions, respectively, with localized production of H₂O₂ facilitating its access to the HRP active site (Figure 3B) [48]. Initially, turnover numbers (k _cat) were compared for uncomplexed SENs (mixed in 500 mM NaCl) and coacervates under conditions matching dye recruitment studies (50 mM sodium phosphate, pH 6.0, 150 mM NaCl) using 5 mM glucose and 1 mM ABTS. Notably, coacervates exhibited 25% higher k _cat values than uncomplexed SENs (Figure 3C). To investigate this enhancement, initial velocities (υ_i ) were measured across a glucose range (0.1–12.5 mM) at constant ABTS. As shown in Figure 3D, coacervates (green) exhibited up to a threefold increase in υ_i at [glucose] <1 mM, likely due to glucose enrichment elevating local concentrations toward GOx saturation (K _Mglucose = 3.7 mM, Figure S15). At higher glucose concentrations, activity increments (1.5–2‐fold) were attributed to H₂O₂ accumulation near HRP@SEN‐. Precipitates (formed at 0 mM NaCl) showed similar activity to uncomplexed SENs below K _Mglucose (4.3 mM, Figure S16), but up to 50% higher activity above K _Mglucose, likely because diffusion barriers dominate at low substrate levels but diminish at high concentrations where intermediate accumulation prevails [48]. Thus, these results indicate that coacervates enhance catalysis at low glucose concentrations primarily via substrate recruitment, while at high concentrations, improvements stem from H₂O₂ accumulation. Importantly, coacervates remained structurally stable throughout the 10‐minute reaction window (Figure 3E).

FIGURE 3.

Catalytic performance of SEN‐based coacervates. (A) Schematic of the enzymatic cascade: GOx oxidizes glucose to gluconic acid and H₂O₂, which HRP subsequently uses to peroxidize ABTS into a chromogenic product. (B) Proposed mechanism for substrate recruitment and in situ intermediate production within coacervates. (C) Turnover numbers (k _cat) of uncomplexed SENs (yellow) and coacervates (purple) under identical conditions. (D) Initial reaction velocities of coacervates and precipitates relative to uncomplexed SENs across glucose concentrations (0.1 to 12.5 mM), and (E) CLSM images of SEN coacervates under reaction conditions at selected time points (0–10 min). The fluorescent emission of coacervates varies as ABTS gets oxidized. Scale bar: 10 µm.

These findings demonstrate that SEN coacervates enable substrate recruitment and facilitate intermediate accumulation, showing their strong potential for multi‐enzyme biocatalysis.

2.4. Catalytic Performance of SEN‐Based Coacervates of Enzymes of Industrial Interest

Building on the principles of SEN coacervation, we investigated industrially relevant enzyme pairs and explored how SEN formulation and phase behavior influence catalytic performance [55, 56]. Ene‐reductases (EREDs) are pivotal in biosynthetic applications for the stereoselective reduction of activated C═C bonds [57], while glucose dehydrogenase (GDH) regenerates NADPH, an essential cofactor for EREDs and other oxidoreductases, using glucose as a sacrificial substrate (Figure 4A) [58]. An ERED from Pseudomonas putida KT2440 and a GDH from Bacillus subtilis were expressed, purified and modified into oppositely charged SENs (Table S1), which resulted in ERED@SEN‐ (Figure S17) and GDH@SEN+ (Figure S18). These SENs exhibited hydrodynamic diameters of 25 and 39 nm and zeta‐potentials of −21.5 and +11.2 mV, respectively. The values are comparable to those obtained for SENs in the GOx‐HRP system under similar synthesis conditions.

FIGURE 4.

Characterization of coacervates formed from ERED@SEN‐ and GDH@SEN+. (A) Biocatalytic cascade of ERED and GDH for cyclohexanone production with cooperative NADPH recycling. (B) Phase diagram showing phase separation across varying SEN concentrations in sodium phosphate buffer (pH 8.0 and 75 mM NaCl). Colored rhombuses indicate formulations selected for subsequent experiments. (C) CLSM overlay of ERED@SEN‐ and GDH@SEN+ formulations displaying coacervate vesicles (CV, purple), nanometric coacervates (NanoC, orange), precipitates (Ppt, gold), and micrometric coacervates (MicroC, green), labelled with sulfoCy5 and sulfoCy3, respectively, in the presence of the catalytic reaction mixture (3 mM cyclohex‐2‐en‐1‐one, 4 mM glucose, 0.2 mM NADP⁺, 75 mM NaCl, and 50 mM sodium phosphate pH 8.0). The scale bar represents 10 µm. (D) Partitioning of low‐molecular‐weight dyes into MicroC. (E) Cyclohex‐2‐en‐1‐one conversions of synthesized formulations.

PhaseXplorer was used to investigate the phase‐separation behavior of ERED@SEN‐ and GDH@SEN+ and under conditions optimized for enzymatic activity (pH 8.0, 75 mM NaCl; Figure 4B) [59]. Particular attention was given to regimes enriched in ERED, owing to the markedly unbalanced turnover numbers of ERED@SEN‐ and GDH@SEN+ (213 ± 19 and 3106 ± 437 min⁻¹, respectively), indicating that ERED is the limiting factor for overall catalytic performance. The dense‐phase region observed for the GDH–ERED system showed that SEN coacervates are enzyme agnostic, phase separating regardless of enzyme identity.

Based on the phase diagram, four formulations were experimentally tested (represented by rhombus symbols in Figure 4B). CLSM imaging in the presence of the reaction mix revealed pronounced morphological variability and structural changes (Figure 4C). The formulation with 0.8 µM ERED@SEN‐ and 1.0 µM GDH@SEN+ (purple rhombus) transformed from spherical coacervates into coacervate vesicles (CV, Figure S19), likely due to charge imbalance and osmotic effects upon contact with the reaction medium [60, 61]. The formulation near the phase boundary (1.5 µM ERED@SEN‐ and 1.0 µM GDH@SEN+; orange rhombus) produced nanometric assemblies (NanoC) [29]. Increasing GDH@SEN+ to 2 µM (golden rhombus) yielded solid precipitates (Ppt), suggesting that excess positive charge promotes stronger interparticle interactions and percolated network formation, driving solid‐liquid phase separation. Finally, rising ERED@SEN‐ to 2.5 µM (with 2.0 µM GDH@SEN+; green rhombus) produced viscous micrometric coacervates (MicroC). Notably, this morphology was reproduced consistently across independent synthesis, demonstrating excellent batch‐to‐batch fidelity (Figure S20). This combination of morphological versatility and fabrication reproducibility positions SEN coacervates as a modular platform for tailoring enzyme organization in applications such as biofuel cells, biosensing, nanomedicine, or remediation.

Protein immobilization yields and the number of proteins incorporated into the phase‐separated structures were quantified (Table S7). ERED@SEN‐ consistently exhibited immobilization yields exceeding 60%, while GDH@SEN+ reached approximately 90%, markedly surpassing the GOx‐HRP system (Figure S12). These improvements are attributed to stronger inter‐SEN ionic interactions at lower salt concentrations (Figure S21), which in some cases even induced precipitate formation under specific SEN ratios.

Among the four evaluated systems, MicroC was selected for further analysis due to its higher incorporation of SENs and its suitability for characterization using CLSM‐related techniques. FRAP analysis of MicroC further confirmed strong interparticle interactions, revealing a high immobile fraction of SENs (0.82 ± 0.02; Figure S22). The internal environment of MicroC was characterized through dye partitioning studies (Figure 4D), revealing a strongly positive interior, likely resulting from the excess and bigger size of GDH@SEN+ over ERED@SEN‐ in the coacervate (Table S7). This is evidenced by the efficient recruitment of pyranine and 5(6)‐carboxyfluorescein, while crystal violet was excluded. Rhodamine 6G also partitioned strongly, likely due to hydrogen‐bonding interactions with the coacervates. Importantly, FRAP analysis of small dyes confirmed high mobility within MicroC (Figure S23). Zeta potential measurements of MicroC (+8 mV) corroborated the net positive environment. Based on these findings, we hypothesize that negatively charged NADP(H) is efficiently recruited via electrostatic interaction, while glucose partitions through hydrogen bonding, as previously confirmed for GOx‐HRP coacervates.

The catalytic performance of the four formulations was assessed in the reduction of cyclohex‐2‐en‐1‐one under limiting NADP⁺ conditions (5 mol% relative to substrate) over a 2 h reaction period, after confirming that the coacervates remained structurally stable for at least this duration (Figure S24). The nature of the phase, SEN ratio, and microenvironment proved decisive for conversion (Figure 4E): CVs reached 38%, NanoC only 7% despite a higher ERED concentration than CVs, whereas MicroC delivered nearly complete conversion (94%). This exceptional performance of MicroC is attributed to its high small‐molecule diffusivity, efficient negatively charged small molecule (e.g., NADPH) recruitment, elevated ERED loading (Table S7), and an optimized SEN ratio (Table S8). Ppts also achieved 38% conversion, confirming that solid assemblies can sustain cascade catalysis, albeit at lower productivity than micrometric coacervates. These results demonstrate that cascade activity is governed not merely by enzyme loading, but by the interplay of morphology, SEN ratios, and the phase‐separated microenvironment, factors that ultimately dictate catalytic performance.

3. Conclusion

In this work, we present a versatile strategy for enabling heteroprotein complex coacervation by engineering the protein surface through single enzyme nanogel (SEN) technology. SENs encapsulate individual enzymes within hydrogel shells bearing tailored charges, resulting in well‐defined, catalytically active nanoparticles. Mixing oppositely charged SENs under broadly tunable pH conditions drives the formation of coacervates with customizable size, morphology, composition, and microenvironment. These assemblies facilitate efficient molecular exchange and high local enzyme concentrations, promoting substrate recruitment, intermediate accumulation, and enhanced cascade catalysis. Nanogels serve as structural elements for coacervate formation, while being catalytically active and eliminating inert structural components. Extending this concept to pharmaceutically relevant enzymes further highlights its universality and positions SEN coacervates as a promising platform for broad biotechnological applications.

Looking ahead, designing SEN‐based coacervates with improved colloidal stability will be key to enabling their reuse across multiple catalytic cycles. The broad chemical flexibility of SENs also opens opportunities for new phase‐separation chemistries and hierarchical, multi‐enzyme assemblies for increasingly sophisticated biocatalytic materials.

Conflicts of Interest

The authors declare no conflicts of interest

Supporting information

Supporting File: adma73688‐sup‐0001‐SuppMat.docx.

Data Availability Statement

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