Engineering bio‐brick protein scaffolds for organizing enzyme assemblies

Abstract

Enzyme scaffolding is an emerging approach for enhancing the catalytic efficiency of multi‐enzymatic cascades by controlling their spatial organization and stoichiometry. This study introduces a novel family of engineered SCAffolding Bricks, named SCABs, utilizing the consensus tetratricopeptide repeat (CTPR) domain for organized multi‐enzyme systems. Two SCAB systems are developed, one employing head‐to‐tail interactions with reversible covalent disulfide bonds, the other relying on non‐covalent metal‐driven assembly via engineered metal coordinating interfaces. Enzymes are directly fused to SCAB modules, triggering assembly in a non‐reducing environment or by metal presence. A proof‐of‐concept with formate dehydrogenase (FDH) and L‐alanine dehydrogenase (AlaDH) shows enhanced specific productivity by 3.6‐fold compared to free enzymes, with the covalent stapling outperforming the metal‐driven assembly. This enhancement likely stems from higher‐order supramolecular assembly and improved NADH cofactor regeneration, resulting in more efficient cascades. This study underscores the potential of protein engineering to tailor scaffolds, leveraging supramolecular spatial‐organizing tools, for more efficient enzymatic cascade reactions.

1. INTRODUCTION

There is a growing need for developing new scaffolds capable of organizing enzymes effectively and, thereby, facilitating cascade reactions. The demand for efficient and adaptive synthetic approaches (Chen & Arnold,  2020) has made enzyme assemblies increasingly important (Gad & Ayakar, 2021). Protein scaffolds have demonstrated their effectiveness in promoting enzyme assembly, enhancing the catalytic efficiency of multi‐enzyme systems (Lin et al., 2014; Ellis et al., ²⁰¹⁹). Enzymes are often linked together to adequately provide or recycle cofactors, such as NADH and NAD(P)H, making fusion enzyme engineering a promising technique for enzyme assembly, especially when in‐situ cofactor recycling is needed (Aalbers & Fraaije, 2019; Yu et al., ²⁰¹⁵).

Although the catalytic efficiency can be significantly increased by bringing two enzymes together, fusion proteins often face drawbacks, such as aggregation into inactive enzyme clusters or a reduction of enzyme intrinsic kinetics (Chen et al., 2013). These limitations can be partially overcome by using more sophisticated methods based on post‐translational assemblies enabled by synthetic protein scaffolds (Vanderstraeten & Briers, 2020; Monterrey et al., ²⁰²²). The most significant advantage of using protein‐based scaffolds for enzyme assembly is that the enzymes are assembled after translation, minimizing the chances of aggregation underlying large fusion proteins (Gad & Ayakar, 2021; Zhang et al., ²⁰¹⁸). Furthermore, the multi‐enzyme systems based on protein‐based scaffolds can be used both in vivo and in vitro, enabling precise control over the nanoscale spatial organization of enzymes (Vanderstraeten & Briers, 2020).

Several synthetic protein scaffolds have been developed to assemble enzymes and enhance the one‐pot transformations of alcohols and aldehydes into amines (Zhang et al., 2019). Zeballos et al. (2021) assembled two enzymes on a synthetic protein scaffold, which, in turn, is immobilized on a solid carrier; a porous agarose microbead. The synthetic protein scaffold used was a cellulosome‐based system based on a type II dockerin–cohesin pair from Clostridium thermocellum. The association between cohesin and dockerin enables the inclusion of individual enzyme subunits into the cellulosome complex (Mechaly et al., 2000). In a more sophisticated approach, Ledesma et al. developed a protein scaffold based on tetratricopeptide repeat affinity proteins (TRAPs), which spatially arrange up to three peptide‐tagged enzymes while locally sequestering the required cofactors. This enables more productive cascades compared their free counterparts (Ledesma‐Fernandez et al., 2023). The aforementioned protein complexes are based on reversible protein–protein interactions where covalent bonds are not involved. In contrast, Zhang et al. (2019), exploited another widely used protein domain as synthetic scaffold, the SpyTag‐SpyCatcher system, to covalently and irreversibly assemble two orthogonal dehydrogneases. The SpyTag‐SpyCatcher protein scaffold (Hatlem et al., 2019) facilitated a dual enzyme cascade for chiral amine synthesis and the scaffolding of the cascade shortened the time needed to accomplish final conversions compared to a free enzyme system (Hatlem et al., 2019).

As an alternative to the described protein–protein interactions, the consensus tetrapeptide repeat (CTPR) module (Mejias et al., 2016; Cortajarena et al., ²⁰¹⁰) was selected for the design and fabrication of a novel set of modular synthetic protein scaffolds (SCABs) that can assemble multi‐enzyme systems. CTPR units present numerous characteristics that make them ideal scaffolds for this application: (1) modular nature (Cortajarena et al., 2011); (2) notable stability (Cortajarena & Regan, 2011); (3) tolerance to mutations; (4) structure defined by few conserved residues within their sequence (Mejias et al., 2016); (5) intrinsic self‐assembly properties to form linear protein arrays (Mejías et al., 2014; Grove et al., ²⁰¹³); and (6) extended structure that displays a large surface area to volume ratio (Uribe et al., 2021). These features have allowed previous demonstrations of the potential of engineered CTPR proteins as scaffolds to template different small molecules and nanoelements, such as, photoactive molecules (Mejías et al., 2016), single‐walled carbon nanotubes (López‐Andarias et al., 2018), electroactive clusters (Mejias et al., 2019), and gold nanoparticles (Mejias et al., 2021).

Among these features, some are particularly relevant for the intended enzyme scaffolding application, which make them highly suitable candidates for use as protein scaffold for enzyme assemblies. In this regard, CTPR proteins are more resistant to the destabilizing effects of mutations as they are thermodynamically more stable than their natural counterparts, such as TPRs (Cortajarena et al., 2010; Cortajarena & Regan, ²⁰⁰⁶), and than TRAP domains engineered based on natural TPR domains (Main et al., 2003; Speltz et al., ²⁰¹⁵; Cortajarena et al., ²⁰⁰⁸). In addition, CTPRs hold inherent self‐assembly properties. Identical CTPR units assemble into linear arrays spanning up to hundreds of nanometers, facilitated by their well‐defined inter‐repeat packing interface. Furthermore, the insertion of cysteines at the N‐ and C‐ terminal ends of CTPR units allowed the generation of disulfide bonds that stitched those interactions, creating stable and covalently linked linear nanofibers, as well as ordered solid films (Mejías et al., 2014; Phillips et al., ²⁰¹²). Given that the assemblies of arrayed proteins occur at the nanometric scale, this biomolecular platform hold promise for organizing multi‐enzyme systems, particularly when intermediate transport between active sites represents a rate‐limiting step (i.e., in presence of competing enzymes). Furthermore, CTPR protein sequence is more versatile compared to TRAP proteins (Speltz et al., 2015), which are nearly identical to their natural counterparts; TPRs (Uribe et al., 2021; Aires et al., ²⁰²⁰). The CTPR unit is a helix‐turn‐helix motif consisting of 34 amino acids, of which only 8 are necessary to enable proper structural folding. Due to the limited number of conserved residues, it is easier and more reproducible to introduce functional mutations to endow the protein with desired features. For example, the addition of metal‐binding residues, such as histidines and cysteines enables the coordination of metal ions and the subsequent stabilization of novel nanomaterials within the CTPR proteins (Mejias et al., 2016; Uribe et al., ²⁰²¹; Aires et al., ²⁰¹⁹; Cortajarena et al., ²⁰²¹). The coordination of certain amino acids (histidines, cysteines, or tyrosines) with transition metals like copper, nickel, cobalt, or zinc is a method that is gaining popularity for engineering well‐defined protein–protein assemblies (Salgado et al., 2009; Salgado et al., ²⁰¹⁰), and is suitable to be applied to CTPR units.

The present work is based on the development of two protein–protein assemblies of two fusion enzymes formed by a CTPR domain (scaffolding unit, SCAB) and a dehydrogenase (biocatalytic unit); either the formate dehydrogenase from Candida boidinii (FDH) or the alanine dehydrogenase from Bacillus stearothermophilus (AlaDH), giving raise to the constructs SCAB‐FDH and SCAB‐AlaDH, by which new organized multi‐enzyme systems are obtained. The assembly approaches rely on selective head‐to‐tail interactions, promoted by either single engineered cysteine at the N‐ and C‐terminal ends, resulting in the production of disulfide covalent bonds (Mejías et al., 2014), or metal‐driven assembly, in which tailored histidines makes coordination bonds with cooper (II) ion forming a metal‐protein interaction that controls the protein–protein assembly (Salgado et al., 2010). To accomplish this objective, we selectively assembled FDH and AlaDH by post‐translational scaffolding through the fusion to SCAB proteins. Enzyme cascades that require NADH in situ recycling are excellent examples to assess the efficiency of these protein‐based scaffolds in enhancing the cascade performance. Numerous methods relating to NADH cofactor recycling via enzyme assembly have been studied during the past 10 years due to the high demand for this class of cofactors (Ledesma‐Fernandez et al., 2023; Rocha‐Martín et al., ²⁰¹²; Klein et al., ²⁰¹⁹; Hartley et al., ²⁰¹⁹). In the selected enzyme cascade, while AlaDH catalyzes the asymmetric reduction of α‐ketoacids to enantiopure α‐amino acids at the expense of NADH, FDH uses formic acid as an ancillary substrate to in situ and concomitantly recycle NADH. Through a variety of analytical tools, we confirm that the synthetic CTPR‐based scaffolds successfully assemble the orthogonal FDH‐AlaDH bi‐enzymatic cascade and report on the enhanced performance of the assembled systems.

2. RESULTS AND DISCUSSION

2.1. Design of the SCAffolding bricks

In this work, we explored the potential of using genetically programmed intermolecular staples that selectively bridge engineered CTPR domains and assemble multi‐enzyme systems in a spatially organized manner. To achieve this, two strategies were employed to design orthogonal inter‐molecular staples that could stabilize the assemblies formed through the intrinsic head‐to‐tail interactions between CTPR modules (Mejías et al., 2014). Specifically, we used cysteine‐mediated covalent disulfide bonds and non‐covalent metal‐mediated assembly as stapling chemistries to lock the supramolecular assemblies.

In our first strategy, we were inspired by previous work of our group, where head‐to‐tail interactions were stapled by selective disulfide bonds between unique N‐ and C‐terminal cysteines within CTPR20 proteins (CTPR wild‐type protein with 20 identical CTPR repeats), leading to the formation of ordered protein nanofibers (Mejías et al., 2014). Hence, we devised orthogonal scaffolding units to encode directional order within the assembly. For this aim, we exploited for the first time the two orthogonal interfaces of a CTPR protein (the intra‐ and inter‐repeat packing interfaces) to design two unique orthogonal modules that can assemble in a directional manner, avoiding the pseudo‐infinite self‐polymerization already described (Figure 1a) (Mejías et al., 2014). Each CTPR unit is composed of two helices; A and B, that are genetically fused and interact with each other to form a A‐B intra‐repeat interface. Furthermore, CTPR proteins present a second inter‐repeat packing interface composed mostly of the interaction between the helix B, and the helix A′ of the following repeat to form a B‐A′ inter‐repeat interface (Figure 1a). These two interfaces can be further employed to drive oligomerization between CTPR units. We designed two novel CTPR‐based modules with orthogonal unsatisfied interfaces (intra‐ and inter‐), whose sequences are A‐B‐A‐B‐A for SCAB1 and B‐A‐B‐A‐B for SCAB2 (Figure 1a), based on interfaces observed in the crystal structure of a long CTPR superhelix (CTPR8, PDB ID: 2HYZ (Kajander et al., 2007). To ensure stability, each module comprises two and a half repeats. Consecutive alternating SCAB1 and SCAB2 modules, we hypothesize their assemble into linear arrays to recapitulate the CTPR super helical extended structure (Figure 1a). To stabilize that assembly, we introduced a unique cysteine residue at both the C‐ and N‐terminal ends of the SCAB1 and SCAB2 modules, giving rise to SCAB1_C and SCAB2_C, respectively. Thus, they are able to bind each other through an inducible covalent stapling based on the formation of a disulfide bond at the A‐B′ inter‐molecular interface (Figure 1b).

FIGURE 1.

Schematic representation of the two SCAffolding Bricks (SCAB) systems designed for their selective and orthogonal assembly through disulfide bonds or metal coordination. (a) A CTPR unit is comprised of two helices (A‐B). CTPR proteins comprise of arrays of CTPR units can be splitted in two CTPR modules (SCAB1: A‐B‐A‐B‐A and SCAB2: B‐A‐B‐A‐B) with unsatisfied orthogonal interfaces (intra‐ and inter‐) that will drive directional assembly based on intrinsic head‐to‐tail interactions. (b) SCAB1_C and SCAB2_C modules in which two cysteines have been introduced at C‐ and N‐ terminal positions, respectively. Scheme of SCAB1_C and SCAB2_C assembly promoted by the formation of a disulfide bond (shown in red), resulting in SCAB_C. (c) The engineered His module (H) display four histidines for metal coordination at positions 2, 6, 9 and 13 positions (shown in cyan). SCAB1_H and SCAB2_H modules comprise three repeats, which combine two WT modules (pink and light pink) and a H module (cyan) at the C‐terminal or N‐terminal end repeat, respectively. Scheme of the assembly of SCAB1_H and SCAB2_H promoted by copper‐driven interaction through the histidine residues, resulting in SCAB_H. The 3D structures of the SCAB modules have been modeled based on the crystal structure of the CTPR8‐WT (PDB ID: 2HYZ) (Kajander et al., 2007).

In our second approach, we applied a similar rationale centered on head‐to‐tail interactions to design a metal‐driven assembly within CTPR modules. This strategy aims to stabilize the previously mentioned head‐to‐tail inter‐molecular interactions, as previously used to drive complex protein assemblies (Salgado et al., 2010). The formation of such a metal‐driven staple, which brings together two SCABs modules, relies on the strength and selectivity of metal‐protein interactions (Salgado et al., 2009). For the design of two new SCAB modules, we engineered two metal‐coordination bi‐histidine sites within SCAB modules with three CTPR repeats (A‐B‐A‐B‐A‐B sequence). Using previous design principles for metal coordination (Aires et al., 2019; Salgado et al., ^{2009, 2010}), four coordinating histidines (His) were incorporated at positions 2, 6, 9, and 13 of the A helix in a CTPR domain (His module, A_H helix) (Figure 1c). The side chain conformations and backbone geometry of the His were modeled and found to be compatible with the CTPR structure (see Section 4). To enhance scaffold stability, modules comprising three repeats were generated by combining the newly engineered His module A_H with wild‐type CTPR units to obtain SCAB1_H (A‐B‐A‐B‐A_H‐B) and SCAB2_H (A_H‐B‐A‐B‐A‐B) modules with the metal coordination interface at the C‐ or N‐ terminal ends, respectively. The inter‐repeat coordination interface that emerged upon the metal‐directed interaction of the SCAB1_H and SCAB2_H modules was simulated by CheckMyMetal online server (Zheng et al., 2014, 2017) (Figure S1), and the metal‐coordination distances were validated computationally with the aforementioned server to ensure compatibility with the sites described for metal‐coordinating proteins. The resulting inter‐repeat interface comprises two tetra‐histidine affinity sites that will link the two SCAB_H modules in presence of divalent metals.

These newly designed sets of SCAB pair modules, SCAB1_C and SCAB2_C, and SCAB1_H and SCAB2_H (Tables S1 and S2) were expressed, purified, and characterized. All SCAB modules are expressed as his‐tag fusions, and his‐tag are cleaved or not according to the experimental needs, making them suitable for their selective purification or solid‐phase assembly on agarose‐based supports. The purity and structural integrity of the modules were validated using SDS‐PAGE gel electrophoresis, MALDI‐TOF spectra, and circular dichroism spectra (Figures S2–S4). Next the assembly of the two pairs of SCAB modules was evaluated through the two distinct assembly procedures.

The assembly of SCAB_C was performed stepwise in solid phase using a Ni‐NTA affinity resin. The aim was to develop a methodology to achieve the assembled systems already immobilized and to avoid non‐specific polymerization of modules with two cysteines in potential future applications of multi‐modular assemblies with more than two components. First, SCAB1_C was immobilized on the resin through its His‐tag, and then an excess of SCAB2_C without His‐tag was added and incubated overnight at 37°C. After washing the excess of SCAB2_C, the final assembly was eluted and analyzed (Figure 2a,b). Electrophoresis analysis under non‐reducing conditions showed the presence of SCAB_C dimer and the disassembly after reduction of the disulfide bond with dithiothreitol (DTT) (Figure S2a). MALDI‐TOF spectra demonstrate that the size of the SCAB_C assembly, determined by spectra is 25.7 kDa, in agreement with the expected mass of the dimer, and the stoichiometry of the assembly is 1:1 (SCAB1_C:SCAB2_C), thus validating the proposed scaffolding strategy assembled in solid‐phase. After addition of DTT, we demonstrated the reversibility of assembly as both SCAB1_C and SCAB2_C monomers (13.5 and 12.1 kDa) were also detected by mass spectrometry (Figures S2a and S3a). Circular dichroism analysis revealed that the α‐helical secondary structure of the SCAB_C is preserved within the assembly, and the system presents similar stability to the individual components as determined by the thermal melting temperature (T _m (SCAB1_C) = 59.03°C and T _m (SCAB2_C) = 63.03°C) (Figure S4a). Finally, the SCAB_C assembly was analyzed by size exclusion chromatography (SEC). As expected, this analysis revealed a significant increase in the elution peak volume for SCAB_C assembly (11.15 mL) compared to SCAB_C modules separately (12.89 mL and 13.09 mL, specifically for SCAB1_C and SCAB2_C), illustrating the orthogonal assembly of the SCAB_C modules (Figure 2b).

FIGURE 2.

Assembly SCAffolding Bricks SCAB_C and SCAB_H. (a) Schematic representation of SCAB_C assembly‐based scaffold composed by SCAB1_C (in gray) and SCAB2_C (in green), resulting in the SCAB_C assembled system. (b) SEC chromatograms through a Superdex S75 (10/300) GL column monitored at 280 nm wavelength for the SCAB_C modules (SCAB1_C and SCAB2_C), and SCAB_C assembled system. (c) Schematic representation of the SCAB_H assembly‐based scaffold composed by SCAB1_H (in bright pink and His module in cyan) and SCAB2_H (in light pink and His module in cyan), resulting in the SCAB_H assembled system. d) SEC chromatograms through a Superdex S75 (10/300) GL column monitored at 280 nm wavelength by SEC for the SCAB_H modules (SCAB1_H and SCAB2_H), and SCAB_H assembled system.

The metal‐driven assembly of SCAB1_H and SCAB2_H was carried out in solution using different metal salts, including CuSO₄, NiSO₄, and CoSO₄. This assembly was not performed in solid phase since the design SCAB_H modules, in principle, should not be able to self‐polymerize, and the presence of engineered His in both modules will interfere with the selective immobilization of the priming SCAB unit (SCAB1_H) through its N‐terminus fused His‐tag. Fifty micromolars of untagged SCAB1_H and SCAB2_H modules were incubated in solution for 1 h at 50°C with two equivalents of the metal salt per protein to form the metal‐driven biomolecular assembly (Figure 2c). SEC analysis showed that the largest shift in size occurred in the presence of the Cu²⁺ metal ion (Figure S5). Therefore, copper was selected as coordination metal for the assembly of SCAB_H using a 1:1 molar ratio of SCAB_H modules at 50 μM protein concentration, with two equivalents of Cu, and incubated for 1 h at 50°C. The SCAB_H assembly was purified by SEC, which showed a clear shift in the elution peak volume (8.92 mL) when compared with the individual SCAB_H modules (10.87 mL and 11.37 mL, respectively for SCAB1_H and SCAB2_H), indicating the assembly of the SCAB_H modules (Figure 2d). However, in SDS‐electrophoresis gel, the band corresponding to the assembled system was not detected, probably because the metal was diluted, leading to the disassembly of the complex (Figure S2b). Finally, by inductively coupled plasma mass spectrometry (ICP‐MS) we quantified 1.88 ± 0.94 Cu atoms per SCAB_H assembly (Figure S6), in accordance with the two metal‐binding sites introduced in each SCAB_H.

2.2. Assembly of multi‐enzyme complexes through engineered SCAB bio‐bricks

Once we confirmed the effective orthogonal assembly of SCAB_C and SCAB_H bricks, they were used as scaffolding platforms to form multi‐enzyme systems arrays with nanometric precision. To this end, we selected a two‐enzyme system comprised of NADH‐dependent L‐alanine dehydrogenase from Bacillus stearothermophilus (AlaDH) and NADH cofactor regenerating formate dehydrogenase from Candida boidinii (FDH) that perform the biosynthesis of L‐alanine with in‐situ NADH recycling. For the assembly of the bi‐enzymatic systems, the previously validated SCAB modules were fused to the corresponding enzymes. AlaDH, a hexameric protein, was fused to the C‐terminus of SCAB2_C and SCAB2_H, whereas FDH, a dimeric protein, was fused to the C‐terminus of SCAB1_C and to the N‐terminus of SCAB1_H, since FDH fusion at the C‐terminus of SCAB1_H interfered with the metal coordination. This steric hindrance seems to have a considerably lower influence on cysteine stapling. In silico predictions of the inter‐enzyme distance based on the 3D model of the inter‐repeated stapled CTPRs domains suggest that these two enzymes were separated by approximately 3.2 nm and 5.3 nm when assembled into the SCAB_C (Figure 3a) and SCAB_H (Figure 3c), respectively. Thus, we compared the assembly efficiency of the two‐scaffold architectures to further study their effect on the performance of the model enzyme cascade mentioned above. The resulting enzymes variants fused with the SCAB modules were named as AlaDH@SCAB2_C, AlaDH@SCAB2_H, FDH@SCAB1_C, and FDH@SCAB1_H (Tables S1 and S2). These SCAB‐enzyme fusions were purified, and their purity and masses were verified by SDS‐PAGE gel (Figures S2c,d) and mass spectrometry (Figure S3b,c). The observed masses matched with the calculated mass from their corresponding amino acid sequence. Furthermore, circular dichroism analysis revealed that the secondary structures of the SCAB and the enzymes were negligibly affected by their direct fusion compared to individual SCAB modules and enzymes, which were also stable and showed similar thermal denaturation transitions (Figure S4c,d).

FIGURE 3.

Schematic representation and characterization of FDH/AlaDH@SCAB_C and FDH/AlaDH@SCAB_H assemblies. (a) Schematic representation of SCAB_C assembly composed of SCAB1_C (in gray) and SCAB2_C (in green), and their corresponding fused enzymes in their monomeric structural state, FDH (in turquoise, PDB ID: 5DNA), and AlaDH (in purple, PDB ID: 1PJB). The resulting FDH/AlaDH@SCAB_C assembled system, based on head‐to‐tail‐interactions, is shown, and the distance between the enzyme anchoring points on SCAB_C scaffolds is approximately 3.2 nm. (b) SEC chromatograms through a Superdex S200 (10/300) GL column monitored at 280 nm wavelength by SEC for the SCAB_C‐enzyme fusions (FDH@SCAB1_C and AlaDH@SCAB2_C), and the assembled FDH/AlaDH@SCAB_C system. (c) Schematic representation of SCAB_H assembly composed of SCAB1_H (WT modules in bright pink, and His module in cyan) and SCAB2_H (WT modules in light pink, and His module in cyan), and their corresponding fused enzymes in their monomeric structural state, FDH (in turquoise, PDB ID: 5DNA), and AlaDH (in purple, PDB ID: 1PJB). The resulting FDH/AlaDH@SCAB_H assembled system, based on metal‐driven assembly, is shown, and the distance between the enzyme anchoring points on SCAB_C scaffolds is approximately 5.3 nm. (d) SEC chromatograms through a Superdex S200 (10/300) GL column monitored at 280 nm wavelength for the SCAB_H‐enzyme fusions (FDH@SCAB1_H and AlaDH@SCAB2_H), and the assembled FDH/AlaDH@SCAB_H system.

Once the SCAB‐enzyme fusions were characterized, their assembly was carried out following the two strategies described above for the SCAB modules. The assembly of FDH/AlaDH@SCAB_C was performed as described for SCAB_C. However, in this scenario, the large size of the fused enzymes hinders the formation of the SCAB‐driven assembly when the priming unit is already immobilized. Therefore, it appears that the fusion of the enzymes to SCAB_C induces steric hindrances, requiring all the modules to be in solution for their successfully assembly. Thereby, SCAB_C‐enzyme fusions (FDH@SCAB1_C and AlaDH@SCAB2_C) were first incubated in solution at 50 μM protein concentration at 4°C overnight. Additionally, a lower assembly temperature was implemented to preserve enzyme activity and slow down the assembly process of the fusion proteins. Then the FDH/AlaDH@SCAB_C assembly was immobilized and purified through affinity Ni‐NTA resin by His‐tag FDH@SCAB1_C and further purified by SEC. SEC analysis showed a notable increase in the elution peak volume of the assembly, which eluted at the exclusion volume (~8 mL), indicating a large MW assembly (>600 kDa) when compared to the well‐resolved individual SCAB_C‐enzyme fusions (10.81 and 13.11 mL, for FDH@SCAB1_C and AlaDH@SCAB2_C, respectively) (Figure 3b). The assembly was analyzed by SDS‐PAGE gel and MALDI‐TOF mass spectrometry (Figures S2c and S3b). Gel electrophoresis did not show a clear band for the FDH/AlaDH@SCAB_C assembly, with a band above 250 kDa, probably due to its large molecular weight already observed by SEC. However, two bands corresponding to the FDH@SCAB1_C and AlaDH@SCAB2_C monomers could be detected after addition of DTT to the assembled fraction (Figures S2c and S3b) by both gel electrophoresis and mass spectrometry, which confirms the disulfide stapling of the SCAB_c‐based scaffolds, enabling the complexation of the multi‐enzyme system. Circular dichroism spectra showed that the structure of the FDH/AlaDH@SCAB_C was not significantly affected when compared with the individual elements (Figure S4b, left). The thermal denaturation curves show that all the individual elements as well as the assembly are stable and well folded since they show a coorperative denaturation curve (Figure S4b, right). Finally, dynamic light scattering (DLS) measurements revealed a hydrodynamic diameter for FDH/AlaDH@SCAB_C assembly of 199.2 ± 21.9 nm, in agreement with SEC and SDS‐PAGE analysis (Figure S7). These large assemblies may be attributed to the oligomeric state of the enzymes, as AlaDH and FDH are hexamers and dimers, respectively. We hypothesize that the oligomeric states are assembled instead of the monomers, as the enzymes need to be oligomeric to preserve their activity. These oligomers may nucleate larger supramolecular assemblies beyond the SCAB‐based scaffolding favored by the covalent nature of the disulfide bond that may shift the assembly equilibrium towards higher oligomeric orders.

In parallel, we conducted the metal‐driven assembly of FDH and AlaDH using the SCAB_H pair described above for the standalone SCAB_H modules (50 μM protein concentration of each SCAB_H‐enzyme fusions and 2 equivalents of CuSO₄ per protein concentration). In this case, a temperature of 30°C was selected as it allows the histidine‐cooper coordination without adversely affecting the stability and activity of the assembled enzymes (Figure 3c). Under these conditions, we obtained FDH/AlaDH@SCAB_H metal‐driven assembly, which was purified by SEC. The SEC‐purified FDH/AlaDH@SCAB_H assembly showed a significant change in elution peak volume (8.55 mL) compared to the individual SCAB_H‐enzyme fusions (10.89 and 9.42 mL, for FDH@SCAB1_H and AlaDH@SCAB2_H, respectively), demonstrating the high efficiency of the orthogonal assembly of the SCAB_H‐enzyme fusions (Figure 3d). In contrast to the large assemblies observed for the SCAB_C‐based assemblies, this assembly eluted within the resolution volume, and an approximate size of 522 kDa extrapolated from the MW calibration curve was estimated (Figure S7). DLS measurements confirmed this smaller assembly size for the metal‐driven non‐covalent assembly with a hydrodynamic diameter of 11.5 ± 0.8 nm for the FDH/AlaDH@SCAB_H assembly (Figure S8). Finally, ICP‐MS analysis confirmed the expected stoichiometry with 1.98 ± 0.24 Cu atoms per FDH/AlaDH@SCAB_H assembly (Figure S6).

2.3. Catalytic activity, and biosynthesis of L‐alanine of scaffolded multi‐enzyme complexes

Next, we evaluated the effect of the assembly on the activity of the scaffolded enzymes. For this purpose, we measured the enzymatic activity of both free and assembled enzymes (Figure S9) using UV–vis based assays (see Materials and Methods) and determined the Michaelis–Menten curves (Figure S10) and parameters (Table S3 and Figure 4).

FIGURE 4.

Michaelis–Menten steady‐state kinetic parameter (catalytic efficiency, k _cat/k _M ratio) of free and scaffolded enzymes. (a) k _cat/k _M ratio data for FDH in the free SCAB‐enzyme fusions, FDH@SCAB1_C and FDH@SCAB1_H, and in the scaffolded systems, FDH/AlaDH@SCAB_C and FDH/AlaDH@SCAB_H, towards formate and NAD⁺ as distinct substrates. (b) k _cat/k _M ratio data for AlaDH in SCAB‐enzyme fusions free, AlaDH@SCAB2_C and in AlaDH@SCAB2_H, and SCAB‐enzyme fusions scaffolded, FDH/AlaDH@SCAB_C and FDH/AlaDH@SCAB_H, towards pyruvate and NADH as distinct substrates.

In the case of FDH, its fusion to SCAB_H presented a much lower catalytic efficiency than its fusion to SCAB_C regardless of if this enzyme was scaffolded or not. Such low catalytic efficiency is due to the high K _M values of this FDH@SCAB_H1 towards formate. In contrast, the catalytic efficiencies of the AlaDH fusions to SCABs were negligibly affected. Nonetheless, the two enzymes fused to SCAB modules exhibited sufficient catalytic activity to evaluate the impact of scaffolding on their catalytic performance. When these chimeras were scaffolded together with FDH in both scaffold architectures, the AlaDH suffered a dramatic decrease of the catalytic efficiency towards pyruvate. In the case of AlaDH scaffolded with FDH on SCAB_C, this drop in catalytic efficiency is mainly due to the high K _M value of the scaffolded AlaDH towards pyruvate. Although there is not a clear trend on the effect of the molecular assembly on the kinetic parameters of both enzymes, we suggest that AlaDH is negatively affected by its assembly along FDH rather than by the scaffold architecture, while the fusion with SCAB_H negatively affects FDH to a higher extent than the assembly itself.

Finally, we assessed the performance of the biosynthetic cascade when the enzymes were either scaffolded or free. As controls, we used in all cases the enzyme fused to the corresponding SCAB unit. We used 1 equivalent of pyruvate, 1.3 equivalents of sodium formate, 6.6 equivalents of ammonium chloride and sub‐stoichiometric amounts of NADH (150 times less than pyruvate) to synthesize L‐alanine (Figure 5a). Figure 5b shows that the scaffolded FDH/AlaDH@SCAB_C multi‐enzyme system is 3.6 times faster than its free counterpart, achieving a specific productivity (g_product g_enzyme ⁻¹ h⁻¹) of 1.05 g_product g_enzyme ⁻¹ h⁻¹ and a titer of 9 mM of L‐alanine after 24 h, corresponding to a chromatographic yield (CY) of 12%. Meanwhile, under the same conditions and incubation period, the non‐assembled system (free FDH@SCAB1_C and AlaDH@SCAB2_C) achieves 2.6 times lower CY. Contrastingly, the scaffolded FDH/AlaDH@SCAB_H multi‐enzyme system displays the same catalytic performance as the free one, attaining a specific productivity of 0.17 g_product g_enzyme ⁻¹ h⁻¹ and a titer of 3.23 mM of L‐alanine after 24 h, result 4.3% CY. Under the same conditions, free FDH@SCAB1_H and AlaDH@SCAB2_H produced a CY of just 4.6% (Figure 5c).

FIGURE 5.

Catalytic performance of SCAB scaffolded enzymes for L‐alanine synthesis in batch‐mode. (a) Scheme of the catalytic cycle as tested by HPLC. (b) L‐alanine yields at 1:1 enzyme ratio of FDH:AlaDH for the free SCAB_C enzymes (FDH@SCAB1_C and AlaDH@SCAB2_C) and for the scaffolded SCAB_C‐enzymes (FDH/AlaDH@SCAB_C). (c) L‐alanine yields at 1:1 enzyme ratio of FDH:AlaDH for the free SCAB_H‐enzymes (FDH@SCAB1_H and AlaDH@SCAB2_H) and for the scaffolded SCAB_H‐enzymes (FDH/AlaDH@SCAB_H). Reaction mixture: 100 mM formate, 75 mM pyruvate, 500 mM ammonium chloride, and 0.5 mM NADH. (d) Schematic representation of potential supramolecular assemblies of the scaffolded enzymes on SCAB_C protein scaffolds driven the oligomeric state of AlaDH and FDH enzymes.

The volumetric productivity and the yield determined for the L‐alanine biosynthesis indicate that only the scaffolding of the multi‐enzyme system in SCAB_C exerts a positive effect on the overall throughput of the enzyme cascade when compared to the effect of the scaffolding on the SCAB_H units. Given that FDH is the rate‐limiting enzyme in this cascade (Figure 4), we hypothesize that the transport of NAD⁺/NADH between AlaDH and FDH when they are covalently scaffolded through SCAB_C is more efficient than when they are assembled into SCAB_H. The enhanced NADH recycling efficiency when these two enzymes have been scaffolded through protein–protein interaction has been previously reported for other biomolecular designs (Ledesma‐Fernandez et al., 2023; Bauler et al., ²⁰¹⁰). This increase in the cascade productivity is difficult to explain by the apparent catalytic efficiency of each scaffolded enzymes. In particular, the apparent K _M of FDH (rate‐limiting enzyme) assembled into SCAB_C towards NAD⁺ (intermediate) is 2.6‐fold higher than that of the same enzyme assemble into SCAB_H (Table S3). Therefore, the observed enhancement in the cascade performance when catalyzed by the system assembly into SCAB_C is not supported by the apparent kinetic parameters determined for each enzyme under steady‐state conditions. Hence, we suggest that geometrical and clustering effects should explain the productivity differences found between the two scaffolds. First, we observe that FDH and AlaDH were spatially arranged in a closer distance when scaffolded in SCAB_C than in SCAB_H. However, this difference in inter‐enzyme distance (3.2 nm vs. 5.3 nm) is too small to explain any kinetic benefit due to close‐proximity channeling. Indeed, theoretical studies state that intermediate concentration gradients are not observed at inter‐enzyme distance lower than 1–5 μm under non‐restricted diffusion conditions (Ellis et al., 2019; Wheeldon et al., ²⁰¹⁶). Another plausible explanation is the formation enzyme aggregates of different sizes, as observed in the characterization of the assemblies by SEC and gel electrophoresis. For SCAB_C‐based assemblies, large aggregates were detected by SDS‐PAGE, SEC and DLS, and the absence of a peak in MALDI‐TOF analysis also supports the formation of protein aggregates of large molecular weight. In this scenario, when FDH releases NADH, this cofactor will encounter AlaDH in its way heading towards the bulk solution. This type of intermediate channeling is probabilistic and depends more on the size of the aggregate and their concentration in the bulk than on the inter‐enzyme distance within the scaffold. It has been theoretically reported that the cascade productivity increases 6 times compared to the non‐aggregated and delocalized enzymes for a two‐enzyme scaffold forming aggregates of 260 nm with an optimal separation between aggregates of 6.5 μm. In our case, the productivity enhancement was the half of that theoretically proposed. However, it still suggests that SCAB_C aggregates of about 200 nm, due to oligomeric nature of the assembled enzymes, are responsible for the enhancement of the overall cascade throughput (Figure 5d). This conclusion is supported by previous reports indicating that enzyme clustering benefits from colocalizing multiple enzyme molecules that process an intermediate with multiple molecules of the enzyme producing it. Thus, enzyme clustering increases the rate of intermediate processing, thereby enhancing the efficiency of enzymatic pathways (Castellana et al., 2014; Breger et al., ²⁰²³). In this context, the differences in the catalytic productivity between the two scaffolding strategies could be attributed mainly to the different supramolecular assembly states. In fact, FDH/AlaDH@SCAB_H showed discrete assemblies (about 11 nm) when compared with FDH/AlaDH@SCAB_C system (Figure S8). The formation of larger assemblies for FDH/AlaDH@SCAB_C may be originated by the covalent nature of the disulfide staple that shifts the equilibrium towards higher ordered assemblies as illustrated in Figure 5d. In contrast, FDH/AlaDH@SCAB_H assemblies are mediated by weaker non‐covalent interactions dominated by the equilibrium with a K _D in the order of μM for copper:bi‐histidine interactions (Pirro et al., 2023), which, as shown experimentally results in discrete assemblies.

3. CONCLUSIONS

In conclusion, we have developed a methodology for enzyme assembly on newly engineered scaffolds based on tetratricopeptide consensus repeat (CTPR) proteins, named as SCAB modules. In this study, we present two modular methodologies that involve selective head‐to‐tail interactions and orthogonal stapling strategies, based on covalent disulfide bonds and metal‐driven assembly, to generate ordered assemblies. We have successfully developed and validated two sets of scaffolding units with control and reversible assembly triggers, that is, non‐reducing conditions and the presence of metal, and demonstrated their efficient assembly.

These scaffolding units have been demonstrated as useful tools for the development of functional scaffolded multi‐enzyme systems. Through the two engineered scaffolding strategies, we constructed a two‐enzyme cascade comprised of AlaDH and FDH enzymes. As observed, both enzymes retained their activity within the scaffolded assemblies, with a significant enhancement of the cascade throughput when they were assembled on the SCAB_C units, resulting in a 3.6‐fold increase in specific productivity. This improvement upon enzyme scaffolding is notable compared to previously reported examples. In a comparable cascade to the one presented in this study, Zhang et al. reported a two‐fold improvement in the cascade throughput when two alcohol dehydrogenases were scaffolded into EutM self‐assembly protein (Zhang et al., 2018). Additionally, we examined the same cascade organized in a different scaffold, which included further interactions between the scaffold and the reaction intermediates, and observed only a slightly higher fold increase, 4.25‐fold (Ledesma‐Fernandez et al., 2023). Thus, the system presented herein falls in the current state of the art. This catalytic improvement can be attributed to the proximity of scaffolded enzymes within the SCAB_C units and primarily to the size of the supramolecular assemblies driven by the multimeric state of the scaffolded enzymes, and the covalent nature of the disulfide‐bond‐driven assembly. We suggest that the dominant effect of the supramolecular assembly may explain the major differences between the results obtained using SCAB_C and SCAB_H scaffolds. In fact, the SCAB_H scaffolding strategy did not increase the catalytic efficiency and lacks the ability to form such supramolecular aggregates. This effect is probably due to the non‐covalent equilibrium‐driven interactions between these SCAB_H units, when compared with the covalent disulfide‐based assemblies.

When large supramolecular enzyme aggregates are formed, there is a cofactor channeling effect mediated by enzyme clustering. In this process, the NADH cofactor encounters AlaDH after being released from FDH, leading to a more efficient reuse of the NADH cofactor compared to the free enzyme system. This process promotes the regeneration of NADH cofactor by efficiently transferring the byproduct from one enzyme to a neighboring enzyme cascade, thereby enhancing the catalytic efficiency. In this case, this effect is determined not by the distances between enzymes but by the size of enzyme assemblies and their concentration in the bulk. Proximity channeling can also be achieved by clustering enzymes fused to protein fibrils or immobilized on nanoparticles. Similar findings were reported by Berger et all using a seven‐enzyme system immobilized on nanoparticles (CdS quantum dots). The oligomeric nature of the enzymes promoted the nanoparticle aggregation provoking a proximity channeling effect that enhances the cascade throughput of complex metabolic pathways (i.e., glucolysis) in vitro (Breger et al., 2023). Multi‐enzyme clusters can also be genetically programmed using dock‐and‐lock peptides (Kang et al., 2019), or curli proteins (Chen et al., 2023). While the dock‐and‐lock peptides forms aggregates of roughly 20 nm, the curli protein does fibers of dozens of microns. In both cases, the enzymatic cascade yield increased by the same order of magnitude as that observed in this work using the SCAB_C units, which result in aggregates of 200 nm. Thus, cascade performance has been improved in all the cases despite the size of enzyme aggregates vary between the different assembly approaches.

When conducting a thorough comparison between the SCAB‐based scaffolding systems presented here and the previously reported TRAP‐based scaffolding (Ledesma‐Fernandez et al., 2023), we highlight different aspects that reveal the suitability of each system for particular applications. Firstly, TRAP scaffolding limits the number of enzymes that can be scaffolded to the number of orthogonal TRAP systems developed, currently at three. This approach would need additional protein engineering of the TRAPs to scaffold larger systems. Conversely, SCABs offer the advantage of assembling more complex cascades involving a greater number of enzymes by fusing different enzymes to the already designed SCAB modules and conducting sequential assembly. In terms of enzyme modification, TRAP‐scaffolding involves fusion to short tag peptides, resulting in minimal enzyme modification with fewer potential deleterious effects on enzyme activity compared to the fusion to SCAB modules. Regarding spatial assembly control, both systems offer similar nanometric spatial control of assembled enzymes, when the supramolecular assembly is not considered. SCAB‐based assembly, encoded by the formation of disulfide bonds, and metal‐driven coordination in both cases allow for responsive and reversible assembly/de‐assembly controlled by the reducing state of the environment and the presence of metal ions, respectively. Implementing responsive assembly in TRAP‐based assemblies would be more challenging and may require the use of expensive peptides as competitors of the enzyme‐scaffold interaction. Regarding catalysis, while TRAP‐based scaffolding enables diffusion‐limited channeling, SCABs can achieve proximity channeling. TRAP‐based scaffolding may encounter greater steric hindrances when assembling enzymes of high molecular weight with subunits formed by large polypeptides, as the peptide–TRAP interaction imposes spatial constraints on assembled enzymes. TRAPs can be heterogenized through immobilization techniques, whereas heterogenization in porous microparticles is more challenging with SCABs due to the large size of the supramolecular assemblies achieved. This latter aspect makes SCABs less versatile for integration into different reactor designs (e.g., stirred batch reactor or packed‐bed reactor).

This work also illustrates the potential of using different engineered scaffolds to encode different properties of the scaffolded enzymatic systems, such the inter‐enzyme distance and the size of the assemblies, parameters that directly impact the catalytic productivity of the scaffolded cascades. The protein‐based scaffolds outlined in the present work possess the capability to facilitate the development of more efficient multi‐enzyme systems, ensuring that scaffolded enzymes are closely positioned. Besides, its application could be extended to other fields where the arrangement of organized complexes is required to achieve efficient functional systems, beyond the obvious applications in the field of biocatalysis.

4. METHODS

4.1. Protein design, cloning and molecular biology

The 3D structures of the designed SCAB modules were initially modeled using PyMol and Swiss‐model tools to assess the models and potential effects on the structure and potential clashes of the introduced mutations. SCAB modules were modeled based on the structure of the CTPR wild type structure (PDB ID: 2HYZ). Subsequently, after energy minimization of the models using Chimera, the generated structures were evaluated using MOLProbity (Davis et al., 2007). The model structures obtained for the variants confirmed their compatibility with the preservation of the CTPR structure, as expected, since they are in non‐conserved positions of the CTPR consensus sequence.

For the assembly of the SCAB_C modules, the wild‐type (WT) CTPR3 (CTPR protein with three repetitions) gene cloned into the pProEx‐HTa vector was used as a starting point. This DNA encodes a CTPR with three identical repeat modules. A protocol based on the overlapping PCR technique was used to incorporate a coordinating cysteine at the C‐terminal end into a CTPR3‐WT gene and SCAB1_C was created. The same process was used for the creation of SCAB2_C. The only difference is that the coordinating cysteine was introduced at the N‐terminal end. Then, the amplified fragments encoding the SCAB_C modules were cloned into the pProEx‐HTa vector using the BamHI and HindIII restriction sites. Finally, the cysteine insertion in SCAB1_C and SCAB2_C was verified by sequencing.

For the assembly of the SCAB_H modules, single point mutations were made in the CTPR1 WT gene cloned in the pProEx‐HTa vector. A sequential protocol based on rapid site‐directed mutagenesis was used to incorporate four coordinating histidines into the CTPR1 WT gene and created CTPR1‐4His. Within the 34 amino acids that the CTPR1 WT module is comprised of, mutations were introduced that included the substitution of CTPR1 WT residues for the coordinating histidines at positions E2H, N6H, N9H, and K13H. Three rapid targeted mutagenesis protocols were performed to introduce the required mutations. In the first, the E2H mutation was introduced. In the second, the N6H and N9H mutations were introduced. In the latter, the K13H mutation was introduced. Then, amplified fragment encoding CTPR1‐4His was cloned into pProEx‐HTa vector using BamHI and HindIII restriction sites. Finally, after the cloning process, the clones were sequenced to verify that the mutations had been introduced correctly. Histidines have been introduced to assemble two proteins by means of coordination between histidines and metal ions. After verifying the mutations in CTPR1‐4His module, a modular cloning strategy, based on the restriction enzymes BamHI and BglII that generate compatible sticky ends (Speltz et al., 2015), was used to construct two modules (SCAB1_H and SCAB2_H) with three repeats by combining the mutated CTPR1‐4His with the coordination residues and the unmutated CTPR2 WT unit. SCAB1_H gene was created by ligation of the mutated fragment of CTPR1‐4His with the coordination histidines into the pProEx‐HTa vector, which already contained the wild type of fragment of a CTPR2 gene only digested with BamHI. SCAB2_H gene was created by ligation of the wild type of fragment of a CTPR2 WT into the pProEx‐Hta vector, which already contained the mutated CTPR1‐4His gene with the coordination histidines only digested with BamHI. The insertion of the histidine‐mutated fragment in SCAB1_H and SCAB2_H was verified by sequencing.

Bioassays, an industrial partner of the HOMBIOCAT project, fused the SCAB_C and SCAB_H modules to FDH and AlaDH enzymes and cloned into pet‐28b (+) plasmid using our previously constructed plasmids containing each part of these SCAB‐enzyme fusions (FDH@SCAB1_C, AlaDH@SCAB2_C, FDH@SCAB1_H and AlaDH@SCAB2_H). To clone the enzymes into pet‐28b (+) the fragment encoding the enzyme was digested with HindIII/BamHI to be subsequently inserted into the scaffold plasmid with the SCAB modules which was digested with HindIII/BglII to make a fusion protein harboring the scaffold domain at the C‐terminus.

4.2. Protein expression and purification

SCAB modules and SCAB‐enzyme fusions were overexpressed in Escherichia coli C41 cells. An overnight saturated cell culture was grown to an OD600 = 0.6–0.7 at 37°C after being diluted in 1 L of LB. The following step was adding 0.6 mM IPTG, followed by overnight growth at 20°C, to induce overexpression. Proteins were purified as His‐tagged fusions using nickel nitrilotriacetic acid affinity chromatography (Ni‐NTA) in using standard protocols (Crowe et al., 1996). Subsequently the His‐tag was cleaved with Tobacco Etch Virus (TEV) protease. The proteins were dialyzed against buffer containing 150 mM NaCl, 50 mM sodium phosphate buffer pH 7.4 with membranes with cutoff of 10 kDa (SpectraPor®) and further purified by a second Ni‐NTA affinity chromatography purification to separate the TEV and His‐tag from the pure protein fraction. Based on the extinction coefficients computed from their amino acid composition and the absorbance at 280 nm, the protein concentration was measured. SCAB modules were stored at −20°C. SCAB‐enzyme fusions were stored at 4°C, to preserve their activity without requiring the presence of glycerol. Only SCAB1_C modules maintained the histidine tag to perform the assembly on solid support (Ni‐NTA resin).

4.3. MALDI‐TOF mass spectrometry

Using the Voyager‐DE PRO Biospectrometry Workstation mass spectrometer, MALDI‐TOF experiments were carried out. In PB buffer (Phosphate buffer 10 mM pH 7.4), the samples were prepared. Four microliters of matrix and 1 μL of protein were mixed. The matrix solution was made by combining sinapinic acid matrix with acetonitrile–water at a ratio of 50:50, 0.1% TFA (trifluoroacetic acid), and a final concentration of 10 mg·mL⁻¹ (Cohen & Chait, 1997).

4.4. Circular dichroism

Using a Jasco J‐815 spectrophotometer and PBS buffer (10 mM NaCl, 10 mM phosphate buffer, pH 7.4), circular dichroism (CD) (Greenfield, 2007; Miles & Wallace, ²⁰¹⁵) studies were carried out. From 190 to 260 nm, protein spectra were measured. Additionally, by capturing data every 1‐degree, thermal denaturation curves in the range of 15–100°C were obtained.

4.5. SCAB_C modules and SCAB_C ‐enzyme fusions assembly

The formation of orthogonal biobricks as intra‐repeat interactions was stabilized with stapled disulfide bonds based on SCAB_C modules (Cortajarena et al., 2010). The process was performed in the solid state using a Ni‐NTA resin and was based on the head‐tail‐interactions between the complementary scaffold modules SCAB1_C and SCAB2_C (Mejías et al., 2014). The first step was activation of the cysteine at the C‐terminus of the SCAB1_C module. This was done by adding 10 μL of 1 M dithiothreitol (DTT) (CAS number: 3483‐12‐3, Merck) into 1 mL volume of SCAB1_C module at 70 μM concentration in PBS buffer (150 mM NaCl, 50 mM sodium phosphate buffer pH 7.4) to a final activation concentration of 50 μM and incubating at room temperature for 20 min. Excess DTT was removed with the PD‐10 column. Then, in order to avoid self‐dimerization, incubation with 100 μL of excess aldrithiol (20 mM in DMSO) (CAS number: 2127‐03‐9, Merck) into 1 mL volume of SCAB1_C module at 50 μM concentration in PBS buffer (150 mM NaCl, 50 mM sodium phosphate buffer pH 7.4) to a final activation concentration of 20 μM was performed for 1 h at 40°C (Lin et al., 2006). Excess aldrithiol was removed using a PD‐10 column. Finally, 1 mL at 20 μM protein concentration in PBS buffer (150 mM NaCl, 50 mM sodium phosphate buffer pH 7.4) of His‐tagged SCAB1_C module was incubated with a Ni‐NTA resin for 1 h at room temperature to promote solid‐phase assembly. The cysteine at the N‐terminus of the SCAB2_C module (without His‐Tag) was activated with DTT with the same methodology and buffers mentioned above. Excess DTT was removed using the PD‐10 column. In the next step, 1 mL at 200 μM protein concentration in PBS buffer (150 mM NaCl, 50 mM sodium phosphate buffer pH 7.4) of the cleaned SCAB2_C module was incubated with the previously immobilized SCAB1_C module at a 1:10 SCAB1_C:SCAB2_C ratio to form on the “head‐to‐tail interaction” a disulfide covalent bond that clamps the interaction overnight at room temperature. After overnight incubation, excess non‐immobilized SCAB2_C module was cleaned up using PBS buffer (150 mM NaCl, 50 mM sodium phosphate buffer, pH 7.4). Finally, the assembly formed by SCAB_C was eluted in 1 mL in PBS buffer with imidazole (150 mM NaCl, 300 mM imidazole, 50 mM sodium phosphate buffer pH 7.4).

However, in the case of the SCAB_C‐enzyme fusions (FDH@SCAB1_C and AlaDH@SCAB2_C) the process was quite similar, but there were some differences. First, FDH@SCAB1_C and AlaDH@SCAB2_C modules were submitted to cysteines activation and clean‐up processes using the same methodology described for SCAB1_C and SCAB2_C modules. Next, the activated modules were incubated for disulfide bond formation in 0.5 mL with 50 μM protein concentration for each SCAB_C‐enzyme fusions in PBS buffer (150 mM NaCl, 50 mM sodium phosphate buffer pH 7.4) and in solution at 4°C overnight to promote assembly between these proteins of considerable dimensions. As the assembly process was done in solution, activation with aldrithiol was omitted. After overnight, 1 mL the assembly of the SCAB_C‐enzyme fusions at 25 μM final concentration was incubated 1 h at room temperature in Ni‐NTA resin in order to clean and purify the assembled proteins from the unassembled ones using PBS buffer (150 mM NaCl, 50 mM sodium phosphate buffer pH 7.4). Finally, the assembly formed with FDH/AlaDH@SCAB_C was eluted in 1 mL in PBS buffer with imidazole (150 mM NaCl, 300 mM imidazole, 50 mM sodium phosphate buffer pH 7.4).

4.6. SCAB_H modules and SCAB_H ‐enzyme fusions assembly

The formation of orthogonal biobricks as intra‐repeated interactions was stabilized with a coordination between the introduced histidines in the engineered SCAB_H modules with metals (López‐Andarias et al., 2018; Mejias et al., ²⁰¹⁹; Mejias et al., ²⁰²¹). The process was performed in solution and was based on the metal‐driven assembly between complementary SCAB1_H and SCAB2_H scaffolding modules with designed coordination histidines and copper metal. The first step was the incubation of 0.25 mL of each SCAB_H module at 100 μM concentration in Tris–HCl buffer (20 mM Tris–HCl buffer pH:7.4) and 100 μM copper (II) sulfate pentahydrate salt (CAS number: 7758‐99‐8, Merck) concentration for 1 h, at 50°C and 850 rpm. Then, the 0.5 mL of the assembly through metal driven assembly at final concentration of 50 μM of each SCAB_H module was cleaned and purified by FPLC gel filtration over a Superdex 75 (10/300) GL column using an ÄKTA Pure protein purification system. The chromatography was carried out in Tris–HCl buffer (20 mM Tris–HCl buffer pH 7.4) at a flow rate of 0.5 mL/min, and the chromatogram recorded at 280 nm.

In the case of the SCAB_H‐enzyme fusions (FDH@SCAB1_H and AlaDH@SCAB2_H) the synthesis process was quite similar, but there were some differences. 0.25 mL each SCAB_H‐enzyme fusion at 100 μM protein concentration in Tris–HCl buffer (20 mM Tris–HCl buffer pH 7.4) were incubated with 100 μM copper (II) sulfate pentahydrate salt (CAS number: 7758‐99‐8, Merck) concentration at 30°C to preserve enzyme activity. In the final reaction volume (0.5 mL) the concentration of each SCAB_H‐enzyme fusion was of 50 μM. Then, the assembly reaction was cleaned and purified by FPLC gel filtration over a Superdex 200 (10/300) GL column using an ÄKTA Pure protein purification system. The chromatography was carried out in Tris–HCl buffer (20 mM Tris–HCl buffer pH 7.4) at a flow rate of 0.5 mL/min, and the chromatogram recorded at 280 nm.

4.7. Size determination by size exclusion chromatography

The size increases of the SCAB_C, SCAB_H, FDH/AlaDH@SCAB_C and FDH/AlaDH@SCAB_H relative to unassembled elements were studied by size exclusion chromatography (SEC) using fast protein liquid chromatography (FPLC) (Crowe et al., 1996). In order to analyze the differences in the elution peaks volume of the assembled and non‐assembled elements, the SCAB modules and the SCAB‐enzyme fusions with cysteines and histidines were analyzed by SEC. Once the individual elements were analyzed, the increments were corroborated in relation to the elution peaks volume of the 4 assemblies obtained: SCAB_C, SCAB_H, FDH/AlaDH@SCAB_C and FDH/AlaDH@SCAB_H. Assembled and non‐assembled SCAB modules were injected into SEC on a Superdex 75 (10/300) GL size exclusion column, while assembled and non‐assembled SCAB‐enzyme fusions were injected into SEC on a Superdex 200 (10/300) GL size exclusion column. Runs were performed at 0.5 mL/min in PBS buffer (150 mM NaCl, 50 mM PB buffer pH 7.4) and Tris–HCl buffer (20 mM Tris–HCl buffer pH 7.4) for the cysteine and histidine protein variants, respectively, at 4°C. The volume used for all samples analyzed by SEC was 500 μL. The protein concentration of each SCAB_C modules and SCAB_C‐enzyme fusions was 25 μM, while the protein concentration of each SCAB_H modules and SCAB_H‐enzyme fusions was 50 μM. The measurements of the molecular weight of FDH/AlaDH@SCAB_H assembly were estimated based on calibration curves performed using an AKTA prime plus fast protein liquid chromatography (FPLC) equipment and commercial calibration kit (Gel Filtration Calibration Kit‐HMW from Cytiva).

4.8. Inductively coupled plasma mass spectrometry

One hundred microliters of each of the SEC‐purified SCAB_H and FDH/AlaDH@SCAB_H assemblies sets at a concentration of 13.07 and 11.65 μM respectively, were mixed with 300 μL of 37% HCl and the resulting suspension was sonicated for 45 min at 60°C. Finally, 2700 mL of distilled water was added (Aires et al., 2019). The Cu metal concentration of the assembled systems previously purified by SEC was determined by measuring the sample by ICP‐MS.

4.9. Dynamic light scattering

The hydrodynamic diameter of the FDH/AlaDH@SCAB_C and FDH/AlaDH@SCAB_H assembled systems were determined using the Zetasizer Ultra Red Malvern. The measurements were conducted at a temperature of 25°C, employing a 173° backscatter angle. The equilibration time for the system was 120 s. For each measurement, 100 μL of each assembly at 2.5 μM protein concentration were added in a ZEN0040 cuvette, with water as a dispersant. The ZS Xplorer software was utilized to monitor the hydrodynamic diameter measurements. For each sample, measurements were performed in triplicate (n = 3). The average of the hydrodynamic diameter values obtained from the triplicates was used for data representation.

4.10. Enzymatic activity assays

The enzyme activity of free enzymes, SCAB‐enzyme fusions, and scaffolded complexes were evaluated in solution by UV–vis spectrophotometry using Synergy H1 Hybrid Multi‐Mode Microplate Reader from BioTeK Instrument in 96‐well UV–vis transparent plates. For the AlaDH activity, the consumption of NADH was determined by the decrease in the absorbance at 340 nm under the following reaction conditions: 0.2 μM AlaDH, 0.5 mM NADH, 75 mM pyruvate, and 500 mM ammonium chloride in 25 mM potassium phosphate pH 8.0 at 30°C. For the FDH activity, the production of NADH was evaluated by the rise in absorbance at 340 nm under the following reaction conditions; 2 μM FDH,1 mM NAD⁺, 100 mM sodium formate in a 25 mM sodium phosphate buffer pH 7.0 at 30°C. In both enzyme assays, the reaction volume was 200 μL and the absorbance was monitored along the time to calculate the slope of absorbance as a function of time. To calculate the enzyme activity, we used the Lamber‐beer equation and a linear fit of the initial time points of each reaction. The NADH concentration was calculated for FDH and AlaDH using the molar extinction coefficient (Ɛ) Ɛ _{NADH‐340 nm} = 6200 M⁻¹ cm⁻¹. Under the specified conditions, one unit of FDH is defined as the quantity of enzyme required to make 1 μmol of NADH per minute. While one unit of AlaDH is defined as the quantity of enzyme required to utilize 1 μmol of NADH per minute. The specific activity (U mg⁻¹) was estimated by adjusting the activity units in the experiment to the enzyme concentration.

4.11. Enzyme kinetic parameters

Michaelis constant (K _M), maximum rate (V _max), turnover number (k _cat), and catalytic efficiency (k _cat/K _M) of SCAB‐enzyme fusions free and scaffolded were determined through the colorimetric enzyme assays described above. The K _M and V _max were calculated using a variety of concentration ranges for the cofactors NADH (0–0.66 mM) and NAD⁺ (0–8.33 mM), pyruvate (0.9–200 mM), formate (0–62.5 mM), and ammonium chloride (500 mM). The Michaelis–Menten equation was used to fit experimental data (Srinivasan, 2022):

$$V=\left(\frac{V_{\max }\times \left[\mathrm{S}\right]}{K_{\mathrm{M}}+\left[\mathrm{S}\right]}\right)$$

where V _max is the maximal reaction velocity, [S] is the substrate concentration, and K _M is the Michaelis–Menten constant.

The K _M and V _max kinetic parameters' stated errors were computed using the average of three repetitions. The following equation was used to get the turnover numbers:

$$k_{\mathrm{cat}}=V_{\max }/\left[\text{enzyme}\right]$$

The catalytic efficiency of the systems is determined by the ratio k _cat/K _M.

4.12. Biotransformation of L‐alanine

A biocatalyst molar ratio of 1:1 FDH:AlaDH and concentrations of each monomer of 0.28:0.28 μM was prepared in 1 mL of a reaction mixture composed by 75 mM pyruvate, 100 mM sodium formate, 500 mM ammonium chloride, and 0.5 mM NADH in nanopure water. Reactions were incubated with orbital agitation at 500 rpm and 25°C for 24 h. At different time points (1, 2, 4, 8, and 24 h), the reaction was stopped by ultrafiltration using Amicon Ultra‐0.5 Centrifugal Filter Units at 13,100 rfc for 30 min. The flowthrough was collected to analyze the L‐alanine concentration by HPLC (Velasco‐Lozano et al., 2018), prior derivatization with Marfey's reagent (Bhushan & Brückner, 2004). Briefly, 20 μL of reaction samples diluted 1:10 were combined with 15 mM Marfey's reagent (Cat. No. 48895, Thermo Scientific) in acetone, 8 μL of sodium 1 M bicarbonate, and 20 μL of the mixture was incubated for 1 h at 50°C and 400 rpm. After stopping the derivatized process with 8 μL of 2 M HCl, the mixture was centrifuged at 6700 rfc for 15 min. To conduct the HPLC analysis, the supernatant was further filtered. Derivatized samples were analyzed in an Agilent Technologies 1120 Compact LC using an EC‐C18 2.7 μm column (4.6 × 100 mm, Agilent) and the mobile phases A (0.1% TFA in water) and B (Acetonitrile) at a flow rate of 1 mL·min⁻¹. Analytes were detected at 340 nm and eluted using the gradient shown below: Starting at 90%–80% A from 0 to 17 min, the mobile phase was maintained at 60% from 20 to 30 min, recovered to baseline conditions 90% A in 1 minute, and sustained at 90% from 31 to 40 min. By using a calibration curve to suit the peak's size, the conversion degree of L‐alanine was estimated.

AUTHOR CONTRIBUTIONS

Alba Ledesma‐Fernandez: Investigation; methodology; writing – original draft. Susana Velasco‐Lozano: Investigation; writing – review and editing; methodology. Pedro Campos‐Muelas: Investigation; methodology. Ricardo Madrid: Funding acquisition; supervision; writing – review and editing. Fernando López‐Gallego: Project administration; resources; writing – review and editing; funding acquisition; conceptualization. Aitziber L. Cortajarena: Conceptualization; investigation; funding acquisition; writing – original draft; supervision; resources; project administration.

Supporting information

Appendix S1: Supporting information.

Ledesma‐Fernandez A, Velasco‐Lozano S, Campos‐Muelas P, Madrid R, López‐Gallego F, Cortajarena AL. Engineering bio‐brick protein scaffolds for organizing enzyme assemblies. Protein Science. 2024;33(5):e4984. 10.1002/pro.4984