Development of a Peptide-Mediated Multienzyme Assembly System in Bacillus licheniformis: Screening, Characterization, and Application in Dual-Enzyme Cascade Reaction

Abstract

As synthetic biology advances, prokaryotic microorganisms have become critical platforms for heterologous biosynthesis in cell factory applications. However, conventional free enzyme systems encounter substantial challenges, including inefficient intermediate transfer, toxic intermediate accumulation, and vulnerability to temperature and pH fluctuations. Enzyme complex catalytic systems offer promising solutions to these limitations. Bacillus licheniformis, a Generally Recognized as Safe (GRAS) host with exceptional protein secretion capacity, represents an ideal chassis for enzyme complex construction. This study developed a peptide-mediated platform in B. licheniformis to enable enzyme complex self-assembly and evaluated its effects on metabolic pathway performance. Five peptide elements were screened through fusion with enhanced orange/green fluorescent proteins (eOFP/eGFP) and transglutaminase (TGase). Effective peptide pairs were identified by measuring fluorescence intensity, visualizing complex formation via laser confocal microscopy, and assessing TGase activity. Subsequently, recombinant strains expressing peptide-fused key metabolic enzymes (gadTt and KdgA) were constructed for whole-cell biotransformation using gluconate as substrate to investigate the impact of peptide-mediated enzyme complexes on pyruvate synthesis. In the fluorescent protein system, P18/D18—amphipathic peptides that drive enzyme self-assembly via intermolecular hydrophobic interactions—increased extracellular fluorescence intensity of eOFP and eGFP by 31.11% and 25.21%, respectively. The D18 peptide significantly elevated TGase activity by enhancing structural stability to over 1.3-fold that of the control. For pyruvate synthesis, the peptide-mediated enzyme complex exhibited remarkable advantages in substrate conversion rate (up to 53.08%) and thermostability, confirming the platform’s ability to enhance substrate channeling despite no optimization for absolute yield. This study established a novel peptide-mediated multienzyme self-assembly platform in B. licheniformis, providing a valuable strategy for artificial metabolic channel design in synthetic biology.

1. Introduction

In synthetic biology, engineering efficient microbial cell factories is a core strategic goal for the green biomanufacturing of high-value chemicals, with precise metabolic pathway design as the key to overcoming metabolic efficiency bottlenecks [1]. A critical theoretical basis for optimizing artificial metabolic pathways comes from natural systems: enzymes often form multi-enzyme complexes that use spatial compartmentalization to achieve directional and efficient substrate channeling, minimizing intermediate diffusion loss and suppressing side reactions [2]. Current spatial enzyme assembly strategies mainly include scaffold-free assembly, fusion proteins, synthetic scaffold-mediated assembly, and physical compartmentalization [3,4,5,6,7,8]. Among these, peptide-mediated scaffold-free assembly has attracted extensive attention due to its high spatial flexibility and minimal interference with enzyme structure [9]. For instance, in eukaryotic systems, the RIAD-RIDD peptide pair achieves precise enzyme localization through specific molecular interactions [10], while in bacteria, the SsrA tag is recognized by SspB protein and directed to the ClpXP protease for degradation [11,12]. In hydrophobic self-assembly strategies, amphipathic peptides P18 and D18 leverage their hydrophobic amino acid-rich domains to drive spontaneous aggregation of fusion proteins via intermolecular hydrophobic interactions, forming highly active nanoscale aggregates or organelle-like structures intracellularly. This enables local enzyme enrichment and spatial organization without requiring complex scaffold proteins [13,14]. Despite these advances, peptide assembly research is mostly focused on model microbes like Escherichia coli and Saccharomyces cerevisiae, with limited exploration in industrial strains [10]. Moreover, existing studies often focus on single-type peptides, lacking systematic screening of mechanistically diverse peptides in hosts. For secretory-dependent industrial enzymes, whether peptide-mediated aggregation interferes with transmembrane translocation or protein folding via steric hindrance remains unresolved. Bacillus licheniformis possesses numerous advantageous characteristics, including broad substrate specificity, a rich enzymatic repertoire, contamination resistance, and exceptional thermostability [15,16]. Furthermore, this organism features a highly efficient protein secretion system coupled with flexible metabolic network regulation capabilities. It can effectively secrete heterologous recombinant proteins into the extracellular space or periplasm to alleviate intracellular accumulation stress [17], while simultaneously accommodating and adapting to various artificially engineered metabolic pathways [18,19]. These attributes establish B. licheniformis as an ideal chassis strain for high-efficiency heterologous protein expression and microbial metabolic engineering. However, current metabolic engineering modifications of B. licheniformis primarily rely on gene-level expression regulation, lacking sophisticated intervention strategies at the protein spatial organization level [18,19,20]. Traditional free-enzyme systems fail to address bottlenecks like intermediate diffusion loss, pathway flux imbalance, and toxic intermediate accumulation in complex heterologous biosynthetic pathways. Compared with complex protein scaffolds, peptide-mediated enzyme assembly systems have advantages of small molecular weight, design flexibility, and minimal host physiological burden [10]. Therefore, developing a peptide-based enzyme complex self-assembly system on this industrial platform holds certain engineering significance for overcoming the limitations of metabolic efficiency and for constructing the next generation of intelligent microbial cell factories.

2. Results

2.1. Screening of Peptide Elements

To achieve efficient and controllable assembly of artificial enzyme complexes in B. licheniformis, a demand-oriented screening strategy was implemented, leveraging the host’s Generally Recognized as Safe (GRAS) status, robust protein secretion capacity, and metabolic pathway construction requirements. Given the industrial application context of B. licheniformis, candidate peptides must satisfy three fundamental criteria: (i) cellular compatibility—peptides must exhibit no cytotoxicity, avoid interference with host growth metabolism and protein secretion systems (e.g., the twin-arginine translocation (Tat) pathway) [21], and possess codon-optimized sequences compatible with B. licheniformis codon usage to prevent translation efficiency loss caused by rare codons [18,22]; (ii) controllable interaction—peptides must demonstrate well-defined interaction mechanisms (e.g., specific binding or self-assembly) capable of stably mediating target protein complex formation with interaction strengths matching enzymatic reaction demands; and (iii) enzyme functional compatibility—peptide fusion must preserve the catalytic center, signal peptide functionality, and proper folding of target enzymes to maintain catalytic activity.

Based on these principles, we implemented a literature-based screening strategy from a validated enzyme assembly peptide library in synthetic biology [9,23,24,25,26]. The specific steps were: (1) Systematic literature retrieval of peptide elements previously reported to mediate enzyme assembly, with strict exclusion of unvalidated candidates or those with documented cytotoxicity in Gram-positive bacteria; (2) Preliminary filtering based on the three aforementioned criteria (cellular compatibility, controllable interaction, enzyme functional compatibility), prioritizing peptides with clear mechanisms and experimental validation in microbial hosts; (3) Selection of mechanistically diverse types to ensure the platform’s versatility. Through this systematic process, we prioritized peptide candidates with potential compatibility for B. licheniformis, including: (i) specific binding peptide pairs (RIAD/RIDD) [24], which achieve precise recognition through hydrogen bonding and hydrophobic interactions with nanomolar-range dissociation constants, matching the ordered substrate channeling requirement in metabolic pathways; (ii) hydrophobic self-assembling peptides (P18, D18) [23], where P18 contains 18 amino acids (45% hydrophobic residues) and D18 features a hydrophobic core with weak charge regions—both form multimers via intermolecular hydrophobic interactions, enabling construction of localized high-enzyme-concentration microdomains suitable for secreted protein aggregation in extracellular or peri-membrane regions of B. licheniformis; (iii) membrane-anchoring self-assembly peptides (SPFH domain), containing conserved membrane-binding motifs that anchor target enzymes to B. licheniformis functional membrane microdomains (FMMs), utilizing the host’s membrane system for spatial enzyme enrichment while reducing cytoplasmic diffusion losses; and (iv) encapsulation-type peptides (EncSig), whose C-terminal α-helix drives protein self-encapsulation into nanoscale aggregates, simultaneously increasing local enzyme concentration and restricting intermediate diffusion to minimize intracellular side reactions in B. licheniformis. This literature-based screening strategy avoids the uncertainty of de novo screening and ensures inherent compatibility with B. licheniformis’ physiological characteristics (

Table 1. Short peptide elements obtained from the initial screening.

Peptide Element	Fusion Strategy and Position	Representative Example
P18	N-terminal fusion; capable of both homo- and hetero-interactions	N-terminal fusion of these two targeting peptides to key enzymes in the 1,2-propanediol biosynthetic pathway enabled enzyme aggregation, substantially enhancing 1,2-propanediol production [14,27]
D18
SPFH domain	N-terminal fusion	N-terminal fusion of the SPFH domain to key enzymes in the mevalonate pathway facilitated enzyme recruitment to functional membrane microdomains, significantly improving isoprenoid production [28]
EncSig	C-terminal fusion	C-terminal fusion of EncSig to enzymes in the violacein biosynthetic pathway induced protein encapsulation into nanoscale compartments, enhancing pathway flux and violacein yield [25]
RIAD-RIDD	N-terminal or C-terminal fusion; hetero-specific binding pair	Fusion of RIAD and RIDD to complementary enzyme pairs in the terpenoid pathway enabled precise enzyme co-localization, improving product titers through enhanced substrate channeling [26]

).

2.2. Interaction of Peptide Elements with Single-Gene Expression of Fluorescent Proteins

Building upon our preliminary systematic screening of peptide elements, we successfully identified five distinct classes of peptide elements with varying mechanisms of action. These include hydrophobic self-assembling peptides (e.g., P18 and D18), membrane-anchoring domains (SPFH), encapsulating peptides (EncSig), and specific binding peptide pairs (RIAD-RIDD). To elucidate the universal mechanisms governing peptide-mediated regulation of extracellular secretion and single-protein expression, we constructed a series of recombinant strains expressing enhanced orange fluorescent protein (eOFP) and enhanced green fluorescent protein (eGFP). The eOFP series (BLeO1–BLeO8) and eGFP series (BLeG1–BLeG8) incorporated these peptide variants (P18, D18, SPFH, EncSig, RIAD, RIDD, etc.) fused at either the N- or C-terminus. Previous studies in our lab revealed that the either eGFP or eOFP could be secreted in B. licheniformis. Using strains BLeO and BLeG (lacking peptide fusions) as controls, we systematically monitored the dynamic changes in intracellular and extracellular fluorescence intensity over a 72-h fermentation period. This approach allowed for a comprehensive analysis of how peptide type and fusion orientation influence protein secretion efficiency and expression levels.

2.2.1. Structural Considerations for Fluorescent Protein-Peptide Fusions

According to the structural modeling results (Figure 1), the fluorescence activity of eOFP/eGFP relies on two critical factors: the integrity of the β-barrel structure and the stability of the local microenvironment surrounding the chromophore. Notably, our modeling results confirm that peptide elements including SPFH, RIAD, and RIDD are inherently prone to forming aberrant hydrogen bonds with the fluorescent protein: SPFH’s membrane-anchoring motif contains polar amino acid side chains that the model predicts will disrupt the β-barrel’s native hydrogen bond network; RIAD/RIDD’s hetero-specific binding domains carry charged residues that modeling shows can invade the chromophore’s microenvironment. Such non-native interactions may compromise the integrity of the β-barrel structure or destabilize the chromophore, ultimately leading to reduced fluorescence intensity or complete quenching.

2.2.2. Effect of Short Peptides on Orange Fluorescent Protein Expression

We dynamically monitored the extracellular fluorescence intensity of the eOFP recombinant strain, and the results are shown in Figure 2a. Compared with the strains BLeO1 and BLeO2 that were fused with the P18/D18 self-assembly peptide, the extracellular fluorescence intensity of this strain continuously increased during the fermentation process and reached its peak at 72 h: at 72 h, BLeO2 reached 50,454 A.U. (an increase of 31.11% compared to the control strain BLeO), BLeO1 reached 43,139 A.U. (an increase of 12.10%), and it maintained a rapid fluorescence growth rate after 48 h, while the control strain showed a significant deceleration phenomenon. In contrast, strains BLeO3, BLeO4, BLeO6, and BLeO8 had almost no extracellular fluorescence throughout the fermentation process, indicating that these peptides severely inhibited the secretion of eOFP. Figure 2c also shows the corresponding intracellular fluorescence intensity trend: BLeO1 and BLeO2 still maintained a moderate level of fluorescence in the cells, while the intracellular fluorescence of BLeO3, BLeO4, BLeO6, and BLeO8 was extremely low, indicating that these peptides not only hindered the secretion of eOFP but also affected its expression or structural stability within the cells.

2.2.3. Effect of Short Peptides on Green Fluorescent Protein Expression

The extracellular fluorescence intensities of the eGFP recombinant strains were quantified, with results presented in Figure 2b,d. Strain BLeG1 achieved an extracellular fluorescence intensity of 229,894 A.U. at 72 h, a 6.81% increase relative to the control BLeG (215,226 A.U.). Strain BLeG2 reached 269,487 A.U., corresponding to a significant 25.21% enhancement. Conversely, the membrane-anchoring peptide strain BLeG3 yielded only 1765 A.U. at 72 h (0.82% of control), and the encapsulating peptide strain BLeG4 produced 1770 A.U. (0.82% of control). Specific binding peptide strains BLeG6 and BLeG7 generated intensities of 2410 A.U. and 2176 A.U., respectively (1.12% and 1.01% of control), while BLeG8 remained at extremely low fluorescence levels. These results indicate that membrane-anchoring, encapsulating, and specific binding peptides severely impair the extracellular secretion and expression of eGFP, P18/D18 peptides do not disrupt the folding or secretion of reporter proteins.

2.3. Characterization of Dual-Gene Fluorescent Protein Recombinant Strains Based on Peptide Screening

Previous investigations of single fluorescent protein expression revealed marked differential effects of peptide types on protein expression and secretion. Self-assembling peptides P18 and D18 exhibited distinctive advantages, whereas membrane-anchoring SPFH domains, encapsulation EncSig, and other peptide categories strongly suppressed fluorescent protein expression. Based on these findings, we selected P18 and D18 peptides for further investigation to elucidate their regulatory mechanisms governing fluorescent protein colocalization and secretion in dual-gene co-expression systems.

Recombinant strains were cultivated following the protocol described in Section 4.2.2. Fermentation broth samples were collected at 24, 48, and 72 h to quantify intracellular and extracellular fluorescence intensities, with results shown in Figure 3. For eOFP, strains BLOG2 and BLOG5 achieved extracellular fluorescence intensities of 60,453 A.U. and 53,315 A.U. at 72 h, respectively, representing increases of 31.93% and 16.35% relative to control BLOG; strains BLOG3 and BLOG4 reached 39,168 A.U. and 44,578 A.U. at 72 h, corresponding to reductions of 14.52% and 2.72%, respectively. For eGFP, strain BLOG4 attained an extracellular fluorescence intensity of 153,813 A.U. at 72 h, representing an 86.20% enhancement over control BLOG (82,605 A.U.); strains BLOG2, BLOG3, and BLOG5 exhibited fluorescence intensities of 37,930 A.U., 50,175 A.U., and 9858 A.U. at 72 h, corresponding to decreases of 54.08%, 39.26%, and 88.07% relative to the control, respectively. Notably, BLOG4 demonstrated optimal eGFP expression and secretion performance, suggesting that its peptide combination confers superior compatibility with eGFP folding and secretion machinery.

To elucidate the assembly dynamics of dual fluorescent protein genes mediated by peptide tags, we employed laser scanning confocal microscopy. Microscopic examination (Figure 4) revealed pronounced colocalization of green and orange fluorescence signals in experimental groups relative to controls, characterized by abundant and uniformly distributed yellow puncta resulting from spectral overlap. This visualization directly demonstrates peptide-mediated spatial proximity and assembly of the dual-gene products, providing compelling experimental validation for our hypothesis that short peptides facilitate gene product colocalization—a critical foundation for mechanistic investigation and downstream applications.

2.4. Regulatory Effects of Peptide Tags on Transglutaminase Catalytic Activity and Stability

Construction of functional enzyme complexes in B. licheniformis necessitates evaluation of whether peptide tag fusion compromises target gene expression or enzymatic catalytic performance. We selected transglutaminase (TGase) as a reporter enzyme due to its distinctive maturation mechanism: TGase is initially synthesized as an inactive zymogen (Pro-TGase) containing an N-terminal pro-peptide essential for proper folding and transmembrane secretion via the Sec pathway. Following extracellular translocation, proteolytic cleavage of this pro-peptide by host proteases is obligatory to expose the catalytic center and generate mature, active TGase [29,30,31] (Figure 5a). This maturation cascade poses a critical challenge for peptide tag engineering. To systematically investigate fusion effects at this sensitive locus and comprehensively delineate peptide-mediated modulation of enzyme function, we designed a combinatorial fusion strategy encompassing both N- and C-terminal attachment sites, generating a panel of recombinant expression constructs (Figure 5b). This multi-positional approach enables assessment of how peptide insertion sites influence TGase secretion efficiency, zymogen processing, and ultimate catalytic competence. Beyond identifying potential deleterious effects, this design framework aims to uncover enhancement mechanisms that may inform rational optimization of enzyme complex architectures.

2.4.1. Impact of Peptide Tags on Transglutaminase Catalytic Activity

Recombinant strains were cultivated following Section 4.2.4 (1), with fermentation broth sampled at 48 and 72 h for TGase activity determination. Results (Figure 6a) revealed that select peptide fusions significantly modulated mature TGase activity. Relative to the untagged parental strain BLpT, strain BLpT2 (N-terminal D18 fusion) and BLpT7 (C-terminal RIAD fusion) exhibited elevated enzymatic activity at 48 h, with BLpT2 demonstrating the most pronounced enhancement at 151.19% of control levels and BLpT7 achieving 133.03% of baseline activity. The remaining constructs (BLpT1, BLpT3, BLpT4, BLpT5, BLpT6, BLpT8) displayed TGase activities statistically indistinguishable from the BLpT control.

2.4.2. Influence of Peptide Tags on Transglutaminase Stability

To evaluate the regulatory effects of short peptides on enzyme stability, samples fermented for 48 h and 72 h were maintained at room temperature, and their residual enzyme activities were measured at 0 h, 24 h, and 48 h to systematically analyze enzyme activity decay patterns under room temperature storage conditions. The experimental results (Figure 6b,c) demonstrated that enzyme activity of all samples gradually decreased with prolonged storage; however, the rate and magnitude of decline varied depending on peptide type and fermentation duration. For 48 h-fermented samples, after 24 h of room temperature storage, recombinant strains BLpT7 and BLpT8 harboring C-terminal RIAD and RIDD peptide fusions exhibited superior stability, with enzyme activity retention rates of 62.83% and 72.67%, respectively—significantly higher than the control group’s 45.59%. However, when storage time extended to 48 h, enzyme activity retention rates across all experimental groups declined to 35.67–52.38%, with no significant inter-group differences. For 72 h-fermented samples, stability trends remained essentially consistent with 48 h samples: after 24 h storage, enzyme activity retention ranged from 73.64% to 87.32%, further decreasing to 34.86–39.11% after 48 h. Enzyme activity decay followed the general pattern of progressive decline with extended storage time; nevertheless, fermentation duration and peptide type influenced decay kinetics and absolute retention values, with peptide elements demonstrating more pronounced enhancement of short-term (≤24 h) enzyme stability. Notably, although 72 h samples exhibited similar decay trends to 48 h samples, their absolute enzyme activity values after 24 h at room temperature were substantially higher.

2.5. Whole-Cell Conversion of Gluconate to Pyruvate

Based on the efficient peptide pair (P18/D18) identified in previous work and its capacity to mediate enzyme complex assembly, recombinant strains harboring peptide-fused key metabolic enzymes—gadTt (sodium gluconate dehydrogenase) and KdgA (2-keto-3-deoxygluconate aldolase)—were constructed. Whole-cell conversion using sodium gluconate as substrate was performed to produce pyruvate, investigating the effect of peptide-mediated enzyme complexes on pyruvate synthesis efficiency and providing empirical support for applying this self-assembly platform in metabolic pathway engineering.

Experimental results (Figure 7) revealed that although the absolute pyruvate yields of the two recombinant strains were lower than the traditional control, they exhibited significant advantages in substrate conversion rate and thermostability. In temperature gradient experiments spanning 70–80 °C, strain BLgK1 maintained a maximum substrate conversion rate of ~53%, which was higher than the control strain BLgK. For strain BLgK2, whose yield was closer to the control, the substrate conversion rate was still higher than BLgK. This convincingly demonstrates that P18/D18 peptide-mediated spatial aggregation of gadTt and KdgA successfully achieved an efficient substrate channeling effect, significantly minimizing diffusion losses of metabolic intermediates and optimizing the reaction microenvironment. Collectively, successful construction of BLgK1 and BLgK2 validates the core advantages of peptide self-assembly strategies in enhancing enzymatic reaction specificity and thermostability.

3. Discussion

The construction of efficient microbial cell factories increasingly relies on the spatial organization of enzymes to mimic the compartmentalization observed in eukaryotic organelles [21,32,33]. In this study, we developed a peptide-mediated enzyme assembly platform in the Generally Recognized As Safe (GRAS) strain Bacillus licheniformis, which is renowned for its high secretory capacity. A systematic evaluation of peptide tags revealed significant differences in their compatibility with the host’s secretion machinery. While amphipathic self-assembling peptides (P18/D18) successfully promoted extracellular secretion of reporter proteins—resulting in fluorescence enhancements of up to 31.11%—peptides designed for membrane anchoring (SPFH) or encapsulation (EncSig) caused severe secretion defects [34].This divergence can be mechanistically explained by the constraints of the general secretory (Sec) pathway, the primary route for protein export in B. licheniformis. A key requirement of the Sec translocon is that substrate proteins must remain largely unfolded and translocation-competent before traversing the membrane channel [21,35]. The SPFH domain, which naturally targets functional membrane microdomains (FMMs), exhibits high hydrophobicity and a strong tendency to associate with lipids, likely leading to premature folding or membrane interaction that impedes Sec-dependent translocation.

This study challenges the long-standing conclusion regarding enzyme precursor modification, proving that certain peptide fusions (particularly those fusing the D18 peptide to the N-terminal of transglutaminase) can enhance rather than inhibit enzyme activity, unexpectedly increasing the enzyme activity to 151.19% of the control group. Although the traditional view holds that N-terminal fusions usually hinder the processing of signal peptides or the removal of pro-peptides, the D18 fusion was not hindered but rather promoted. Since the maturation process of the Bacillus licheniformis transglutaminase is a strictly regulated process that relies on the hydrolysis of the N-terminal pro-peptide to achieve, this pro-peptide may act as an intramolecular chaperone to prevent premature activation and lethal cross-linking. Structural analysis using PyMOL v3.0 (Figure S1) confirmed that the D18, RIAD, and RIDD fusions maintained a high degree of structural consistency with the natural transglutaminase (RMSD₁ = 0.102, RMSD₇ = 0.117, RMSD₈ = 0.166), preserving the overall tertiary/quaternary structure and functional activity. All peptides induced unique conformational changes near the conserved Pro 57–Asp 59 motif, which is crucial for enzyme precursor activation [31]. Observations using BLpT7 (the C-terminal RIAD fusion) revealed significant local structural changes near the native α-helical structure, but did not disrupt the overall folding structure. In summary, these structural changes indicate that the peptide fusions can regulate the conformation of the activation region, thereby affecting the maturation efficiency of TGase while not damaging the integrity of the structure.

The P18/D18 scaffold was applied to the pyruvate synthesis pathway with the aim of verifying the functionality of this platform rather than pursuing high pyruvate production. Although the absolute yield was relatively low, the substrate conversion rate was higher, which directly indicates that substrate channeling is the key driving factor. The engineered strain BLgK1 achieved an outstanding substrate conversion rate of approximately 53%. In contrast, the conversion rate of the control group was lower. This provides strong evidence for substrate channeling. By connecting gadTt and KdgA, the P18/D18 scaffold reduced the diffusion distance of metabolic intermediates, ensuring that the intermediate products of the first enzyme could immediately be provided to the second enzyme. The relatively low absolute pyruvate yield was caused by the overexpression of aggregated hydrophobic peptides and the formation of large protein complexes, which may consume cellular resources and lead to a lower absolute pyruvate yield [36,37]. Moreover, although our study indicates that peptide-mediated assembly can improve activity and stability, future work can include combining the standardization of catalytic results with the abundance of enzymes. Measuring the concentration of enzymes will help distinguish the effects caused by spatial organization from those resulting from differences in expression levels, thereby providing a more rigorous verification of assembly efficiency. This demonstrates that the peptide-mediated assembly platform can effectively increase the flux of the metabolic pathway, laying the foundation for the efficient production of high-value chemicals through subsequent optimization.

This study demonstrates that the constructed enzyme complex exhibits extremely high thermal stability—a crucial advantage for industrial applications. Within the range of 75 to 80 degrees Celsius, the conversion rate of the assembled strain BLgK2 was significantly higher than that of the free enzyme control group. We propose that the P18/D18-mediated assembly makes the tetrameric structure of the gadTt–KdgA complex more stable. The multiple point hydrophobic interactions within the assembly core are likely to act as structural anchors, restricting the conformational freedom of the enzyme subunits. This reduces the entropy increase during unfolding, thereby increasing the activation energy required for thermal denaturation. This stabilization mechanism is similar to those observed in natural multi-enzyme complexes (such as the pyruvate dehydrogenase complex) and artificial immobilized enzyme systems. However, unlike the solid support immobilization methods that typically introduce mass transfer limitations, this self-assembly system retains the advantages of liquid-phase reactions. This thermal stability makes the engineered Aspergillus cinereus strain particularly suitable for harsh industrial processes, providing a reliable platform for the production of high-value chemical products [38,39].

4. Materials and Methods

4.1. Materials

4.1.1. Strains and Plasmids

The strains and plasmids used in this study are shown in Table S1.

4.1.2. Major Instruments and Reagents

Microplate reader (TECAN); benchtop high-speed centrifuge (Sigma-Aldrich Trading Co., Ltd., Shanghai, China); PCR thermal cycler and nucleic acid gel electrophoresis system (Bio-Rad, Hercules, CA, USA); high-performance liquid chromatograph (Agilent Technologies, Santa Clara, CA, USA).

Restriction endonucleases (Thermo Fisher Scientific, Waltham, MA, USA); 2 × Taq/Phanta PCR Master Mix, plasmid DNA extraction kit, and DNA purification kit (Nanjing Vazyme Biotech Co., Ltd., Nanjing, China); DNA molecular weight marker (Takara Bio., Kusatsu, Japan); kanamycin, ampicillin, and tetracycline (Merck Sigma-Aldrich, St. Louis, MO, USA); peptone, yeast extract, and agar powder (OXOID, Basingstoke, UK). All other reagents were of analytical grade from domestic or imported sources.

4.1.3. Culture Media and Cultivation Conditions

Luria–Bertani (LB) medium (g/L): peptone 10.0, yeast extract 5.0, sodium chloride 10.0; solid LB medium required addition of 2% agar powder. Nutrient-rich medium (g/L): dipotassium phosphate 10.00, potassium dihydrogen phosphate 1.36, ammonium phosphate 5.00, urea 10.00, magnesium sulfate heptahydrate 2.00, monosodium glutamate 10.00, trace elements 2.00, calcium carbonate 5.00, peptone 20.00, yeast extract 10.00. Synthetic medium (g/L): dipotassium phosphate 10.00, potassium dihydrogen phosphate 1.36, ammonium sulfate 5.00, urea 10.00, magnesium sulfate heptahydrate 2.00, monosodium glutamate 10.00, trace elements 2.00. Natural medium (g/L): dipotassium phosphate 10.00, potassium dihydrogen phosphate 1.36, ammonium sulfate 5.00, urea 10.00, magnesium sulfate heptahydrate 2.00, monosodium glutamate 10.00, trace elements 2.00, peptone 20.00, yeast extract 10.00. Trace elements (g/L): ferric ammonium citrate 54.4, manganese chloride tetrahydrate 9.8, cobalt chloride hexahydrate 1.6, copper chloride dihydrate 1.0, boric acid 1.9, zinc sulfate heptahydrate 9.0, sodium molybdate dihydrate 1.1, sodium selenate 1.5, nickel sulfate hexahydrate 1.5. All media were sterilized at 115 °C for 20 min. Final concentrations of ampicillin, kanamycin, and tetracycline added to media were 100, 50, and 50 μg/mL, respectively.

4.1.4. Primers

Primer sequence information used in this study is provided in Table S2. Primers were designed using SnapGene version 6.0.2 (Insightful Science, San Diego, CA, USA), and all primer synthesis and sequencing were performed by Sangon Biotech (Sangon Biotech Co., Ltd., Shanghai, China).

4.2. Methods

4.2.1. Construction of Recombinant Strains

(1)

Construction of recombinant strains expressing eOFP and eGFP

The empty vector ePWBN was digested with restriction endonucleases Xho I and BamH I, and the linearized vector fragment was recovered by gel electrophoresis. Using plasmids harboring eOFP and eGFP genes (preserved in our laboratory) as templates, PCR amplification was performed with the universal forward primer OFP-F paired with different reverse primers (O4-R, O7-R, O8-R) and the universal reverse primer OFP-R paired with different forward primers (OFP-F, O1-F, O2-F, O3-F, O5-F, O6-F), yielding gene fragments OFP, OFP1, OFP2, OFP3, OFP4, OFP5, OFP6, OFP7, and OFP8. The same approach was applied to obtain GFP, GFP1, GFP2, GFP3, GFP4, GFP5, GFP6, GFP7, and GFP8 gene fragments. The PCR reaction system consisted of 25 μL 2 × Phanta Max Master Mix, 2 μL each of forward and reverse primers (10 μmol/L), 1 μL template DNA, and ddH₂O to a final volume of 50 μL. PCR conditions were as follows: initial denaturation at 95 °C for 5 min; 30 cycles of 95 °C for 15 s, 55 °C for 15 s, and 72 °C at 1 kb/min; final extension at 72 °C for 10 min. The purified PCR products were ligated with the linearized ePWBN vector using homologous recombination enzyme. The ligation products were transformed into E. coli JM109 competent cells. Ten to twenty transformants were randomly selected for colony PCR identification, and plasmid DNA from correct transformants was extracted and verified by Sanger sequencing to confirm sequence accuracy. This yielded expression plasmids eO, eO1, eO2, eO3, eO4, eO5, eO6, eO7, and eO8. The constructed expression plasmids were electroporated into B. licheniformis with aprE and amyL genes deleted according to the method of Xiao et al. [15], generating recombinant strains BLeO, BLeO1, BLeO2, BLeO3, BLeO4, BLeO5, BLeO6, BLeO7, BLeO8, BLeG, BLeG1, BLeG2, BLeG3, BLeG4, BLeG5, BLeG6, BLeG7, and BLeG8.

(2)

Construction of strains co-expressing eOFP and eGFP fusion proteins

Using plasmids harboring eOFP and eGFP genes (preserved in our laboratory) as templates, overlap extension PCR was performed with forward primers OFP-F, O1-F, O2-F, O4-F and reverse primers GFP-R, G1-R, G2-R, G4-R. The templates were the previously obtained OFP, OFP1, OFP2, OFP4, GFP, GFP1, GFP2, and GFP4 fragments. Vector preparation, homologous recombination, and transformation procedures followed those described in Section 4.2.1 (1). After verification, recombinant strains BL, BLOG1, BLOG2, BLOG3, BLOG4, and BLOG5 were obtained.

(3)

Construction of TGase Recombinant Strains

The empty vector pHY was digested with restriction endonucleases Xho I and Nhe I, and the linearized vector fragment was recovered by gel electrophoresis. Using a plasmid harboring the TGase gene (preserved in our laboratory) as template, PCR amplification was performed with the universal forward primer TG-F paired with different reverse primers (TG-R, TG4-R, TG7-R, TG8-R) to obtain gene fragments TGase, TGase^EncSig, TGase^RIAD, and TGase^RIDD. Similarly, different forward primers (TG1-F, TG2-F, TG3-F, TG5-F, TG6-F) were paired with the universal reverse primer TG-R to obtain ^P18TGase, ^D18TGase, ^SPFHTGase, ^RIADTGase, and ^RIDDTGase gene fragments. Homologous recombination and transformation procedures followed those described in Section 4.2.1 (1), ultimately yielding recombinant strains BL, BLpT1, BLpT2, BLpT3, BLpT4, BLpT5, BLpT6, BLpT7, and BLpT8.

(4)

Construction of Strains Co-expressing gadTt and KdgA

Using plasmids harboring gadTt and KdgA genes (preserved in our laboratory) as templates, fragment PCR was performed with forward primers gK-1-F, gK-2-F, gK1-1-F, gK1-2-F, gK2-1-F, gK2-2-F and reverse primers gK-1-R, gK-2-R, gK1-1-R, gK2-1-R to obtain fragments gadTt, KdgA, ^P18gadTt, ^D18KdgA, ^D18gadTt, and ^P18KdgA. Using the above fragments as templates, overlap extension PCR was performed with forward primers gK-1-F, gK1-1-F, gK2-1-F and the universal reverse primer gK-2-R to obtain fragments gadTt-KdgA, ^P18gadTt-^D18KdgA, and ^D18gadTt-^P18KdgA. Vector preparation, homologous recombination, and transformation procedures followed those described in Section 4.2.1 (1). After verification, recombinant strains BLgK, BLgK1, and BLgK2 were obtained.

4.2.2. Expression of eOFP and eGFP and Measurement of Fluorescence Intensity

Verified recombinant strains were inoculated into test tubes containing 15 mL seed medium and cultivated at 37 °C with shaking at 220 rpm for 20 h to prepare seed cultures. Subsequently, 3% (v/v) of the seed culture was transferred into 250-mL shake flasks containing 30 mL synthetic medium and cultivated under identical conditions for 72 h. Fermentation broth samples were collected at 24-h, 48-h, and 72-h intervals. Aliquots of the fermentation broth were centrifuged at 10,000 rpm for 10 min at 4 °C. The supernatants were designated as extracellular samples for measuring extracellular fluorescence intensity, while the cell pellets were washed 2–3 times with phosphate buffer and resuspended in the same buffer for intracellular fluorescence measurements. Fluorescence intensity was determined using a multimode microplate reader [40]. For eOFP, the excitation wavelength was set at 532 nm and emission at 578 nm; for eGFP, excitation was set at 480 nm and emission at 520 nm. Each sample was measured in triplicate, and mean values were recorded as final fluorescence intensity data.

4.2.3. Fluorescence Intensity Determination of Co-Expressed Orange and Green Fluorescent Proteins

Samples of dual-gene recombinant strains BLOG, BLOG1, BLOG2, BLOG3, BLOG4, and BLOG5 were collected at 48 h and 72 h of fermentation, appropriately diluted with phosphate buffer, and deposited onto confocal microscopy-compatible glass slides. Coverslips were carefully applied to avoid air bubble formation. The slides were mounted on the stage of a laser scanning confocal microscope with optimized scanning parameters. For eOFP orange fluorescence, a 532-nm laser was used for excitation; for eGFP green fluorescence, a 480-nm laser was employed. Microscope focus, gain, and other parameters were adjusted to ensure clear imaging and distinct fluorescence signals [41,42].

4.2.4. Expression and Activity Assay of Transglutaminase

(1)

Transglutaminase Expression

Verified positive single colonies were inoculated into 15 mL LB liquid medium and cultivated overnight at 37 °C with shaking at 220 rpm to prepare seed cultures. The following day, 3% (v/v) of the overnight culture was transferred into fresh nutrient-rich medium supplemented with appropriate antibiotic concentrations and cultivated at 37 °C with shaking at 220 rpm. Fermentation broth samples were collected at 48 h and 72 h for TGase activity determination.

(2)

Transglutaminase Activity Assay

TGase activity was determined according to GB/T 34795–2017 [43] “Detection Method for Transglutaminase Activity” using the hydroxamate colorimetric method. TGase catalyzes the reaction between the substrate N-benzyloxycarbonyl-L-glutaminylglycine and hydroxylamine to produce L-glutamic acid-γ-monohydroxamate. One unit of enzyme activity is defined as the amount of enzyme required to catalyze the formation of 1.0 μmol/L glutamic acid-γ-monohydroxamate per minute at 37 °C and pH 6.0, expressed as U/mL.

4.2.5. Whole-Cell Biotransformation

Cells were cultivated in natural medium at 37 °C with shaking at 220 rpm for 48 h. Following fermentation, cells were harvested and washed twice with 0.1 M HEPES buffer, then resuspended in the same buffer to an OD600 of 100. A 1.25 mL aliquot of the cell suspension was transferred into a 1.5 mL microcentrifuge tube and preheated in a heating block for 30 min. The reaction was initiated by adding 0.14 mL of 300 g/L sodium gluconate stock solution. Samples were collected at 0 h and 4 h to terminate the reaction.

4.2.6. HPLC Detection Methods

Pyruvate detection conditions: UPLC system; UV detector at 210 nm; XD-C18AQ column; mobile phase A: 2% methanol + B: 98% formic acid (0.1%); flow rate 0.15 mL/min; column temperature 30 °C; run time 12 min. Sodium gluconate detection conditions: HPLC system; differential refractive index detector; Shodex 1011H column; mobile phase: 5 mM sulfuric acid; flow rate 0.6 mL/min; column temperature 55 °C; run time 26 min.

4.2.7. Optimization of Whole-Cell Biotransformation Conditions

Temperature optimization was performed for whole-cell biotransformation. Under controlled reaction pH, cell density, and substrate concentration, three temperature conditions (70, 75, and 80 °C) were evaluated to investigate their effects on conversion efficiency. The substrate conversion rate was calculated according to Equation (1):

R1/% = (C1/C2) × 100%

(1)

where R1 represents the mass conversion rate of pyruvate (%), C1 denotes the mass concentration of pyruvate solution (g/L), and C2 indicates the mass concentration of sodium gluconate (g/L).

4.2.8. Statistical Analysis Methods

All experiments were independently repeated three times, and the average value was taken as the final result. The differences between two sets of data were analyzed using a 2-tailed Student’s t-test, while the differences between multiple sets of data were compared using one-way ANOVA and Tukey’s test. “*” and “***” were used to indicate the significance of p < 0.05 and p < 0.001, respectively.

5. Conclusions

This study systematically screened five classes of peptide elements with distinct mechanisms of action. Through fusion expression with transglutaminase and fluorescent reporter proteins, we discovered that specific peptide fusions effectively enhance enzyme activity and stability, while self-assembling peptides significantly promote extracellular fluorescent intensities. Furthermore, through a dual-fluorescence protein co-expression system, we directly validated peptide-mediated protein co-localization effects. In metabolic pathway applications, the optimized peptide pair (P18/D18) was applied to assemble key metabolic enzymes (gadTt and KdgA). Whole-cell catalysis experiments confirmed that this strategy enhances local conversion efficiency through substrate channeling effects. By successfully employing B. licheniformis as an industrial chassis, we developed a novel peptide-mediated multi-enzyme complex self-assembly strategy, providing a reference for constructing artificial metabolic channels in synthetic biology.