NeuAc (N-acetylneuraminic acid) synthesis relies on the collaborative action of enzymes AGE and NAL, with two reversible reactions: one between GlcNAc (N-acetylglucosamine) and ManNAc (N-acetylmannosamine), and the other between ManNAc and NeuAc (Figure 1). This process requires their precise coordination to drive substrate-to-product conversion. The project constructed plasmids (pETDuet-AGE-NAL, pET28a-AGE-Linker-NAL, pET28a-Scaffold) via homologous recombination, which served as carriers for expressing AGE, NAL, and Scaffold. SDS-PAGE validation confirmed the successful synthesis and purification of these proteins, including the AGE-Linker-NAL fusion protein and components of the combined system. Additionally, functional tests (notably TLC, ELISA, and HPLC) characterized the enzyme activities of AGE and NAL under varied reaction time and temperature. Collectively, these efforts laid the foundation for understanding and optimizing the NeuAc synthesis pathway mediated by AGE and NAL.

Figure 1 The pathway to synthesize NeuAc by using AGE and NAL

1.1 PCR Amplification

We constructed pETDuet-AGE-NAL, pET28a-AGE-Linker-NAL, and pET28a-Scaffold by using homologous recombination. The AGE, NAL, Scaffold, and AGE-Linker-NAL sequences were amplified by PCR. Figure 2 shows the agarose gel electrophoresis results of the PCR. We can find out that the DNA fragment of AGE was approximately 1262 bp in length, NAL was 987 bp, Scaffold was 775 bp, AGE-Linker-NAL was 2117 bp, which were consistent with the predicted size. This indicated that the target gene was successfully amplified.

Figure 2 Agarose gel electrophoresis results of PCR

1.2 Plasmid linearization

The pETDuet-1 plasmid was used as a vector to ligate the AGE and NAL, with a length of 7563 bp after ligating both target genes to the plasmid. The pET28a plasmid was used as a vector to ligate Scaffold and AGE-Linker-NAL, with lengths of 5995 bp and 7342 bp, respectively, after ligation. Figure 3 shows the bands of these plasmids after double enzyme digestion, which were consistent with the size. These results indicate a successful linearization of the pETDuet-1 and pET28a plasmid.

Figure 3 The gel electrophoresis validation of the linearized plasmids

1.3 Homologous Recombination

Prior to constructing the pETDuet-AGE-NAL plasmid, we first completed the construction of pETDuet-AGE. This construction employed a homologous recombination approach. The colonies obtained after transformation are shown in Figure 4, indicating successful recombination and the acquisition of positive clones. Subsequently, we selected a single colony for colony PCR verification. Electrophoresis revealed a specific band at 1262 bp, matching the expected size, preliminarily confirming that the AGE fragment had been correctly recombined into the pETDuet vector. To further validate the construction accuracy, we extracted the recombinant plasmid and submitted it for sequencing. Sequence alignment confirmed that the AGE gene was correctly inserted with a fully intact reading frame, showing no frameshifts or mutations. This indicates the successful construction of the pETDuet-AGE plasmid, laying the foundation for subsequent introduction of the NAL gene and further dual-enzyme co-expression studies.

Figure 4 pETDuet-AGE Plasmid Verification Diagram

Building upon the successful construction of the pETDuet-AGE plasmid, we further employed homologous recombination to insert the NAL gene into the vector, thereby constructing the pETDuet-AGE-NAL recombinant plasmid. Clear single colonies were visible on the transformation plates, indicating successful recombination. Randomly selected colonies were subjected to PCR validation. Electrophoresis analysis of the amplified products revealed a specific band at the expected size (approximately 2249 bp), preliminarily confirming the correct insertion of the NAL gene. Subsequently, positive clone plasmids were extracted for sequencing analysis. The sequencing results fully matched the expected sequence, confirming that both the AGE and NAL genes were correctly integrated into the pETDuet vector with intact reading frames and no frameshift mutations. These results indicate the successful construction of the pETDuet-AGE-NAL plasmid, providing a reliable vector for subsequent dual-enzyme co-expression and catalytic function studies.

Figure 5 pETDuet-AGE-NAL Plasmid Verification Diagram

We further employed homologous recombination to sequentially assemble the AGE gene, Linker sequence, and NAL gene into the pET28a vector, constructing the pET28a-AGE-Linker-NAL recombinant plasmid. Following the transformation, uniformly distributed single colonies were observed on the agar plate, indicating successful recombinant transformation. Several colonies were randomly selected for colony PCR verification. Electrophoresis results showed a specific band at the expected size (approximately 2303 bp for the AGE-Linker-NAL fragment), preliminarily confirming the correct insertion of the three-gene fusion into the vector. To further validate construction accuracy, positive clone plasmids were extracted for sequencing analysis. Sequence results fully matched the expected sequence, confirming that the AGE, Linker, and NAL elements were correctly integrated into the pET28a vector as designed, with precise codon transitions and no frameshift mutations. These results confirm the successful construction of the pET28a-AGE-Linker-NAL triple fusion expression vector, providing essential material for subsequent studies investigating the regulatory role of the Linker sequence on the spatial conformation and catalytic efficiency of the dual enzymes.

Figure 6 pET28a-AGE-Linker-NAL Plasmid Verification Diagram

We proceeded to employ homologous recombination to clone the artificial scaffold protein gene into the pET28a vector, constructing the pET28a-Scaffold recombinant plasmid. Clear single colonies were observed on the plate post-transformation, indicating successful transformation. Subsequently, multiple colonies were randomly selected for colony PCR validation. Electrophoresis results showed a specific band at the expected size (approximately 755 bp), providing preliminary evidence that the Scaffold gene had been successfully inserted into the vector. To further confirm construction accuracy, we extracted positive clone plasmids for sequencing analysis. The sequencing results fully matched the expected sequence, confirming that the Scaffold gene was correctly integrated into the pET28a vector with an intact coding frame, free of frameshifts or mutations. These results indicate the successful construction of the pET28a-Scaffold plasmid, laying the foundation for subsequent studies investigating the regulatory role of scaffold proteins in multi-enzyme assembly and metabolic pathway efficiency.

Figure 7 pET28a-Scaffold Plasmid Verification Diagram

2.1 Heat shock conversion into BL21(DE3)

Each of the recombinant plasmids was independently transformed into E. coli BL21(DE3) competent cells using the heat shock method. The competent cells were first incubated with plasmid DNA on ice and then exposed to a brief heat shock at 42 ℃ to facilitate DNA uptake. Following recovery in LB medium, the cells were plated on selective LB agar plates containing the corresponding antibiotic and incubated overnight at 37 ℃.

Co-transform pET28a-Scaffold plasmid and pETDuet-AGE-NAL plasmid into E. coli BL21(DE3) competent cells. Simultaneously, individually transform pET28a-AGE-Linker-NAL and pETDuet-AGE-NAL single plasmids into the same host strain. The co-transformed colonies grew on plates containing both ampicillin and kanamycin, while the remaining colonies grew on ampicillin-only plates. Clear, single colonies were observed on all plates post-transformation, indicating successful transformation for each combination.

Figure 8 Agar plate showing BL21 colonies after transformation

Notes: A: pETDuet-AGE-NAL, B: pET28a-AGE-Linker-NAL, C: pET28a-Scafford + pETDuet-AGE-NAL.

Single colonies from each experimental group were selected for colony PCR verification. Results showed: The total transformation group simultaneously amplified the Scaffold characteristic band (approximately 755 bp) and the AGE-NAL fusion fragment band (approximately 2249 bp); The pET28a-AGE-Linker-NAL group amplified the expected-size Linker-AGE-NAL band (approximately 2303 bp), while the pETDuet-AGE-NAL group exhibited the characteristic AGE-NAL band (approximately 2249 bp). Electrophoresis results for all conditions aligned with expectations. These findings confirm that all three transformation combinations successfully yielded positive clones, establishing a reliable experimental foundation for subsequent comparisons of functional differences between the scaffold-mediated multienzyme assembly system and the direct fusion expression system.

Figure 9 PCR Identification of Colonies from Three Plasmids

2.2 Lysate protein concentration

Each PCR-verified colony was inoculated into liquid LB medium and cultured overnight, followed by scale-up fermentation. After cell harvest via centrifugation, bacterial pellets were subjected to ultrasonic lysis to release intracellular contents.

2.3 SDS-PAGE verification

After protein purification using a Nickel column, SDS-PAGE analysis was performed on both crude and purified samples. Proteins were separated on a 15% polyacrylamide gel and stained with Coomassie Brilliant Blue.

For pET28a-AGE-Linker-NAL, as shown in Figure 10, a band corresponding to AGE-Linker-NAL was detected in the elution solution (band 5), confirming the successful expression of this fusion protein and its effective separation and identification after purification.

Figure 10 SDS-PAGE for AGE-Linker-NAL

For pETDuet-AGE-NAL, Figure 11 showed that clear bands corresponding to AGE and NAL (at their expected positions) were observed in the elution solutions 1 and 2 (bands 5, 6), indicating the successful expression of both proteins and their effective enrichment after purification.

Figure 11 SDS-PAGE for AGE and NAL

Additionally, for the co-expression of pET28a-Scaffold and pETDuet-AGE-NAL, Figure 12 also revealed that in the elution solution 1 and 2 (bands 5, 6), not only bands of AGE and NAL but also a visible band corresponding to Scaffold were present, demonstrating that all proteins in the Scaffold + AGE + NAL fusion system were expressed and purified.

Figure 12 SDS-PAGE for Scaffold, AGE, and NAL

3.1 Thin-Layer Chromatography (TLC)

The activities of AGE and NAL were qualitatively analyzed using TLC at two time points: 30 and 60 minutes. As shown in Figure 13, in the AGE-catalyzed reactions, the band corresponding to the non-denatured enzyme at 30 minutes migrated similarly to the product reference, while the denatured enzyme at 60 minutes co-migrated with the substrate. This indicates that 30 minutes was sufficient for optimal AGE activity under these conditions. For NAL, the TLC results showed separation primarily at the substrate positions, suggesting that the current mobile phase may not be ideal for resolving NAL-derived products due to their high polarity.

Figure 13 TLC for AGE and NAL under 30 and 60 minutes

To further validate the generation of ManNAc and NeuAc, we optimized the developing solvent system for thin-layer chromatography (TLC). After reacting at 42°C for 30 minutes, samples underwent two duplicate developments using freshly prepared developing solvent, with TLC plates thoroughly dried using a hair dryer after each development. Subsequently, TLC plates were immersed in a developing agent and heated at 130°C for 5 minutes for visualization.

Results showed that in the three distinct enzyme structures—AGE-NAL (A-N), AGE-Linker-NAL (A-L-N), and Scaffold + NAL-AGE (S+A-N)—no bands corresponding to ManNAc or NeuAc were detected in the A-N group. In contrast, both A-L-N and S+A-N groups exhibited distinct target product bands. This result indicates that introducing Linker or Scaffold modules into the enzyme structure effectively prevents reversible reactions during metabolism, thereby enhancing the synthetic efficiency of ManNAc and NeuAc.

Figure 14 TLC for AGE and NAL under 30 minutes

We conducted systematic TLC analysis of the enzymatic activities of three enzyme constructs—AGE-NAL (A-N), AGE-Linker-NAL (A-L-N), and Scaffold+NAL-AGE (S+A-N)—reacted for one hour at three temperature conditions: 25°C, 37°C, and 42°C. As shown in Figure 15A, the A-N construct yielded the highest amount of ManNAc in the 1-hour incubation group, with no detectable NeuAc product formation at this time point. This indicates that A-N tends to accumulate the intermediate ManNAc over shorter reaction times.

When incubation was extended to 2 hours, as shown in Figure 15B, both the A-L-N and S+A-N constructs exhibited distinct NeuAc bands whose intensity approached that of the target product reference band. This indicates that these constructs efficiently catalyze NeuAc synthesis upon extended reaction time. Compared to the 1-hour incubation group, the 2-hour incubation group exhibited more pronounced product bands at 37°C, indicating that appropriately extending the reaction time significantly enhances catalytic efficiency.

Collectively, these results indicate that the A-N construct favors ManNAc accumulation in short-duration reactions, while the A-L-N and S+A-N constructs promote final NeuAc production after extended reaction times, with 37°C and 2 hours representing optimal conditions. This phenomenon indicates significant differences in catalytic properties among different enzyme constructs. The introduction of linkers or scaffolds may promote the conversion of reaction intermediates into end products, thereby enhancing overall catalytic efficiency.

Figure 15 TLC for NAL-AGE (A-N), AGE-Linker-NAL (A-L-N), and Scaffold + NAL-AGE (S+A-N) under 25°C, 37°C, and 42°C for 1 hour,and 2 h

3.2 Enzyme-linked immunosorbent assay (ELISA)

To evaluate the catalytic performance of the dual-enzyme system, an ELISA-based quantification assay was conducted.

Table 1 The standard table of Neu5Ac

A 4-parameter logistic regression (4PL) model was employed for standard curve fitting, with the equation shown in Figure 16, which exhibited a high correlation coefficient (R² = 0.9999).

Figure 16 The equation and the standard curve for the ELISA experiment

The absorbance measured in the experiment was substituted into the equation for calculation to obtain the initial detected concentration of sialic acid. The actual content of sialic acid in the samples was then derived by multiplying by the dilution factor (5 times), as shown in Table 2.

Table 2

The final results were plotted in Figure 17. For the enzyme construct A-N, the sialic acid synthesis efficiency was the highest at a concentration of 1 mg/mL, and decreased at medium to high concentrations (5-10 mg/mL). At a low concentration, the rate of the first-step reaction in sialic acid synthesis was slightly higher than that of the second-step reaction, resulting in the overall reaction proceeding toward sialic acid synthesis. Under high concentration conditions, the product sialic acid was decomposed into intermediate products by the NAL enzyme, which were further decomposed by the AGE enzyme.

For the enzyme construct A-L-N, the content of sialic acid showed an upward trend as the concentration increased. Compared with the maximum value of 1 mg/mL in the A-N group, the detection values at various concentrations in the A-L-N group were relatively stable. This indicated that the role of the Linker is to stabilize enzyme function rather than overactivate it.

For the enzyme construct S+A-N, all detection indicators were generally higher than those in the A-L-N group. At the concentration of 10 mg/mL, the detected value was second only to that of the A-N group at 1 mg/mL. Through the clustering effect, the Scaffold formed high-concentration A-N enzyme clusters locally, significantly improving the reaction efficiency. Meanwhile, by binding to the AGE and NAL enzymes, the Scaffold induced their conversion from an inactive conformation to an active conformation, which could directly enhance the catalytic efficiency of individual enzymes. This was the core reason why the detected values of the S+A-N group were higher than those of the A-L-N group.

In conclusion, the enzyme construction Scaffold+AGE-NAL can maximize the efficiency of sialic acid synthesis at medium to high concentrations, making it the optimal solution for efficiently increasing sialic acid production.

Figure 17 Concentration of Neu5Ac tested by the ELISA Kit calculated by the standard curve

3.3 High-Performance Liquid Chromatography (HPLC)

The catalytic experiment of the A-N enzyme construct was illustrated in Figure 18A. The sialic acid content reached its highest level at a reaction time of 60 minutes across all three reaction temperatures. As the reaction time prolonged, the sialic acid yield exhibited a decreasing trend. We hypothesize that this phenomenon was associated with the reversibility of the catalytic reaction. When the content of the intermediate product ManNAc-6P accumulated to a certain extent, the reaction proceeded in the reverse direction. Consequently, the substrate for the second step of the reaction was reduced, leading to a decrease in the final yield of sialic acid. Under the condition of a reaction time of 60 minutes, the reaction temperature of 42 °C yielded the highest accumulation of sialic acid. Overall, the optimal reaction system for enzyme construct A-N is as follows: temperature 42 °C + reaction time 60 minutes.

The similar results were also observed when enzyme constructs A-L-N and S+A-N were used, as shown in Figure 18B and 18C. Notably, when the S+A-N enzyme construct was used, the accumulation of sialic acid rebounded under the conditions of 42 °C and a reaction time of 240 minutes. This may be attributed to the fact that the protein scaffold effectively reduced the occurrence of reversible reactions under such reaction conditions. Overall, the optimal reaction systems for both enzyme constructs A-L-N and S+A-N were as follows: temperature 42 °C + reaction time 60 minutes.

Figure 18 Sialic acid production by each enzyme construct under different reaction times and temperatures

The sialic acid production of the three enzyme constructs under the condition of 42 °C, 240 minutes was illustrated in Figure 19. Although the sialic acid content decreased as the reaction time prolonged, the enzyme construct S+A-N exhibited the smallest degree of decrease and achieved the highest final accumulated product concentration. This indicated that construct S+A-N can effectively reduce the occurrence of reversible reactions and was thus the optimal enzyme construct for producing sialic acid.

Figure 19 Sialic acid production by three enzyme constructs under 42 ℃, 240 minutes

Through detailed analysis of the HPLC peak chart in Figure 20, the successful generation of sialic acid and the significant catalytic advantage of the S+A-N construct can be further confirmed.

In the chromatogram of the S+A-N construct, a distinct and sharp peak appears at approximately 11 minutes retention time, which aligns perfectly with the characteristic peak of sialic acid in the standard sample. This peak exhibits symmetry and a large area, indicating high product concentration and excellent separation with no significant tailing.

In contrast, although product peaks are detectable at the same retention time for other constructs (e.g., A-N and A-L-N), their response intensity is significantly lower than that of S+A-N. These peaks are smaller and accompanied by minor impurity peaks. Particularly in the A-N sample, a chromatographic peak belonging to the intermediate ManNAc (retention time approximately 3 minutes) appeared before the target product peak, further indicating incomplete reaction and lower catalytic efficiency.

In summary, the HPLC quantitative results align with previous activity assessment conclusions: the S+A-N construct most effectively catalyzes sialic acid synthesis, yielding the highest product yield and optimal purity, fully demonstrating its superiority as a highly efficient catalytic module.

Figure 20 Chromatograms of sialic acid content under 42 ℃, 240 minutes

Our experimental data have clearly demonstrated the successful enzymatic synthesis of sialic acid, confirming the catalytic functionality of the enzyme constructs under the applied conditions. Nevertheless, several aspects of the methodology offer opportunities for further refinement that would enhance the robustness and interpretability of the results. For instance, the developing solvent system and coloring reagents used in the thin-layer chromatography (TLC) assays could be systematically optimized to improve resolution between substrates and products, thereby allowing more accurate qualitative and quantitative analysis. In addition, future studies would benefit from more comprehensive functional assessments of the enzyme constructs—such as kinetic studies, stability tests under various conditions, and structural analyses—to better understand their catalytic mechanisms and performance limitations. Implementing these improvements would not only strengthen the validity of our current conclusions but also provide deeper insights into the enzymatic synthesis process, further establishing a reliable foundation for subsequent research and potential applications.