Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2025 Jan 31;10(5):4792–4800. doi: 10.1021/acsomega.4c09986

Development of a Multienzyme Cascade for Salvianolic Acid A Synthesis from l-Tyrosine

Mingxi Zhang †, Jiayi Zhong †, Yuekai Zhang †, Weijie Wang †, Weirui Zhao †,*, Sheng Hu †, Changjiang Lv §, Jun Huang §, Lehe Mei †,‡,∥,*
PMCID: PMC11822487  PMID: 39959076

Abstract

graphic file with name ao4c09986_0006.jpg

Salvianolic acid A (SAA) has an important application value for preventing and treating cardiovascular diseases. In this study, we developed a novel multienzyme cascade system for the efficient biosynthesis of SAA, utilizing l-tyrosine (l-Tyr) as a cost-effective and stable starting material. The cascade system incorporated four enzymes: membrane-bound l-amino acid deaminase from Proteus vulgaris (Pvml-AAD), d-lactate dehydrogenase from Pediococcus acidilactici (Pad-LDH), 4-hydroxyphenylacetate 3-hydroxylase from Escherichia coli (EcHpaBC), and formate dehydrogenase from Mycobacterium vaccae N10 (MvFDH). All reaction steps in the cascade system were thermodynamically favorable. In addition, to avoid generating an unstable intermediate (3,4-dihydroxyphenyl-pyruvate, DHPPA), which was produced owing to the promiscuity of EcHpaBC and Pad-LDH, we performed the cascade system according to the reaction sequence of deamination, chiral reduction, and ortho-hydroxylation. Under optimized conditions, the developed cascade system yielded 81.67 mM SAA from an initial concentration of 100 mM l-Tyr, corresponding to a yield of 81.67%.

1. Introduction

Salvianolic acid A (SAA), (R)-(+)–3,4-dihydroxyphenyllactic acid, is a key water-soluble constituent of Salvia miltiorrhiza (danshen), a widely recognized Chinese medicinal herb.1,2 SAA exhibits a range of pharmacological activities, including antioxidant,3 antithrombotic, blood circulation-promoting,4 and antiatherosclerotic activities.5 These bioactivities have supported its clinical application in treating various cardiovascular diseases, such as coronary artery disease and angina.4,6 Additionally, SAA has shown other important pharmacological effects, such as neuroprotection against tat-induced neurotoxicity,7 antianxiety,8 antitumor effects,9 prevention of DNA damage,10 inhibition of apoptosis,11 and promotion of cell proliferation.12

SAA can be obtained through physical extraction, chemical synthesis, and biosynthesis. Traditionally, SAA has been extracted from the rhizomes of S. miltiorrhiza. However, this method is limited by the low concentration of SAA in the crude root (0.045%),13 the scarcity of S. miltiorrhiza resources, and an unsatisfactory yield, making it insufficient to meet the growing market demand. Chemical synthesis methods have been developed to overcome these limitations, but they result in a racemic mixture of SAA, requiring complex isomer separation, and present drawbacks such as harsh reaction conditions, multistep procedures, environmental concerns, and high material costs.6,14 Consequently, there is a strong demand for stereoselective and ecofriendly methods for SAA production. Multienzyme cascade reactions, considered as a promising platform for next-generation biomanufacturing,15 have attracted considerable attention due to their potential to achieve near-theoretical product yields, increase mass transfer efficiency, and offer innovative engineering opportunities.16−18 Recently, several efficient multienzyme cascade systems for SAA synthesis have been developed.19−21 For example, Han et al. developed a one-pot multienzyme cascade pathway using tyrosine aminotransferase from Escherichia coli (EcTyrB), d-isomer-specific 2-hydroxyacid dehydrogenase from Lactobacillus frumenti (LfD2-HDH), and glutamate dehydrogenase from Clostridium difficile (CdGluD) to synthesize SAA from l-DOPA, yielding 98.3% at 20 mM l-DOPA and 51.5% at 100 mM l-DOPA under optimal conditions.19 Similarly, Hu et al. reported a two-step enzymatic cascade for SAA production from l-DOPA, achieving 48.3 mM SAA from 50 mM l-DOPA.20 To date, to our knowledge, all reported multienzyme cascade systems for SAA synthesis have utilized l-DOPA as the initial substrate. However, l-DOPA is both costly and unstable. It is noteworthy that the decomposition of l-DOPA under alkaline conditions leads to a progressive darkening of the reaction mixture,19 which could potentially pose challenges for further downstream applications and process optimization. Therefore, some methods, such as the addition of ascorbic acid,20 should be implemented to prevent it from being oxidized during its usage. Thus, an efficient multienzyme cascade system for SAA synthesis with a more stable and cheaper substrate is highly expected.

The promiscuity of enzymes allows them to act on a broad range of substrates. Incorporating promiscuous enzymes into synthetic pathways can be advantageous for producing a range of valuable compounds with similar structures. However, this versatility can also result in multiple reaction pathways, leading to the accumulation of unwanted byproducts, such as unstable intermediates, causing the loss of the carbon sources.22,23 To achieve high yields of the target product, it is essential to avoid the disadvantages of promiscuous enzymes and steer multienzyme cascade reactions toward pathways that favor product formation.

In this study, we developed an in vitro multienzyme cascade system to synthesize SAA from l-tyrosine (l-Tyr), a more stable and cost-effective substrate than l-DOPA. This system comprised membrane-bound l-amino acid deaminase from Proteus vulgaris (Pvml-AAD), d-lactate dehydrogenase from Pediococcus acidilactici (Pad-LDH), 4-hydroxyphenylacetate 3-hydroxylase from E. coli (EcHpaBC), and formate dehydrogenase from Mycobacterium vaccae N10 (MvFDH). Initially, l-Tyr was efficiently deaminated into 4-hydroxyphenylpyruvic acid (4-HPPA) by Pvml-AAD. The subsequent conversion of 4-HPPA to SAA involved chiral reduction of the keto groups and ortho-hydroxylation of the phenol ring, catalyzed by Pad-LDH and EcHpaBC, respectively. Due to the promiscuity of both enzymes, the conversion of 4-HPPA to SAA could proceed through two pathways (Figure 1). In pathway 1, Pad-LDH catalyzed the reduction of 4-HPPA to 4-hydroxyphenyllactic acid (4-HPLA), which was then ortho-hydroxylated by EcHpaBC to produce SAA. In pathway 2, EcHpaBC first ortho-hydroxylated 4-HPPA to 3,4-dihydroxyphenylpyruvic acid (DHPPA), which was subsequently reduced to SAA by Pad-LDH. However, the intermediate DHPPA in pathway 2 was unstable under reaction conditions, leading to the loss of the carbon sources, whereas the metabolites in pathway 1 were more stable. To optimize SAA production, we performed a multistep cascade reaction strategy that favored pathway 1, avoiding the formation of unstable intermediates in pathway 2. This approach resulted in a highly efficient SAA synthesis, demonstrating an excellent productivity.

Figure 1.

Figure 1

Open in a new tab

Pathways for SAA synthesis used in this study. The components include Pvml-AAD (membrane-linked l-amino acid deaminase), Pad-LDH (d-lactate dehydrogenase), MvFDH (formate dehydrogenase), EcHpaBC (4-hydroxyphenylacetic acid-3-hydroxylase), l-Tyr (l-tyrosine), 4-HPPA (4-hydroxyphenylpyruvic acid), 4-HPLA (4-hydroxyphenyllactic acid), DHPPA (3,4-dihydroxyphenylpyruvic acid), and SAA (salvianic acid A).

2. Materials and Methods

2.1. Chemicals

l-Tyr, 4-HPPA, 4-HPLA, DHPPA, and SAA were sourced from Yuanye Biotech Co., Ltd. (Shanghai, China). NADH was purchased from Sangon Co., Ltd. (Shanghai, China). FADH2 was purchased from Macklin Co., Ltd. (Shanghai, China). 2,4-Dinitrophenylhydrazine (DNP) was obtained from China Medicine Co., Ltd. Isopropyl-β-d-thiogalactopyranoside, ampicillin, and kanamycin were obtained from Sangon Co., Ltd. (Shanghai, China).

2.2. Construction of Plasmid and Strain

The primers, strains, and plasmids utilized are detailed in Table S1. The DNA fragments of EcHpaBC from E. coli (GenBank accession no. CP053602.1) and Pad-LDH from P. acidilactici DSM 20284 (GenBank accession no. AEEG01000002) were inserted into the pETDuet-1 vector between the BamH I and Nco I and the pRSFDuet-1 vector between the Nco I and Not I using restriction enzyme digestion and ligation techniques, respectively. For the coexpression of FDH from M. vaccae N10, MvFDH (GenBank accession no. AB072394.1) was codon optimized, synthesized, and inserted into the pETDuet-hpabc or pRSFDuet-dldh vector between the Nde I and Xho I multiple cloning sites. DNA sequencing verification was performed, resulting in the formation of plasmids pET-28a-mlaad, pETDuet-hpabc-fdh, and pRSFDuet-dldh-fdh. The plasmids were transformed into E. coli BL21 (DE3) strains using standard heat shock methods.

2.3. Preparation of a Recombinant E. coli Biocatalyst

The recombinant strains BL21(DE3)-pET-28a-mlaad, BL21(DE3)-pETDuet-hpabc-fdh, and BL21(DE3)-pRSFDuet-dldh-fdh were cultured in 5 mL of the Luria–Bertani broth (LB) medium on a rotary shaker at 37 °C and 200 rpm overnight. The seed mixture was then inoculated into LB media supplemented with the corresponding antibiotic (50 μg/mL kanamycin, 100 μg/mL ampicillin) at a 2% inoculum volume and cultured on a shaker at 37 °C and 200 rpm until the OD600 reached 0.6–0.8. IPTG was added at a final concentration of 0.05 mM, and the culture was maintained at 28 °C and 150 rpm for 6 h. After the induction culture process was completed, the cells were harvested by centrifugation at 10,000 g for 1 min at 4 °C and washed with 1× phosphate-buffered saline (PBS) buffer at pH 7.4.

2.4. Characterization of the BL21(DE3)-pET-28a-mlaad Whole-Cell Catalyst for 4-HPPA Synthesis

The induced bacteria were collected and added to the reaction system for catalysis. The reaction details were as follows: a specific amount of the BL21(DE3)-pET-28a-mlaad whole-cell catalyst was mixed with 3 mL of a substrate mixture containing 5 mM l-Tyr and 50 mM KPi buffer. The mixture was placed in a constant-temperature oscillator set at 200 rpm for 15 min. The optimized conditions included a pH range of 6.5–10.5, a temperature range of 23–51 °C, and a bacterial concentration with OD600 values between 1.5 and 9. The reaction were quenched by adding an equal volume of 20% trichloroacetic acid (TCA) solution, and the DNP method was used to measure the 4-HPPA content.

2.5. Biotransformation of 4-HPPA to SAA by a Permeabilized Biocatalyst

To overcome the cell envelope barrier and prevent the diffusion of substrates and products, BL21(DE3)-pRSFDuet-dldh-fdh and BL21(DE3)-pETDuet-hpabc-fdh cells were permeabilized with 1% hexane (v/v) for 10 min at the end of the induction period before initiating the reaction.20 Subsequently, the cells were washed with 1× PBS buffer at pH 7.4 and then centrifuged at 10,000 g for 1 min at 4 °C. The effects of pH and temperature on product conversion were systematically investigated. For BL21(DE3)-pRSFDuet-dldh-fdh, a reaction mixture containing 50 mM KPi buffer, 25 mM 4-HPPA or DHPPA, 50 mM sodium formate, and 1 mM NAD+ was used to optimize the reaction conditions. Biotransformation reactions were conducted at varying pH values and temperatures: DHPPA (16–44 °C, pH 4.5–8.5) and 4-HPPA (23–51 °C, pH 4.5–8.5), using a bacterial concentration of OD600 = 1.

Similarly, for BL21(DE3)-pETDuet-hpabc-fdh, the permeabilized biocatalyst was added to a 3 mL substrate mixture containing 50 mM KPi buffer, 25 mM 4-HPPA or 4-HPLA, 50 mM sodium formate, 1 mM NAD+, and 1 mM FADH2. Reactions were conducted under varying pH and temperature conditions: for 4-HPLA (23–51 °C, pH 6.5–10.5) and for 4-HPPA (23–51 °C, pH 5.5–9.5), with a bacterial concentration of OD600 = 7.5. All reactions were performed in a temperature-controlled shaking incubator at 200 rpm for 1 h and were quenched by adding an equal volume of 1 M HCl. The concentrations of the corresponding products in the reaction mixture were determined by using high-performance liquid chromatography (HPLC).

2.6. Stability of Intermediates and Products

A 20 mL portion of each solution was prepared at a concentration of 50 mM in a 100 mL shake flask. The flasks were then placed on a shaker set to 200 rpm to observe the degradation of each substance over a period of 12 h.

2.7. Analysis Methods

The DNP method was used to quantify 4-HPPA accurately in the reaction mixture. The method involved the following steps: first, a 100 μL aliquot of the reaction mixture was diluted to an appropriate concentration. Then, 90 μL of a 20% TCA solution and 40 μL of a 20 mM 2,4-dinitrophenylhydrazine solution were added. The contents were thoroughly mixed and allowed to react in the dark at room temperature for 15 min. Next, 800 μL of a 0.8 M NaOH solution was added to induce color development. After incubation for 15 min in the dark, the absorbance of the solution was measured at 520 nm using a spectrophotometer.24

The concentrations of 4-HPLA and SAA in the reaction mixture were quantified using HPLC with an LC-2030 system equipped with a Hypersil ODS2 C18 column (5 μm, 250 mm × 4.6 mm, ELITE, China) purchased from Dalian Elite Company. The detection method involved using 0.5% acetic acid in water (solvent A) and methanol (solvent B) at a ratio of 80:20 (v/v) at a flow rate of 1.0 mL·min–1. The analytical wavelength was set at 281 nm, the column temperature at 30 °C, and the injection volume at 10 μL; the detection time was 15 min. For DHPPA detection, its concentration was measured by HPLC using the same LC-2030 system and column. A linear gradient elution was employed, utilizing 0.05% trifluoroacetic acid in water (solvent A) and 0.05% trifluoroacetic acid in methanol (solvent B), at a flow rate of 1.0 mL·min–1. The A/B ratios were 90:10, 0:100, 0:100, and 90:10 at respective time points of 0, 20, 23, and 26 min. The analysis was conducted at a wavelength of 281 nm, with the column maintained at 30 °C, an injection volume of 10 μL, and a total detection time of 26 min.25

3. Results

3.1. Thermodynamic Analysis from l-Tyr to SAA

We performed thermodynamic analyses of the synthetic pathways for SAA production from l-Tyr using eQuilibrator, as shown in Table 1.26 The overall Gibbs free energy change (ΔrG′total) for the entire multienzyme cascade reaction was calculated to be a significantly negative value of −856.1 kJ/mol, indicating a strong thermodynamic driving force for the conversion of l-Tyr to SAA, making the process nearly irreversible. Furthermore, the Gibbs free energy change (ΔrG′) for each individual step in both synthetic pathways was negative. Our system exhibits more thermodynamic advantages than the multienzyme system developed by Han et al. for SAA production from l-DOPA. In their system, while the standard Gibbs free energy changes for EcTyrB and LfD2-HDH were favorable (−8.2 and −19.5 kJ/mol, respectively), the Gibbs free energy change of CdgluD (36.5 kJ/mol) was thermodynamically unfavorable.19

Table 1. Thermodynamic Characterization of the Two SAA Biosynthetic Pathways and Corresponding Reactionsa.

Reactions ΔrG′ (kJ/mol) J+ (%) J– (%)
Step 1 l-Tyr + O2 + 2H+ → 4-HPPA + NH3 + H2O –356.2 ∼1 ∼0
Pathway 1 4-HPPA + NADH + H+ → 4-HPLA + NAD+ –4.5 ∼0.86 ∼0.14
4-HPLA + O2 + FADH2 → SAA + H2O + FAD –475.8 ∼1 ∼0
Pathway 2 4-HPPA + O2 + FADH2 → DHPPA + H2O + FAD –416.7 ∼1 ∼0
DHPPA + NADH + H+ → SAA + NAD+ –63.6 ∼1 ∼0
Overall reaction l-Tyr +2O2 + 3H+ + NADH + FADH2 → NH3 + 2H2O + NAD+ + FAD + SAA –836.5 ∼1 ∼0
Open in a new tab
a

Thermodynamic profiles of the Gibbs free energy change for individual steps of the pathways were determined under physiological conditions (pH 7.0, ionic strength I = 0.20, pMg = 3.0, substrates = 1 mM) using eQuilibrator 3.0; ΔrG′= −RT ln(J+/J–); J+ represents the forward flux in SAA synthesis, J– represents the backward flux in SAA cleavage, and the fraction in the enzymatic flux is indicated. The gas constant (R) is 8.314 J/(mol kilocalories), and the temperature (T) is 303.15 K (30 °C). The symbol “∼” represents “close to”.

3.2. Production of 4-HPPA from l-Tyr

The initial step in the cascade involves the deamination of l-Tyr to 4-HPPA. Four kinds of enzymes are capable of generating α-keto acids from l-amino acids, including amino acid transferases (AATs), amino acid dehydrogenases (ADHs), l-amino acid oxidases (LAAOs), and membrane-bound l-amino acid deaminases (ml-AADs).27 Among these, ml-AAD is the preferred enzyme due to its high efficiency, as it does not require coenzymes or amino acid receptors, which are essential for AATs and ADHs. Additionally, unlike LAAOs, ml-AAD does not produce harmful hydrogen peroxide byproducts that could denature the enzyme or degrade newly formed α-keto acids.28,29 Another advantage of ml-AAD is its membrane-bound nature, anchoring proteins to the outer cytomembrane, which avoids the cell mass transfer resistance of the cell envelope toward the substrate and product, making it ideal for whole-cell catalysis.30 Among the reported ml-AADs, Pvml-AAD demonstrates superior catalytic activity toward aromatic amino acids such as l-Tyr and l-phenylalanine.24,31 Thus, a whole-cell E. coli biocatalyst expressing Pvml-AAD (BL21(DE3)-pET-28a-mlaad) was used to catalyze this reaction.

To enhance the efficiency of the BL21(DE3)-PET-28a-mlaad system, key catalytic parameters were optimized. The optimal pH for the whole-cell biocatalyst was determined to be 6.5–10.5. Experimental results showed that the catalyst exhibited the maximum activity at a pH value of 8.5, retaining over 80% of its activity within the range of pH 7.5–9.5 (Figure 2a). The reason for this phenomenon was that the alkaline environment likely facilitated the deprotonation of the amino group in l-Tyr, making it more susceptible to deamination by Pvml-AAD. The optimal temperature of the biocatalyst was found to be 23–51 °C, with the highest activity observed at 37 °C. The catalytic activity increased progressively as the temperature rose from 23 to 37 °C but decreased above 37 °C (Figure 2b). To optimize cell density for the maximum biotransformation efficiency, reactions were conducted with cell concentrations ranging from 1.5 to 9 OD600/mL, and at a cell density of 6 OD600/mL, as the bacterial concentration increased, the growth trend of 4-HPPA production exhibited a significant decline (Figure 2c) considering the economic cost of producing biocatalysts, the condition where 6 OD600/mL was performed in this study. Based on these findings, the optimal conditions for whole-cell catalysis were determined to be a cell concentration of 6 OD600/mL, a pH of 8.5, and a temperature of 37 °C.

Figure 2.

Figure 2

Open in a new tab

Effects of pH, temperature, and bacterial concentration on the catalytic activity of Pvml-AAD-engineered bacteria are shown in (a), (b), and (c), respectively.

3.3. Synthesis of SAA from 4-HPPA through Chiral Reduction and ortho-Hydroxylation Reactions

The production of SAA from 4-HPPA requires two reactions: the chiral reduction of keto groups and ortho-hydroxylation of the phenol ring. For the reduction of the ketone group, NAD(P)H-dependent lactate dehydrogenases (LDHs) are commonly employed in biocatalysis due to their superior catalytic efficiency toward various keto-based substrates. These enzymes have been widely utilized in the synthesis of chiral α-hydroxycarboxylic acids.24,32 Since SAA is a d-optically active isomer, the use of a d-specific LDH is necessary to reduce the keto groups of 4-HPPA. In this study, we selected the d-LDH from P. acidilactici, which has demonstrated high catalytic activity toward a series of aromatic α-keto acids, including 4-HPPA.33 To regenerate NADH for Pad-LDH, we coexpressed Pad-LDH with formate dehydrogenase (FDH) from M. vaccae N10 in E. coli BL21(DE3).34 The resulting recombinant E. coli BL21(DE3)-pRSFuet-dldh-fdh was used for the chiral reduction of keto groups in this study.

For the ortho-hydroxylation of the phenol ring, 4-hydroxyphenylacetate-3-hydroxylase (4HPA3H) was selected as the most suitable enzyme due to its high expression levels and catalytic efficiency.35 This enzyme consists of two components: HpaB (monooxygenase) and HpaC (FAD oxidoreductase). HpaB is the larger subunit and functions as an FADH2-dependent monooxygenase, while HpaC is responsible for supplying FADH2 to HpaB by oxidizing NAD(P)H.36,37 Among the various 4HPA3Hs, the E. coli 4HPA3H (EcHpaBC) exhibited strong ortho-hydroxylation activity not only toward 4HPA analogs but also a variety of structurally complex phenolic compounds. Therefore, EcHpaBC was used in this study to perform the ortho-hydroxylation of the phenol ring.38 Additionally, to regenerate NADH for EcHpaBC, the EcHpaBC was coexpressed with MvFDH in E. coli BL21(DE3), and the resulting recombinant strain, BL21(DE3)-pETDuet-hpabc-fdh, was used for the ortho-hydroxylation reaction.

Due to the promiscuity of both the ortho-hydroxylase EcHpaBC and the reductase Pad-LDH, two distinct reaction pathways were established for converting 4-HPPA to SAA, depending on whether hydroxylation or reduction occurred first (Figure 1; pathways 1 and 2). Both pathways were thermodynamically favorable. To optimize the efficiency of the multienzyme cascade system, it was crucial to determine the appropriate reaction conditions and assess the stability of intermediate metabolites in each pathway.

The catalytic performance of BL21(DE3)-pRSFDuet-dlhd-fdh and BL21(DE3)-pETDuet-hpabc-fdh was evaluated on their respective substrates in different pathways (Figure 3). To reduce the permeability barrier of the cell membrane and enhance the catalytic activity, the cell envelope was permeabilized using 1% hexane. For BL21(DE3)-pRSFDuet-dlhd-fdh, which catalyzed the chiral reduction reaction, when 4-HPPA was used as the substrate (Figure 1; pathway 1), the highest product yield was achieved at 44 °C (Figure 3a) and pH 5.5 (Figure 3b). However, when DHPPA (Figure 1; pathway 2) was used as the substrate, the optimal conditions shifted to 30 °C (Figure 3a) and pH 6.5 (Figure 3b). Notably, the optimal temperature for the reduction step differed significantly between the two pathways, likely due to the instability of DHPPA, the intermediate in pathway 2. For BL21(DE3)-pETDuet-hpabc-fdh, which was responsible for the ortho-hydroxylation reaction, when 4-HPPA was used as the substrate (Figure 1; pathway 2), optimal product yields were exhibited at 37 °C (Figure 3c) and pH 7.5 (Figure 3d). In contrast, when 4-HPLA was used as the substrate (Figure 1; pathway 1), the optimal pH shifted to 9.5 (Figure 3d), while the optimal temperature remained constant at 37 °C (Figure 3c). It was interesting to note that despite the structural similarity between 4-HPLA and 4-HPPA, the pH requirements for their ortho-hydroxylation reactions were significantly different. We observed that during the conversion of 4-HPPA to DHPPA in pathway 2, the reaction mixture appeared brown at high pH, suggesting that DHPPA was unstable and prone to degradation under neutral and alkaline conditions, a finding confirmed by subsequent experiments.

Figure 3.

Figure 3

Open in a new tab

(a) The effect of the temperature on the yield of reduction products was investigated. Reactions were performed at pH 5.5 with 25 mM either 4-HPPA (pathway 1 substrate) or DHPPA (pathway 2 substrate) in reaction mixtures containing recombinant cells at an OD600 of 1. The temperature range for pathway 1 was 23–51 °C, and for pathway 2, it was 16–44 °C. (b) The effect of pH on the yield of reduction products was investigated. Reactions were performed at 30 °C with reaction mixtures containing recombinant cells at an OD600 of 1 and 25 mM either 4-HPPA (pathway 1 substrate) or DHPPA (pathway 2 substrate). The pH range tested was 4.5–8.5. (c) The effect of temperature on the yield of hydroxylation products was investigated. Reactions were performed at pH 7.5 with recombinant cells at an OD600 of 7.5 and 25 mM either 4-HPLA (pathway 1 substrate) or 4-HPPA (pathway 2 substrate). The temperature range tested was 23–51 °C. (d) The effect of pH on the yield of hydroxylation products was investigated. Reactions were performed at 37 °C with 25 mM either 4-HPPA (pathway 1 substrate) or DHPPA (pathway 2 substrate) in reaction mixtures containing recombinant cells at an OD600 of 7.5. The pH range for pathway 1 was 6.5–10.5, and for pathway 2, it was 5.5–9.5.

Minimizing carbon source loss is crucial for the stable and efficient production of target compounds, necessitating an evaluation of the stability of the intermediates. Moreover, the darkening observed in the reaction mixture during pathway 2 prompted further investigation of the stability of intermediate metabolites. To address this, we assessed the stability of 4-HPPA, the intermediate products 4-HPLA and DHPPA, and the final product SAA under their respective optimized reaction conditions for each pathway (Figure 4). Our analysis revealed that both 4-HPPA and SAA retained over 85% of their initial concentrations after 12 h under their respective reaction conditions in both pathways 1 and 2, indicating that these compounds are stable under the examined conditions. In pathway 1, the intermediate 4-HPLA demonstrated good stability during both the chiral reduction step (pH 5.5, 44 °C) and the subsequent ortho-hydroxylation reaction step (pH 9.5, 37 °C). However, in pathway 2, the intermediate DHPPA was found to be highly unstable, degrading significantly under both the optimized ortho-hydroxylation reaction conditions (pH 7.5, 37 °C) and the subsequent reduction conditions (pH 6.5, 30 °C). This degradation resulted in carbon source loss and complications in the purification process. Due to the instability of DHPPA and its associated challenges, pathway 2 was deemed unsuitable for the in vitro multienzyme cascade synthesis of SAA.

Figure 4.

Figure 4

Open in a new tab

Degradation curves of 4-HPPA, DHPPA, 4-HPLA, and SAA observed under their respective optimal reaction conditions in both pathways 1 and 2. Specifically, (a) degradation of 4-HPPA over 12 h under the following conditions: pH 5.5, 37 °C; pH 5.5, 44 °C; pH 7.5, 37 °C; pH 8.5, 37 °C. (b) Degradation of 4-HPLA over 12 h under different conditions: pH 5.5, 37 °C; pH 5.5, 44 °C; pH 9.5, 37 °C. (c) Degradation of DHPPA over 12 h under various conditions: pH 6.5, 30 °C; pH 6.5, 37 °C; pH 7.5, 37 °C. (d) Degradation of SAA after 12 h under different conditions: pH 6.5, 30 °C; pH 6.5, 37 °C; pH 9.5, 37 °C.

3.4. Three-Step Synthesis of SAA from l-Tyr

Based on the stability analysis, pathway 1, which utilizes 4-HPLA as the intermediate product, was identified as the more suitable option for the synthesis of SAA. In contrast, pathway 2, which involves DHPPA as the intermediate, was deemed unsuitable due to its poor intermediate stability. Additionally, the promiscuity of EcHpaBC and Pad-LDH made them unsuitable for a one-pot reaction, as this could lead to the formation of undesired byproducts through multiple pathways. Although enzyme specificity could potentially be improved through protein engineering, this approach poses significant challenges, especially for substrates with highly similar structures.39 As a result, we performed a multistep cascading reaction process for the synthesis of SAA based on pathway 1. This approach allowed each reaction step to proceed under its optimal conditions: BL21(DE3)-pET28a-mlaad (pH 8.5, 37 °C), BL21(DE3)-pETDuet-dlhd-fdh (pH 5.5, 44 °C), and BL21(DE3)-pETDuet-hpabc-fdh (pH 9.5, 37 °C).

The first step of the reaction, converting l-Tyr to 4-HPPA, was carried out using BL21(DE3)-pET28a-mlaad under the optimal conditions (6 OD600/mL cells, pH 8.5, and 37 °C). When 100 mM l-Tyr was used as a substrate, the insoluble tyrosine gradually dissolved, and the reaction mixture became clearer as the process progressed. After 6 h, the highest yield of 4-HPPA (87.6 mM) was obtained, corresponding to a yield of 87.6% (Figure 5a). At this point, no significant insoluble tyrosine remained in the reaction mixture. Extending the reaction time led to a slight decrease in 4-HPPA concentration, indicating that 6 h was the optimal reaction time for this step. After the reaction, the BL21(DE3)-pET28a-mlaad cells were removed by centrifugation, and the pH of the mixture was adjusted to 5.5.

Figure 5.

Figure 5

Open in a new tab

(a) Time profile of 4-HPPA production from l-Tyr using the BL21(DE3)-pET-28a-mlaad whole-cell catalyst under the optimal conditions. (b) Effects of the permeabilized BL21(DE3)-pRSFDuet-dldh-fdh cell concentration on 4-HPLA. (c) Effect of the permeabilized BL21(DE3)-pETDuet-hpabc-fdh cell concentration on SAA production.

In the second step, 175.2 mM sodium formate (1:2 molar ratio of 4-HPPA to sodium formate), 1 mM NAD+, and 2–8 OD600/mL permeabilized BL21(DE3)-pETDuet-dlhd-fdh cells were added to convert 4-HPPA to 4-HPLA at 44 °C. The conversion of 4-HPLA increased with cell concentration, achieving a maximum yield of 83.4 mM after 6 h, corresponding to a yield of 95.3% from 4-HPPA (Figure 5b).

In the final step, the recombinant cells were removed, and the pH of the reaction mixture was adjusted to 9.5. Next, 1 mM FAD, 166.8 mM sodium formate (1:2 molar ratio of 4-HPLA to sodium formate), and 6–24 OD600/mL permeabilized BL21(DE3)-pETDuet-hpabc-fdh cells were added to catalyze the ortho-hydroxylation reaction at 37 °C. The optimal cell concentration was OD600 = 12, at which 81.7 mM SAA was produced after 5 h (Figure 5c). Higher cell concentrations resulted in reduced SAA production, likely due to mass transfer limitations. Thus, in our developed three-step biotransformation process, 100 mM l-Tyr was converted to 81.7 mM SAA, in a yield of 81.7%.

In recent years, the development of multienzyme cascade systems for the synthesis of SAA has emerged as a prominent research focus. In our study, l-Tyr was strategically employed as the starting substrate for the SAA synthesis. Remarkably, compared to l-DOPA, which had been prevalently used in prior studies, l-Tyr is more stable. Additionally, as a common and widely available amino acid, l-Tyr represents a more accessible and economically viable option in contrast to l-DOPA. In addition, both intermediates and substrates showed good stability under our preparation conditions, effectively avoiding the loss of carbon sources. Furthermore, our devised pathway for SAA synthesis originating from l-Tyr manifests negative Gibbs free energy alterations at each reaction stage, indicating its thermodynamic superiority. What’s more, for operating the developed multienzyme cascade system, we used a multistep strategy. Although this method increases operating costs and complexity, such as separating cells from the reaction solution, it effectively avoids the issue of the promiscuity of certain enzymes, reducing the loss of the carbon source, and ensures that each reaction in the multienzyme cascade occurs under the optimal conditions. In view of the pronounced price discrepancy between SAA and l-Tyr, our proposed strategy demonstrates considerable potential for large-scale production. One major challenge in constructing multienzyme cascade reactions is the promiscuity of certain enzymes, which can lead to the formation of undesired intermediates. Although spatially or temporally isolating the enzymes from nontarget substrates can solve the problem, this measure complicates the process. Therefore, it is well worth improving EcHpaBC specificity toward different substrates (4-HPPA and 4-HPLA) through protein engineering in future studies, even if it is a challenging task.

4. Conclusion

In this study, we developed a novel in vitro multienzyme cascade synthesis based on thermodynamic feasibility analysis and stability assessments of various pathways. Using the stable and cost-effective substrate l-Tyr, the system produced 81.67 mM from 100 mM l-Tyr, with a substrate conversion rate of 81.67%. This demonstrates the system’s potential for industrial-scale SAA production.

Acknowledgments

This work was supported by grants from the National Natural Science Foundation of China (31971372 and 32371542) and the Zhejiang Natural Science Foundation (LY21B060003).

Glossary

Abbreviations

SAA

salvianolic acid A

l-Tyr

l-tyrosine

4-HPPA

4-hydroxyphenylpyruvic acid

4-HPLA

4-hydroxyphenyllactic acid

DHPPA

3,4-dihydroxyphenylpyruvic acid

Pvml-AAD

membrane-linked l-amino acid deaminase from Proteus vulgaris

Pad-LDH

d-lactate dehydrogenase from Pediococcus acidilactici

EcHpaBC

hydroxyphenylacetate 3-hydroxylase from Escherichia coli

MvFDH

formate dehydrogenase from Mycobacterium vaccae N10

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.4c09986.

  • Strains used were E. coli BL21(DE3), used as a host for recombinant enzyme expression; primer sequences used to amplify the Pad-LDH and EcHpaBC genes; plasmids used were pETDuet-1, pRSFuet-1, and pET-28a-mlaad; construction of plasmids pETDuet-hpabc-fdh and pRSFDuet-dldh-fdh; analysis of the expression of enzymes Pvml-AAD, Pad-LDH, EcHpaBC, and MvFDH via SDS-PAGE; gene sequences of the four enzymes used in this study were: Pvml-AAD, Pad-LDH, EcHpaBC, and MvFDH; the mechanism involves the recycling of FADH2 with the assistance of NADH-dependent MvFDH (PDF)

Author Contributions

M.X.Z. and W.R.Z. conceived and designed the research. M.X.Z., J.Y.Z., Y.K.Z., and W.J.W. conducted the experiments. M.X.Z. and J.Y.Z. analyzed the data. M.X.Z. and W.R.Z. performed the investigation and wrote the manuscript. W.R.Z., S.H., C.J.L., J.H., and L.H.M. contributed to manuscript revision and reading. All authors read and approved the manuscript.

The authors declare no competing financial interest.

Supplementary Material

ao4c09986_si_001.pdf (208.5KB, pdf)

References

  1. Zhao G.; Zhang H.; Ye T.; Xiang Z.; Yuan Y.; Guo Z.; Zhao L. Characterization of the radical scavenging and antioxidant activities of danshensu and salvianolic acid B. Food Chem. Toxicol. 2008, 46, 73–81. 10.1016/j.fct.2007.06.034. [DOI] [PubMed] [Google Scholar]
  2. Zhou X.; Chan S. W.; Tseng H. L.; Deng Y.; Hoi P. M.; Choi P. S.; Or P. M. Y.; Yang J.; Lam F. F. Y.; Lee S. M. Y.; Leung G. P. H. Danshensu is the major marker for the antioxidant and vasorelaxation effects of Danshen (Salvia miltiorrhiza) water-extracts produced by different heat water-extractions. Phytomedicine 2012, 19 (14), 1263–1269. 10.1016/j.phymed.2012.08.011. [DOI] [PubMed] [Google Scholar]
  3. Wang X.; Wang Z.; Xu J. Effect of salvianic acid A on lipid peroxidation and membrane permeability in mitochondria. J. Ethnopharmacol. 2005, 97, 441–445. 10.1016/j.jep.2004.11.036. [DOI] [PubMed] [Google Scholar]
  4. Zhou X.; Chan K.; Yeung J. H. K. Herb-drug interactions with Danshen (Salvia miltiorrhiza): A review on the role of cytochrome P450 enzymes. Drug Metabolism Personalized Therapy 2012, 27, 9. 10.1515/dmdi-2011-0038. [DOI] [PubMed] [Google Scholar]
  5. Song Q.; Zhang Y.; Han X.; Zhang Y.; Zhang X.; Gao Y.; Zhang J.; Chu L.; Zhao S. Potential mechanisms underlying the protective effects of salvianic acid A against atherosclerosis in vivo and vitro. Biomed. Pharmacother. 2019, 109, 945–956. 10.1016/j.biopha.2018.10.147. [DOI] [PubMed] [Google Scholar]
  6. Sidoryk K.; Filip K.; Cmoch P.; Łaszcz M.; Cybulski M. Efficient synthesis and physicochemical characterization of natural danshensu, its S isomer and intermediates thereof. J. Mol. Struct. 2018, 1153, 135–148. 10.1016/j.molstruc.2017.09.118. [DOI] [Google Scholar]
  7. Bao D.; Wang J.; Liu J.; Qin T.; Liu H. The attenuation of HIV-1 Tat-induced neurotoxicity by Salvianic acid A and Danshen granule. Int. J. Biol. Macromol. 2019, 124, 863–870. 10.1016/j.ijbiomac.2018.11.146. [DOI] [PubMed] [Google Scholar]
  8. Kwon G.; Kim H. J.; Park S. J.; Lee H. E.; Woo H.; Ahn Y. J.; Gao Q.; Cheong J. H.; Jang D. S.; Ryu J. H. Anxiolytic-like effect of danshensu [(3-(3, 4-dihydroxyphenyl)-lactic acid)] in mice. Life Sci. 2014, 101, 73–78. 10.1016/j.lfs.2014.02.011. [DOI] [PubMed] [Google Scholar]
  9. Wang M.; Tang W.; Gong N.; Liu P. Sodium Danshensu inhibits the progression of lung cancer by regulating PI3K/Akt signaling pathway. Drug Dev. Res. 2022, 83, 88–96. 10.1002/ddr.21846. [DOI] [PubMed] [Google Scholar]
  10. Zhang Y.; Guo J.; Qi Y.; Shao Q.; Liang J. The prevention of radiation-induced DNA damage and apoptosis in human intestinal epithelial cells by salvianic acid A. J. Radiation Res. Applied Sci. 2014, 7, 274–285. 10.1016/j.jrras.2014.05.003. [DOI] [Google Scholar]
  11. Yang G.; Zhang H.; Lin R.; Wang W.; Shi X.; Liu Y.; Ji Q. Down-Regulation of CD40 Gene Expression and Inhibition of Apoptosis with Danshensu in Endothelial Cells. Basic Clin. Pharmacol. Toxicol. 2009, 104, 87–92. 10.1111/j.1742-7843.2008.00342.x. [DOI] [PubMed] [Google Scholar]
  12. Chen Y.; Lee S.; Lin Y.; Chiang S.; Lin C. Effects of danshensu and salvianolic Acid B from Salvia miltiorrhiza Bunge (Lamiaceae) on cell proliferation and collagen and melanin production. Molecules 2014, 19, 2029–2041. 10.3390/molecules19022029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Lam F. F. Y.; Yeung J. H. K.; Chan K. M.; Or P. M. Y. Relaxant effects of danshen aqueous extract and its constituent danshensu on rat coronary artery are mediated by inhibition of calcium channels. Vasc. Pharmacol. 2007, 46 (4), 271–277. 10.1016/j.vph.2006.10.011. [DOI] [PubMed] [Google Scholar]
  14. Bai Y.; Zhang Q.; Jia P.; Yang L.; Sun Y.; Nan Y.; Wang S.; Meng X.; Wu Y.; Qin F.; Sun Z. Improved process for pilot-scale synthesis of Danshensu ((±)-DSS) and its enantiomer derivatives. Org. Process Res. Dev. 2014, 18 (12), 1667–1673. 10.1021/op4002593. [DOI] [Google Scholar]
  15. Santacoloma P. A.; Sin G. R.; Gernaey K. V.; Woodley J. M. Multienzyme-catalyzed processes: Next-generation biocatalysis. Org. Process Res. Dev. 2011, 15, 203–212. 10.1021/op1002159. [DOI] [Google Scholar]
  16. Benítez-Mateos A. I.; Roura Padrosa D.; Paradisi F. Multistep enzyme cascades as a route towards green and sustainable pharmaceutical syntheses. Nat. Chem. 2022, 14, 489–499. 10.1038/s41557-022-00931-2. [DOI] [PubMed] [Google Scholar]
  17. Isabel O. G.; Eduardo G. J. Enzyme catalysed tandem reactions. Curr. Opin. Chem. Biol. 2013, 17, 236–249. 10.1016/j.cbpa.2013.02.015. [DOI] [PubMed] [Google Scholar]
  18. Ricca E.; Brucher B.; Schrittwieser J. H. Multi-enzymatic cascade reactions: Overview and perspectives. Adv. Synth. Catal. 2011, 353, 2239–2262. 10.1002/adsc.201100256. [DOI] [Google Scholar]
  19. Han R.; Gao K.; Jiang Y.; Zhou J.; Xu G.; Dong J.; Schwaneberg U.; Ji Y.; Ni Y. Self-Sufficient in vitro multi-enzyme cascade for efficient synthesis of danshensu from l-DOPA. ACS Synth. Biol. 2023, 12, 277–286. 10.1021/acssynbio.2c00552. [DOI] [PubMed] [Google Scholar]
  20. Hu K. S.; Chen C. L.; Ding H. R.; Wang T. Y.; Zhu Q.; Zhou Y. C.; Chen J. M.; Mei J. Q.; Hu S.; Huang J.; Zhao W. R. Production of salvianic acid A from L-DOPA via biocatalytic cascade reactions. Molecules 2022, 27 (18), 6088. 10.3390/molecules27186088. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Xiong T.; Jiang J.; Bai Y.; Fan T.; Zhao Y.; Zheng X.; Cai Y. Biosynthesis of D-danshensu from L-DOPA using engineered Escherichia coli whole cells. Appl. Microbiol. Biotechnol. 2019, 103, 6097–6105. 10.1007/s00253-019-09947-0. [DOI] [PubMed] [Google Scholar]
  22. Porokhin V.; Amin S. A.; Nicks T. B.; Gopinarayanan V. E.; Nair N. U.; Hassoun S. Analysis of metabolic network disruption in engineered microbial hosts due to enzyme promiscuity. Metabolic Eng. Commun. 2021, 12, e00170 10.1016/j.mec.2021.e00170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Zhou Z.; Zhang X.; Wu J.; Li X.; Li W.; Sun X.; Wang J.; Yan Y.; Shen X.; Yuan Q. Targeting cofactors regeneration in methylation and hydroxylation for high level production of ferulic acid. Metab. Eng. 2022, 73, 247–255. 10.1016/j.ymben.2022.08.007. [DOI] [PubMed] [Google Scholar]
  24. Zhao W.; Ding H.; Lv C.; Hu S.; Huang J.; Zheng X.; Yao S.; Mei L. Two-step biocatalytic reaction using recombinant Escherichia coli cells for efficient production of phenyllactic acid from l-phenylalanine. Process Biochem. 2018, 64, 31–37. 10.1016/j.procbio.2017.09.019. [DOI] [Google Scholar]
  25. Bai C.; Zhao G. Separation of salvianic acid A from the fermentation broth of engineered Escherichia coli using macroporous resins. J. Sep. Sci. 2015, 38, 2833–2840. 10.1002/jssc.201500416. [DOI] [PubMed] [Google Scholar]
  26. Beber M. E.; Gollub M. G.; Mozaffari D.; Shebek K. M.; Noor E.; Milo R.; Noor E. eQuilibrator 3.0--a platform for the estimation of thermodynamic constants. arXiv 2021, 50 (D1), D603–D609. 10.48550/arXiv.2103.00621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Schadewaldt P.; Adelmeyer F. Coupled Enzymatic Assay for Estimation of Branched-Chainl-Amino Acid Aminotransferase Activity with 2-Oxo Acid Substrates. Anal. Biochem. 1996, 238, 65–71. 10.1006/abio.1996.0252. [DOI] [PubMed] [Google Scholar]
  28. Niu P.; Dong X.; Wang Y.; Liu L. Enzymatic production of α-ketoglutaric acid from l-glutamic acid via l-glutamate oxidase. J. Biotechnol. 2014, 179, 56–62. 10.1016/j.jbiotec.2014.03.021. [DOI] [PubMed] [Google Scholar]
  29. Song Y.; Li J.; Shin H.-D.; Du G.; Liu L.; Chen J. One-step biosynthesis of α-ketoisocaproate from l-leucine by an Escherichia coli whole-cell biocatalyst expressing an l-amino acid deaminase from Proteus vulgaris. Sci. Rep. 2015, 5, 12614. 10.1038/srep12614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Hossain G. S.; Li J.; Shin H.-D.; Du G.; Wang M.; Liu L.; Chen J. One-Step Biosynthesis of α-Keto-γ-Methylthiobutyric Acid from L-Methionine by an Escherichia coli Whole-Cell Biocatalyst Expressing an Engineered L-Amino Acid Deaminase from Proteus vulgaris. PLoS One 2014, 9, e114291 10.1371/journal.pone.0114291. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Ding H.; Zhao W.; Lü C.; Huang J.; Hu S.; Yao S.; Mei L.; Wang J.; Mei J. Biosynthesis of 4-hydroxyphenylpyruvic acid from l-tyrosine using recombinant Escherichia coli cells expressing membrane bound l-amino acid deaminase. Chin. J. Chem. Eng. 2018, 26, 380–385. 10.1016/j.cjche.2017.08.009. [DOI] [Google Scholar]
  32. Wang M.; Zhu L.; Xu X.; Wang L.; Yin R.; Yu B. Efficient production of enantiomerically pure d-phenyllactate from phenylpyruvate by structure-guided design of an engineered d-lactate dehydrogenase. Appl. Microbiol. Biotechnol. 2016, 100, 7471–7478. 10.1007/s00253-016-7456-1. [DOI] [PubMed] [Google Scholar]
  33. Mu W.; Yu S.; Zhu L.; Jiang B.; Zhang T. Production of 3-phenyllactic acid and 4-hydroxyphenyllactic acid by Pediococcus acidilactici DSM 20284 fermentation. Eur. Food Res. Technol. 2012, 235, 581–585. 10.1007/s00217-012-1768-x. [DOI] [Google Scholar]
  34. Yamamoto H.; Mitsuhashi K.; Kimoto N.; Kobayashi Y.; Esaki N. Robust NADH-regenerator: Improved α-haloketone-resistant formate dehydrogenase. Appl. Microbiol. Biotechnol. 2005, 67, 33–39. 10.1007/s00253-004-1728-x. [DOI] [PubMed] [Google Scholar]
  35. Prieto M. A.; Garcia J. L. Molecular characterization of 4-hydroxyphenylacetate 3-hydroxylase of Escherichia coli. A two-protein component enzyme. J. Biol. Chem. 1994, 269, 22823–22829. 10.1016/S0021-9258(17)31719-2. [DOI] [PubMed] [Google Scholar]
  36. Louie T. M.; Xie X. S.; Xun L. Coordinated Production and Utilization of FADH2 by NAD(P)H–Flavin Oxidoreductase and 4-Hydroxyphenylacetate 3-Monooxygenase. Biochemistry 2003, 42, 7509–7517. 10.1021/bi034092r. [DOI] [PubMed] [Google Scholar]
  37. Xun L.; Sandvik E. R. Characterization of 4-Hydroxyphenylacetate 3-Hydroxylase (HpaB) of Escherichia coli as a Reduced Flavin Adenine Dinucleotide-Utilizing Monooxygenase. Appl. Environ. Microbiol. 2000, 66 (2), 481–486. 10.1128/AEM.66.2.481-486.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Chen W.; Yao J.; Meng J.; Han W.; Tao Y.; Chen Y.; Guo Y.; Shi G.; He Y.; Jin J.; Tang S. Promiscuous enzymatic activity-aided multiple-pathway network design for metabolic flux rearrangement in hydroxytyrosol biosynthesis. Nat. Commun. 2019, 10, 960. 10.1038/s41467-019-08781-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Chen Z.; Sun X.; Li Y.; Yan Y.; Yuan Q. Metabolic engineering of Escherichia coli for microbial synthesis of monolignols. Metab. Eng. 2017, 39, 102–109. 10.1016/j.ymben.2016.10.021. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao4c09986_si_001.pdf (208.5KB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES