Characterization of a New Marine Leucine Dehydrogenase from Pseudomonas balearica and Its Application for L-tert-Leucine Production

Abstract

Leucine dehydrogenase (LeuDH) has emerged as the most promising biocatalyst for L-tert-leucine (L-Tle) production via asymmetric reduction in trimethylpyruvate (TMP). In this study, a new LeuDH named PbLeuDH from marine Pseudomonas balearica was heterologously over-expressed in Escherichia coli, followed by purification and characterization. PbLeuDH possessed a broad substrate scope, displaying activities toward numerous L-amino acids and α-keto acids. Notably, compared with those reported LeuDHs, PbLeuDH exhibited excellent catalytic efficiency for TMP with a K_m value of 4.92 mM and a k_cat/K_m value of 24.49 s⁻¹ mM⁻¹. Subsequently, L-Tle efficient production was implemented from TMP by whole-cell biocatalysis using recombinant E. coli as a catalyst, which co-expressed PbLeuDH and glucose dehydrogenase (GDH). Ultimately, using a fed-batch feeding strategy, 273 mM (35.8 g L⁻¹) L-Tle was achieved with a 96.1% yield and 2.39 g L⁻¹ h⁻¹ productivity. In summary, our research provides a competitive biocatalyst for L-Tle green biosynthesis and lays a solid foundation for the realization of large-scale L-Tle industrial production.

1. Introduction

The versatile small molecules, unnatural amino acids (UAAs) and their derivatives, play prominent roles in cosmetics, pharmaceuticals, food additives manufacturing, chemical synthesis, etc. [1,2,3]. Notably, optically pure UAAs are critical scaffold intermediates to synthesize chiral drugs [4]. L-tert-Leucine (L-Tle), one of the most typical chiral UAAs, is extensively applied as the template of organocatalyst for asymmetric synthesis owing to its hydrophobic, bulky, and inflexible tert-butyl side chain [5]. Meanwhile, L-Tle is the key precursor for active pharmaceutical ingredients, such as tumor-fighting agents and anti-HIV protease inhibitors [6,7,8]. Moreover, L-Tle is employed as a nutrition fortifier and animal feed additive in cosmetic and food fields [4]. Considering the high value and tremendous market demand of L-Tle, its efficient production has received increasing attention.

In industry, L-Tle is mainly synthesized by chemical or enzymatic methods. Unluckily, many deficiencies, including tedious processes, harsh reaction conditions, severe pollution, poor enantioselectivity, and low yield in traditional chemical methods result in reduced competitiveness [9]. Alternatively, the green biomanufacturing of L-Tle via enzymatic catalysis became more popular owing to its high enantioselectivity and environmental friendliness [10]. Various enzymes such as lipase, penicillin acylase, and hydantoinase can catalyze the kinetic resolution of racemic DL-Tle to synthesize enantiopure L-Tle. However, the maximum theoretical yield of these resolution strategies was merely 50% [11]. Additionally, transaminases can catalyze trimethylpyruvate (TMP) into L-Tle with glutamate as an amino donor. Nonetheless, the low equilibrium constant and severe product inhibition of transaminase hinder its large-scale industrial applications [6]. In contrast, leucine dehydrogenase (LeuDH, EC 1.4.1.9) is the most promising biocatalyst for L-Tle production via asymmetric reduction in TMP because of its high atom utilization and 100% theoretical yield. As shown in Figure 1, with NAD(H) as the cofactor, LeuDH reversibly catalyzes the oxidative deamination of L-amino acids and the reductive amination of corresponding α-keto acids using NH₄⁺ or NH₃ as amino donors [2,12,13].

So far, many LeuDHs from Bacillus [12,14], Exiguobacterium [15,16], Labrenzia [10], Laceyella [13], Lysinibacillus [17], Planifilum [4], Sporosarcina [18] species, etc have been identified and characterized for L-Tle production from TMP. Among them, several LeuDHs with excellent catalytic efficiencies exhibited huge industrial potential for L-Tle production. For instance, EsLeuDH with good thermostability from Exiguobacterium sibiricum showed 12.7 U mg⁻¹ and 37.7 U mg⁻¹ of specific activity toward L-leucine and TMP as substrates, respectively. Encouragingly, a 945 g L⁻¹ d⁻¹ space-time yield of L-Tle was achieved by using the whole-cell biocatalyst co-expressing EsLeuDH and glucose dehydrogenase in Escherichia coli [16]. Additionally, a thermostable LsLeuDH with 183 U mg⁻¹ of specific activity toward L-leucine was obtained from Laceyella sacchari [13], and the engineered E. coli co-expressing LsLeuDH and formate dehydrogenase could completely catalyze 1 M TMP into L-Tle with 99.9% ee [19]. Nevertheless, all the above-mentioned LeuDHs were obtained from the terrestrial microbes. In fact, the marine or aquatic microbes which can adapt to special living environments (high salinity and pressure, low temperature) are still deemed as valuable untapped resources [20]. To our knowledge, only three LeuDHs, including psychrophilic SpLeuDH from Sporosarcina psychrophile [18], cold-adapted AdLeuDH from Alcanivorax dieselolei [21], and salt-tolerant PrLeuDH from Pseudoalteromonas rubra [22] from marine or aquatic microbes, have been reported. However, the two LeuDHs showed low catalytic efficiency for L-Tle production. For AdLeuDH, the L-Tle titer was only 1.58, 3.56, 6.89, and 11.98 g L⁻¹ after 4, 12, 24, and 32 h of reaction, when the substrate (TMP) concentration was 3.25, 6.51, 13.01, and 26.03 g L⁻¹, respectively [22]. Taken together, screening new LeuDHs with extraordinary catalytic properties from marine or aquatic microbes is a worthwhile task.

In this study, a new LeuDH (PbLeuDH) was cloned from our previous isolated marine Pseudomonas balearica and was successfully over-expressed in E. coli BL21 (DE3). Subsequently, the biochemical properties of PbLeuDH were systematically investigated. Moreover, a recombinant E. coli whole-cell biocatalyst was constructed, in which PbLeuDH was co-expressed with glucose dehydrogenase for NADH regeneration. Finally, L-Tle was efficiently produced via the reductive amination of TMP by fed-batch whole-cell catalysis.

2. Results and Discussion

2.1. Sequence Analysis of PbLeuDH

The putative Pbleudh gene from P. balearica had a 1086 bp-length ORF encoding a putative PbLeuDH enzyme with 361 amino acids, and the isoelectric point of 6.05 and molecular weight of 37.9 kDa were predicted by using the Compute pI/Mw tool (https://web.expasy.org/compute_pi/, accessed on 5 July 2022). BLASTp analysis showed that the putative PbLeuDH enzyme belongs to the Glu/Leu/Phe/Val dehydrogenase family. The amino acid sequence identities of PbLeuDH were analyzed with that of other known functional LeuDHs. The result in Figure 2 indicated that the PbLeuDH enzyme shared the highest sequence identity of 48.4% with AdLeuDH from A. dieselolei MCCC 1A02288, which suggested that the putative PbLeuDH may be a new leucine dehydrogenase. As shown in Figure 2, four key active site residues (K68, K80, D115, and N267) were highly conserved in all aligned LeuDHs, which was consistent with the previous studies [22,23]. Among them, one residue (D115) and two residues (K68 and N267) are responsible for interacting with the amine group and the carboxyl group of the L-amino acid substrate, respectively, while residue K80 is responsible for activating the water molecule around the active site. In addition, nine highly conserved residues (S152, T155, G185, G187, V189, D208, C242, A243, and N266) were predicted as the coenzyme NAD(H) binding site by the InterProScan web server (http://www.ebi.ac.uk/interpro/search/sequence/, accessed on 5 July 2022).

2.2. Expression and Purification of Recombinant PbLeuDH

E. coli (pET28a-Pbleudh) was induced by IPTG for the overexpression of recombinant PbLeuDH. As shown in Figure 3, SDS-PAGE analysis on the crude enzyme extract of E. coli (pET28a-Pbleudh) revealed the presence of soluble expressed PbLeuDH with a molecular weight of approximately 40.0 kDa. The crude enzyme extract was subsequently purified by a HisTrap-HP affinity column and the enzyme yield of 67.30% was achieved with 3.57 of purified fold (

Table 1. PbLeuDH activity toward L-leucine before and after purification.

Enzyme	Total Activity (U)	Total Protein (mg)	Specific Activity (U mg−1)	Purification Yield (%)	Purification Factor
Crude extract	106.72	14.38 ± 0.52	7.42 ± 0.15	100.00	1.00
Purified PbLeuDH	71.82	2.71 ± 0.06	26.50 ± 0.27	67.30	3.57

). The recombinant enzyme activity was assayed by using L-leucine as the substrate under standard conditions (room temperature, 10 mM L-leucine, 0.5 mM NAD⁺, 100 mM glycine-NaOH, pH 10.0 buffer). As shown in Table 1, the recombinant PbLeuDH exhibited its activity of 7.42 U mg⁻¹ and 26.50 U mg⁻¹ for crude enzyme extract and purified enzyme, respectively, in the presence of L-leucine, which was obviously higher compared to other reported LeuDHs such as the BsLeuDH from B. sphaericus ATCC 4525 [14], PfLeuDH from P. fimeticola [4], and EsLeuDH from E. sibiricum ECU9271 [16] that exhibited specific activities of 12.19 U mg⁻¹, 3.49 U mg⁻¹, and 12.7 U mg⁻¹, respectively.

2.3. Biochemical Characterization of PbLeuDH toward L-Leucine

The effect of pH value on the activity of PbLeuDH was investigated in the range of 7.0–13.0. The results showed that the highest activity of PbLeuDH at pH 11.5 could be observed, and its relative activity was over 75% at pH 10.5–12.0 (Figure 4A). Concerning the pH effect on stability, the remaining activity of PbLeuDH was over 84% after 3 h incubation at pH 7.0–11.0 and over 73% after 10 h incubation at pH 7.0–10.0, and more than 45% of the activity was still retained even in pH 6.0 and 11.0 after 10 h of incubation (Figure 4B). Therefore, the PbLeuDH enzyme possessed broad reaction pH when compared with ever reported LeuDHs such as LsLeuDH from L. sacchari (pH 9.5–11.0) [13], PrLeuDH from P. rubra DSM 6842 (pH 9.5–12.0) [22], and CfLeuDH from Citrobacter freundii JK-91 (pH 9.0–11.0) [24].

The reaction temperature and stability of PbLeuDH were investigated as shown in Figure 4C–E. The highest activity of the purified enzyme was determined at 60 °C, which was similar to that of L. sacchari [13], and had a relative activity of about 80% at 50 °C. When the reaction temperatures were below 50 °C and above 60 °C, the enzyme activity appeared to rapid decrease (Figure 4C). The enzyme thermal stability assay showed that over 97% of the initial activity remained after incubation at 30−40 °C for 0.5 h, while a rapid decrease could be observed when the temperature was over 40 °C (Figure 4D), indicating that the PbLeuDH enzyme possessed weak thermostability. As shown in Figure 4E, the residual enzyme activity of over 80% could be retained only at 30 °C for 12 h, while only about 40% of the residual activity could be detected when the enzyme was incubated at 40 °C for 6 h. Therefore, the PbLeuDH enzyme was not suitable for in vitro biocatalysis to synthesize L-Tle. To eliminate the influence of weak thermostability, whole-cell biocatalysis was used to perform L-Tle production from TMP at a mesophilic temperature.

Moreover, we investigated the effects of metal ions or ethylenediaminetetraacetic acid (EDTA) at the final concentration of 1 mM on the activity of PbLeuDH. As shown in Figure 4F, all the tested metal ions showed no activation of the PbLeuDH. In contrast, the enzyme activity could be remarkably inhibited by the trivalent ions (Al³⁺, Fe³⁺), which was similar to that for LeuDHs from E. sibiricum [15,16].

2.4. Substrate Specificity and Kinetic Parameters of PbLeuDH

LeuDHs can reversibly catalyze the oxidative deamination of L-amino acids and the reductive amination of corresponding α-keto acids using NH₄⁺ or NH₃ as an amino donor. Here, the substrate specificity of PbLeuDH was studied (Figure 5). For reductive amination, PbLeuDH exhibited the highest activity toward TMP and showed no activity toward 2-Oxoglutaric acid among the five aliphatic α-keto acids tested. For oxidative deamination, PbLeuDH exhibited strict substrate preference for nonpolar aliphatic L-amino acids, while weak activities for polar amino acids (L-Lysine, L-Cysteine) and aromatic L-amino acids (L-Phenylalanine) as substrates could be detected. Among the nonpolar aliphatic L-amino acids tested, PbLeuDH exhibited relatively high activities toward L-Leucine, L-Valine, and L-Isoleucine. Similar results were also observed for most of the previously reported LeuDHs [4,13,14,15,16]. It is worth emphasizing that the relative activity of PbLeuDH toward L-Tle is only 5.7%, which suggested that the PbLeuDH enzyme favored the conversion from TMP to L-Tle. Considering the reversible reaction between TMP and L-Tle, the PbLeuDH enzyme possessed an obvious advantage for L-Tle synthesis via reductive amination of TMP.

As shown in

Table 2. Kinetic parameters of PbLeuDH for TMP and L-leucine.

Substrate	Vmax (U mg−1)	Km (mM)	kcat (s−1)	kcat/Km(s−1 mM−1)
TMP	91.78 ± 6.41	4.92 ± 0.31	120.49 ± 6.04	24.49
L-leucine	23.68 ± 2.37	0.40 ± 0.03	14.97 ± 1.27	37.43

and Figure S1, the K_m value of PbLeuDH toward L-leucine was merely 0.40 mM, suggesting that L-leucine was the natural substrate of PbLeuDH. Moreover, the K_m and k_cat/K_m values of PbLeuDH toward TMP were 4.92 mM and 24.49 s⁻¹ mM⁻¹, respectively. The kinetic parameters of PbLeuDH toward TMP showed lower K_m value and higher catalytic efficiency in comparison to those of EsLeuDH (5.96 mM and 7.62 s⁻¹ mM⁻¹) from E. siribicum [15], AdLeuDH (100 mM and 0.10 s⁻¹ mM⁻¹) from A. dieselolei MCCC 1A02288 [21], and BcLeuDH (13.20 mM and 1.61 s⁻¹ mM⁻¹) from B. coagulans NL01 [12], which demonstrated that the PbLeuDH enzyme possessed high affinity and excellent catalytic efficiency toward TMP and could be employed as a promising biocatalyst for TMP bioconversion.

2.5. Construction of Whole-Cell Biocatalyst Coexpressing PbLeuDH and BmGDH

The results above implied that PbLeuDH, which had an excellent catalytic performance toward TMP, would be an effective catalyst for converting TMP into L-Tle. As shown in Figure 6A, the process of L-Tle production via reductive amination reaction by LeuDHs needed NADH as a cofactor to drive the reaction toward product formation [25,26,27]. Considering that the cofactor NADH is expensive, the in situ NADH regeneration system was deemed as the ideal strategy to reduce production costs. Hence, glucose dehydrogenase (GDH) or formate dehydrogenase (FDH), which use glucose or formate as co-substrates, are widely applied to drive the cofactor cycle by coupling with LeuDHs, thus improving the yield of L-Tle [4,9,15,16,28].

In the present study, a whole-cell biocatalyst of E. coli (pET28a-Pbleudh-Bmgdh) carrying the genes of Pbleudh and Bmgdh (GenBank ID: ADE67871.1) from Bacillus megaterium were constructed for the co-expression of both enzymes, and a ribosomal binding site (RBS) of GTTAAAAAGGAGATATA was designed at the upstream of the Bmgdh gene. In the biocatalytic system, PbLeuDH catalyzes TMP into L-Tle using NH₄⁺ as an amino donor, while BmGDH catalyzes the oxidation of glucose into gluconate, coupled with the in situ reduction of NAD⁺ to NADH (Figure 6A). SDS-PAGE analysis demonstrated that two specific bands with a molecular weight of about 40.0 kDa and 28.0 kDa could be clearly observed, indicating that two enzymes of PbLeuDH and BmGDH in the recombinant E. coli (pET28a-Pbleudh-Bmgdh) strain was successfully co-expressed (Figure 6B). Moreover, the crude enzyme activities of PbLeuDH and BmGDH from E. coli (pET28a-Pbleudh-Bmgdh) were determined at room temperature in the presence of NADH and NAD⁺ as the coenzyme by using TMP and glucose as substrates, respectively. The crude enzyme extract of E. coli (pET28a-Pbleudh-Bmgdh) exhibited the activities of 13.41 U mg⁻¹ for TMP and 15.27 U mg⁻¹ for glucose, suggesting that a whole-cell biocatalyst was successfully constructed.

2.6. Effects of pH, Temperature, WCW, and Substrate Concentration on L-Tle Production by E. coli (pET28a-Pbleudh-Bmgdh)

To obtain a high yield of L-Tle, the whole-cell biocatalysis conditions of E. coli (pET28a-Pbleudh-Bmgdh) were systematically investigated. First, the effects of pH on biocatalytic performance were conducted in the reaction system containing 10 g L⁻¹ of wet cells and 100 mM of TMP at 30 °C for 3 h using different pH values. Figure 7A showed that the L-Tle titer was markedly increased with the pH increase from 7.5 to 8.5, and a sharp decrease could be observed at the pH of 9.0 and 9.5. Therefore, the optimal pH during the biocatalytic process was controlled at 8.5, resulting in a maximum L-Tle titer of 87.15 mM. A major reason might be attributed to the rapid decrease in the BmGDH activity in E. coli (pET28a-Pbleudh-Bmgdh) from 15.27 U mg⁻¹ of pH 7.0 to 1.66 U mg⁻¹ of pH 9.0 (Figure S2), which limited the effective regeneration of the cofactor NADH and thus led to a significantly negative impact on L-Tle yield. Then, the effect of temperature (25−42 °C) on biocatalytic performance was studied at pH 8.5. Figure 7B showed that the maximum L-Tle titer of 87.74 mM was obtained from 100 mM TMP for 3 h at 30 °C, and thus the subsequent studies were performed at 30 °C. Subsequently, biocatalyst concentrations (WCW) for bioconversion reaction were optimized. As shown in Figure 7C, 95.02 mM of L-Tle was obtained when 20 g L⁻¹ of the WCW was used. The substrate concentration is also a crucial factor in the bioconversion reaction; many previous studies have witnessed that high-concentration TMP would generate an inhibitory effect in LeuDHs [15,16,19,21]. As shown in Figure 7D, 48.77 mM and 95.02 mM of L-Tle could be obtained with a conversion rate of over 95% under the TMP concentrations of 50 mM and 100 mM, respectively. However, when the TMP concentration reached 200 mM, although 116.84 mM L-Tle was achieved, only 58.4% of the conversion rate was obtained with a residual TMP up to 179.17 mM, suggesting that high-concentration TMP would generate strong substrate inhibition and result in low L-Tle yield.

2.7. Fed-Batch Synthesis of L-Tle from TMP by E. coli (pET28a-Pbleudh-Bmgdh)

We found that high-concentration TMP would strongly inhibit the bioconversion reaction to reduce the L-Tle yield. Therefore, the substrate fed-batch strategy was implemented to alleviate the TMP inhibition effect. During the whole reaction process by fed-batch, the TMP concentration was kept at a low level (no more than 110 mM). As shown in Figure 8, a continuous increase in the L-Tle titer could be observed with the rapid consumption of TMP before 7 h, and thereafter the rate of TMP consumption gradually slowed down. Ultimately, a total of 284 mM of TMP was consumed at 15 h, resulting in an L-Tle titer up to 273 mM (35.8 g L⁻¹) with a 96.1% yield and 2.39 g L⁻¹ h⁻¹ productivity. In addition, a moderate decrease in cell viability (from 98.43% at 0 h to 83.74% at 15 h) was observed during the reaction process, likely due to bacterial decay.

Compared to previous reports, the biosynthesis efficiency of the L-Tle developed in this study belonged to the middle level. In previous studies, the highest space-time yield (STY) of L-Tle (2136 g L⁻¹ d⁻¹) by GDH–R3–LeuDH whole cells could be achieved [28]. Additionally, a 945 g L⁻¹ d⁻¹ STY of L-Tle was obtained by a recombinant whole-cell biocatalyst, in which EsLeuDH was co-expressed with GDH in E. coli [16]. Nevertheless, this study was competitive when compared with those in the following reports. For example, an engineered E. coli co-expressing PfLeuDH and FDH could catalyze 0.1 M TMP into L-Tle for 25 h, with a yield and STY of L-Tle reaching 87.38% and 10.90 g L⁻¹ d⁻¹ [4]. In another example, a two-enzymes coupled system using EsLeuDH and FDH was established to enzymatically produce 0.5 M (65.6 g L⁻¹) L-Tle from 0.5 M TMP with 2.19 g L⁻¹ h⁻¹ productivity after 30 h reaction [15]. Taken together, this work provides a promising alternative approach for efficient L-Tle biosynthesis from TMP. In future work, the catalytic efficiency and substrate TMP tolerance of PbLeuDH can be improved by molecular modification to satisfy industrial applications.

3. Materials and Methods

3.1. Strains, Plasmids, Primers, and Chemicals

Table 3. Bacterial strains and plasmids used in this study.

Strains or Plasmids	Characteristic	Source
Strains
P. balearica	Wild type	Lab stock
B. megaterium ATCC 14581	Wild type	Lab stock
E. coli DH5α	General cloning host	Tiangen Biotech
E. coli BL21(DE3)	General expression host	Tiangen Biotech
E. coli (pET28a-Pbleudh)	E. coli BL21(DE3) with plasmid pET28a-Pbleudh	This study
E. coli (pET28a-Pbleudh-Bmgdh)	E. coli BL21(DE3) with plasmid pET28a-Pbleudh-Bmgdh	This study
Plasmids
pET28a	Kmr; expression vector	Laboratory stock
pET28a-Pbleudh	Kmr; Pbleudh in pET28a	This study
pET28a-Pbleudh-Bmgdh	Kmr; Pbleudh and Bmgdh in pET28a	This study

and

Table 4. The primers used in this study.

Primers	Sequences	Source
P1	CAGCAAATGGGTCGCGGATCCATGGCAGGGCGCCAGACCA	BamHI
P2	CTCGAGTGCGGCCGCAAGCTTTCAGGCGTGCAAGCGCAGG	HindIII
P3	CAGCAAATGGGTCGCGGATCCATGGCAGGGCGCCAGACCA	BamHI
P4	TATATCTCCTTTTTAACTCAGGCGTGCAAGCGCAG	RBS
P5	GTTAAAAAGGAGATATAATGTATAAAGATTTAGAAGG	RBS
P6	CTCGAGTGCGGCCGCAAGCTTTTATCCGCGTCCTGCTTGG	HindIII

listed the strains, primers, and plasmids used in this study. All strains in this study were cultured in Luria-Bertani (LB) broth. Kanamycin (50 μg/mL final concentration) was used in recombinant E. coli strains. The primer synthesis and commercial sequencing were completed by Sangon Biotech (Shanghai, China). Restriction endonucleases (BamHI and HindIII) and Q5 DNA polymerase were obtained from NEB Co. (Ipswich, MA, USA). ClonExpress Kit for the construction of recombinant plasmids was from Vazyme Biotech (Nanjing, Jiangsu, China). Genomic DNA and plasmid DNA were obtained by Mini-prep Kits (OMEGA, Shanghai, China). HisTrap-HP affinity column and HiTrap desalting column for protein purification were from GE Co. (Boston, MA, USA). The chemical standards of L-Tle and TMP were from Sigma Co. (St. Louis, MO, USA). All other chemicals were obtained from Aladdin (Shanghai, China) at analytical grade.

3.2. Cloning, Expression, and Purification of PbLeuDH

The Pbleudh gene encoding PbLeuDH was amplified via PCR with the genomic DNA of P. balearica as a template using the primer pairs P1/P2 designed according to the LeuDH sequence of the annotated genome from P. balearica DSM 6083 (Genbank ID: CP007511.1). Subsequently, the PCR-products were ligated into the BamHI-HindIII sites of the pET28a (+) vector using the ClonExpress Kit, and then the resulting pET28a-Pbleudh plasmid was transformed (heat shock) into competent E. coli BL21(DE3) cells to obtain the recombinant expression strain E. coli (pET28a-Pbleudh).

The E. coli (pET28a-Pbleudh) was cultured (37 °C, 180 rpm) for 2.5 h in a conical flask (250 mL) with 50 mL LB broth containing 50 μg/mL kanamycin, and subsequently, the protein expression was induced at 18 °C and 180 rpm for 14 h by adding 0.1 mM isopropyl-beta-D-thiogalactoside (IPTG). After induction, wet cells were collected via centrifugation (4 °C, 10 min, 10,000× g) and then resuspended in binding buffer (20 mM imidazole, 20 mM phosphate, 500 mM NaCl, pH 7.4). The crude enzyme after sonication in an ice bath for 10 min was prepared via centrifugation (4 °C, 20 min, 13,000× g) to remove debris for further purification.

For the recombinant PbLeuDH purification, the crude enzyme was loaded onto the affinity column (HisTrap-HP), pre-equilibrated with binding buffer. Next, the column was eluted with 20−500 mM linear gradient imidazole in binding buffer at 4 °C. Subsequently, the eluates with the target protein PbLeuDH were desalted and concentrated by a HiTrap desalting column with 20 mM Tris-HCl, pH 8.0 buffer. Finally, the purified enzyme was evaluated by SDS-PAGE, and its concentration was determined by the Bradford method. The obtained pure enzyme was stored at 4 °C in 20 mM Tris-HCl, pH 8.0 buffer (containing 100 mM NaCl) before usage.

3.3. Enzyme Activity Assay

The PbLeuDH activity was measured by monitoring the change in the NAD(H) absorbance (ε₃₄₀ = 6.22 mM⁻¹ cm⁻¹) at 340 nm and room temperature in the first 3 min of the enzymatic reaction. The reaction system (1 mL) contained 10 mM L-amino acid (or 10 mM α-keto acid), 0.5 mM NAD⁺ (or 0.25 mM NADH), 100 mM glycine-NaOH, pH 10.0 buffer (or 1 M NH₄Cl-NH₃‧H₂O, pH 10.0 buffer), and an appropriate amount of enzyme. One unit (U) of PbLeuDH activity was defined as the amount of enzyme catalyzing the reduction of 1 μmol NAD⁺ per minute or the oxidation of 1 μmol NADH per minute [16].

The BmGDH activity was determined by a similar method. The reaction system (1 mL) contained 10 mM glucose, 1 mM NAD⁺, 100 mM potassium phosphate, pH 7.0 buffer, and an appropriate amount of enzyme. One unit (U) of BmGDH activity was defined as the amount of enzyme catalyzing the reduction of 1 μmol NAD⁺ per minute [16].

3.4. Biochemical Characterization of PbLeuDH toward L-Leucine

The optimum pH of the PbLeuDH activity was detected at room temperature in different pH buffers (100 mM) ranging from 7.0 to 13.0 (pH 7.0–8.0 Na₂HPO₄-citric acid buffer; pH 8.0–9.0 Tris-HCl buffer; pH 9.0–13.0 glycine-NaOH buffer). To investigate the pH stability of PbLeuDH, its residual activity after incubation at 25 °C for 3 h and 10 h in pH 6.0–11.0 buffers was measured. Similarly, the optimum temperature of PbLeuDH activity was examined via performing reactions at temperatures of 30–70 °C in 100 mM glycine-NaOH, pH 10.0 buffer. The residual activity of PbLeuDH was measured after incubation at pH 10.0 and temperatures of 30–60 °C for 0.5 h to evaluate its thermostability. The impacts of various metal ions (Cu²⁺, Al³⁺, Fe³⁺, Co²⁺, Mn²⁺, Mg²⁺, Zn²⁺, and Na⁺) or EDTA on the PbLeuDH activity were investigated by adding them to the reaction system (pH 10.0) with the final concentration of 1 mM. The percent relative (or residual) activity of PbLeuDH represents the enzyme activity relative to the highest (or initial) enzyme activity, which was defined as 100%.

3.5. Substrate Specificity and Kinetic Parameters of PbLeuDH

The PbLeuDH substrate specificity was tested with different substrates including L-amino acids for the oxidative deamination reaction and α-keto acids for the reductive amination reaction. The reaction mixture (1 mL) contained 10 mM L-amino acid (or 10 mM α-keto acid), 0.5 mM NAD⁺ (or 0.25 mM NADH), 100 mM glycine-NaOH, pH 10.0 buffer (or 1 M NH₄Cl-NH₃‧H₂O, pH 10.0 buffer) and appropriate amount of enzyme. The enzyme activity of PbLeuDH for L-leucine (oxidative deamination reaction) and TMP (reductive amination reaction) was defined as 100%.

The kinetic parameters of PbLeuDH toward L-leucine, including the Michaelis-Menten constant (K_m) and the maximal reaction rate (V_max), were obtained by Michaelis-Menten analysis, in which the Lineweaver–Burk plots were created by measuring the activities of PbLeuDH in a reaction mixture (1-mL) containing different concentrations of L-leucine (0.1–4 mM), 0.5 mM NAD⁺, 100 mM glycine-NaOH, pH 10.0 buffer, and purified 2.0 μg PbLeuDH. Meanwhile, based on the PbLeuDH molecular weight, the values of k_cat and k_cat/K_m were calculated. Similarly, the reaction mixture (1 mL) for TMP contained 0.05–50 mM TMP, 0.25 mM NADH, 1 M NH₄Cl-NH₃‧H₂O, pH 10.0 buffer, and purified 2.0 μg PbLeuDH.

3.6. Construction of Whole-Cell Biocatalyst Coexpressing PbLeuDH and BmGDH

The Pbleudh and Bmgdh genes encoding PbLeuDH and BmGDH (GenBank ID: ADE67871.1) were amplified by PCR with P. balearica and B. megaterium ATCC 14581 genomic DNA as a template by using two primer pairs, P3/P4 and P5/P6, respectively. The detailed information on the gene sequences is presented in Table S1. Subsequently, the amplified products were assembled into a BamHI-HindIII-linearized pET28a vector by ClonExpress Kit to get the recombinant pET28a-Pbleudh-Bmgdh plasmid. The obtained plasmid was transformed (heat shock) into E. coli BL21 (DE3) to generate the PbLeuDH-BmGDH-coexpressing strain E. coli (pET28a-Pbleudh-Bmgdh).

3.7. Whole-Cell Biocatalyst Preparation

The E. coli (pET28a-Pbleudh-Bmgdh) was cultured (37 °C, 180 rpm) for 2.5 h in a conical flask (250-mL) with 50-mL LB broth containing 50 μg mL⁻¹ kanamycin, and subsequently protein expression was induced at 18 °C and 180 rpm for 14 h by adding 0.1 mM IPTG. After induction, the wet cells were collected via centrifugation (4 °C, 10 min, 10,000× g) and then washed twice using NaCl (0.85%). The resting cells were stored at 4 °C before bioconversion experiments.

3.8. L-Tle Production from TMP by Whole-Cell Biocatalysis

The initial L-Tle biosynthesis from TMP by resting cells of E. coli (pET28a-Pbleudh-Bmgdh) was conducted at 30 °C and 200 rpm for 3 h in flasks (100-mL) containing 10 mL reaction system (100 mM TMP, 120 mM glucose, 100 mM NH₄Cl, and 10 g L⁻¹ of wet cells). The reaction system pH was set at 8.5 using 4 M NaOH and maintained at 8.5 via the addition of 50% NH₃·H₂O during the reaction process. For the process optimization, five factors including whole-cell biocatalysis, pH (7.5–9.5), temperature (25–42 °C), wet cell weight (WCW, 10–40 g L⁻¹), and TMP concentration (50–400 mM) were investigated by single factor tests. The supernatant was sampled and filtered through a 0.22 μm filter and then quantified by HPLC. The cell viability was measured by PI staining and flow cytometry according to the method established by Wang et al. [29].

3.9. Analytical Method

Based on the method established previously [10], the TMP and L-Tle were quantified by an HPLC platform (Ultimate 3000, Thermo Fisher, Waltham, MA, USA) equipped with an Agilent ZORBAX SB-Aq C18 column (4.6 mm × 250 mm, 5 μm) at a wavelength of 205 nm (UV-VIS detector) using a mobile phase consisting of KH₂PO₄ (20 mM, pH 2.6): methanol with the 95:5 (v/v) ratio at a 0.8 mL min⁻¹ flow rate with 10 μL of sample volume and 30 °C of column temperature.

4. Conclusions

In conclusion, we successfully cloned, expressed, and characterized a new leucine dehydrogenase named PbLeuDH, which exhibited excellent catalytic performance toward TMP and thus was deemed as the potential biocatalyst for converting TMP into L-Tle. Subsequently, the L-Tle efficient biosynthesis was implemented from TMP via whole-cell biocatalysis using recombinant E. coli as the catalyst, in which PbLeuDH was co-expressed with BmGDH. Ultimately, 273 mM (35.8 g L⁻¹) L-Tle was successfully achieved with a 96.1% yield and 2.39 g L⁻¹ h⁻¹ productivity. Collectively, this study has initially synthesized L-Tle by whole-cell biocatalysis, laying a solid foundation for the realization of large-scale L-Tle industrial production.