Coproduction of 5-Aminovalerate and δ-Valerolactam for the Synthesis of Nylon 5 From L-Lysine in Escherichia coli

The compounds 5-aminovalerate and δ-valerolactam are important building blocks that can be used to synthesize bioplastics. The production of 5-aminovalerate and δ-valerolactam in microorganisms provides an ideal source that reduces the cost. To achieve efficient biobased coproduction of 5-aminovalerate and δ-valerolactam in Escherichia coli, a single biotransformation step from L-lysine was constructed. First, an equilibrium mixture was formed by L-lysine α-oxidase RaiP from Scomber japonicus. In addition, by adjusting the pH and H2O2 concentration, the titers of 5-aminovalerate and δ-valerolactam reached 10.24 and 1.82 g/L from 40 g/L L-lysine HCl at pH 5.0 and 10 mM H2O2, respectively. With the optimized pH value, the δ-valerolactam titer was improved to 6.88 g/L at pH 9.0 with a molar yield of 0.35 mol/mol lysine. The ratio of 5AVA and δ-valerolactam was obviously affected by pH value. The ratio of 5AVA and δ-valerolactam could be obtained in the range of 5.63:1–0.58:1 at pH 5.0–9.0 from the equilibrium mixture. As a result, the simultaneous synthesis of 5-aminovalerate and δ-valerolactam from L-lysine in Escherichia coli is highly promising. To our knowledge, this result constitutes the highest δ-valerolactam titer reported by biological methods. In summary, a commercially implied bioprocess developed for the coproduction of 5-aminovalerate and δ-valerolactam using engineered Escherichia coli.


INTRODUCTION
Over the years, mounting global environmental, climate change, economic concerns, and fossil fuel sources are leading to a shift in the production of traditional bulk chemicals toward more green, renewable, economic, and sustainable routes Gordillo Sierra and Alper, 2020;Wendisch, 2020). In many cases, the need has been partially met by biorefineries, in which microbial cell factories convert renewable feedstock resources into high-value and useful chemicals Klenk et al., 2020;Youn et al., 2020;Zhang et al., 2021). While many chemicals are being developed via biotechnology, polyamide monomers are an important class of compounds Prell et al., 2020;Osire et al., 2021). 5-Aminovalerate (5AVA) and δ-valerolactam are attractive monomers for the production of biopolyamides, serving as raw materials for clothes, architecture, and disposable goods.
Lactams are important chemicals used in the manufacture of commercial polyamides. However, there are few reports on the direct bioproduction of lactams from engineered microorganisms. Zhang et al. confirmed that 1.1 g/L γ-butyrolactam was formed from L-glutamate by identifying a newly 2-pyrrolidone synthase ORF26 from Streptomyces aizunensis, with a yield of 0.14 g/g . Then, Zhang et al. further revealed the catalytic promiscuity of ORF26, which cyclized ω-amino acids to produce of γ-butyrolactam, δ-valerolactam, and ε-caprolactam (Zhang et al., 2017b). However, the titers of δ-valerolactam and ε-caprolactam achieved were relatively low; 705 mg/L δ-valerolactam and 2.02 mg/L ε-caprolactam were produced, respectively. Chae et al. reported that β-alanine CoA transferase could activate ω-amino acids to produce 54.14 g/L γ-butyrolactam, 29 mg/L δ-valerolactam, and 79.6 μg/L ε-caprolactam, respectively (Chae et al., 2017). In addition, a novel route for δ-valerolactam was discovered through the direct oxidative decarboxylation of L-pipecolic acid by DavB in Xu et al.'s research (Xu et al., 2020). 90.3 mg/L δ-valerolactam was achieved from L-pipecolic acid by DavB expressed in E. coli . However, the titer of δ-valerolactam generated was rather low ( Table 1).
In this study, 5AVA and δ-valerolactam were coproduced from an equilibrium mixture by adjusting pH and H 2 O 2 in E. coli ( Figure 1). The α-amino group of L-lysine was oxidized by RaiP from S. japonicus to form the equilibrium mixture. 2K6AC, P2C, and 2 P2C in this equilibrium mixture were oxidized to generate 5AVA, δ-valerolactam, and δ-valerolactam, respectively. In addition, the ratio of 5AVA and δ-valerolactam could be regulated by pH. The route of coproduction of 5AVA and δ-valerolactam was first proposed in this study. As a result, a promising strategy for coproducing 5AVA and δ-valerolactam in a single biotransformation step by adjusting the pH and H 2 O 2 was established.

Cultivation Conditions
The engineering strains were streaked onto Luria-Bertani (LB) agar plates with 100 mg/L Amp at 37°C for overnight. Engineering strains used for biotransformation in the shake flask were cultured in LB medium with 100 mg/L Amp. After the OD 600 reached 0.6, 0.2 mM of isopropyl β-D-thiogalactoside (IPTG) and 6.5 g/L of L-lysine HCl were added. The pH was controlled at 5.0, 6.0, 7.0, 8.0, and 9.0 by NH 3 ·H 2 O and 10% H 2 SO 4 at 30°C after 12 h. H 2 O 2 was added after 12 h.

Biotransformation
Biotransformation was performed in a 5.0-L fermenter. The medium consisted of 55 g/L of glucose, 0.004 g/L of CoCl 2 ·6H 2 O, 0.02 g/L of Na 2 SO 4 , 1.6 g/L of MgSO 4 ·7H 2 O, 0.0064 g/L of ZnSO 4 , 0.0006 g/L of Cu 2 SO 4 ·5H 2 O, 1.6 g/L of (NH 4 ) 2 SO 4 , 0.00756 g/L of FeSO 4 ·7H 2 O, 2 g/L of citric acid, 7.5 g/L of K 2 HPO 4 ·3H 2 O, and 250 μl of antifoam 289. The pH was controlled at 7.0 by the automatic addition of NH 3 ·H 2 O and 10% H 2 SO 4 at 30°C. After the OD 600 reached 20, 0.2 mM IPTG was added to the broth. When the OD 600 reached 80, the pH was controlled at 5.0, 6.0, 7.0, 8.0, and 9.0 by the automatic addition of NH 3 ·H 2 O and 10% H 2 SO 4 . L-lysine HCl was added to at an initial concentration of 40 g/L. H 2 O 2 was added after 24 h.

RESULTS AND DISCUSSION
Construction of a Synthetic Route for the Simultaneous Synthesis of 5-Aminovalerate and δ-Valerolactam in E. coli A synthetic route for the concurrent synthesis of 5AVA and δ-valerolactam from L-lysine was constructed from an equilibrium mixture (Figure 1). The designed route for the coproduction of 5AVA and δ-valerolactam consists of two steps: 1) the deamination of α-amino group in L-lysine to generate an equilibrium mixture by RaiP from S. japonicus, with this equilibrium mixture containing P2C, 2HP2C, 2 P2C, 2K6AC, 6A2HH2E, and 6A2DHC; 2) the decarboxylation of 2K6AC, P2C, and 2 P2C in this equilibrium mixture to produce 5AVA and δ-valerolactam via H 2 O 2 , respectively. First, a plasmid pET21a-raiP was constructed and introduced into E. coli BL21(DE3) to obtain the strain BL21-raiP. As shown FIGURE 2 | Feasibility for the coproduction of 5AVA and δ-valerolactam in a single biotransformation step. pH was controlled at 7.0. 6.5 g/L of L-lysine HCl was added as the substrate. Statistics were performed by two-tailed Student's t-test. *p < 0.05. All experiments were performed a minimum of three independent sets. in Figure 2, engineering strain BL21-raiP produced 0.26 g/L 5AVA and 0.08 g/L δ-valerolactam in the absence of H 2 O 2, and 0.45 g/L 5AVA and 0.24 g/L δ-valerolactam in pH 7.0 and 5 mM H 2 O 2 , respectively. The specific activity of RaiP was 5.14 units/mg. These results demonstrated the feasibility of the coproduction of 5AVA and δ-valerolactam in E. coli. In addition, the strain ML03-raiP with cadA knocked out was constructed. The engineered strain ML03-raiP produced 0.58 g/L 5AVA and 0.29 g/L δ-valerolactam, nearly about 0.29-fold and 0.21-fold increase compared to control strain BL21-raiP at pH 7.0 and 5 mM H 2 O 2 ( Figure 2). In the past, some studies of the concurrent bioproduction of bulk chemicals were investigated (Li et al., 2017). Few examples are the simultaneous synthesis of 5AVA and glutarate by Corynebacterium glutamicum (Rohles et al., 2016;Haupka et al., 2020), β-glucan and pullulan by engineering Aureobasidium pullulans (Wang G.-L. et al., 2020), acetoin and succinic acid by Enterobacter cloacae (Su et al., 2021), polyhydroxyalkanoates and exopolysaccharides by Yangia sp. ND199 (Romero Soto et al., 2021), and xylitol and ethanol by yeast strains (Shankar et al., 2020). Lopez-Hidalgo et al. reported that the engineered strain increased 30% the coproduction of ethanol and hydrogen used wheat straw and corn stover as feedstock (Lopez-Hidalgo et al., 2021). And 11.0 g/L polyhydroxybutyrate and 1.5 g/L violacein pigment were successfully co-synthesized in Iodobacter sp. PCH194 (Kumar et al., 2021). 7,12-dioxolithocholate and L-tert-leucine were simultaneously produced in a cofactor self-sufficient cascade system for enhancing the atom efficiency (You et al., 2021). Chae et al. found that only 29 mg/L δ-valerolactam was produced by β-alanine CoA transferase (Chae et al., 2017). Xu et al. reported that 90.3 mg/L δ-valerolactam was successfully obtained by an oxidative decarboxylase DavB . However, the low titers limit the prospect of industrial application.

The Effect of H 2 O 2 on the Simultaneous Synthesis of 5-Aminovalerate and δ-Valerolactam
The effect of H 2 O 2 on the simultaneous synthesis of 5AVA and δ-valerolactam in engineering strain ML03-raiP-katE at pH 7.0 is shown in Figure 3. It showed that the addition of H 2 O 2 had a significant effect on the titers of 5AVA and δ-valerolactam. Engineering E. coli ML03-raiP-katE was cultured in LB medium to form an equilibrium mixture containing P2C, 2HP2C, 2 P2C, 2K6AC, 6A2HH2E, and 6A2DHC. At 5 mM H 2 O 2 addition, recombinant ML03-raiP-katE produced 0.58 g/ L 5AVA and 0.29 g/L δ-valerolactam after 24 h, respectively, increased about 0.87-fold and 2.22-fold compared to the control group without H 2 O 2 . With the continuous increase in H 2 O 2 concentration to 10 mM, the titers of 5AVA and δ-valerolactam both were further increased to 0.96 g/L 5AVA and 0.42 g/L δ-valerolactam, with a yield increase of 2.13-fold and 3.67-fold compared to the control without H 2 O 2 addition, respectively. However, with the increase in H 2 O 2 concentration to 15 mM, the titers of 5AVA and δ-valerolactam decreased dramatically (Figure 3).
H 2 O 2 is an important reactive oxygen species in organisms and is produced in response to signal transduction, growth, and development (Oldroyd, 2013;Sies and Jones, 2020). H 2 O 2 enters cells to regulate signaling and cellular processes through aquaporin membrane proteins and covalently modifies cytoplasmic proteins (Sies and Jones, 2020). Wu et al. found that H 2 O 2 sensor HPCAI is a receptor kinase (Wu et al., 2020). However, excess of H 2 O 2 could inhibit cell growth and affect the production of target compounds, resulting in low OD 600 (Cheng et al., 2018b). Therefore, in this study, a strategy was proposed that the H 2 O 2 produced by RaiP was decomposed by overexpression of catalase in E. coli in the early stage, and then H 2 O 2 was added in the later stage to produce 5AVA and δ-valerolactam. The specific activity of KatE was 23.58 units/mg. The H 2 O 2 that is generated by RaiP can affect the cell growth and the titers of products (Cheng et al., 2021b). The coexpression of RaiP and KatE in E. coli might provide a more convenient and effective method for the production of 5AVA and δ-valerolactam. As shown in Supplementary Figure S1, the coexpressed E. coli BL21 (DE3) strain harboring pET21a-raiP-katE showed another distinct 84-kDa band on SDS-PAGE, which was consistent with the calculated molecular weight of catalase.

The Effect of pH on the Ratio of 5-Aminovalerate and δ-Valerolactam
The effect of pH on the ratio of 5AVA and δ-valerolactam in engineering strain ML03-raiP-katE with 10 mM H 2 O 2 addition is shown in Table 3. It showed that the pH had a great effect on the ratio of 5AVA and δ-valerolactam; 1.12 g/L 5AVA and 0.25 g/L δ-valerolactam were generated at pH 5.0 after adding H 2 O 2 for 12 h. The maximum ratio of 5AVA and δ-valerolactam was FIGURE 3 | Effect of H 2 O 2 on the coproduction of 5AVA and δ-valerolactam by strain ML03-raiP-katE in flasks. pH was controlled at 7.0. 6.5 g/L of L-lysine HCl was added as the substrate. All experiments were performed a minimum of three independent sets.
Frontiers in Bioengineering and Biotechnology | www.frontiersin.org September 2021 | Volume 9 | Article 726126 reached 4.48:1 at pH 5.0. With the increase in pH, the titer of δ-valerolactam increased gradually, resulting in a decrease in the ratio of 5AVA and δ-valerolactam; 1.08 g/L 5AVA and 0.33 g/L δ-valerolactam were obtained at pH 6.0. When the pH value was 7.0, recombinant ML03-raiP-katE could produce 0.96 g/L 5AVA and 0.42 g/L δ-valerolactam after 24 h from the equilibrium mixture. In addition, the titer of δ-valerolactam increased significantly to 0.56 g/L at pH 8.0, with a titer increase of 0.33fold compared with pH 7.0. Interestingly, the titer of δ-valerolactam was higher than 5AVA at pH 9.0, and the ratio of 5AVA and δ-valerolactam was 0.91. As a result, the flux of the equilibrium mixture would shift to 5AVA under acidic condition and to δ-valerolactam under alkaline condition. These findings are consistent with Kamio's research (Kamio et al., 2009). However, their specific ratio has not been reported (Ko et al., 2008;Kamio et al., 2009).

Biotransformation for the Coproduction of 5-Aminovalerate and δ-Valerolactam
Time profiles for the simultaneous synthesis of 5AVA and δ-valerolactam were investigated by biotransformation of engineered strain ML03-raiP-katE at pH 5.0 ( Figure 4A) and pH 9.0 ( Figure 4B) in a 5-L fermenter. The catalase KatE was overexpressed to remove H 2 O 2 , which significantly improved OD 600 and the titer of products in the 5-L fermenter (Cheng et al., 2021b). The titers of 5AVA and δ-valerolactam were very low before the addition of H 2 O 2 . In this process, the main accumulation was the equilibrium mixture produced by RaiP from lysine. Although H 2 O 2 was produced by RaiP, its low concentration leads to low production of 5AVA and δ-valerolactam. After adding H 2 O 2 for 12 h, the titers of 5AVA and δ-valerolactam increased significantly to 8.88 and 1.56 g/L at pH 5.0. Finally, 10.24 g/L 5AVA and 1.82 g/L δ-valerolactam were obtained, with a total molar yield of 0.52 mol/mol lysine, and its ratio was 5.63:1 at pH 5.0. The difference was that the titers of 5AVA and δ-valerolactam were 3.42 and 5.12 g/L after adding H 2 O 2 for 12 h at pH 9.0. Finally, 3.98 g/L 5AVA and 6.88 g/L δ-valerolactam were obtained, with a total molar yield of 0.51 mol/mol lysine, and its ratio was 0.58:1 at pH 9.0. The previous results showed that the ratio of 5AVA and δ-valerolactam was significantly regulated by pH. δ-Valerolactam would be the main component in alkaline condition.
We have previously reported the production of 5AVA by overexpression of RaiP, but the titer and change in δ-valerolactam were not noticed in this process. At the same time, the addition of ethanol improved the expression level of RaiP, which increases the cost and leads to uneconomical (Cheng et al., 2018b;Cheng et al., 2020;Cheng et al., 2021a). Xu et al. reported that the expression of DavB from P. putida could synthesize 90.3 mg/L of δ-valerolactam from L-pipecolic acid . Interestingly, the coexpression of RaiP, glucose dehydrogenase GDH, P2C reductase DpkA, and LysP could produce more δ-valerolactam from lysine, up to 242 mg/L . This may be due to the fact that part of δ-valerolactam does not originate from the oxidative decarboxylation of L-pipecolic acid but from this equilibrium mixture in this study. Compared with other biotransformation for production of 5AVA, the advantage in this study was to realize the simultaneous synthesis of 5AVA and δ-valerolactam. In terms of biotransformation mechanism, the simultaneous synthesis of 5AVA and δ-valerolactam mainly includes two steps: 1) the formation of an equilibrium mixture by RaiP from lysine and 2) the oxidization of the equilibrium mixture to 5AVA and δ-valerolactam by H 2 O 2 at different pH values.

CONCLUSION
Many important monomers of polyamides, such as adipate, cadaverine, and 3-hydroxybutyrate, have been extensively studied in microbes. The results presented here demonstrated that engineering E. coli also has the potential to be used as a promising alternative to produce monomers of polyamides derived from petrochemicals. In this study, the strategy for coproducing 5AVA and δ-valerolactam by adjusting the pH and H 2 O 2 in E. coli was proposed. H 2 O 2 was regulated to improve the synthesis efficiency of δ-valerolactam in E. coli in different a Statistical analysis of the 5AVA production was performed with every two separated lines. 6.5 g/L L-lys HCl and 0.2 mM IPTG were added. 10 mM H 2 O 2 was added after 12 h.
Frontiers in Bioengineering and Biotechnology | www.frontiersin.org September 2021 | Volume 9 | Article 726126 6 pH environments, which also increased 5AVA accumulation. The ratio of 5AVA and δ-valerolactam was significantly affected by pH value. δ-Valerolactam would be the main component in alkaline condition. The titers of 5aminovalerate and δ-valerolactam reached 3.98 and 6.88 g/ L from 40 g/L L-lysine HCl at pH 9.0, with a total yield of 0.51 mol/mol lysine. The present findings indicated a promising strategy for the simultaneous synthesis of two commercial products in a single biotransformation step. These strategies could be widely applied for sustainable production of many commercially monomers of polyamides.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material; further inquiries can be directed to the corresponding authors.

AUTHOR CONTRIBUTIONS
JC, WT, and ZL performed the experiments and analyzed the data. LL, XG, and XW analyzed data. JC and CL drafted the FIGURE 4 | Time profiles of 5AVA and δ-valerolactam production were investigated by biotransformation of engineered strain ML03-raiP-katE at pH 5.0 (A) or pH 9.0 (B) in a 5-L fermenter. The experiments were conducted at 40 g/L L-lysine HCl, 37°C and 250 rpm. 10 mM H 2 O 2 was added after reaction 24 h. All experiments were performed a minimum of three independent sets.
Frontiers in Bioengineering and Biotechnology | www.frontiersin.org September 2021 | Volume 9 | Article 726126 7 manuscript. JC, CL, and GZ coordinated the study and finalized the manuscript.