Pulcherriminic acid is a cyclodipeptide derived from cyclo(l-Leu–l-Leu), which shares the same iron chelation group with hydroxamate sidephores. Generally, pulcherriminic acid-producing strains could be the perfect candidates for antibacterial and anti-plant-pathogenic fungal agents. In this study, we obtained the promising W4/pHY-yvmA pulcherriminic acid-producing strain via a multistep metabolic modification. The engineered W4/pHY-yvmA strain is able to achieve 556.1 mg/liter pulcherriminic acid production, which is the highest yield so far to the best of our knowledge.
KEYWORDS: Bacillus licheniformis, pulcherriminic acid, cyclodipeptides, metabolic engineering
ABSTRACT
The cyclodipeptide pulcherriminic acid, produced by Bacillus licheniformis, is derived from cyclo(l-Leu–l-Leu) and possesses excellent antibacterial activities. In this study, we achieved the high-level production of pulcherriminic acid via multistep metabolic engineering of B. licheniformis DWc9n*. First, we increased leucine (Leu) supply by overexpressing the ilvBHC-leuABCD operon and ilvD, involved in Leu biosynthesis, to obtain strain W1, and the engineered strain W2 was further attained by the deletion of gene bkdAB, encoding a branched-chain α-keto acid dehydrogenase in W1. As a result, the intracellular Leu content and pulcherriminic acid yield of W2 reached 147.4 mg/g DCW (dry cell weight) and 189.9 mg/liter, which were 227.6% and 48.9% higher than those of DWc9n*, respectively. Second, strain W3 was constructed through overexpressing the leucyl-tRNA synthase gene leuS in W2, and it produced 367.7 mg/liter pulcherriminic acid. Third, the original promoter of the pulcherriminic acid synthetase cluster yvmC-cypX in W3 was replaced with a proven strong promoter, PbacA, to produce the strain W4, and its pulcherriminic acid yield was increased to 507.4 mg/liter. Finally, pulcherriminic acid secretion was strengthened via overexpressing the transporter gene yvmA in W4, resulting in the W4/pHY-yvmA strain, which yielded 556.1 mg/liter pulcherriminic acid, increased by 337.8% compared to DWc9n*, which is currently the highest pulcherriminic acid yield to the best of our knowledge. Taken together, we provided an efficient strategy for enhancing pulcherriminic acid production, which could apply to the high-level production of other cyclodipeptides.
IMPORTANCE Pulcherriminic acid is a cyclodipeptide derived from cyclo(l-Leu–l-Leu), which shares the same iron chelation group with hydroxamate sidephores. Generally, pulcherriminic acid-producing strains could be the perfect candidates for antibacterial and anti-plant-pathogenic fungal agents. In this study, we obtained the promising W4/pHY-yvmA pulcherriminic acid-producing strain via a multistep metabolic modification. The engineered W4/pHY-yvmA strain is able to achieve 556.1 mg/liter pulcherriminic acid production, which is the highest yield so far to the best of our knowledge.
INTRODUCTION
Cyclodipeptides, also known as 2,5-diketopiperazine or 2,5-dioxopiperazine, are a class of peptides catalyzed by cyclodipeptide synthetase (CDPS) (1, 2). In recent years, a variety of biological properties of cyclodipeptides have been reported (3). For example, cyclo(Phe-Pro), cyclo(Tyr-Phe), and cyclo(Leu-Tyr) act as quorum-sensing signaling molecules. Cyclo(d-Pro–l-Phe) exhibits antibacterial activity, and cyclo(l-Phe–l-Pro) and cyclo(l-Phe–trans-4-OH-Pro) antagonize fungi. Moreover, cyclodipeptides can act as molecular fragments for molecular drug design (4). Therefore, increasing attention has been paid to the exploration of naturally novel bioactive cyclodipeptides and the construction of recombinant strains with high-level production of cyclodipeptides.
Pulcherriminic acid is a cyclodipeptide derived from cyclo(l-Leu–l-Leu) (5) and is synthesized mainly by yeasts and Bacillus species (6, 7). Iron chelation transforms pulcherriminic acid into insoluble pulcherrimin and further results in iron deletion in the environment, which might be the reason for the antimicrobial activity exhibited by some bacteria and yeasts (8). The pulcherriminic acid-producing strains could restrain the growth of various bacteria and pathogenic fungi (6, 9) and, thus, have been considered great candidates for biocontrol and bacteriostatic agents (10–12).
In recent years, the synthetic pathway of pulcherriminic acid has been extensively studied in Bacillus (13), which removed the barrier to enhancing pulcherriminic acid production. As shown in Fig. 1, leucine (Leu), the precursor of pulcherriminic acid, is formed by catalyzing IlvBH (acetohydroxyacid synthase), IlvCD (keto-acid reductoisomerase and dihydroxy-acid dehydratase), and LeuABCD (2-isopropylmalate synthase, 3-isopropylmalate dehydrogenase, and dehydratase) (14). Leucyl-tRNA synthetase LeuS then catalyzes Leu into leucyl-tRNA (15). The CDPS YvmC then uses two molecules of leucyl-tRNA to generate cyclo(l-Leu–l-Leu) (cLL). A cytochrome P450 encoded by cypX next oxidizes cLL to form pulcherriminic acid (13). Ultimately, pulcherriminic acid is secreted from the cell through major facilitator superfamily (MFS) transporters (16, 17).
FIG 1.
Metabolic engineering of B. licheniformis for enhanced production of pulcherriminic acid. Red arrows indicate the overexpressed pathways, multiple arrows represent multistep reactions, and the red X indicates a deletion.
Due to the repression of CodY and TnrA on the ilvBHC-leuABCD operon and the negative feedback of Leu on 2-isopropylmalate synthase LeuA, the supplementing of Leu in B. licheniformis might be insufficient, indicating that the substrate supply is one of the limiting factors for pulcherriminic acid production (18). The transformation of Leu into Leu-tRNA also is considered to affect peptide and protein synthesis efficiency (15). Based on our previous research, the transcription of the pulcherriminic acid synthesis gene cluster yvmC-cypX and secretion protein gene yvmA were regulated by multiple regulators (AbrB, YvnA, and YvmB) in an iron-rich environment (17).
Previously, an efficient genome-editing tool mediated by CRISPR/Cas9 nickase was established in B. licheniformis DW2, resulting in the strain DWc9n* (19). In this work, we aimed to engineer an efficient pulcherriminic acid-producing strain based on the original strain DWc9n*. Here, we employed a multistep metabolic engineering strategy, including redirecting the carbon flux toward precursor Leu and leucyl-tRNA biosynthesis, cyclodipeptide synthetase overexpression, and transportation enhancement. Collectively, this study provided an exemplary strategy as well as a promising B. licheniformis strain for high-level production of pulcherriminic acid.
RESULTS
Strengthening Leu synthesis and its effect on pulcherriminic acid production.
Leucyl-tRNA generated from Leu serves as the sole substrate for pulcherriminic acid synthesis (20). In this study, to evaluate the effect of Leu supply on pulcherriminic acid production, different concentrations of Leu were added to the pulcherriminic acid production medium. Based on our results, the addition of Leu benefited pulcherriminic acid production, and the addition of 2.0 g/liter Leu resulted in the best performance, the yield of which was increased by 24.0% (Fig. 2A). This result indicated that the inadequate Leu synthesis of DWc9n* is one of the limiting factors in pulcherriminic acid synthesis.
FIG 2.
ilvBHC-leuABCD and ilvD overexpression and bkdAB deletion lead to increased pulcherriminic acid production. (A) The yield of pulcherriminic acid and OD600 under different Leu concentrations in DWc9n*. (B) Gene transcription of ilvC in the ilvBHC-leuABCD operon and ilvD at 20 h. (C) The concentrations of intracellular Leu among these recombinant strains at 24 h. (D) Pulcherriminic acid yield and OD600. All experiments were performed in three replicates, and data are presented as the means ± standard deviations for each sample point. All data were analyzed for the variance at P values of <0.05 and <0.01, and a t test was applied to compare the mean values using the software package Statistica 6.0. One (P < 0.05) and two (P < 0.01) asterisks indicate the significance levels between recombinant strains and the control.
To relieve the repression of CodY and TnrA on the promoters of ilvD and the ilvBHC-leuABCD operon and strengthen the Leu synthetic pathway, the promoters of ilvD and ilvBHC-leuABCD both were replaced with the widely used strong promoter PbacA successively, resulting in the strain W1. The transcription of ilvD and the ilvBHC-leuABCD operon of W1 was evaluated by quantitative reverse transcription-PCR (qRT-PCR) at 20 h. Our results showed that the transcription of ilvD and ilvC was increased by 141.6% and 91.7%, respectively, compared with those of DWc9n* (Fig. 2B). In addition, the intracellular Leu concentration was increased 173.6% compared to that of DWc9n* (45.1 mg/g dry cell weight [DCW]) at 24 h (Fig. 2C). As a result, the yield of pulcherriminic acid in W1 reached 149.9 mg/liter, an increase of 17.6% compared to that of DWc9n* (127.5 mg/liter) (Fig. 2D).
The bkdAB gene encodes the E1 subunit of the branched-chain α-keto acid dehydrogenase complex, which converts branched-chain amino acids (Leu, Ile, and Val) to branched-chain α-ketoacyl-coenzyme A starters (BCCSs), involved in the synthetic pathways of branched-chain fatty acids (21). To further improve the accumulation of intracellular Leu, the bkdAB gene was deleted in W1 to obtain strain W2. Based on the results shown in Fig. 2D, the deletion of bkdAB was not conducive to cell growth (22), and cell density was decreased by 23.8% in W2 compared to that of DWc9n*. A possible reason for this is that branched-chain fatty acids are important components of the cell membrane (21). While the concentration of intracellular Leu reached 147.4 mg/g DCW in W2, it was 19.7% higher than that of W1 (Fig. 2C). As a result, the deletion of bkdAB benefited pulcherriminic acid production (189.9 mg/liter), which was increased 26.7% compared to that of W1 (Fig. 2D). Taken together, these results inferred that enhancing the Leu supply promoted the production of pulcherriminic acid.
Strengthening the expression of leucyl-tRNA synthase LeuS increased pulcherriminic acid production.
CDPS takes the activated amino acids in the form of amino acid tRNAs (aa-tRNAs) to form cyclodipeptides, and aa-tRNA was attained under the catalysis of aa-tRNA synthetase (23). Hence, the overexpression of leuS might strengthen the conversion of Leu to leucyl-tRNA and further benefit pulcherriminic acid production. Based on that possibility, we replaced the native promoter of leuS with the strong promoter PbacA in the W2 strain to construct the W3 strain. The results showed that the transcription of leuS in W3 was 141.6% higher than that of DWc9n* (Fig. 3A), and the intracellular Leu concentration was decreased by 30.2% compared to that of W2 (Fig. 2C). The pulcherriminic acid yield of W3 reached 367.7 mg/liter, increases of 93.6% and 189.5% compared to those of W2 and DWc9n*, respectively (Fig. 2D and 3B). Therefore, these results demonstrated that the overexpression of leuS promoted the transformation of Leu to pulcherriminic acid.
FIG 3.
Effect of leuS and yvmC-cypX overexpression on pulcherriminic acid production. (A) Gene transcription of the leuS gene and yvmC in the yvmC-cypX operon at 20 h. (B) Pulcherriminic acid yield and OD600. All experiments were performed in three replicates, and data are presented as the means ± standard deviations for each sample point. All data were analyzed for the variance at P values of <0.05 and <0.01, and a t test was applied to compare the mean values using the software package Statistica 6.0. One (P < 0.05) and two (P < 0.01) asterisks indicate the significance levels between recombinant strains and the control.
Enhancing the expression of yvmC-cypX gene cluster promoted pulcherriminic acid production.
Although gene leuS was overexpressed to increase precursor leucyl-tRNA accumulation for pulcherriminic acid synthesis, the concentration of intracellular Leu in strain W3 remained higher than that of DWc9n*. To achieve more efficient Leu utilization for pulcherriminic acid production, the pulcherriminic acid biosynthesis gene cluster yvmC-cypX was overexpressed via promoter replacement based on the strain W3, resulting in strain W4.
The results of qRT-PCR showed that the yvmC transcription in W4 strain was 452.1% higher than that in DWc9n* (Fig. 3A). Meanwhile, the concentration of intracellular Leu in W4 (55.3 mg/g DCW) was decreased by 46.3% compared to that of W3 (102.9 mg/g DCW) (Fig. 2C). As expected, the yield of pulcherriminic acid produced by W4 reached 507.4 mg/liter, increases of 38.0% and 299.5% compared with those of W3 and DWc9n*, respectively (Fig. 3B). Taken together, the results suggested that the yvmC-cypX gene cluster acted as the rate-limiting step for pulcherriminic acid synthesis.
Overexpression of transporter gene yvmA promoted pulcherriminic acid secretion.
To further improve pulcherriminic acid production, the major facilitator superfamily (MFS)-like transporter YvmA, which has been proven to be the pulcherriminic acid exporter (17), was overexpressed in W4, resulting in the W4/pHY-yvmA strain. In the W4/pHY-yvmA strain, the transcription of yvmA was 34.5-fold higher than that of DWc9n*/pHY300 (Fig. 4A), and the concentration of intracellular Leu was 42.7 mg/g DCW at 24 h, significantly lower than that of DWc9n* (Fig. 2C). Moreover, the yield of pulcherriminic acid in the W4/pHY-yvmA strain reached 556.1 mg/liter, increases of 9.0% and 327.8% compared to those of W4/pHY300 and DWc9n*/pHY300, respectively (Fig. 4B). In addition, our results showed that the intracellular pulcherriminic acid content in the W4/pHY-yvmA strain was 8.4 mg/g DCW, significantly lower than that of W4/pHY300 (31.5 mg/g DCW) (Fig. 4C). Therefore, our results demonstrated that the efficient secretion of pulcherriminic acid benefited its biosynthesis.
FIG 4.
Effect of yvmA overexpression on pulcherriminic acid production. (A) Gene transcription of yvmA at 20 h. (B) Pulcherriminic acid yield and OD600. (C) The concentrations of intracellular pulcherriminic acid at 44 h. All experiments were performed in three replicates, and data are presented as the means ± standard deviations for each sample point. All data were analyzed for the variance at P values of <0.05 and <0.01, and a t test was applied to compare the mean values using the software package Statistica 6.0. One (P < 0.05) and two (P < 0.01) asterisks indicate the significance levels between recombinant strains and the control.
Furthermore, the time curves of B. licheniformis DWc9n*/pHY300 and W4/pHY-yvmA strains were measured, and the residual substrates and main by-products were determined. As shown in Fig. 5A, the maximum cell density of DWc9n*/pHY300 was attained at 40 h, followed by a slight decrease from 44 h to 48 h. The growth trend of the W4/pHY-yvmA strain was similar to that of DWc9n*/pHY300 from 0 to 36 h, but the growth rate and cell density declined gradually after 36 h. Meanwhile, pulcherriminic acid produced by DWc9n*/pHY300 increased slightly throughout the process and reached 130.5 mg/liter at 48 h. The maximum yield of pulcherriminic acid in the W4/pHY-yvmA strain reached 556.1 mg/liter, which was 327.8% higher than that of DWc9n*/pHY300. Additionally, the high growth rate of DWc9n*/pHY300 resulted in fast glucose consumption in the early growth stage. After 24 h, the yield of pulcherriminic acid in the W4/pHY-yvmA strain increased a great deal, and glucose levels in the culture medium dropped (Fig. 5B). In addition, the highest total concentration of the main by-products, acetoin and 2,3-butanediol (2,3-BD), of the W4/pHY-yvmA strain was 8.5 g/liter, decreased by 19.1% compared to that of DWc9n*/pHY300 (10.4 g/liter) (Fig. 5C). All in all, these results confirmed that the metabolic engineering strategy used in this research could effectively strengthen the carbon flux toward pulcherriminic acid biosynthesis.
FIG 5.
Time curves of B. licheniformis DWc9n*/pHY300 and W4/pHY-yvmA strains during pulcherriminic acid production. (A) Pulcherriminic acid and OD600. (B) Residual glucose and citric acid. (C) Acetoin and 2,3-butanediol. All experiments were performed in three replicates, and data are presented as the means ± standard deviations for each sample point.
DISCUSSION
Bacillus-based biological control agents possess great potential in integrated pest management systems (24). The cyclodipeptide pulcherriminic acid, synthesized by B. licheniformis, works as an iron chelator to antagonize certain pathogens (25). In this study, the capability of pulcherriminic acid synthesis was significantly increased by rewiring the metabolic pathways of B. licheniformis, and the recombinant strain W4/pHY-yvmA was constructed, giving the highest yield of pulcherriminic acid to date.
Precursor supply plays a critical role in metabolite biosynthesis (26). Here, Leu, derived from pyruvate, acted as the initial precursor for pulcherriminic acid synthesis. Because of the low concentration of intracellular Leu in B. licheniformis DW2 (27), Leu supply might be a bottleneck for pulcherriminic acid production. Vogt et al. (28) have obtained an efficient l-Leu production strain of Corynebacterium glutamicum by removing the feedback resistance and deleting the repressor LtbR for l-Leu synthesis, and Leu yield was increased to 24 g/liter under optimized fed-batch fermentation. In addition, Zhu et al. (27) improved the accumulation of intracellular branched-chain amino acids (BCAAs) by overexpressing the BCAA importer BrnQ, which led to a 22.4% increase in the bacitracin yield. In our work, the concentration of intracellular Leu was increased by 227.6% via overexpressing ilvBHC-leuABCD and deleting bkdAB, and a 49.0% increase of pulcherriminic acid yield was achieved. At the same time, overexpressing aminoacyl tRNA synthetase was proved an efficient strategy for metabolite synthesis. Xia et al. (29) upregulated the expression of glycyl-tRNA synthetase to enhance the glycyl-tRNA pool for producing a spider silk-like protein. Here, the aminoacyl tRNA synthetase LeuS was overexpressed to increase the formation of leucyl-tRNA, which led to a 189.5% increase of pulcherriminic acid production compared to that of DWc9n* (Fig. 3B) and a 30.6% decrease of intracellular Leu compared to that of W2 (Fig. 2C). In summary, our results showed that precursor supply played a vital role in pulcherriminic acid synthesis, and pulcherriminic acid yield could be increased via strengthening the supplies of Leu and leucyl-tRNA.
However, a high concentration of intracellular Leu might trigger Leu feedback inhibition on the 2-isopropylmalate synthetase LeuA (28). To verify whether the intracellular Leu reached the concentration threshold of feedback inhibition on LeuA activity, we introduced the leuA mutations R529H and G532D into the W2 genome, resulting in W2(HD). The results showed that the pulcherriminic acid production and intracellular Leu concentration of W2(HD) did not increase significantly (see Fig. S1 and S2 in the supplemental material) compared with those of W2. This result indicated that the Leu concentration was not high enough to trigger the LeuA feedback inhibition.
The recognition of substrate by CDPS is dependent on the molecular structure of the amino acids and corresponding aa-tRNAs, which resulted in the lower substrate specificity of CDPS (23). Therefore, several cyclodipeptide by-products synthesized in this process significantly decreased the synthesis efficiency of target cyclo-amino acids and related metabolites (30). Although most CDPS shows a degree of nonspecific substrate selectivity, it mainly incorporates five hydrophobic amino acids (Phe, Leu, Tyr, Met, and Trp) into cyclodipeptides. For instance, the cyclodipeptides catalyzed by CDPS AlbC from Streptomyces sinensis showed at least 12 different amino acid combinations, such as cyclo(l-Phe–l-Leu), cyclo(l-Phe–l-Tyr), cyclo(l-Phe–l-Phe), cyclo(l-Phe–l-Met), and cyclo(l-Tyr–l-Met) (15). Six cyclodipeptides, including cyclo(l-Tyr–l-Tyr), cyclo(l-Tyr–l-Phe), cyclo(l-Tyr–l-Leu), cyclo(l-Tyr–l-Ala), cyclo(l-Tyr–l-Met), and cyclo(l-Tyr–l-Trp), were synthesized by CDPS Rv2275 of Mycobacterium tuberculosis (31). Based on previous research, CDPS YvmC from Bacillus also showed a low substrate specificity, and approximately 40% of products were cyclo(l-Leu–l-Phe) and cyclo(l-Leu–l-Met) instead of the target product cLL (13). In this research, our results showed that the concentration of intracellular Phe was decreased significantly in the yvmC-cypX overexpression strain (Fig. S3), synchronously with that of intracellular Leu, which suggested that cyclo(l-Leu–l-Phe) also is accumulated in strain W4. Unfortunately, the concentration of intracellular Met has not been detected in this research. Despite this, the catalysis specificity of YvmC should be engineered to improve pulcherriminic acid production in our subsequent work.
Unlike antibiotic-producing strains, which are self-resistant to excreted antibiotics, the growth of the pulcherriminic acid-producing strains were repressed by iron starvation (17). Thus, the synthesis of cyclic dipeptides has to be rigorously regulated in microorganisms (16). Based on previous research, the synthesis of pulcherriminic acid is regulated by the transcriptional regulators AbrB, YvnA, and YvmB to reach iron homeostasis under a range of iron stresses (17). Similar dual-regulation systems were also confirmed in the biosynthesis of several other secondary metabolites (15, 32). In this study, the strategy of promoter replacement might have broken the rigorous regulatory system of B. licheniformis and allowed a pulcherriminic acid yield of 556.1 mg/liter in the W4/pHY-yvmA strain. To the best of our knowledge, the highest pulcherriminic acid yield of B. licheniformis was found to be 331.2 mg/liter via optimized fermentation engineering strategies (25). In our study, the pulcherriminic acid yield of 556.1 mg/liter produced by the W4/pHY-yvmA strain has been the highest yield to date.
MATERIALS AND METHODS
Strains, plasmids, and growth conditions.
The strains and plasmids used in this study are presented in Table 1. The oligonucleotide primers used in this study are listed in Table 2. Escherichia coli DH5α was used for vector construction, and B. licheniformis DWc9n* was used as the host strain to construct recombinant strains (19). The plasmid pHY300 was used for constructing the gene expression vector. Luria-Bertani (LB) medium was applied for cultivating E. coli and B. licheniformis; if necessary, antibiotics (20 μg/liter tetracycline, 20 μg/liter kanamycin; Sigma-Aldrich) were added to the medium. The seed culture was cultivated at 37°C for 12 h and then transferred into the pulcherriminic acid production medium [in grams per liter: 40 glucose, 12 sodium citrate, 6.20 (NH4)2SO4, 0.50 K2HPO4·3H2O, 0.50 MgSO4·7H2O, 0.20 CaCl2·2H2O, 0.25 FeCl3·6H2O, and 0.02 MnSO4·H2O, pH 7.50] in 250-ml flasks on a rotary shaker (230 rpm) at 37°C for 48 h.
TABLE 1.
Strains and plasmids used in this research
| Strain or plasmid | Descriptiona | Source |
|---|---|---|
| Strains | ||
| E. coli DH5α | supE44 ΔlacU169 (ϕ80 lacZΔM15) hsdR17 recA1 gyrA96 thi1 relA1 | TaKaRa Co., Ltd. |
| B. licheniformis DWc9n* | B. licheniformis DW2 derivative with Cas9n integrated expression and bacABC operon knocked out | Laboratory collection |
| W1 | DWc9n* derivative, PbacA-ilvBHC-leuABCD-ilvD | This study |
| W2 | W1 derivative, ΔbkdAB | This study |
| W3 | W2 derivative, PbacA-leuS | This study |
| W4 | W3 derivative, PbacA-yvmC-cypX | This study |
| W4/pHY300 | W4 with pHY300 | This study |
| W4/pHY-yvmA | W4 with pHY-yvmA | This study |
| Plasmids | ||
| pHY300 | E. coli-Bacillus shuttle vector; Ampr in E. coli, Tcr in both E. coli and B. licheniformis | Laboratory collection |
| pHY-yvmA | pHY300 derivative harboring yvmA expression cassette for over expression of yvmA; Ampr, Tcr | This study |
| pGRNA | pHY300 + P43(noRBS) + gRNA + TamyL(target yvmC) | Laboratory collection |
| pGRNA-PbacA(ilvB) | pHY300 + P43(noRBS) + gRNA(PilvB) + TamyL + LH + PbacA + RHilvB | This study |
| pGRNA-PbacA(ilvD) | pHY300 + P43(noRBS) + gRNA(PilvD) + TamyL + LH + PbacA + RHilvD | This study |
| pGRNA-PbacA(leuS) | pHY300 + P43(noRBS) + gRNA(PleuS) + TamyL + LH + PbacA + RHleuS | This study |
| pGRNA-PbacA(yvmC) | pHY300 + P43(noRBS) + gRNA(PyvmC) + TamyL + LH + PbacA + RHyvmC | This study |
| pGRNA-ΔbkdAB | pHY300 + P43(noRBS) + gRNA(bkdAB) + TamyL + LH + RHbkdAB | This study |
Ampr, ampicillin resistance; Tcr, tetracycline resistance.
TABLE 2.
Primers used for strain construction and qRT-PCR in this study
| Name and function | Sequencea (5′ to 3′) |
|---|---|
| PCR | |
| PilvB-sgRNA-F (EcoRI) | CGGAATTCGAAAAGCCTTTTCGCCCCTT GTTTTAGAGCTAGAAATAG |
| PilvB-sgRNA-Tamyl-R | CGATATAGCCCGGAGGCGTAATGATGACGGTCCAGC |
| ilvB-AF | GCTGGACCGTCATCATTACGCCTCCGGGCTATATCG |
| ilvB-PbacA-AR | ATTAGACCCGGGCTTGAATCTCACCACCCTTGTTCGCGGC |
| PbacA-F | GCCGCGAACAAGGGTGGTGAGATTCAAGCCCGGGTCTAAT |
| PbacA-R | CTGTACATTCGTCCCCATATAAAAATTCTCCTTTTTGATA |
| ilvB-PbacA-BF | TATCAAAAAGGAGAATTTTTATATGGGGACGAATGTACAG |
| ilvB-BR | GCTCTAGAAATGTCGTGGTCATAGCGG |
| PilvD-sgRNA-F (EcoRI) | CGGAATTCAGCAATCATATGGAATGTGGGTTTTAGAGCTAGAAATAG |
| PilvD-sgRNA-Tamyl-R | GTTTTCGGGTCATTGAGCTAATGATGACGGTCCAGC |
| ilvD-AF | GCTGGACCGTCATCATTAGCTCAATGACCCGAAAAC |
| ilvD-PbacA-AR | ATTAGACCCGGGCTTGAATCTCCAGCTTGTTTTTCAACAAT |
| PbacA-F | ATTGTTGAAAAACAAGCTGGAGATTCAAGCCCGGGTCTAAT |
| PbacA-R | ACTGCGTAAACCTGTCATATAAAAATTCTCCTTTTTGATA |
| ilvD-PbacA-BF | TATCAAAAAGGAGAATTTTTATATGACAGGTTTACGCAGT |
| ilvD-BR (XbaI) | GCTCTAGATCGAGTCCTTTCTCATCG |
| PleuS-sgRNA-F (EcoRI) | CGGAATTCGTACCATTGAAGCCGTACTCATATTCAAGTTTTAGAGCTAGAAATAG |
| PleuS-sgRNA-Tamyl-R | ATCACCTTCCTAAAGCCCTAATGATGACGGTCCAGC |
| leuS-AF | GCTGGACCGTCATCATTAGGGCTTTAGGAAGGTGAT |
| leuS-PbacA-AR | ATTAGACCCGGGCTTGAATCTCTTTCACCAGCCGTCAACT |
| PbacA-F | AGTTGACGGCTGGTGAAAGAGATTCAAGCCCGGGTCTAAT |
| PbacA-R | TTTGATGGTCAAAACTCAAATAAAAATTCTCCTTTTTGATA |
| leuS-PbacA-BF | TATCAAAAAGGAGAATTTTTATTTGAGTTTTGACCATCAAA |
| leuS-BR (XbaI) | GCTCTAGAGCTTCTAGAGCAAGAACGGTTCCAAGC |
| PyvmC-sgRNA-F (EcoRI) | CGGAATTCAAGTCATTCTATGTCCGTTAGTTTTAGAGCTAGAAATAG |
| PyvmC-sgRNA-Tamyl-R | CGCCGTCGCACGGGACTTTTAATGATGACGGTCCAGC |
| yvmC-AF | GCTGGACCGTCATCATTAAAAGTCCCGTGCGACGGCG |
| yvmC-PbacA-AR | ATTAGACCCGGGCTTGAATCTCGTCAACACAAATCAAATTC |
| PbacA-F | GAATTTGATTTGTGTTGACGAGATTCAAGCCCGGGTCTAAT |
| PbacA-R | TCTCCATTGTAAGCTCTGTCATATAAAAATTCTCCTTTTTGATA |
| yvmC-PbacA-BF | TATCAAAAAGGAGAATTTTTATATGACAGAGCTTACAATGGAGA |
| yvmC-BR (XbaI) | GCTCTAGAGTATTCAACAGCCAGCTCCA |
| bkdAB-sgRNA-KF (EcoRI) | CGGAATTCGCGGGTGATGGACACGCCTTGTTTTAGAGCTAGAAATAG |
| bkdAB-sgRNA-Tamyl-KR | TCCCGAATTCGGATCCTCTAATGATGACGGTCCAGC |
| bkdAB-KAF | GCTGGACCGTCATCATTAGAGGATCCGAATTCGGGA |
| bkdAB-KAR | CGCTATATTAAAACTCGGATAACTGGCATTGTTCTTC |
| bkdAB-KBF | GAAGAACAATGCCAGTTATCCGAGTTTTAATATAGCG |
| bkdAB-KBR (XbaI) | GCTCTAGACGCTCCCGCGGCATTTGT |
| PbacA-yvmA-F (BamHI) | CGGGATCCAGATTCAAGCCCGGGTCTAAT |
| PbacA-yvmA-R | TTAATTTCATTGGTCTCTAATAAAAATTCTCCTTTTTGATA |
| yvmA-F | TAAGAGAGGAATGTACACATGAGACCAATGAAATTAA |
| yvmA-R | ATCCGTCCTCTCTGCTCTTTTAAGACCTGGCAGCTTT |
| amyL-yvmA-F | AAAGCTGCCAGGTCTTAAAAGAGCAGAGAGGACGGAT |
| amyL-yvmA-R (XbaI) | GCTCTAGACGCAATAATGCCGTCG |
| ilvB-PbacA-YF | CGAAGGGTCTGCGGATAA |
| ilvD-PbacA-YF | ACAAATGGATCCTACTGA |
| leuS-PbacA-YF | TGCTCGATGGCCTGAAGT |
| yvmC-PbacA-YF | GGAAAGCAGCATTTGATAAG |
| bkdAB-KYF | TAATGCTGTCGGCGTTTG |
| bkdAB-KYR | CCACTTCCATCATCGTCCAG |
| pHY-F | CAGATTTCGTGATGCTTGTC |
| pHY-R | GTTTATTATCCATACCCTTAC |
| qPCR | |
| 16S rRNA-F | ACCTAACCAGAAAGCCACGG |
| 16S rRNA-R | GTTTACGGCGTGGACTACCA |
| ilvB-F | CGGTTATTTTAGCGGGTGCG |
| ilvB-R | GCATACGTTCCGTGCATTCC |
| ilvD-F | CGACGGACGAAAAATCTCGC |
| ilvD-R | TGTTTGGCAAATTCCCTGCG |
| leuS-F | CATATGCCGTTTTGGCTCCG |
| leuS-R | ACGGCATAAGCACCTGTGAA |
| yvmC-F | CATACCAGCGTTTGCGGATG |
| yvmC-R | AAAATATCAGGGGCGGCGAT |
| yvmA-F | GCGTCGTCTGTTTTGGCTTT |
| yvmA-R | TCATCGTAGAGAACAGCGGC |
Underlining indicates an overlap region for SOE-PCR; generated restriction sites are in boldface.
Construction of the promoter replacement strains.
The promoter of the bacitracin synthetase cluster PbacA has been proven to be a strong promoter in our previous research (33). Here, the promoter PbacA was employed to replace the original promoters of the ilvBHC-leuABCD operon, ilvD, the yvmC-cypX operon, and leuS to construct the gene overexpression strains using the CRISPR-Cas9n toolkit. The construction procedure for the ilvBHC-leuABCD cluster promoter replacement strain served as an example. Briefly, the ribosome-binding site (RBS)-free P43 promoter coupled with the single guide RNA (sgRNA) fragment, the PbacA promoter, as well as up- and downstream homology arms of PilvBHC-leuABCD (0.5 kb) were amplified with the corresponding primers (Table 2). These fragments then were fused by splicing overlap extension (SOE)-PCR and inserted into pHY300 by using the EcoRI and XbaI restriction sites, constructing the plasmid named pGRNA-PbacA (ilvBHC-leuABCD).
pGRNA-PbacA (ilvBHC-leuABCD) then was electrotransformed into DWc9n*, and the positive colonies were verified by diagnostic PCR and DNA sequencing. Colonies were cultivated in LB medium without tetracycline at 37°C for several generations to remove the remaining plasmids in positive colonies. The tetracycline-sensitive colonies then were verified with diagnostic PCR, and the DNA sequence was applied to confirm the ilvBHC-leuABCD promoter replacement strain (33). Similarly, the ilvD, yvmC-cypX operon, and leuS promoter replacement strains were obtained by the same method (see Fig. S4 in the supplemental material).
Construction of the gene deletion strain.
The bkdAB deletion strain of B. licheniformis was obtained by using the CRISPR-Cas9n toolkit (19). The procedure for the construction of the deletion strain is shown in Fig. S5 and was the same as that of the promoter replacement strain. Briefly, the P43 promoter and up- and downstream homology arms of bkdAB (0.5 kb) were amplified, and the gene deletion plasmid pGRNA-ΔbkdAB was constructed. pGRNA-ΔbkdAB then was transformed into B. licheniformis by electroporation, and the bkdAB deletion strain was verified by diagnostic PCR and DNA sequencing.
Construction of the yvmA overexpression strains.
The gene expression vector was constructed according to the previously reported research (34). Briefly, the PbacA promoter, the yvmA gene, and the amyL terminator from B. licheniformis DW2 were amplified and fused by SOE-PCR. The fused fragment was inserted into the pHY300 vector at the restriction sites BamHI and XbaI. Diagnostic PCR and DNA sequencing confirmed that the YvmA expression vector, named pHY-yvmA, was constructed successfully. pHY-yvmA was electrotransferred into B. licheniformis strains to construct YvmA overexpression strains. In addition, pHY300 was transformed into B. licheniformis to serve as the control strain.
Determinations of pulcherriminic acid concentration and cell density.
The concentration of pulcherriminic acid was measured according to our previously reported research (25). The purification and standard curve method was performed on pulcherriminic acid according to the standard protocols (25). Briefly, the volume of 1 ml culture broth was evenly mixed with 1 ml 2 M NaOH solution to dissolve pulcherriminic acid and centrifuged at 10,000 × g for 2 min, and the absorbance of the supernatant was determined at 410 nm (including intracellular and extracellular pulcherriminic acid). The concentration of pulcherriminic acid was calculated via a standard curve. Meanwhile, the cell density was measured by determining the optical density at 600 nm (OD600).
To detect the intracellular pulcherriminic acid, 1 ml cells was harvested by centrifugation at 10,000 × g for 5 min. The cell pellet then was washed with distilled water to remove the extracellular pulcherrimin. Afterwards, the cells were disrupted by sonication on ice for 20 min. Cell debris was mixed with 1 ml 2 M NaOH solution to dissolve pulcherriminic acid centrifuged at 10,000 × g for 2 min, and the absorbance of the supernatant was determined at 410 nm.
Quantitative real-time PCR.
B. licheniformis cells were collected at 12 h for RNA extraction according to our previous report (18). The RNA concentration was measured on a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). The first-strand cDNA was synthesized from 50 ng total RNA using RevertAid first-strand cDNA synthesis kits (Thermo Scientific). The resulting cDNA was used as the template for qRT-PCR with primers listed in Table 2 using the iTaqTM universal SYBR green supermix (Bio-Rad, USA). The 16S rRNA gene served as the reference gene, and all assays were performed in triplicate.
Determination of the intracellular concentrations of amino acids.
The intracellular precursor amino acids were measured using gas chromatography (GC) according to our previously reported method (14). Concentrations were calculated via standard curves made with the corresponding amino acids.
Supporting information.
The pulcherriminic acid yield of W2 and W2(HD) is shown in Fig. S1. The intracellular Leu concentrations of W2 and W2(HD) at 24 h are shown in Fig. S2. The concentrations of intracellular Phe among those recombinant strains at 24 h are provided in Fig. S3. The procedure for the construction of the promoter replacement strain is shown in Fig. S4. The procedure for the construction of the deletion strain is shown in Fig. S5.
Supplementary Material
ACKNOWLEDGMENTS
This work was supported by the National Key Research and Development Program of China (2018YFA0900300) and the Technical Innovation Special Fund of Hubei Province (2018ACA149).
We have no competing interests to declare.
D. Wang and S. Chen designed the study. S. Wang, D. Wang, X. Li, and D. Zhang carried out the molecular biology studies and constructed engineering strains. S. Wang, H. Wang, D. Zhang, and J. Zhu carried out the fermentation studies. S. Wang, D. Wang, Y. Zhan, D. Cai, Q. Wang, X. Ma, and S. Chen analyzed the data and wrote the manuscript. All authors read and approved the final manuscript.
Footnotes
Supplemental material is available online only.
REFERENCES
- 1.Belin P, Moutiez M, Lautru S, Seguin J, Pernodet JL, Gondry M. 2012. The nonribosomal synthesis of diketopiperazines in tRNA-dependent cyclodipeptide synthase pathways. Nat Prod Rep 29:961–979. doi: 10.1039/c2np20010d. [DOI] [PubMed] [Google Scholar]
- 2.Lautru S, Gondry M, Genet R, Pernodet JL. 2002. The albonoursin gene cluster of S. noursei biosynthesis of diketopiperazine metabolites independent of nonribosomal peptide synthetases. Chem Biol 9:1355–1364. doi: 10.1016/s1074-5521(02)00285-5. [DOI] [PubMed] [Google Scholar]
- 3.Mishra AK, Choi J, Choi SJ, Baek KH. 2017. Cyclodipeptides: an overview of their biosynthesis and biological activity. Molecules 22:1796. doi: 10.3390/molecules22101796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kamal R, Gusain YS, Kumar V, Sharma AK. 2015. Disease management through biological control agents: an eco-friendly and cost effective approach for sustainable agriculture–a review. Agric Rev 36:37. doi: 10.5958/0976-0741.2015.00004.5. [DOI] [Google Scholar]
- 5.Cryle MJ, Bell SG, Schlichting I. 2010. Structural and biochemical characterization of the cytochrome P450 CypX (CYP134A1) from Bacillus subtilis: a cyclo-L-leucyl-L-leucyl dipeptide oxidase. Biochemistry 49:7282–7296. doi: 10.1021/bi100910y. [DOI] [PubMed] [Google Scholar]
- 6.Sipiczki M. 2006. Metschnikowia strains isolated from botrytized grapes antagonize fungal and bacterial growth by iron depletion. Appl Environ Microbiol 72:6716–6724. doi: 10.1128/AEM.01275-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Uffen RL, Canale-Parola E. 1972. Synthesis of pulcherriminic acid by Bacillus subtilis. J Bacteriol 111:86–93. doi: 10.1128/JB.111.1.86-93.1972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gore-Lloyd D, Sumann I, Brachmann AO, Schneeberger K, Ortiz-Merino RA, Moreno-Beltran M, Schlafli M, Kirner P, Santos Kron A, Rueda-Mejia MP, Somerville V, Wolfe KH, Piel J, Ahrens CH, Henk D, Freimoser FM. 2019. Snf2 controls pulcherriminic acid biosynthesis and antifungal activity of the biocontrol yeast Metschnikowia pulcherrima. Mol Microbiol 112:317–332. doi: 10.1111/mmi.14272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kántor A, Hutková J, Petrová J, Hleba L, Kačániová M. 2016. Antimicrobial activity of pulcherrimin pigment produced by Metschnikowia pulcherrima against various yeast species. J Appl Microbiol 5:282–285. doi: 10.15414/jmbfs.2015/16.5.3.282-285. [DOI] [Google Scholar]
- 10.Kurtzman PC, Droby S. 2001. Metschnikowia fructicola, a new ascosporic yeast with potential for biocontrol of postharvest fruit rots. Syst Appl Microbiol 24:395–399. doi: 10.1078/0723-2020-00045. [DOI] [PubMed] [Google Scholar]
- 11.Saravana D, Ciavorella A, Spadaro D, Garibaldi A, Gullino M. 2008. Metschnikowia pulcherrima strain MACH1 outcompetes Botrytis cinerea, Alternaria alternata and Penicillium expansum in apples through iron depletion. Postharvest Biol Tec 49:121–128. doi: 10.1016/j.postharvbio.2007.11.006. [DOI] [Google Scholar]
- 12.Talibi I, Boubaker H, Boudyach EH, Ait Ben Aoumar A. 2014. Alternative methods for the control of postharvest citrus diseases. J Appl Microbiol 117:1–17. doi: 10.1111/jam.12495. [DOI] [PubMed] [Google Scholar]
- 13.Bonnefond L, Arai T, Sakaguchi Y, Suzuki T, Ishitani R, Nureki O. 2011. Structural basis for nonribosomal peptide synthesis by an aminoacyl-tRNA synthetase paralog. Proc Natl Acad Sci U S A 108:8. doi: 10.1073/pnas.1019480108/-/DC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Liu Z, Yu W, Nomura CT, Li J, Chen S, Yang Y, Wang Q. 2018. Increased flux through the TCA cycle enhances bacitracin production by Bacillus licheniformis DW2. Appl Microbiol Biotechnol 102:6935–6946. doi: 10.1007/s00253-018-9133-z. [DOI] [PubMed] [Google Scholar]
- 15.Sauguet L, Moutiez M, Li Y, Belin P, Seguin J, Le Du MH, Thai R, Masson C, Fonvielle M, Pernodet JL, Charbonnier JB, Gondry M. 2011. Cyclodipeptide synthases, a family of class-I aminoacyl-tRNA synthetase-like enzymes involved in non-ribosomal peptide synthesis. Nucleic Acids Res 39:4475–4489. doi: 10.1093/nar/gkr027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Randazzo P, Aubert-Frambourg A, Guillot A, Auger S. 2016. The MarR-like protein PchR (YvmB) regulates expression of genes involved in pulcherriminic acid biosynthesis and in the initiation of sporulation in Bacillus subtilis. BMC Microbiol 16:190. doi: 10.1186/s12866-016-0807-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wang D, Zhan Y, Cai D, Li X, Wang Q, Chen S. 2018. Regulation of the synthesis and secretion of the iron chelator cyclodipeptide pulcherriminic acid in Bacillus licheniformis. Appl Environ Microbiol 84:e00262-18. doi: 10.1128/AEM.00262-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cai D, Zhu J, Zhu S, Lu Y, Zhang B, Lu K, Li J, Ma X, Chen S. 2019. Metabolic engineering of main transcription factors in carbon, nitrogen, and phosphorus metabolisms for enhanced production of bacitracin in Bacillus licheniformis. ACS Synth Biol 8:866–875. doi: 10.1021/acssynbio.9b00005. [DOI] [PubMed] [Google Scholar]
- 19.Li K, Cai D, Wang Z, He Z, Chen S. 2018. Development of an efficient genome editing tool in Bacillus licheniformis using CRISPR-Cas9 nickase. Appl Environ Microbiol 84:e02608-17. doi: 10.1128/AEM.02608-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lincecum TL Jr, Tukalo M, Yaremchuk A, Mursinna RS, Williams AM, Sproat BS, Van Den Eynde W, Link A, Van Calenbergh S, Grotli M, Martinis SA, Cusack S. 2003. Structural and mechanistic basis of pre- and posttransfer editing by leucyl-tRNA synthetase. Mol Cell 11:12. doi: 10.1016/s1097-2765(03)00098-4. [DOI] [PubMed] [Google Scholar]
- 21.Bentley GJ, Jiang W, Guaman LP, Xiao Y, Zhang F. 2016. Engineering Escherichia coli to produce branched-chain fatty acids in high percentages. Metab Eng 38:148–158. doi: 10.1016/j.ymben.2016.07.003. [DOI] [PubMed] [Google Scholar]
- 22.Wu Q, Zhi Y, Xu Y. 2019. Systematically engineering the biosynthesis of a green biosurfactant surfactin by Bacillus subtilis 168. Metab Eng 52:87–97. doi: 10.1016/j.ymben.2018.11.004. [DOI] [PubMed] [Google Scholar]
- 23.Gondry M, Sauguet L, Belin P, Thai R, Amouroux R, Tellier C, Tuphile K, Jacquet M, Braud S, Courcon M, Masson C, Dubois S, Lautru S, Lecoq A, Hashimoto S, Genet R, Pernodet JL. 2009. Cyclodipeptide synthases are a family of tRNA-dependent peptide bond-forming enzymes. Nat Chem Biol 5:414–420. doi: 10.1038/nchembio.175. [DOI] [PubMed] [Google Scholar]
- 24.Biver S, Steels S, Portetelle D, Vandenbol M. 2013. Bacillus subtilis as a tool for screening soil metagenomic libraries for antimicrobial activities. J Microbiol Biotechnol 23:850–855. doi: 10.4014/jmb.1212.12008. [DOI] [PubMed] [Google Scholar]
- 25.Li X, Wang D, Cai D, Zhan Y, Wang Q, Chen S. 2017. Identification and high-level production of pulcherrimin in Bacillus licheniformis DW2. Appl Biochem Biotechnol 183:1323–1335. doi: 10.1007/s12010-017-2500-x. [DOI] [PubMed] [Google Scholar]
- 26.Lynch S, Eckert C, Yu J, Gill R, Maness PC. 2016. Overcoming substrate limitations for improved production of ethylene in E. coli. Biotechnol Biofuels 9:3. doi: 10.1186/s13068-015-0413-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Zhu J, Cai D, Xu H, Liu Z, Zhang B, Wu F, Li J, Chen S. 2018. Enhancement of precursor amino acid supplies for improving bacitracin production by activation of branched chain amino acid transporter BrnQ and deletion of its regulator gene lrp in Bacillus licheniformis. Synth Syst Biotechnol 3:236–243. doi: 10.1016/j.synbio.2018.10.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Vogt M, Haas S, Klaffl S, Polen T, Eggeling L, van Ooyen J, Bott M. 2014. Pushing product formation to its limit: metabolic engineering of Corynebacterium glutamicum for L-leucine overproduction. Metab Eng 22:40–52. doi: 10.1016/j.ymben.2013.12.001. [DOI] [PubMed] [Google Scholar]
- 29.Xia XX, Qian ZG, Ki CS, Park YH, Kaplan DL, Lee SY. 2010. Native-sized recombinant spider silk protein produced in metabolically engineered Escherichia coli results in a strong fiber. Proc Natl Acad Sci U S A 107:14059–14063. doi: 10.1073/pnas.1003366107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Jacques IB, Moutiez M, Witwinowski J, Darbon E, Martel C, Seguin J, Favry E, Thai R, Lecoq A, Dubois S, Pernodet JL, Gondry M, Belin P. 2015. Analysis of 51 cyclodipeptide synthases reveals the basis for substrate specificity. Nat Chem Biol 11:721–727. doi: 10.1038/nchembio.1868. [DOI] [PubMed] [Google Scholar]
- 31.Vetting MW, Hegde SS, Blanchard JS. 2010. The structure and mechanism of the Mycobacterium tuberculosis cyclodityrosine synthetase. Nat Chem Biol 6:797–799. doi: 10.1038/nchembio.440. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Florence H, Christine OD, Frédérique VG, Sandra C, Sandrine L, Dominique E, William N, Sylvie R. 2008. PecS is a global regulator of the symptomatic phase in the phytopathogenic bacterium Erwinia chrysanthemi 3937. J Bacteriol 190:7508. doi: 10.1128/JB.00553-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Shi J, Zhan Y, Zhou M, He M, Wang Q, Li X, Wen Z, Chen S. 2019. High-level production of short branched-chain fatty acids from waste materials by genetically modified Bacillus licheniformis. Bioresour Technol 271:325–331. doi: 10.1016/j.biortech.2018.08.134. [DOI] [PubMed] [Google Scholar]
- 34.Cai D, Chen Y, He P, Wang S, Mo F, Li X, Wang Q, Nomura CT, Wen Z, Ma X, Chen S. 2018. Enhanced production of poly-gamma-glutamic acid by improving ATP supply in metabolically engineered Bacillus licheniformis. Biotechnol Bioeng 115:2541–2553. doi: 10.1002/bit.26774. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.