Design of 1-OH-PHZ biosynthetic pathway in P. chlororaphis H18
We have isolated a strain P. chlororaphis H18 that can secrete reddish-brown pigments from the plant rhizosphere soil, and its colorful metabolites were purified and their structures elucidated to be PCA (1), 2-OH-PHZ (2), and 2-hydroxy-phenazine-1-carboxylic acid (3). Using genome-mining and sequence alignment, their biosynthetic gene cluster was confirmed (Fig. 1A). Since the basic biosynthetic pathway of phenazines has already been clearly resolved, proteins PhzABCDEFGO catalyze the shikimate pathway end product chorismate to form the phenazine common precursor PCA (Fig. 1B). Then, a monooxygenase PhzO converts PCA into 2-OH-PHZ, and 2-hydroxy-phenazine-1-carboxylic acid is the intermediate of this process . A previous study showed that FAD-dependent monooxygenase PhzS of P. aeruginosa PAO1 can catalyze the oxidative decarboxylation of PCA to generate 1-OH-PHZ . So we planned to integrate this phzS gene into P. chlororaphis H18 genome to construct 1-OH-PHZ biosynthetic pathway.
Biosynthesis and identification of 1-OH-PHZ
When cultured in KB medium for 48 h, P. chlororaphis H18 produces 86.6 mg/L PCA, 71.8 mg/L 2-OH-PHZ, and 14.8 mg/L 2-hydroxy-phenazine-1-carboxylic acid. The formation of 2-OH-PHZ and 2-hydroxy-phenazine-1-carboxylic acid catalyzed by PhzO consumed a significant amount of PhzS substrate PCA. Therefore, we synthesized gene phzS from P. aeruginosa PAO1 and replaced phzO with homologous recombination to obtain mutant PH18-1 (Additional file 1: Figs. S3 and S4). In the PH18-1 fermentation broth, 72.3 mg/L of 1-OH-PHZ (4) generation was detected and the production of 2-OH-PHZ (2) and 2-hydroxy-phenazine-1-carboxylic acid (3) was abolished (Fig. 2A, B). 1-OH-PHZ shows outstanding biological activities against crop pathogens, for instance, Bipolaris maydis, Alternaria solani, and Fusarium graminearum (Additional file 1: Table S5).
Semi-rational design and remodeling to improve the catalytic activity of PhzS
The concentration of 1-OH-PHZ in PH18-1 fermentation broth is relatively low (72.3 mg/L), which limited its industrial application. To address this issue, the promotion of the 1-OH-PHZ conversion rate is an essential job. We know that enzymes are the most critical factors in the biosynthesis process of natural products, especially those rate-limiting step enzymes. In the 1-OH-PHZ biosynthetic pathway, the flavin-dependent hydroxylase PhzS catalyzes the crucial step of decarboxylation and hydroxylation at PCA 1-position. We believe that optimizing PhzS will be an effective way to improve pathway efficiency. Semi-rational design and molecular remodeling are feasible methods for promoting protein catalytic performance. First, PCA was docked into the crystal structure of PhzS (PDB: 2RGJ) as a ligand to predict the PCA-binding hydrophobic pocket or activity site dominated by N48, L76, M205, V207, R215, V217, Y219, M238, P317, M318, G319  (Fig. 4A). To facilitate the detection of PhzS mutants’ activity, we expressed PhzS protein in E. coli BL21 Rosetta (DE3) and verified its activity by whole-cell catalysis (Fig. 3). Then, alanine scanning was employed to identify the function of these residues by detecting mutant enzyme activities on PCA in vitro. N48A, L76A, Y219A, and M238A mutants completely lost their catalytic abilities, only M318A presented a 24% activity increase. Next, amino acids M318 and V217 were subjected to saturation mutagenesis, M318T, M318I, and V217I resulted in 40%, 5%, and 35% higher activity than the wild type enzyme (Additional file 1: Fig. S1). More site-directed mutants in M205, V207, R215, P317, and G319 were constructed using residues I, S, T, and V respectively, among which better activity (1.4-fold of the WT strain) was observed in PhzSV207I (Additional file 1: Fig. S2). Afterward, we used M318A, M318T, V217I, and V207I to construct double and triple mutants. In these mutants, the resulting PhzSM318A/V207I achieved the highest reactivity (an increase of 70% compared with PhzS) to catalyze PCA (Fig. 4B). These mutations are likely to remodel PhzS conformation and consequently influence the substrate-binding pocket, which enhances PhzS catalytic activity. When protein PhzSM318A/V207I was introduced into P. chlororaphis H18, the obtained strain PH18-1–2 presented 151.8 mg/L 1-OH-PHZ concentration, which is 1.1 times higher than PH18-1 (Fig. 4C). This indicates that the engineering of PhzS is an effective way to increase the fermentation yield of 1-OH-PHZ.
The PH18-1–2 fermentation result showed that there is still a certain amount of precursor PCA in the fermentation broth. Afterward, we try to enhance the conversion of PCA to 1-OH-PHZ by adding PhzS copy numbers. When we inserted two copies of the PhzS M318A/V207I gene to replace PhzS in PH18-1, the constructed strain PH18-1–3 produced 202.4 mg/L 1-OH-PHZ and more PCA is consumed.
Disruption of negative regulatory genes and enrichment of precursor supply to enhance 1-OH-PHZ production
In order to improve 1-OH-PHZ yield, we first chose to increase the supply of phenazine substrate pool in P. chlororaphis H18. Chorismate, a shikimate pathway end product is the key phenazine derivative precursor. Several enzymes such as phosphoenolpyruvate (PEP) synthetase PpsA, transketolase TktA, quinate/shikimate dehydrogenase AroE, and dehydroquinic acid synthase AroB have been studied to enhance the shikimate pathway and chorismate [9, 24]. Through sequence alignment, we found that, PpsA, TktA, AroB, and AroE homologous proteins genes all exist in the in P. chlororaphis H18 genome. Hence, we planned to introduce one more copy of ppsA, tktA, aroB, and aroE genes into P. chlororaphis H18. Besides, RpeA, RsmE, PsrA, and Lon have been reported as negative regulators in Pseudomonas chlororaphis strains [29,30,31,32]. The disruption of these regulatory genes has been proven to effectively improve phenazines titer. Thus, we cloned aroE, tktA, aroB, and ppsA genes from P. chlororaphis H18 to replace the possible negative regulatory genes repA, rsmE, psrA, and lon to obtain strains PH18-1-4, PH18-1-5, PH18-1-6, PH18-1-7 respectively (Fig. 5A). According to the HPLC detection results, the 1-OH-PHZ concentration in the constructed final strain PH18-1-7 was 361.4 mg/L, a fourfold increase over PH18-1 when fermented in KB medium.
Further improvement of 1-OH-PHZ production
Research suggested that GacA is a positive global regulator for phenazine biosynthesis in Pseudomonas chlororaphis 30–84 , but negatively regulated phenazines production in Pseudomonas sp. M18 and Pseudomonas chlororaphis GP72 . When we knocked out the gacA gene from P. chlororaphis H18, the phenazine generation was almost abolished, which means GacA plays a role in positive regulation in P. chlororaphis H18 on PCA synthesis (Fig. 5B). Afterward, an additional copy of gacA was introduced into PH18-1–7 to replace the secondary metabolite pyrrolnitrin biosynthetic gene prnA. The resultant mutant PH18-1–8 had a 20% increase in 1-OH-PHZ titer (Fig. 5A and Additional file 1: Fig. S4), which is fivefold more than mutant PH18-1 (Fig. 6A).
To further optimize 1-OH-PHZ titer, we tried different kinds of fermentation medium, for instance, KB, PPM, PCM, PPN, and SCM. The HPLC detection data illustrated that PPM is the best for 1-OH-PHZ production, in which the 1-OH-PHZ level is 70% higher than in KB broth (Fig. 6B).
Moreover, shikimic acid is a chief precursor of chorismate in the shikimate pathway, and all known phenazines are converted from chorismate. Hence, we suppose that directly increasing the supply of upstream intermediate or precursor substrate may affect the enhancement of 1-OH-PHZ production. Given the high price of chorismate on the market, the economic shikimic acid is suitable for feeding. When 100 mg/L, 200 mg/L, 400 mg/L, or 800 mg/L of shikimic acid was fed into 50 mL PH18-1–8 strain fermentation broth respectively in 12 h. HPLC detection demonstrates that the addition of shikimic acid can indeed improve the 1-OH-PHZ level (Fig. 6C). Among them, 200 mg/L shikimic acid feeding showed the best effect (20% increase over no feeding). Although higher concentrations of shikimic acid can also increase 1-OH-PHZ titer, the burden of metabolism or feedback inhibition of the synthetic pathway probably influences the yield. Otherwise, engineering feedback resistance of DHAP synthase is another effective approach to improve the yield of shikimate derived products, which could be used in our future work.
Optimization of fed-batch 1-OH-PHZ fermentation
To combine all the above engineering modifications, mutant strain PH18-1-8 (Fig. 5A and Additional file 1: Fig. S4) could generate 938.5 mg/L of 1-OH-PHZ in 50 mL shake flasks. Further improvement of the 1-OH-PHZ production level was carried out by fed-batch fermentation. Strain PH18-1-8 was first employed in a 5 L fermenter containing 2 L PPM liquid medium. During the fermentation process, glucose was also added to meet the demand of the carbon source. Production of 1-OH-PHZ continued to increase until 54 h, and the highest production was 3.6 g/L. The time-course analysis of glucose consumption, OD value, and 1-OH-PHZ concentration during fermentation are illustrated in Fig. 7, respectively.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.