The Gene paaZ of the Phenylacetic Acid (PAA) Catabolic Pathway Branching Point and ech outside the PAA Catabolon Gene Cluster Are Synergistically Involved in the Biosynthesis of the Iron Scavenger 7-Hydroxytropolone in Pseudomonas donghuensis HYS

The newly discovered iron scavenger 7-hydroxytropolone (7-HT) is secreted by Pseudomonas donghuensis HYS. In addition to possessing an iron-chelating ability, 7-HT has various other biological activities. However, 7-HT’s biosynthetic pathway remains unclear. This study was the first to report that the phenylacetic acid (PAA) catabolon genes in cluster 2 are involved in the biosynthesis of 7-HT and that two genes, paaZ (orf13) and ech, are synergistically involved in the biosynthesis of 7-HT in P. donghuensis HYS. Firstly, gene knockout and a sole carbon experiment indicated that the genes orf17–21 (paaEDCBA) and orf26 (paaG) were involved in the biosynthesis of 7-HT and participated in the PAA catabolon pathway in P. donghuensis HYS; these genes were arranged in gene cluster 2 in P. donghuensis HYS. Interestingly, ORF13 was a homologous protein of PaaZ, but orf13 (paaZ) was not essential for the biosynthesis of 7-HT in P. donghuensis HYS. A genome-wide BLASTP search, including gene knockout, complemented assays, and site mutation, showed that the gene ech homologous to the ECH domain of orf13 (paaZ) is essential for the biosynthesis of 7-HT. Three key conserved residues of ech (Asp39, His44, and Gly62) were identified in P. donghuensis HYS. Furthermore, orf13 (paaZ) could not complement the role of ech in the production of 7-HT, and the single carbon experiment indicated that paaZ mainly participates in PAA catabolism. Overall, this study reveals a natural association between PAA catabolon and the biosynthesis of 7-HT in P. donghuensis HYS. These two genes have a synergistic effect and different functions: paaZ is mainly involved in the degradation of PAA, while ech is mainly related to the biosynthesis of 7-HT in P. donghuensis HYS. These findings complement our understanding of the mechanism of the biosynthesis of 7-HT in the genus Pseudomonas.


Introduction
Pseudomonas is a genus of ubiquitous Gram-negative γ-proteobacterium, the members of which are known for their capacity to colonize various ecological niches; in addition, they are highly versatile and adaptable [1][2][3].This adaptability is considered to be associated with their diverse and highly elaborate siderophore systems [4][5][6][7].Iron is an essential trace element for virtually all living organisms to maintain a normal life as it is involved in several key metabolic processes [8][9][10][11].However, owing to its low solubility under physiological pH and aerobic conditions, iron is poorly bioavailable in the environment [12][13][14][15].To survive in an iron-deficient environment, Pseudomonas can produce a variety of siderophores [16,17].P. donghuensis HYS, isolated from Donghu Lake, was the type of strain of this newly classified species [18,19] and exhibited a high iron-chelating capacity [20][21][22].P. donghuensis HYS secretes two types of iron scavengers: the fluorescent pyoverdine and a large quantity of nonfluorescent iron scavenger 7-hydroxytropolone (7-HT), which is a symmetrical seven-membered heteroatomic carbon ring containing a carbonyl group and two hydroxyl groups; 7-HT contributes to the notable iron-chelating ability in the culture supernatant of P. donghuensis HYS [20][21][22].P. donghuensis HYS was more virulent to C. elegans than P. aeruginosa [23][24][25].Other P. donghuensis strains also exhibited broad antibacterial activity, antifungal activity, and plant growth-promoting (PGP) activities [26][27][28][29], which are closely related to the 7-HT produced by these strains.
Siderophores are low-molecular-mass compounds (200-2000 Da) synthesized and secreted by bacteria to respond to iron deficiency.Siderophores possess a high affinity for iron (III) and can chelate the iron from different environments to ensure their survival [8, 15,30].The biosynthesis of siderophores is regulated by extracellular iron concentrations [31,32].Among the variety of siderophores of Pseudomonads, pyoverdines (PVDs) are the primary siderophore and have been extensively studied due to their high affinity for Fe (III) and because they are widespread in Pseudomonas [33,34].Further, Pseudomonas often produces a wide variety of secondary siderophores of lower affinity, such as pyochelin, pseudomonine, quinolobactin, and thioquinolobactin [16,[35][36][37][38].In addition to iron-chelating characteristics, secondary siderophores are often endowed with interesting characteristics, such as forming complexes with other metals, pathogenicity, antimicrobial activity, and biocontrol, which could well contribute to the competitiveness of these bacteria and better reflect the high adaptability of Pseudomonads to diverse environments and niche colonization [5,7,37,39].The newly discovered nonfluorescent iron scavenger 7-hydroxytropolone (7-HT) is secreted in large quantities by the P. donghuensis HYS strain [20,22].In addition, 7-HT is the main metabolite antagonistic to phytopathogenic fungi and plant-probiotic properties in other P. donghuensis strains [26,40].
As identified in our previous work [20,22], to define the essential genes involved in the biosynthesis of the novel iron scavenger 7-HT, a random transposon insertion mutation was carried out in P. donghuensis HYS.The mutants that are deficient in siderophore yields were obtained by screening.The mutant genes were mainly located in two gene clusters, referred to as cluster 1 and cluster 2. Cluster 1 was composed of 12 genes related to the production of 7-HT.Among these genes was an nfs cluster (orf6-orf9) composed of four synthetases that participated in the synthesis of 7-HT.The deletion of these four genes eliminated 7-HT production.Furthermore, the production of 7-HT and the expression of orf6-9 in relation to 7-HT biosynthesis was regulated by extracellular iron concentrations [20,22].
In P. donghuensis HYS, the biosynthetic pathway of 7-HT remains unclear, preventing its function from being fully applied and characterized.Thus, in this study, we focused on investigating the biosynthetic pathway of 7-HT.Whether these screened PAA genes orf17-orf21 (paaEDCBA) and orf26 (paaG) arranged in gene cluster 2 are actually involved in 7-HT biosynthesis, and the role of paaZ in the biosynthesis of 7-HT remain unclear.These questions urgently needs to be answered.Revealing the roles of PAA genes in 7-HT biosynthesis will not only provide further insight into the biosynthetic pathways of 7-HT, but will also more generally, improve understanding of the biosynthetic mechanisms for troplones and siderophores in the genus Pseudomonas.

The Genes orf17-orf21 and orf26 in Cluster 2 Are Essential for the Production of the Iron Scavenger 7-HT
As shown in Table S1, the transposon mutants corresponding to orf17-21 and orf26 produced significantly decreased total amounts of siderophores.The five genes related to the production of siderophores (orf17, orf19-orf21, and orf26) were identified in cluster 2 of P. donghuensis HYS (Table S1 and Figure 1A).Among these genes, orf13-orf21 constituted a transcription operon.There was a cotranscriptional relationship between orf26 and orf27 (Figure S1).
To determine whether orf17-21 and orf26 affect the production of the iron scavenger 7-HT, we constructed an in-frame homologous deletion of mutants ∆orf17, ∆orf18, ∆orf19, ∆orf20, ∆orf21, ∆orf26, and ∆orf17-21, respectively.Each knockout strain showed markedly reduced siderophore production, both on CAS agar plates and in a liquid MKB medium, compared to that in the wild-type strain HYS (Figure 1B-D).On CAS agar plates, the yellow chelated halos of deleted mutants were much smaller than the WT strain HYS (Figure 1B), and the chelating ability of the liquid MkB culture supernatant of each knockout strain to the CAS detection solution was also remarkably lower (approximately one-sixth to seven times) than that of the WT strain HYS (Figure 1C).Moreover, the two characteristic absorption peaks of 7-HT, at 330 nm and 392 nm, disappeared, indicating a deficiency in the 7-HT yield.By contrast, only the characteristic absorption peak of pyoverdine, at 405 nm, could be detected in these mutants (Figure 1D).The results showed that the mutants no longer produced 7-HT.These results proved that the genes containing orf17 to orf21 and orf26 in cluster 2 were indeed essential for the biosynthesis of 7-HT.
To determine whether the expression levels of these genes are inhibited by high extracellular iron concentrations, the transcriptional levels of these genes were detected by RTq-PCR during the exponential phase in P. donghuensis HYS inoculated in a liquid MKB medium and in MKB supplemented with 30 µM Fe 2+ , respectively.RTq-PCR results indicated that the transcriptional levels of the related genes orf17-orf21 and orf26 were significantly decreased in the high iron concentration-MKB medium (supplemented with 30 µM of FeSO 4 ); these levels decreased by about 10-fold compared with the limited iron concentration-MKB medium (not supplemented with FeSO 4 ), respectively (Figure 1E).It was confirmed that the expression levels of genes orf17-orf21 and orf26 were inhibited by high concentrations of iron.These results further indicated the production of HYS controls 7-HT by regulating the expression of orf17-orf21and orf26 under different iron conditions and that these genes were also involved in the production of 7-HT.
Subsequently, a BLASTP search was performed with the protein sequences of these genes in NCBI.The results showed that these genes were related to the metabolism of PAA and shared high homologies to PAA cluster genes of E. coli K-12 (43-70%) (Table S2).When cluster 2 of P. donghuensis HYS was aligned with the PAA cluster of E. coli K-12, we found that they were highly similar and showed that the orf17, orf19-orf21, and orf26 genes, particularly, were homologous to paaE, paaCBA, and paaG PAA catabolon genes (Table S2).From the above comparison results, it was preliminarily speculated that cluster 2 might be a PAA metabolic gene cluster.
The sole carbon utilization experiment showed that the orf13, orf17-21, and orf26 genes in cluster 2 were involved in the catabolism of PAA in P. donghuensis HYS (Figure S2) and preliminary indicated that the PAA degradation pathway could be related to the biosynthesis of 7-HT in P. donghuensis HYS.
In brief, these results indicated that the orf17-21 (paaEDCBA) and orf26 (paaG) genes in cluster 2 were homologous to PAA metabolism genes and involved in the PAA catabolism in P. donghuensis HYS.Moreover, they were essential for the production of the iron scavenger 7-HT in P. donghuensis HYS.(B) Siderophore production in wild-type HYS and the ∆orf13, ∆orf17, ∆orf18, ∆orf19, ∆orf20, ∆orf21, ∆orf17-21, and ∆orf26 mutants were tested on CAS agar plates under normal light (upper) or UV light (lower).(C) Siderophore production in wild-type HYS and the derivative mutant strains in liquid MKB medium was determined as siderophore units (percent) using the CAS liquid assay.(D) Absorption spectra of the supernatants of 24 h liquid MKB cultures from wild-type HYS and derivative strains.(E) The relative transcription level of orf13, orf17-orf21, and orf26 in P. donghuensis HYS in high and low extracellular iron concentrations.RNA was isolated from the indicated strains grown at an exponential phase at 30 • C in a liquid MKB culture supplemented with or without 30 µM FeSO 4 •7H 2 O.Each value is the average from three different cultures ± the standard deviation.The error bars indicate standard deviations from 3 replicates (n = 3).** p < 0.01, *** p < 0.001, Student's t-test.
ORF13 in cluster 2 of P. donghuensis HYS was homologous to PaaZ (Table S2).It is unclear whether orf13 (paaZ) is involved in the biosynthesis of 7-HT in P. donghuensis HYS.
2.2.ORF13, a Homologous Protein of PaaZ, Is Not Necessary for the Biosynthesis of 7-HT in P. donghuensis HYS To investigate whether orf13 (paaZ) also participated in the biosynthesis of 7-HT, the orf13 gene was deleted in HYS.Then, the ability of the ∆orf13 strain to produce siderophores, especially the iron scavenger 7-HT, was determined.
On CAS agar plates (under normal or UV light), the yellow chelated halo of the ∆orf13 strain was not significantly different from that of the WT HYS strain (Figure 1B).This phenomenon preliminarily indicated that the deletion of orf13 had no obvious effect on the total amount of siderophores.The yield of siderophores on the ∆orf13 strain in the liquid MKB medium supernatant was reduced only by about one-fifth to one-sixth compared to that of the wild-type HYS strain (Figure 1C).Similarly, the characteristic absorption peaks of 7-HT in the ∆orf13 strain, at 330 nm and 392 nm, were only slightly decreased compared with the WT strain HYS (Figure 1D).The deletion of orf13 could not completely eliminate the synthesis of 7-HT, but interestingly, the RTq-PCR result indicated that the transcriptional level of orf13 was inhibited by high-concentration iron (Figure 1E).
The results showed that orf13 had no significant effect on the yield of fluorescent pyoverdine and nonfluorescent iron scavenger 7-HT in HYS.The sole carbon experiment showed that orf13 was involved in the catabolism of PAA in P. donghuensis HYS (Figure S2).
Through amino acid sequence alignment, it was found that ORF13 (PaaZ) also consisted of two domains, the N-terminal PaaZ-ALDH and the C-terminal PaaZ-ECH domain, and that the active site of the ALDH domain of PaaZ in P. donghuensis HYS was 258 Glu (E).To inactivate the ALDH domain, the site was mutated to Gln (Q) (E258Q).To further investigate the relationship between orf13 (paaZ), and the production of 7-HT, we constructed derived strains of the pBBR2-paaZ and site-mutated pBBR2-paaZE258Q in the HYS strain.The changes in the 7-HT production of paaZ and paaZE258Q overexpressing strains were observed in a liquid MKB medium (Figures 2 and 3B).

ORF13, a Homologous Protein of PaaZ, Is Not Necessary for the Biosynthesis of 7-HT in P. donghuensis HYS
To investigate whether orf13 (paaZ) also participated in the biosynthesis of 7-HT, the orf13 gene was deleted in HYS.Then, the ability of the Δorf13 strain to produce siderophores, especially the iron scavenger 7-HT, was determined.
On CAS agar plates (under normal or UV light), the yellow chelated halo of the Δorf13 strain was not significantly different from that of the WT HYS strain (Figure 1B).This phenomenon preliminarily indicated that the deletion of orf13 had no obvious effect on the total amount of siderophores.The yield of siderophores on the Δorf13 strain in the liquid MKB medium supernatant was reduced only by about one-fifth to one-sixth compared to that of the wild-type HYS strain (Figure 1C).Similarly, the characteristic absorption peaks of 7-HT in the Δorf13 strain, at 330 nm and 392 nm, were only slightly decreased compared with the WT strain HYS (Figure 1D).The deletion of orf13 could not completely eliminate the synthesis of 7-HT, but interestingly, the RTq-PCR result indicated that the transcriptional level of orf13 was inhibited by high-concentration iron (Figure 1E).
The results showed that orf13 had no significant effect on the yield of fluorescent pyoverdine and nonfluorescent iron scavenger 7-HT in HYS.The sole carbon experiment showed that orf13 was involved in the catabolism of PAA in P. donghuensis HYS (Figure S2).
Through amino acid sequence alignment, it was found that ORF13 (PaaZ) also consisted of two domains, the N-terminal PaaZ-ALDH and the C-terminal PaaZ-ECH domain, and that the active site of the ALDH domain of PaaZ in P. donghuensis HYS was 258 Glu (E).To inactivate the ALDH domain, the site was mutated to Gln (Q) (E258Q).To further investigate the relationship between orf13 (paaZ), and the production of 7-HT, we constructed derived strains of the pBBR2-paaZ and site-mutated pBBR2-paaZE258Q in the HYS strain.The changes in the 7-HT production of paaZ and paaZE258Q overexpressing strains were observed in a liquid MKB medium (Figures 2 and 3B).Compared with the HYS strain, the characteristic absorption peaks of 7-HT at 330 nm and 392 nm were only slightly reduced in the ∆paaZ strain (Figure 1D), while, interestingly, in the paaZ overexpressed strain HYS/pBBR2-paaZ, the characteristic absorption peaks of 7-HT disappeared, and only the absorption peak of pyoverdine at 405 nm remained compared with the empty vector strain HYS/pBBR2, indicating that the production of 7-HT was eliminated (Figure 2).In the paaZE258Q-overexpressed strain HYS/pBBR2-paaZE258Q, the characteristic absorption peaks of 7-HT appeared again compared with the HYS/pBBR2-paaZ strain, indicating that 7-HT was produced again (Figure 2).These data show that the overexpression of paaZ in the wild-type HYS strain inhibited the production of 7-HT, and compared with the overexpression of paaZ, the overexpression of paaZE258Q restored the production of 7-HT.Furthermore, these results further indicated that ORF13 (PaaZ) did not play a key role in the biosynthesis of 7-HT in P. donghuensis HYS.
In summary, although ORF13 is a homologous protein of PaaZ, interestingly, ∆orf13 (∆paaZ) could not completely eliminate the production of 7-HT, indicating that orf13 is not essential for the production of 7-HT.Thus, it is speculated that there may be at least one other protein homologous to ORF13 (PaaZ) or a protein homologous to the PaaZ-ECH domain in P. donghuensis HYS, which could be involved in the biosynthesis of 7-HT.

The Gene Ech Is Homologous to the C-Terminal ECH Domain of ORF13
To obtain predicted ORF13 (PaaZ) homologous proteins or PaaZ-ECH-like proteins, we performed a genome-wide BLASTP search in P. donghuensis HYS using the whole ORF13 amino acid sequences or its C-terminal ECH domain as a query.Using the entire ORF13 protein sequence for BLASTP, eight homologous sequences were obtained in P. donghuensis HYS (Table 1, Figure 3A), six of which were aldehyde dehydrogenase family proteins (Sequence Number: 1, 3, 5, 6, 7, 8) homologous to the N-terminal PaaZ-ALDH domain.The other two were MaoC family dehydratases (Sequence Number: 2, 4)-(R)-hydratase [(R)-specific enoyl-CoA hydratase] and NodN (nodulation factor N)-homologous to the C-terminal PaaZ-ECH domain (Table 1, Figure 3A).Consequently, based on the above comparative analysis, the gene UW3_RS0113785 was renamed ech, and UW3_RS0112810 was renamed NodN.Using the C-terminal ECH domain amino acid sequences of ORF13 as a query (Table S3, Figure S3), these two hydratase proteins (Sequence Number: 2, 4) were also obtained.To identify whether the two hydratases, ech, and NodN, were involved in the biosynthesis of 7-HT (Table 1, Figure 3A), the ∆ech and ∆NodN knockout strain was constructed.Compared with the WT HYS strain and in the ∆ech strain, the two characteristic absorption peaks of 7-HT, at 330 nm and 392 nm, disappeared, indicating that the knockout of ech eliminated the production of 7-HT.However, in the ∆NodN strain, the two characteristic absorption peaks of 7-HT did not significantly change compared with HYS, indicating that NodN did not participate in the biosynthesis of 7-HT (Figure 3C).The results suggest that ech, but not NodN, was involved in the synthesis of 7-HT.

ech Is Essential for the Biosynthesis of the Iron Scavenger 7-HT
To identify whether ech was related to the biosynthesis of the iron scavenger 7-HT in P. donghuensis HYS, ech was knocked out in wild-type HYS and ∆paaZ.Then, the ability of ∆ech and ∆paaZ∆ech strains to produce 7-HT was tested in liquid MKB (Figure 4).
Compared with the wild-type HYS, in the ∆ech mutant strain, the two characteristic absorption peaks of 7-HT disappeared, and only the absorption peak of pyoverdine was left (Figure 4A).Similarly, in the ∆paaZ∆ech mutant strain, the characteristic absorption peaks of 7-HT disappeared, and only the absorption peak of pyoverdine was detected compared with the ∆paaZ strain (Figure 4A).The data confirmed that ech was essential for the production of the iron scavenger 7-HT.
The above data show that the complement of ech could restore the ability of the knockout strains to produce 7-HT, which further indicated that ech was necessary for the biosynthesis of 7-HT and played a crucial role in the biosynthesis of 7-HT in P. donghuensis HYS.

ech Is Essential for the Biosynthesis of the Iron Scavenger 7-HT
To identify whether ech was related to the biosynthesis of the iron scavenger 7-HT in P. donghuensis HYS, ech was knocked out in wild-type HYS and ΔpaaZ.Then, the ability of Δech and ΔpaaZΔech strains to produce 7-HT was tested in liquid MKB (Figure 4).Compared with the wild-type HYS, in the Δech mutant strain, the two characteristic absorption peaks of 7-HT disappeared, and only the absorption peak of pyoverdine was left (Figure 4A).Similarly, in the ΔpaaZΔech mutant strain, the characteristic absorption peaks of 7-HT disappeared, and only the absorption peak of pyoverdine was detected compared with the ΔpaaZ strain (Figure 4A).The data confirmed that ech was essential for the production of the iron scavenger 7-HT.

Identification of Key Conserved Residues of Ech in P. donghuensis HYS
To further understand the effect of the gene ech on the biosynthesis of iron-scavenging 7-HT in P. donghuensis HYS, we performed the sequence alignment of ECH with those of other 12 paaZ-ECH domains, including E. coli K12.It was found that the conserved catalytic residues of ECH were Asp(D)-39, His(H)-44, and Gly(G)-62 (Figures 3B and S4), which is similar to the three conserved catalytic residues of the PaaZ-ECH domain in E.coli K12 (Asp-561, His-566, and Gly-584) [42].The 3D (three-dimensional) spatial protein structure of ECH in P. donghuensis HYS was predicted.The key residues Asp39, His44, and Gly62 are represented as stick models; the location of the active sites of the enzyme is indicated (Figure 5B).
The results indicated that, due to the site mutations of the key conserved catalytic sites, ech was inactivated; thus, the production of 7-HT could not be restored.The results also preliminarily identified that these three residues (Asp39, His44, and Gly62) were the key conserved catalytic sites of ech and further proved that ech was necessary for the biosynthesis of 7-HT in P. donghuensis HYS.At present, there is no report describing the role of ech involved in 7-HT biosynthesis.

The Gene paaZ Cannot Complement the Role of Ech in the Production of 7-HT in P. donghuensis HYS
To further explore the relationship between genes ech and paaZ and the production of 7-HT, pBBR2-paaZ, and site-mutated pBBR2-paaZE258Q were introduced into ∆ech strains to observe whether the production of 7-HT was recovered.

donghuensis HYS
To further explore the relationship between genes ech and paaZ and the production of 7-HT, pBBR2-paaZ, and site-mutated pBBR2-paaZE258Q were introduced into Δech strains to observe whether the production of 7-HT was recovered.
Compared with HYS, the characteristic absorption peaks of 7-HT, at 330 nm and 392 nm, were not recovered in the Δech/pBBR2-paaZ and Δech/pBBR2-paaZE258Q strain, indicating that 7-HT production was not recovered (Figure 6A).The transcriptional levels of paaZ in Δech/pBBR2-paaZ and Δech/pBBR2-paaZE258Q were approximately 180 and 160 times higher than those in wild-type HYS, respectively, indicating the successful expression of paaZ and paaZE258Q (Figure 6B).These results indicated that the introduction of pBBR2-paaZ or pBBR2-paaZE258Q plasmid into ∆ech could not restore the 7-HT production of ∆ech, suggesting that paaZ and paaZE258Q (the ECH domain of PaaZ) could not complement the role of ech in 7-HT biosynthesis.Furthermore, it was demonstrated that ech plays a more important role in the production of 7-HT than paaZ, which is consistent with the phenomenon whereby ∆paaZ (∆orf13) affects the production of 7-HT only slightly (Figure 1).In addition, it was demonstrated that PaaZ was involved in 7-HT biosynthesis to a small extent and could not complement the disappearance of 7-HT production due to the deletion of ech.

PaaZ Is Mainly Involved in the PAA Catabolic Pathway, While Ech Is Not
In P. donghuensis HYS, paaZ is a gene in the PAA metabolism gene cluster 2, and ech is an R-type enoyl-CoA with homology to the PaaZ-ECH domain outside gene cluster 2. This preliminary shows that paaZ is related to PAA degradation (Figure S2); thus, to investigate whether ech is involved in PAA metabolism, the ability of ech to utilize PAA was tested by adding PAA to an M9 minimal medium as a single carbon source.
Compared with the wild-type HYS strain, there was no significant growth difference between the ∆paaZ, ∆ech, and ∆paaZ∆ech strains with glucose as the single carbon source, indicating that the deletion of these genes did not affect the glucose utilization of P. donghuensis HYS (Figure 7A).When PAA was used as the sole carbon source, compared with the wild-type HYS, the two knockout strains, ∆paaZ and ∆paaZ∆ech, showed a growth defect, but there was no growth defect of the ∆ech strain (Figure 7B), indicating that the ∆paaZ strain could not utilize PAA and that the deletion of ech did not affect the utilization of PAA.Thus, it could be concluded that paaZ in gene cluster 2 was involved in PAA catabolism, while ech was not involved in PAA metabolism in P. donghuensis HYS.
Compared with the wild-type HYS strain, there was no significant growth difference between the ΔpaaZ, Δech, and ∆paaZΔech strains with glucose as the single carbon source, indicating that the deletion of these genes did not affect the glucose utilization of P. donghuensis HYS (Figure 7A).When PAA was used as the sole carbon source, compared with the wild-type HYS, the two knockout strains, ∆paaZ and ∆paaZ∆ech, showed a growth defect, but there was no growth defect of the ∆ech strain (Figure 7B), indicating that the ΔpaaZ strain could not utilize PAA and that the deletion of ech did not affect the utilization of PAA.Thus, it could be concluded that paaZ in gene cluster 2 was involved in PAA catabolism, while ech was not involved in PAA metabolism in P. donghuensis HYS.In summary, paaZ in gene cluster 2 was mainly related to PAA catabolism, while ech was not involved in PAA metabolism in P. donghuensis HYS, and ech was mainly involved in the biosynthesis of 7-HT in P. donghuensis HYS.These results confirm that ech and paaZ have functional differentiation roles in P. donghuensis HYS, which should direct further studies on the biosynthesis mechanism of 7-HT.

The Expression Differences of paaZ and ech and the Response to Extracellular Iron Concentration in P. donghuensis HYS
The genes paaZ and ech have different roles in the biosynthesis of 7-HT.ech is mainly involved in the biosynthesis of 7-HT, while paaZ is mainly related to PAA catabolism in P. donghuensis HYS.Thus, to investigate whether there were differences in the expression levels of paaZ and ech during the period of 7-HT mass synthesis in P. donghuensis, during the experiment, the HYS strain was inoculated in the MKB medium and cultivated to the exponential phase (8 h), which was the period when a large amount of 7-HT was synthesized in HYS [20,21].
It was found that the gene expression level of ech was about 19 times that of paaZ (Figure 8A), indicating that paaZ and ech had different expression levels in the exponential period when 7-HT was produced in large quantities.This is consistent with the phenomenon that ∆ech does not produce 7-HT and has a greater impact on the production of 7-HT, while ∆paaZ has little effect on the production of 7-HT.
It was found that the gene expression level of ech was about 19 times that of paaZ (Figure 8A), indicating that paaZ and ech had different expression levels in the exponential period when 7-HT was produced in large quantities.This is consistent with the phenomenon that Δech does not produce 7-HT and has a greater impact on the production of 7-HT, while ΔpaaZ has little effect on the production of 7-HT.To explore the relationship between the transcription level of ech, paaZ, and extracellular iron, the HYS strain was cultured to the logarithmic phase in the MKB medium and MKB medium supplemented with 30 μM FeSO4, respectively.The transcription levels of paaZ and ech were detected.Compared to the low-iron environment (iron-limited MKB medium), in the high-iron environment (MKB supplemented with 30 μM FeSO4), the To explore the relationship between the transcription level of ech, paaZ, and extracellular iron, the HYS strain was cultured to the logarithmic phase in the MKB medium and MKB medium supplemented with 30 µM FeSO 4 , respectively.The transcription levels of paaZ and ech were detected.Compared to the low-iron environment (iron-limited MKB medium), in the high-iron environment (MKB supplemented with 30 µM FeSO 4 ), the transcription levels of paaZ and ech were somewhat reduced, and the transcription level of paaZ decreased by about 10 times; the transcription level of ech dropped by about 2 times (Figure 8B).The above results indicate that the transcription levels of paaZ and ech were inhibited by a high iron concentration, and the inhibition degree of iron to ech was weaker than that of paaZ.

Discussion
The novel iron scavenger 7-hydroxytropolone (7-HT) is secreted by P. donghuensis HYS in large quantities and contributes to the notable iron-chelating activity of P. donghuaensis HYS [20][21][22].Further, 7-HT contributes to the pathogenicity of C. elegans in P. donghuensis HYS [23][24][25].In other P. donghuensis strains, 7-HT is also the main metabolite to antifungal and plant protective agent [26,40].However, its biosynthetic pathway remains unclear.In this study, based on gene knockout, and complementation experiments, we identified that the orf17-21 (paaEDCBA) and orf26 (paaG) genes in gene cluster 2 were essential for the production of 7-HT in P. donghuensis HYS.A sole carbon source experiment indicated that orf13 (paaZ), orf17-21 (paaEDCBA), and orf26 (paaG) related to the biosynthesis of 7-HT were involved in the PAA catabolon pathway.The data indicated that the PAA catabolon pathway participated in the production of 7-HT in P. donghuensis HYS.
This mechanism is similar to the biosynthesis of TDA and roseobacticide in P. inhibens [44,45] and the biosynthesis of 3,7-Dihydroxytropolone (DHT) in Streptomyces spp.[43].That is, the PAA metabolic pathway provides the seven-membered ring carbon skeleton of these compounds.
Interestingly, although ORF13 was homologous to PaaZ, orf13 (paaZ) was not essential for the production of 7-HT in P. donghuensis HYS.The gene ech outside of cluster 2 was homologous to the C-terminal PaaZ-ECH domain, and ech was involved in the biosynthesis of 7-HT in P. donghuensis HYS.These results confirm that ech and paaZ have functional differentiation roles and a synergistic effect in P. donghuensis HYS; ech is mainly related to the biosynthesis of 7-HT and paaZ mainly plays a role in PAA catabolism in P. donghuensis HYS.At present, there is no report describing the role of ech in 7-HT biosynthesis.The gene paaZ was found to participate in the PAA metabolic branching point [41,42].Although paaZ was also found to participate in the PAA catabolism in P. donghuensis HYS, it was not necessary for the biosynthesis of 7-HT, and the gene ech located outside cluster 2 is involved in 7-HT biosynthesis-an unexpected finding.
A new feature was observed in P. donghuensis HYS: there was a complete PAA degradation in cluster 2 and an independent ech outside the PAA cluster, and paaZ had complete ALDH and ECH functions, which is different from the feature observed in the production of TDA and roseobacticide in P. inhibens [44,45] and the biosynthesis of 3,7-Dihydroxytropolone (DHT) in Streptomyces spp.[43].In P. inhibens, in addition to the paaZ1 of the PAA catabolon gene cluster, a specific didomain enzyme, PaaZ2, outside the PAA gene cluster was found.The ECH domain of PaaZ2 is highly homologous to PaaZ1 in P. inhibens, while the N-terminal ALDH domain shows very low homology.In Streptomyces, the trlA gene, located outside the paa gene cluster, is involved in the production of 3,7-Dihydroxytropolone (DHT) [43].The trlA gene encodes a single-domain (ECH) protein, homologous to the C-terminal ECH domain of E. coli bifunctional protein PaaZ.TrlA truncates the PAA catabolic pathway and redirects it toward the formation of DHT [43].The differences are that in P. inhibens, there are two paaZs: paaZ1 and paaZ2, while in Streptomyces, there is only one TrlA homologous to the paaZ-ECH domain without PaaZ.In HYS, there is a paaZ and an ech which are synergistically involved in the synthesis of 7-HT and metabolism of PAA.
It was speculated that the coexistence of paaZ (orf13) and ech would rapidly produce more 7-HT to cope with environmental stress and save energy, giving the strain a competitive advantage and ensuring its survival.The paaZ and ech could cooperate to allow P. donghuensis HYS to adapt to challenging environmental conditions.paaZ and ech exist simultaneously and are functionally differentiated, which not only ensures the metabolism of PAA but also ensures the synthesis of 7-HT in large quantities when needed in order to obtain competitive advantages and realize the regulation of the 7-HT synthesis pathway.
The two genes, paaZ, and ech, had different expression levels in the exponential period.The expression level of ech was significantly higher than that of paaZ in P. donghuensis HYS.This is consistent with the phenomenon that ∆ech does not produce 7-HT and has a greater impact on the production of 7-HT, while ∆paaZ has little effect on the production of 7-HT.In addition, it is interesting that iron has a stronger inhibitory effect on paaZ and a weaker inhibitory effect on the transcription of ech.We speculated that the reason for this may be that ech is also involved in other reaction pathways not strictly regulated by iron concentration.Furthermore, 7-HT can also act as an antibiotic [40,46], enhancing the competitive fitness of bacteria against other microorganisms.Therefore, it is speculated that when the external iron condition is not so scarce, ech and its expression at a high level guarantee the mass synthesis of 7-HT, and ech can be synthesized both rapidly and earlier by HYS in order to chelate iron and restrict the growth of competitors, which gives P. donghuensis HYS a competitive advantage.This could provide insight into how P. donghuensis HYS adjusts the expression of its paaZ and ech genes-key branching points in the 7-HT biosynthesis pathway-to match varying levels of iron, competition, and environmental pressure.

Bacterial Strains, Plasmids, and Growth Conditions
The bacterial strains and plasmids used in this work are listed in Tables S4 and S5.P. donghuensis HYS isolated from Donghu Lake was used as the wild-type strain.Escherichia coli S17-1 λpir [47] was used as the host for DNA manipulation by conjugating with Pseudomonas strains.Plasmid pEX18Gm [48] was used as a vector for gene knockout; plasmid pBBR1MCS-2 [49] was used for gene overexpression, and the promoterless lacZ reporter plasmid pBBR5Z [20] was used for promoter assays.E. coli strains were routinely cultured in a Luria-Bertani (LB) medium at 37 • C. P. donghuensis HYS and its derivative strains were grown in an LB medium and iron-deficient MKB medium (15 mL/L glycerol, 2.5 g/L K 2 HPO 4 , pH 7.2 subsequently supplemented with 2.5 g/L MgSO 4 and 5 g/L casamino acid) at 30 • C. When required, a final concentration of 30 µM FeSO 4 was added to the MKB medium.When necessary, antibiotics were added at the following final concentrations: for E. coli strains, 50 µg/mL kanamycin and 10 µg/mL gentamicin; for P. donghuensis HYS and its derivative strains, 25 µg/mL chloramphenicol, 50 µg/mL gentamicin, and 50 µg/mL kanamycin were used.

Construction of In-Frame Deletion Mutants and Complement Strains of P. donghuensis HYS
Routine molecular genetic manipulation, including PCR, agarose gel electrophoresis, gel extraction, restriction enzyme digestion, and transformation, were performed using standard procedures [50].The DNA amplification primers used are listed in Table S5 in the Supplementary Material.
The deletion was performed by the double homologous recombination method [48].To construct the deletion plasmids, fragments containing 400-600 bp regions upstream and downstream of the target genes and several nucleotides of the ORFs were amplified and ligated into the suicide vector pEX18Gm.The correct recombinant plasmids were verified by sequencing and then transferred into HYS and its derivative strains via conjugation from E. coli S17-1 (λ pir).Single recombinants were selected on LB agar plates with chloramphenicol and gentamicin, which were based on their resistance to two antibiotics.These recombinants were further incubated overnight at 30 • C for at least 24 h in 5 mL of the liquid LB medium without antibiotics to produce a second allelic exchange.The appropriate strain culture was then diluted by gradient on LB agar plates with 5% (wt/vol) sucrose.The correct deletion mutants were selected and further confirmed by PCR and DNA sequencing.
The plasmids used for complementation and overexpression were constructed by cloning the Shine-Dalgarno (SD) sequences and open reading frames (ORFs) of target genes into the broad-host-range vector pBBR1MCS-2.The plasmid with the target gene fragment was transferred into E. coli S17-1 (λ pir), and then the E. coli S17-1 (λ pir) carrying recombinant plasmids and containing the correct fragments was conjugated with P. donghuensis HYS mutant strains.After conjugation with the corresponding mutants, correct monoclonal targets were selected with double-antibiotic treatments, plasmid extraction, and enzyme digestion verification.

Construction of Point Mutations in P. donghuensis HYS
The plasmids used for point mutation were constructed by cloning target genes with the point mutation site into the broad-host-range vector pBBR1MCS-2.An overlap PCR was used to amplify target genes containing the point mutation site, and other procedures were similar to those of the gene complementation in this study.Oligonucleotide primers were used to amplify target genes with the point mutation sites shown in Table S5.

Siderophore Determination Assays
The production of siderophores was determined using the following methods: on universal Chrome Azurol S (CAS) blue agar plates [51], in which the MM9 solution was replaced by a pH 6.8 phosphate buffer [22], siderophores were measured using their high affinity for iron (III).For the siderophores secreted by the bacteria chelate iron from the medium, their color turns from blue to orange.The orange halos around the colonies indicated the secretion of siderophores [51].In this study, the strains were inoculated in the MKB medium and incubated at 30 • C for 24 h.Then, 10 µL of each MKB culture (the optical density at 600 nm (OD 600 ) was adjusted to 1.0) was dropped onto a CAS agar plate, and the quantities of siderophores were measured by the chelated halos formed after incubating for 24 h at 30 • C.
For the liquid MKB medium, the appropriate dilution of each filtered supernatant was mixed with an equal volume of the CAS assay solution [51] using double-distilled water (ddH 2 O) as a reference.After 1 h of incubation at room temperature, the absorbances of the samples (As) and reference (Ar) at 630 nm were detected.Siderophore units were calculated by subtracting the sample absorbance values from the reference according to the following equation [52]: 100 × (Ar − As)/Ar = % siderophore units.
To obtain the characteristic absorption peaks of the siderophores in a liquid MKB medium, the filtered supernatants of 24 h MKB cultures were normalized to an OD 600 of 0.5.The characteristic absorption peaks of 7-HT were at approximately 330 and 392 nm, and that of pyoverdine was at approximately 405 nm.Then, the absorption spectra were measured every 0.5 nm using a UV/visible spectrophotometer (UV-2550; Shimadzu, Kyoto, Japan).

Real-Time qPCR
Triplicate PCR reactions were carried out according to the manufacturer's instructions (SYBR Premix Ex Taq Tli RNaseH Plus; TaKaRa, Kusatsu, Japan) in a qPCR machine (CFX96 real-time PCR detection system; Bio-Rad Technologies, Hercules, CA, USA).The reaction system and cycling conditions for qPCR were performed according to the manufacturer's protocol (SYBR Premix Ex Taq Tli RNaseH Plus; TaKaRa).The experimental data were analyzed using Bio-Rad CFX Manager 3.1 software (Bio-Rad Technologies).To quantify relative gene expression, real-time qPCR data were analyzed by the 2 −∆∆CT method [53].The relative expression of the target genes is shown as the ratio of samples to the wild type after normalization to the reference gene rpoB [54,55].

Bioinformatic Analysis
A genome-wide BLASTP search was performed for predicted ORF13 (PaaZ) homologous proteins or PaaZ-ECH-like proteins using the whole ORF13 amino acid sequence or its C-terminal ECH domain as a query in NCBI BLASTP.Standard databases were searched using the following queries: non-redundant (nr) protein sequences, organism optional, and Pseudomonas donghuensis (Taxid: 1163398).Algorithm parameters used the default settings (Expect threshold: 0.05).
The sequences in Figure S4 were obtained from NCBI and compared using the clusterW method by MEGA 11 version 11.0.13software; some sequences refer to Nelson L. Brock et al. [45].

RNA Isolation and Reverse Transcription (RT)-PCR
A total of 1 mL of the culture of P. donghuensis HYS and its derivative strains in the exponential phase (at an OD 600 of 0.6 to 0.8) was harvested in 50 mL of a liquid MKB medium or medium supplemented with 30 µM FeSO 4 •7H 2 O incubating at 30 • C.Then, the supernatant was removed by centrifugation (13,000× g, 1 min) at 4 • C. The total RNA was extracted using the TRIzol reagent (Ambion, Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions.Then, genomic DNA was removed according to the method proposed by Chen et al. [22].
RNA was quantified using a NanoDrop 2000 (Thermo Fisher Scientific, Shanghai, China) and was reverse-transcribed to generate cDNA using the PrimeScript RT Reagent Kit with gDNA Eraser (TaKaRa; Kyoto, Japan).The RNA was stored at −80 • C, and the cDNA was stored at −20 • C. cDNA was used as a template for real-time qPCR.

Conclusions
In summary, this study is the first to report that the paaABCDE and paaG genes in cluster 2 are involved in the biosynthesis of 7-HT and that two genes (paaZ ((orf13)) at the branching point of the PAA metabolic pathway and ech) are synergistically involved in the biosynthesis of 7-HT in the genus Pseudomonas.The gene paaZ mainly plays a role in the degradation of PAA and slightly affects the biosynthesis of 7-HT, while ech mainly participated in the biosynthesis of 7-HT in P. donghuensis HYS.Moreover, ech and paaZ were distributed far apart in the genome of HYS, and ech was encoded outside of the PAA metabolic gene cluster 2, where paaZ was found, which should direct further studies on the biosynthesis mechanism for 7-HT.This study revealed a natural association between the catabolism of PAA and the biosynthesis of 7-HT: genes related to PAA metabolism are involved in the biosynthesis of the siderophore 7-HT in P. donghuensis HYS.This association has not been reported previously in the genus Pseudomonas.This study complements our understanding of the mechanism of the biosynthesis pathway of 7-HT and provides directions for the study of the 7-HT biosynthesis pathway in P. donghuensis HYS.Moreover, this study also enriched research on the biosynthesis of troplones and siderophores in the genus Pseudomonas.

Figure 1 .
Figure 1.Effects of orf13, orf17-orf21, and orf26 on siderophore production in P. donghuensis HYS.(A) Diagram of the gene arrangement of cluster 2 of P. donghuensis HYS.The arrows and their directions indicate the locations and direction, respectively, of the transcription of the predicted genes in cluster 2. The red font indicates the transposon mutant strains and their insertion positions.(B) Siderophore production in wild-type HYS and the ∆orf13, ∆orf17, ∆orf18, ∆orf19, ∆orf20, ∆orf21,

Figure 3 .
Figure 3. Alignment and analysis of orf13 homologous genes obtained from a genome-wide BLASTP search in P. donghuensis HYS and their relationship with 7-HT biosynthesis.(A) Distribution of the eight proteins with homology to the N-terminal ALDH domain or the C-terminal ECH domain of ORF13 (PaaZ)

Figure 4 .
Figure 4. Different effects of the paaZ and ech genes on the production of 7-HT in P. donghuensis HYS.(A) Absorption spectra of the filtered supernatants of 24 h MKB cultures from wild-type HYS, ∆paaZ, ∆ech, and ∆paaZ∆ech strains.(B) Absorption spectra of the filtered supernatants of 24 h MKB cultures

Figure 5 .
Figure 5. Critical residues of ech control with the involvement of ech in 7-HT biosynthesis and the effect of D39A, H44A, and G62A point mutations of these residues on 7-HT production in P. donghuensis HYS.(A) Absorption spectra of the filtered supernatants of 24 h MKB cultures from wildtype HYS, Δech, Δech/pBBR2, Δech/pBBR2-echD39A, Δech/pBBR2-echH44A, Δech/pBBR2-echG62A, and Δech/pBBR2-ech strains.(B) The predicted 3D (three-dimensional) spatial protein structure for the ech of P. donghuensis HYS.The key residues Asp39, His44, and Gly62 are represented as a stick model, which indicates the location of the active sites of the enzyme.The black font shows the position of key residues in the sequence.The figure was prepared with SWISS-MODEL and PYMOL 2.5.2.

Figure 5 .
Figure 5. Critical residues of ech control with the involvement of ech in 7-HT biosynthesis and the effect of D39A, H44A, and G62A point mutations of these residues on 7-HT production in P. donghuensis HYS.(A) Absorption spectra of the filtered supernatants of 24 h MKB cultures from wild-type HYS, ∆ech, ∆ech/pBBR2, ∆ech/pBBR2-echD39A, ∆ech/pBBR2-echH44A, ∆ech/pBBR2-echG62A, and ∆ech/pBBR2-ech strains.(B) The predicted 3D (three-dimensional) spatial protein structure for the ech of P. donghuensis HYS.The key residues Asp39, His44, and Gly62 are represented as a stick model, which indicates the location of the active sites of the enzyme.The black font shows the position of key residues in the sequence.The figure was prepared with SWISS-MODEL and PYMOL 2.5.2.

Figure 7 .
Figure 7. Sole carbon source tested for wild-type HYS, ΔpaaZ, Δech, and ΔpaaZΔech mutant strains at the stationary phase.(A) Growth in M9 minimal medium with 0.4% glucose as a sole carbon source (black bar).(B) Growth in M9 minimal medium with 0.6 mg/mL PAA as a sole carbon source (grey bar).Each value is the average from three different cultures ± the standard deviation.The growth is expressed as OD600.OD600, optical density at 600 nm.*** p < 0.001, Student's t-test.

Figure 7 .
Figure 7. Sole carbon source tested for wild-type HYS, ∆paaZ, ∆ech, and ∆paaZ∆ech mutant strains at the stationary phase.(A) Growth in M9 minimal medium with 0.4% glucose as a sole carbon source (black bar).(B) Growth in M9 minimal medium with 0.6 mg/mL PAA as a sole carbon source (grey bar).Each value is the average from three different cultures ± the standard deviation.The growth is expressed as OD 600 .OD 600 , optical density at 600 nm.*** p < 0.001, Student's t-test.

Figure 8 .
Figure 8. Relative expression levels of the paaZ and ech genes in the P. donghuensis HYS strain.(A) Relative transcriptional differences between the paaZ and ech genes in the exponential phase.Transcriptional levels are shown as the relative expression of the paaZ and ech compared to the expression of the rpoB gene at the exponential phase.(B) Effect of iron concentration on the expression of the paaZ and ech genes in the exponential phase.Quantitative fluorescence PCR was used to detect the transcription levels of paaZ and ech at MKB or MKB+ 30 μM FeSO4.Transcriptional levels are shown as the relative expression of the paaZ and ech compared to the expression of the rpoB gene at the exponential phase, as measured by qPCR.The error bars indicate the standard deviations (n = 3).** p < 0.01, *** p < 0.001, Student's t-test.

Figure 8 .
Figure 8. Relative expression levels of the paaZ and ech genes in the P. donghuensis HYS strain.(A) Relative transcriptional differences between the paaZ and ech genes in the exponential phase.Transcriptional levels are shown as the relative expression of the paaZ and ech compared to the expression of the rpoB gene at the exponential phase.(B) Effect of iron concentration on the expression of the paaZ and ech genes in the exponential phase.Quantitative fluorescence PCR was used to detect the transcription levels of paaZ and ech at MKB or MKB+ 30 µM FeSO 4 .Transcriptional levels are shown as the relative expression of the paaZ and ech compared to the expression of the rpoB gene at the exponential phase, as measured by qPCR.The error bars indicate the standard deviations (n = 3).** p < 0.01, *** p < 0.001, Student's t-test.
and orf26 in P. donghuensis HYS in high and low extracellular iron concentrations.RNA was isolated from the indicated strains grown at an exponential phase at 30 °C in a liquid MKB culture supplemented with or without 30 μM FeSO4•7H2O.Each value is the average from three different cultures ± the standard deviation.

Table 1 .
Characteristics of PaaZ homologous protein a in P. donghuensis HYS.
a Similarity values are for the most similar protein, determined by BLASTP analysis.b Percentage identity.c Amino acid length of the accession protein.d Gene id in P. donghuensis HYS.