A Novel Estrone Degradation Gene Cluster and Catabolic Mechanism in Microbacterium oxydans ML-6

ABSTRACT

Global-scale estrone (E1) contamination of soil and aquatic environments results from the widespread use of animal manure as fertilizer, threatening both human health and environmental security. A detailed understanding of the degradation of E1 by microorganisms and the associated catabolic mechanism remains a key challenge for the bioremediation of E1-contaminated soil. Here, Microbacterium oxydans ML-6, isolated from estrogen-contaminated soil, was shown to efficiently degrade E1. A complete catabolic pathway for E1 was proposed via liquid chromatography-tandem mass spectrometry (LC-MS/MS), genome sequencing, transcriptomic analysis, and quantitative reverse transcription-PCR (qRT-PCR). In particular, a novel gene cluster (moc) associated with E1 catabolism was predicted. The combination of heterologous expression, gene knockout, and complementation experiments demonstrated that the 3-hydroxybenzoate 4-monooxygenase (MocA; a single-component flavoprotein monooxygenase) encoded by the mocA gene was responsible for the initial hydroxylation of E1. Furthermore, to demonstrate the detoxification of E1 by strain ML-6, phytotoxicity tests were performed. Overall, our findings provide new insight into the molecular mechanism underlying the diversity of E1 catabolism in microorganisms and suggest that M. oxydans ML-6 and its enzymes have potential applications in E1 bioremediation to reduce or eliminate E1-related environmental pollution.

IMPORTANCE Steroidal estrogens (SEs) are mainly produced by animals, while bacteria are major consumers of SEs in the biosphere. However, the understanding of the gene clusters that participate in E1 degradation is still limited, and the enzymes involved in the biodegradation of E1 have not been well characterized. The present study reports that M. oxydans ML-6 has effective SE degradation capacity, which facilitates the development of strain ML-6 as a broad-spectrum biocatalyst for the production of certain desired compounds. A novel gene cluster (moc) associated with E1 catabolism was predicted. The 3-hydroxybenzoate 4-monooxygenase (MocA; a single-component flavoprotein monooxygenase) identified in the moc cluster was found to be necessary and specific for the initial hydroxylation of E1 to generate 4-OHE1, providing new insight into the biological role of flavoprotein monooxygenase.

INTRODUCTION

Steroidal estrogens (SEs), which are generated naturally in living organisms or administered as medicines to humans and livestock, are secreted into soil and aquatic environments mainly via urine and manure (1, 2). SEs, including estrone (E1), 17β-estradiol (17β-E2), estriol (E3), and 17α-ethinylestradiol (EE2), have been classified as group 1 carcinogens by the World Health Organization and represent a new global challenge (3, 4). Among them, E1 has received much attention since it is present at higher concentrations than other SEs in various environments (5). Bartelt-Hunt et al. (6) found that E1 was present at a maximal concentration of 720 ng/L in cattle feedlot runoff in Nebraska (USA). Ben et al. (7) measured the concentration of E1 in the influent and effluent in 14 municipal wastewater treatment plants (WWTPs) distributed across China, obtaining values ranging from 23.6 to 241.0 ng/L and 0.1 to 15.3 ng/L, respectively. Yang et al. (8) performed a large-scale assessment of E1 occurrence in soils from 1 municipality and 7 provinces in China, and the mean concentration of E1 reached 0.53 ± 0.91 μg/kg. However, the predicted no-effect concentration (PNEC) for E1 was 6 ng/L (9–11). Such high residual concentrations of E1 could pose a moderate or high risk to agricultural soil-water systems (12). Ecotoxicological risk and hazard assessment research has revealed that E1 can be associated with prostate cancer in men, breast carcinoma in women, reproductive dysfunction in livestock and wild animals, and physiological perturbations in fish (13–15). Consequently, it is necessary to investigate appropriate approaches to facilitate the degradation of estrogens present in the environment.

Compared with chemical and physical remediation methods, biodegradation is considered an efficient, low-energy-consuming, and pollution-free method to remove E1 from the environment (16). Several E1-degrading biocatalysts have been reported in previous papers, such as Sphingomonas sp., Novosphingobium sp., Rhodococcus sp. and Pseudomonas sp., and their degradation abilities have been examined (17–25). Based on studies of these E1-degrading bacteria, microbial degradation of E1 can be broadly categorized into three pathways (see Fig. S1 in the supplemental material): (i) E1 is instantly split through the 4,5-seco pathway after the hydroxylation of C-4 of the A ring (26–28), (ii) E1 is cleaved after the B ring is hydroxylated (27, 29), and (iii) E1 is initiated in the D ring to form an unstable lactone TP or 16-hydroxyestrone (30). Although the degradation pathways of E1 have been studied, the catabolic enzymes participating in these pathways are relatively poorly understood. In addition, there is limited information on the detoxification of E1 and its degradation products by the above-mentioned biocatalysts.

The catabolism of E1, particularly the initial biotransformation, has received much attention. However, reports about E1-degrading genes are still rare. To date, only cytochrome P450 monooxygenases (CYP450s), namely, EdcA, EstP, and AedA from the edc cluster of Novosphingobium tardaugens NBRC_16725, the est cluster of Novosphingobium sp. strain ES2-1, and the aed cluster of Rhodococcus sp. strain B50, respectively, have been identified, through heterologous expression or gene knockout, to convert E1 to 4-hydroxyestrone (4-OH-E1) by hydroxylation (18, 21, 31). Moreover, flavoprotein monooxygenase (OecB), located in cluster I of Sphingomonas sp. strain KC8, was proposed as the enzyme responsible for the initial hydroxylation of E1 to 4-OH-E1 based on transcriptomic analysis, however, functional verification of OecB has not been performed at the molecular and genetic levels (28). Notably, OecB shares low amino acid sequence identity with EdcA, EstP, and AedA, although they have similar functions in the biodegradation of E1.

Hence, this work aimed to fully discern the catabolic mechanism for E1 in Microbacterium oxydans ML-6, which exhibits high efficiency in E1 degradation under aerobic conditions and was isolated from an experimental farm. The following studies were carried out to (i) ascertain the optimal conditions for growth and E1 degradation by strain ML-6, (ii) propose an E1 degradation pathway for strain ML-6 based on liquid chromatography-mass spectrometry (LC-MS) analysis and information from relevant literature, (iii) identify the genes associated with E1 biodegradation via whole-genome and transcriptome analyses and quantitative reverse transcription-PCR (qRT-PCR), (iv) identify the function of the gene responsible for the initial hydroxylation of E1 by heterologous expression, gene knockout, and complementation experiments, and (v) assess the suitability of strain ML-6 for detoxification of E1 and its degradation products. This paper offers an in-depth understanding of the E1 microbial catabolic mechanism and proposes an excellent microbial candidate for soil bioremediation.

RESULTS AND DISCUSSION

Kinetics of the degradation of SEs by M. oxydans ML-6.

A bacterium with strong E1 degradation ability was isolated from E1-rich soils for detailed study and was designated ML-6. The detailed identification results are provided in the supplemental material. Further biodegradation kinetics were examined by fitting a pseudo-first-order kinetics model as well as half-lives to explore the degradation efficiency of E1 and other SEs (E2, E3, and EE2) (32). The final degradation efficiencies of E1 at 1, 10, and 30 mg/L by strain ML-6 were 100%, 90.1%, and 84.5%, respectively (Fig. 1). The degradation kinetics of E1 at different concentrations were fitted with regression coefficients (R²) of 0.946, 0.977, and 0.916, and the degradation half-lives (t_1/2) of E1 were 19.8, 23.7, and 38.5 h, respectively (Table 1). In addition to E1, 1 mg/L E2, E3, and EE2 were also rapidly degraded by strain ML-6, indicating that strain ML-6 had potential capacity for SEs degradation. Strain ML-6 could also efficiently degrade E3 and EE2 at higher concentrations, and the degradation effect (k/h) and degradation half-life (t_1/2) are shown in Fig. 1 and Table 1. However, strain ML-6 was less effective at degrading E2 at higher concentrations, probably because E2 is toxic or inhibitory to the bacteria (33). Overall, the broad substrate degradation spectrum of strain ML-6 makes it a candidate for practical biodegradation of SEs in complex contaminated environments.

FIG 1.

Kinetics of the degradation of the SEs E1 (A), E2 (B), E3 (C), and EE2 (D) by strain ML-6 under aerobic laboratory conditions.

TABLE 1.

Kinetic parameters for the degradation of steroid estrogens

Compound	Initial concn (mg/L)	k (h−1)a	R 2 b	t1/2 (h)c
E1	1	0.035	0.946	19.8
	10	0.029	0.977	23.7
	30	0.018	0.916	38.5
E2	1	0.029	0.978	23.9
	10	0.005	0.894	139
	30	0.004	0.847	173
E3	1	0.020	0.903	34.7
	10	0.015	0.944	46.2
	30	0.012	0.781	57.8
EE2	1	0.015	0.942	46.2
	10	0.012	0.912	57.8
	30	0.006	0.757	116

^aKinetic rate constant, calculated using the first-order reaction kinetic model.

^bCorrelation coefficient, which represents the fitness of the modeling data.

^cHalf-life, calculated as ln2/k.

Proposed E1 biodegradation pathway in M. oxydans ML-6.

To confirm the biodegradation intermediates, analytical methods, including high-performance liquid chromatography (HPLC) and LC-MS/MS, were applied to detect the main products. During our preliminary LC-MS experiments, 4-OHE1 was detected during the degradation of E1 by ML-6; 4-OHE1 is the product formed before meta-cleavage of the benzene ring (data not shown). Therefore, based on the preliminary experimental results and the reported literature, we hypothesized that further degradation of 4-OHE1 occurs through cleavage of the A ring (pathway i) in strain ML-6 (28). To verify this hypothesis, we cultivated strain ML-6 with E1 and then added the meta-cleavage inhibitor 3-chlorocatechol (100 mg/L) to prevent cleavage of the benzene ring (A ring) (27). Indeed, the results showed that the E1 degradation efficiency of strain ML-6 decreased significantly to 8.78% (compared with 82.5% in the absence of 3-chlorocatechol), indicating that 3-chlorocatechol prevented cleavage of the A ring, thus causing a decline in the E1 degradation efficiency (see Fig. S2 in the supplemental material). Based on these results, we deduced that E1 degradation started from the A ring. Evaluation of the electrospray ionization-MS (ESI-MS) results was carried out through comparison with spectral libraries and relevant literature, which allowed the identification of 10 metabolites in total from 12 to 96 h (see Fig. S3 and Table S1 in the supplemental material) (34). The mass spectrum of product I exhibited a molecular ion at m/z 271.1652 [M+H]⁺, which was detected as 4-OHE1, the foremost biodegradation product of E1 (C₁₈H₂₂O₂; m/z 271.1686), based on the spectral libraries and previous reports (18, 34, 35). Product II (C₁₈H₂₂O₅; m/z 319.0961) was described in previous reports as the meta-cleavage product of 4-OHE1 (35, 36). In addition, the structures of product III to product IX determined from their fragment spectra (and their target m/z values) were C₁₇H₂₂O₄ (m/z 291.1949), C₁₇H₂₄O₄ (m/z 293.1718), C₁₇H₂₆O₅ (m/z 310.3107), C₁₇H₂₄O₅ (m/z 309.1710), C₁₅H₂₂O₄ (m/z 266.1497), C₁₅H₂₀O₅ (m/z 281.0538), and C₁₃H₁₈O₄ (m/z 239.1247), respectively (18, 34).

Based on these products and previous reports, the degradation pathway for E1 in strain ML-6 was deduced (Fig. 2A) (18, 34, 35, 37). Initially, E1 is hydroxylated at the C-4 position (A ring) to produce 4-OHE1 (product I). 4-OHE1 is then oxidized and cleaved between C-4 and C-5 to generate 4-norestrogen-5(10)-en-3-oxocarboxylic acid (product II) and further decarboxylated by oxidation to produce product III (18, 28, 34, 35). Product III is subsequently cleaved to form product IV. Product IV may be hydrolyzed to metabolite V and converted to product VI, which then undergoes a β-oxidation process to form 3aα-H-4α(3′-propanoate)-7aβ-methylhexahydro-1,5-indanedione (HIP) and eventually enters the HIP degradation pathway (18, 34). To enhance our understanding of the microbial degradation of E1, the catabolic mechanisms and degradation enzymes associated with this pathway need to be further explored.

FIG 2.

(A) Proposed pathway for the degradation of E1 by strain ML-6. The enzymes encoded by the genes in this pathway in strain ML-6 are as follows: 3-hydroxybenzoate 4-monooxygenase (mocA); 3,4-dihydroxyphenylacetate 2,3-dioxygenase (mocB); homoprotocatechuate catabolism bifunctional isomerase/decarboxylase (mocC); 2-oxo-hept-4-ene-1,7-dioate hydratase (mocD); 4-hydroxy-2-oxo-heptane-1,7-dioate aldolase (mocE); 5-carboxymethyl-2-hydroxymuconate semialdehyde dehydrogenase (mocF); acyl-coenzyme dehydrogenase (acad); 2-hydroxycyclohexyl-coenzyme dehydrogenase (bad), and acetyl-coenzyme acyltransferase (thiolase) (fadA). The enzyme-encoding genes of strain ML-6 are indicated in italics. Genes located on the moc cluster for steroid A/B-ring degradation are in red, and genes involved in β-oxidation are in blue. (B) The moc gene cluster involved in E1 degradation in strain ML-6 was compared with similar functional gene clusters in other bacteria. The percentages indicate the shared amino acid sequence identity of MocA with other initial E1 hydroxylation proteins. Similar colors indicate enzymes with similar functions. (C) Relative mRNA levels of functional genes in the moc gene cluster. Significant differences between conditioning treatments are indicated by letters (ANOVA, false discovery rate [FDR]-corrected least-squares means, P < 0.05).

Comparative genomic analysis and genome mining of M. oxydans ML-6.

The genome property analysis of strain ML-6 is shown in the supplemental material. In the ML-6 genome, the genes associated with E1 degradation were annotated and predicted, encoding 3-hydroxybenzoate 4-monooxygenase (ML6_01055) and flavin monooxygenase (ML6_02765) (34). E1 metabolite-degrading genes were also annotated and predicted to encode 3,4-dihydroxyphenylacetate 2,3-dioxygenase (ML6_00370 and ML6_01049), glyoxalase (ML6_00265, ML6_00363, and ML6_03360), homoprotocatechuate catabolism bifunctional isomerase/decarboxylase (ML6_01052), hydratase (ML6_00851, ML6_01048, ML6_01082, ML6_01084, and ML6_03216), aldolase (ML6_01047), dehydrogenase (ML6_01050), acetyl coenzyme A (acetyl-CoA) acyltransferase (ML6_01940), and acyl-CoA dehydrogenase (ML6_00513, ML6_00538, ML6_00998, ML6_01786, ML6_02555, and ML6_03218) (18, 34). Moreover, we discovered a potential E1-degrading gene cluster that performed the upstream step of E1 metabolism. The composition and function of this gene cluster are described in the next section.

As of May 2022, whole-genome sequencing has been completed for a total of 11 strains of bacteria with estrogen degradation capacity or estrogen bioconversion capacity, including Sphingomonas sp. strain KC8, Lysinibacillus sphaericus DH-B01, Stenotrophomonas maltophilia SJTL3, Pseudomonas putida SJTE-1, Acinetobacter radioresistens DSSKY-A-001, Deinococcus actinosclerus SJTR-1, Pseudomonas citronellolis SJTE-3, Stenotrophomonas maltophilia SJTH1, Rhodococcus hoagii DSSKP-R-001, and N. tardaugens NBRC_16725. The basic characteristics of the genomes of the strains mentioned above are presented in Table S2 in the supplemental material. Among them, the high GC content probably gives strain ML-6 better heat resistance, allowing it to adapt to complex polluted environments (38). There is general agreement that bacterial tRNA gene sets are, in conjunction with the rest of the translational machinery, shaped by selection for rapid, accurate, and efficient protein translation, which is an important determinant of bacterial growth rate (39). The more tRNA there is, the higher the efficiency of protein production is and the faster the bacterial growth and division are (39). Therefore, the fact that strain ML-6 contained a large number of tRNA genes indicated that it had the potential for protein synthesis and good growth capacity.

Moreover, genome annotation revealed that most strains contained predicted E1-degrading enzymes, such as flavin-dependent monooxygenase and cytochrome P450 monooxygenase, and E1 metabolite-degrading enzymes, for example, VOC (vicinal oxygen chelate) family protein and indolepyruvate ferredoxin oxidoreductase (18, 34). However, most research has focused on the E1 degradation capacity, and research on the catabolic mechanisms of E1 has been very limited. Among the catabolic enzymes, the initial biotransformation enzyme for E1 has received much attention and has been examined in several studies. Recently, CYP450 monooxygenases were proposed to be responsible for the initial degradation of E1 in strain NBRC_16725, and the function of CYP450 was confirmed using gene expression and knockout techniques (18). Strain DSSKP-R-001 and strain DSSKY-A-001 were shown by PCR and RT-PCR to harbor some monooxygenases (40, 41), however, whether these genes are associated with E1 degradation needs further validation.

Transcriptomic analysis showed that the initial hydroxylation of E1 was catalyzed by flavoprotein monooxygenase (OecB) in strain KC8, however, the catalytic function of OecB has not been confirmed at the molecular and genetic levels (28). In this research, the catalytic activity of MocA toward E1 in strain ML-6 was similar to that of OecB in strain KC8; both enzymes are responsible for the initial hydroxylation of E1 by flavoprotein monooxygenase. BLASTP analysis showed that OecB shares low amino acid sequence identity (18.9%) with MocA in strain ML-6 because they belong to different groups of flavoprotein monooxygenases. Sequence analysis illustrated that MocA contains the flavin adenine dinucleotide (FAD)-binding Rossmann motif GxGxxG(x) with NAD(P)H as an electron donor, which is characteristic of group A flavoprotein monooxygenases. OecB, conversely, contains an acyl-CoA dehydrogenase fold and is classified as a group D flavoprotein monooxygenase (see Fig. S4 in the supplemental material) (37, 42, 43). Despite these results, the molecular mechanisms remain to be explored due to the diversity of enzymes involved in the microbial degradation of E1.

To further explore the relationship between the origin and evolution of E1-degrading enzymes, the amino acid sequence identity of flavoprotein monooxygenases in the genus Microbacterium was examined, including the enzymes from strain HG3 (WP_046748851.1), strain VIU2A (WP_127482008.1), strain TPU 3598 (WP_029260490.1), and strain Y-01 (WP_124894450.1), for which whole-genome sequencing has been completed. The results indicated that the amino acid sequence identity of flavoprotein monooxygenases from strain ML-6 and other Microbacterium spp. ranged from 90.2% to 91.5%. The high amino acid sequence identity indicated that flavoprotein monooxygenases are evolutionarily conserved in Microbacterium spp., showing that these species have the potential to degrade E1, however, to date, except for M. oxydans ML-6 reported in this study, there are no other reports on the degradation of E1 by Microbacterium spp., and their actual function remains to be further confirmed.

Prediction of E1 catabolic genes based on transcriptome and qRT-PCR analyses.

To further investigate the functional genes associated with the degradation of E1, transcriptome analysis of strain ML-6 was carried out utilizing cells grown for 8 h and 12 h under E1 and glucose control conditions. The basic information for transcriptome sequencing (RNA-seq) analysis is described in the supplemental material. For RNA-seq analysis, a 10.4-kb gene cluster comprising 9 genes (ML6_01047 to ML6_01055) was of significant interest (Fig. 2B). The gene cluster included 6 functional genes, one transcriptional regulator gene, one transcriptional repressor gene, and one inner membrane transporter gene. Eight of the nine genes were upregulated; therefore, we predicted that this gene cluster might be involved in E1 degradation and designated it moc.

In the moc gene cluster, the ML6_01055 gene (mocA; 11.1-fold change at 8 h, 9.81-fold change at 12 h) encodes a predicted 3-hydroxybenzoate 4-monooxygenase, which is a single-component flavoprotein monooxygenase and might participate in the initial hydroxylation of E1 (44). Subsequently, we hypothesized that A-ring cleavage at the C-4–C-5 position was catalyzed by the product of the ML6_01049 gene (mocB; 7.20-fold change at 8 h, 7.28-fold change at 12 h), a putative 3,4-dihydroxyphenylacetate 2,3-dioxygenase. The ML6_01052 gene (mocC; 2.43-fold change at 8 h, 1.84-fold change at 12 h) encodes a predicted homoprotocatechuate catabolism bifunctional isomerase/decarboxylase presumed to be involved in the decarboxylation of the meta-cleavage product (product II) (18). The ML6_01048 gene (mocD; 3.72-fold change at 8 h, 6.76-fold change at 12 h) encodes a potential 2-oxo-hept-4-ene-1,7-dioate hydratase predicted to be responsible for hydrolyzing product III to product IV (18, 34). Subsequently, product IV is oxidized to product V by a putative 4-hydroxy-2-oxo-heptane-1,7-dioate aldolase (encoded by ML6_01047 [mocE]; 1.52-fold change at 8 h, 2.20-fold change at 12 h) and then converted to product VI through a potential 5-carboxymethyl-2-hydroxymuconate semialdehyde dehydrogenase (encoded by ML6_01050 [mocF]; 5.86-fold change at 8 h, 4.75-fold change at 12 h) (45). Eventually, product V continues to be degraded by the products of β-oxidation-related genes to produce HIP (IX) (34). A group of genes encoding a series of predicted enzymes, including 2-hydroxycyclohexyl-coenzyme dehydrogenase (ML6_03248 [badH]), acetyl-coenzyme acyltransferase (thiolase) (ML6_01940 [fadA]), and acyl-coenzyme dehydrogenase (ML6_00513, ML6_00538, ML6_00998, ML6_01786, ML6_ 02555, and ML6_03218 [ACAD]), may be responsible for the β-oxidation process participating in the downstream degradation pathway of E1 (18, 34).

Thus, qRT-PCR was performed to further verify the transcription levels of the functional genes located in the moc gene cluster. As shown in Fig. 2C, the mRNA expression levels of mocA, mocB, mocC, mocD, mocE, and mocF in the moc gene cluster were significantly increased in the presence of E1, suggesting that the expression of these putative E1-degrading genes was induced by E1 or other E1 degradation intermediates. Moreover, the mRNA expression profiles of these E1-degrading genes were consistent with the transcriptomic analysis results and the expected degradation pathway. Hence, based on the above results, the moc gene cluster plays a pivotal role in the degradation of E1 by strain ML-6.

Interestingly, we found that strain ML-6 and other estrogen-degrading strains (N. tardaugens NBRC_16725, Sphingomonas sp. strain KC8, Novosphingobium sp. strain ES2-1, and Rhodococcus sp. strain B50) shared the same upstream degradation products of E1, but the degradation gene clusters involved were diverse. The plasmid genome data related to the aed cluster in strain B50 were not uploaded and thus were not compared. Hence, the gene organization of the moc cluster was compared with those of the oec cluster from strain KC8 (37), the edc cluster from strain NBRC_16725 (18), and the est cluster from strain ES2-1 (31), and the following differences were identified. Based on amino acid sequence identity, the MocABCDEF proteins encoded by the moc cluster of strain ML-6 showed low (less than 30%) amino acid sequence identity values of 11.3% to 22.2% (strain KC8), 16.3% to 22.6% (strain NBRC_16725), and 15.9% to 24.0% (strain ES2-1). The initial degradation step (hydroxylation at C-4 of the A ring) for E1 in strain NBRC_16725 and strain ES2-1 has been demonstrated to be catalyzed by the CYP450 monooxygenases EdcA and EstP1, respectively. In contrast, this step is catalyzed by the flavoprotein monooxygenases MocA and OecB in strain ML-6 and strain KC8 (not confirmed) (28). Moreover, comparison of MocA with previously reported flavoprotein monooxygenases and CYP450 monooxygenases revealed limited identity (4.08 to 46.6%) by BLASTP analysis.

To further ascertain the relationship of MocA with the above-mentioned monooxygenases, a phylogenetic tree was constructed (Fig. 3). The alignment of sequences that was used to generate the phylogenetic tree was shown in Table S3 in the supplemental material. Notably, phylogenetic analysis showed that MocA was closely related to hydroxybenzoate hydroxylase (AAR25885; 46.6% identity), which catalyzes the hydroxylation of 3-hydroxybenzoate, however, it formed a separate clade from functionally similar flavoprotein monooxygenase (OecB; 18.9% identity) (Fig. 3). Therefore, we could conclude that the flavoprotein monooxygenase MocA is probably a necessary and specific enzyme in E1 degradation, providing new insight into the biological role of flavoprotein monooxygenase in estrogen degradation, which requires further in-depth study. In conclusion, all of the above differences between the moc cluster and other E1-degrading clusters, especially the diversity in gene cluster composition and lower amino acid sequence identity, indicated that the moc cluster was a novel E1-degrading cluster.

FIG 3.

Neighbor-joining phylogenetic tree established based on the amino acid sequences of MocA and its homologous proteins using MEGA (version 7.0) software. Protein IDs or GenBank accession numbers are given in parentheses.

E1-MocA complex analysis by molecular docking.

Three-dimensional analysis of the MocA protein was performed using homology modeling to better understand the structural bioinformatics results for MocA (Fig. 4). Briefly, a three-dimensional structure of MocA was generated by SWISS-MODEL with the X-ray structure of the 3-hydroxybenzoate hydroxylase (MHBH) (Protein Data Bank [PDB] entry 2DKH; 46.6% sequence identity with MocA) from Comamonas testosteroni KH122-3s as a template (Fig. 4A, panel a) (44). MHBH, a member of the group A flavoprotein monooxygenases (Fig. 4B), catalyzes the conversion of 3-hydroxybenzoate to 3,4-dihydroxybenzoate. Based on the crystal structure of MHBH, the predicted structure of MocA contains a UbiH family domain that is also present in MHBH (residues 33 to 417), indicating it to be a FAD-dependent hydroxylase (monooxygenase) (residues 31 to 401) (Fig. 4C) (44). In addition, MocA contains the same FAD-dependent phenol hydroxylase (PHOX) family domain (residues 461 to 617) as MHBH (residues 470 to 636), with a C-terminal thioredoxin (TRX) fold domain, indicating it to be a flavoprotein monooxygenase that can catalyze the hydroxylation of phenol and phenol derivatives in the ortho position (Fig. 4C) (44).

FIG 4.

Binding mode of MocA with E1. (A) (Panel a) Predicted three-dimensional structure of the E1-MocA complex. The FAD and E1 compounds are rendered in yellow and green, respectively. (Panel b) Detailed binding site and surrounding amino acid residues of the E1-MocA complex. The backbone of the protein is rendered as a tube and colored gray. Yellow dashes represent the hydrogen bond distance or π-stacking distance. (B) Three-dimensional structure of MHBH. (C) Three-dimensional structure comparison of MocA and MHBH.

Moreover, to further determine the function of MocA, we analyzed the individual interactions between MocA and the ligand E1 by computational molecular simulations (Fig. 4A, panel b). The binding energy score was −7.96 kcal/mol for E1 docking to MocA, indicating strong binding and a high degree of matching with E1. The amino acid residues surrounding the E1-MocA complex included Tyr-271, Leu-258, Arg-269, etc. Among them, E1 showed several hydrophobic amino acid residues in its cavity (e.g., Ile-260 and Lys-247), enabling its binding to the active site where catalysis occurs. The hydrogen bond donor and acceptor groups of E1 could form hydrogen-bonding interactions with Tyr-271, Leu-258, Asp-75, and Gly-76 of MocA to improve the stability of the E1-MocA complex. The geometrical arrangement of residues forming the substrate-binding pocket is similar to that in MHBH (44). Tyr-271 may play a crucial role in orienting the substrate for attack at the ortho location; this residue in MocA corresponds to Tyr-271 in MHBH, which forms a hydrogen bond via the hydroxyl moiety of the phenol. In addition, in MocA, the Asp-75 residue also forms a hydrogen bond network that is connected to the solvent region through the side chain of Arg-269, corresponding to Asp-75 and Arg-262 residues in MHBH, respectively. It was hypothesized that the hydrogen bond network would be involved is in proton transfer for substrate deprotonation. These results reveal the active site and binding mode of MocA for E1, further supporting the idea that MocA is an enzyme involved in E1 degradation and providing a useful reference for further validation of MocA function.

Gene expression, purification, and functional analysis of MocA.

To confirm the function of MocA in E1 degradation, the mocA gene was expressed in Escherichia coli BL21(DE3). In the presence of IPTG (isopropyl-β-d-thiogalactopyranoside), a single approximately 68-kDa protein band was observed by SDS-PAGE, which corresponded to the predicted size (68.068 kDa) calculated from the 3-hydroxybenzoate 4-monooxygenase amino acid sequence (ExPASy, ProtParam) (see Fig. S5A-a in the supplemental material). Western blot analysis also illustrated that recombinant MocA reacted positively with the His antibody against the polyhistidine affinity tag (catalog no. D191001 and D110087; Sangon Biotech, Shanghai, China) (see Fig. S5A-a in the supplemental material). The purified MocA protein showed 46.7% degradation of 30 mg/L E1 at 30 min, 95.8% degradation at 180 min, and complete degradation at 240 min (see Fig. S5A-b in the supplemental material). Further characterization of metabolites generated from E1 degradation using purified MocA protein was performed by LC-MS. The metabolite with a molecular ion peak at m/z 271.1686 was E1, and a fragment at m/z 287.1652 (addition of OH) detected by MS was confirmed to be 4-OHE1 (see Fig. S5A-c in the supplemental material). This result corresponds to the first step of our proposed E1 degradation pathway. Furthermore, we investigated the effects of temperature, pH, metal salts, and diverse agents on the relative activity of MocA (see Fig. S5B in the supplemental material). Based on these findings, MocA was identified as a monooxygenase capable of transforming E1 to 4-OHE1 via hydroxylation.

Flavoprotein monooxygenase is especially well known for its role in the metabolism of foreign substances and the degradation of environmental toxins. In addition to catalyzing hydroxylation, flavoprotein monooxygenase can also catalyze epoxidation and halogenation reactions. In addition, it can produce valuable compounds that are of great significance in the chemical, pharmaceutical, and agrochemical industries (46). Identification of the E1-hydroxylating 3-hydroxybenzoate 4-monooxygenase (a single-component flavoprotein monooxygenase) MocA improves the understanding of the properties of flavoprotein monooxygenase family members and provides a basis for the utilization of enzyme resources.

Knockout and complementation of the mocA gene in M. oxydans ML-6.

The first step in the E1 catabolic pathway, hydroxylation at the C-4 site on the A ring to generate 4-OHE1, is the rate-limiting step in the entire E1 catabolic process. To investigate the role of mocA in the initial hydroxylation of E1, the mutant strain ML-6ΔmocA was successfully constructed by deleting the mocA gene (see Fig. S6A and B and Table S4 in the supplemental material), and its capacity to metabolize E1 was then analyzed. Compared with the wild-type strain, the mutant strain ML-6ΔmocA exhibited deficient growth and reduced E1 degradation ability. The E1 degradation efficiency decreased from 82.5% (strain ML-6) to 16.8% (strain ML-6ΔmocA) within 72 h, and introduction of the pNit-mocA plasmid restored the capacity of the mutant strain ML-6ΔmocA to degrade E1 (see Fig. S6C in the supplemental material). This result suggested that MocA is a pivotal enzyme for E1 catabolism in strain ML-6, i.e., E1 was transformed to 4-OH-E1 by MocA.

In addition, by combining the Plackett-Burman design (PBD) results and contour line analysis (see Fig. S7A and B in the supplemental material), the optimal culture conditions for strain ML-6 were determined to be as follows: E1 concentration, 30 mg/L; rotation speed, 175 rpm; temperature, 30°C; inoculum size, 2% (vol/vol); and NaCl concentration, 2.58 g/L. Under the optimal culture conditions, strain ML-6 degraded 30 mg/L E1 by up to 84.3% within 72 h (Fig. S7C). In fact, the E1 concentrations found in contaminated water in the environment were 0.1 to 720 ng/L (6, 7). To explore whether strain ML-6 could reduce the concentration of estrone below the PNEC (6 ng/L), we performed E1 degradation experiments with a series of environmental concentrations (10 ng/L, 100 ng/L, and 1,000 ng/L). The results showed that strain ML-6 could degrade E1 below the PNEC at in situ concentrations, indicating the high environmental bioremediation capacity of strain ML-6 (see Fig. S8 in the supplemental material). Moreover, as shown in Fig. S7D, approximately 58.1% total organic carbon (TOC) removal was observed with strain ML-6 after 120 h, which indicates the carbon dioxide generated by the oxidization of all organic matter, including E1 degradation and assimilation of carbon into bacterial biomass (47). The differences in degradation efficiency between strain ML-6 and other reported E1-degrading strains are shown in Table S5 in the supplemental material. The results indicated that the degradation efficiency of strain ML-6 was higher than that of P. putida SJTE-1, Rhodococcus sp. strain BH2-1, and Novosphingobium sp. strain ES2-1 but lower than that of Rhodococcus sp. strain B50 and Sphingomonas sp. strain KC8 (20–22, 31, 34). Furthermore, based on the published reports, in Rhodococcus sp. strain B50, the key E1-degrading enzyme responsible for the initial hydroxylation of E1, cytochrome P450 monooxygenase (AedA), was identified through gene knockout and was different from the enzyme identified in this study. In Sphingomonas sp. strain KC8, flavoprotein monooxygenase (OecB) was the key enzyme responsible for the initial hydroxylation of E1 based on transcriptomic analysis; nevertheless, the function of OecB has not been confirmed.

Phytotoxicity assessment of E1 solutions treated with M. oxydans ML-6.

The morphological changes and related indexes of E1 and E1 solutions treated with strain ML-6 on germinated wheat seeds are shown in Fig. 5 and in Table S6 in the supplemental material. The germination indexes (GI) of wheat seeds treated with distilled water (CK group) (100%) and those treated with E1 solutions with strain ML-6 (T1 group) (77.0%) were higher than those of the seeds treated with E1-only solutions (T2 group) (4.17%). It has been reported that the toxicity of endocrine-disrupting chemicals (EDCs) can be reduced by biological treatment (48), which is similar to the results obtained in the present study showing that E1 solutions treated with strain ML-6 significantly reduced the toxicity of E1 and its by-products to wheat seeds (49).

FIG 5.

Phytotoxicity of E1 and E1 solutions treated with strain ML-6 on germinated wheat seeds.

Conclusion.

In this study, we demonstrated that M. oxydans ML-6 was effective at E1 degradation, and the degradation kinetics further showed that strain ML-6 also has the ability to degrade SEs, including E2, E3, and EE2. Degradation metabolite analyses suggested that strain ML-6 catabolized the conversion of the E1 intermediates to HIP via the classic 4,5-seco pathway. Mining of the whole genome and transcriptome of strain ML-6 revealed a novel E1-degrading gene cluster (moc). The 3-hydroxybenzoate 4-monooxygenase encoded by the mocA gene was found to be necessary and specific for the initial hydroxylation of E1 to generate 4-OHE1, which was verified through heterologous expression, gene knockout, and complementation. The enzymes that catalyze the other steps in the complete E1-degrading pathway remain putative, and further experimental evidence is needed. Additionally, significantly reduced toxicity of E1 solutions treated with ML-6 was observed. The findings reported here provide a notable candidate for the biodegradation of E1 in E1-contaminated environments, shed new light on the microbial mechanism of E1 catabolism, and provided a theoretical basis for future enzyme resource applications.

MATERIALS AND METHODS

Strains, plasmids, chemicals, and culture media.

M. oxydans ML-6 (GenBank accession no. MW007985.1) was isolated from an experimental farm of Northeast Agricultural University (NEAU), Acheng District, Harbin, Heilongjiang Province, China. The details of the isolation and identification of strain ML-6 are provided in the supplemental material. E1 (C₁₈H₂₂O₂; molecular weight [MW] = 270.37, >98%), E2 (C₁₈H₂₄O₂; MW = 272.38, 98%), E3 (C₁₈H₂₄O₃; MW = 288.38, 98%), and EE2 (C₂₀H₂₄O₂; MW = 296.40; ≥98%) were purchased from Aladdin (Shanghai, China). The enzymes required in the molecular biological experiments were acquired from TaKaRa Biotechnology (Dalian, China). The E1 stock solution was prepared in HPLC-grade methanol and used for experimental purposes. Table 2 shows all the strains and plasmids utilized in this study. The composition of the modified mineral salt medium (MMSM) was as follows (in grams per liter): 0.04 CaSO₄, 0.2 MgSO₄·7H₂O, 0.001 FeSO₄·7H₂O, 2.58 NaCl, 0.1 K₂HPO₄, and 0.1 (NH₄)₂SO₄ at pH 6.5 to 7.0.

TABLE 2.

Strains and plasmids used in this study

Strain or plasmid	Description	Source
Strains
M. oxydans ML-6	E1 utilizer, wild type	This study
M. oxydans ML-6ΔmocA	E1 mutant with mocA deleted	This study
E. coli DH5α	supE44 lacU169(ϕ80lacZAM15) recAl endA1 hsdR17 thi-1 gyrA96 relAl	TaKaRa Biotechnology (Dalian, China)
E. coli BL21(DE3)	F− ompT hsdS(rB− mB−) gal dcm lacYI(DE3)	TaKaRa Biotechnology (Dalian, China)
Plasmids
pUcm-T	Clone vector, Ampr	TaKaRa Biotechnology (Shanghai, China)
pMD18-T	Clone vector, Ampr	TaKaRa Biotechnology (Dalian, China)
pET-RFP	Expression vector, Cmr	Li Ming (Tianjin University of Science and Technology, China)
pNit-QC2	Complementation vector, Cmr	Tomohiro Tamura (National Institute of Advanced Industrial Science and Technology, Japan)
pET-28a(+)	Expression vector, Kanr	Sangon Biotechnology (Shanghai, China)
pUT-mocA	pPUcm-T-Simple derivative carrying mocA, Ampr	This study
pUT-CmR	pPUcm-T-Simple derivative carrying the Cmr gene, Cmr and Ampr	This study
pET-mocA	ApaI fragment containing mocA inserted into pET-28a(+)	This study
pMT-mocA	pMD18-T-Simple derivative carrying mocA, Kanr	This study
pMT-KanR	pMD18-T-Simple derivative carrying the Kanr gene	This study
pMT-mocA1-CmR-mocA2	pMD18-T-Simple derivative carrying the recombinant fragment mocA1-Cmr-mocA, Kanr	This study
pUC-mocA-Kana-mocA	pUC57-Simple derivative carrying the recombinant fragment mocA-Cmr-mocA, Ampr	This study

The E. coli strains DH5α and BL21(DE3) were incubated at 37°C in LB medium (5 g/L yeast extract, 10 g/L tryptone, and 10 g/L NaCl) with a selective concentration of kanamycin (Kan) (70 mg/L) or ampicillin (Amp) (70 mg/L) for prokaryotic expression or gene knockout, respectively, according to the preliminary experimental results (data not shown).

E1 quantification and biomass measurement for M. oxydans ML-6.

The optimization of cultivation conditions for strain ML-6 was performed by Plackett-Burman design (PBD) and central composite design (CCD). The methods for optimization of cultivation conditions are detailed in the supplemental material. Under the optimal culture conditions, the residual concentration of E1 was determined by HPLC (1260 series; Agilent Technologies, Palo Alto, CA, USA) (50), and the TOC content was measured using a TOC analyzer (N5-665-N; Analytik, Jena, Germany). Detailed E1 quantification and biomass measurement for strain ML-6 are available in the supplemental material. The cell density was measured as the optical density at 600 nm (OD₆₀₀) using a spectrophotometer.

Kinetics of SE degradation by M. oxydans ML-6.

The degradation kinetics experiments were carried out with 1, 10, and 30 mg/L SEs (E1, E2, E3, and EE2) as target substrates. The construction of the E2, E3, and EE2 degradation systems was as described for E1. The residual concentrations of E1, E2, E3, and EE2 were determined by HPLC every 12 h (51, 52). See the supplemental material for degradation system construction and HPLC detection methods.

Identification of metabolites during E1 biodegradation by M. oxydans ML-6.

The intermediate products formed during E1 degradation by strain ML-6 were analyzed using LC-MS/MS. The preparation process for sample collection for LC-MS was as described for HPLC in the supplemental material (53). The collected samples were extracted using equal volumes of ethyl acetate, and detailed methods are described in the report by Wu et al. (34). The samples were analyzed using a Thermo Finnigan LCQ Deca LC/MSN system (ESI source). The temperature in both ionizations (positive and negative) was set at 350°C. The dry gas (nitrogen) flow rate was 1 L/min, the atomizer gas pressure was 25 lb/in², and the mass ranged from 80 to 1,500 m/z. The obtained experimental data were analyzed with the Xcalibur data analysis system (Thermo Finnigan, Massachusetts, USA) (54). The intermediate compounds formed during E1 degradation by ML-6 were analyzed under the following conditions: first, full-scan mass spectrometry was performed simultaneously in negative- and positive-ion modes. The base peak corresponding to the preset intensity threshold was then chosen from the mass spectrum. At a relative collision energy of 40% or 50%, MS/MS scanning mass spectrograms with selected peaks in an appropriate mass range were obtained (55).

Whole-genome sequencing and analysis of M. oxydans ML-6.

DNA extraction, sequencing, and genome assembly of strain ML-6 were performed in collaboration with Nanjing Personal Biotechnology Co. (Nanjing, China) as previously described (56). Gene prediction for the genome of the strain ML-6 was conducted, and annotation was performed against the GO, COG, Nr, Swiss-Prot, and KEGG databases (57). Genes related to basic metabolism and E1 degradation in the ML-6 genome were identified.

The genomic features of strain ML-6 were compared with those of other E1-degrading strains and Microbacterium spp. for which whole-genome sequencing data were available. The BLAST tool was used for amino acid sequence identity analysis of E1-degrading enzymes. The evolutionary analysis of strain ML-6 was performed by a phylogenetic study (58). The E1 degradation genes in the genomes of the above-mentioned E1-degrading bacteria and Microbacterium spp. were compared with those of strain ML-6.

Transcriptomics of M. oxydans ML-6 and qRT-PCR analyses of key genes.

Strain ML-6 was cultured on MMSM containing 30 mg/L E1 or glucose for 8 h or 12 h. All samples for RNA sequencing and analysis were centrifuged at 6,869 × g and 4°C for 7 min, rapidly frozen with liquid nitrogen, and preserved at −80°C (59). RNA was extracted, sequenced, and analyzed by Nanjing Personal Biotechnology Co. (Nanjing, China). Briefly, RNA was extracted with TRIzol reagent (Invitrogen, USA), followed by RNA library construction using an Illumina TruSeq RNA sample preparation kit. Analysis of differences in expression was performed with edgeR (59). The expression levels of crucial E1-degrading genes in strain ML-6 were further examined by qRT-PCR based on the transcriptomic results. The total RNA extraction steps were executed following the methods of Wang et al. (60). The methods for qRT-PCR analyses of key genes are detailed in the supplemental material. The differences in relative mRNA expression levels were calculated using the approach described by Zhao et al. (61). Table 3 lists all the qRT-PCR primers utilized.

TABLE 3.

Primers used in this study

Gene target	Sequence (5′→3′)	Purpose
mocA	F: ATCGCGCTGTTAGCGTCAGCGCGCGGGGAGCAGC	Used to amplify mocA to express it in ML-6
	R: AGACAGAACTTAATGATGCAGTTCCACCACCACGG
mocA1	F: GAGCTCGGTACCCGGGGATCATGCAGTTCCACCACCACGG	Used to amplify mocA to knock it out in ML-6
	R: TTTCTCCATTAGCGGATGCCCACCTCGAC
Cmr gene	F: GCATCCGCTAATGGAGAAAAAAATCACTGG	Used to amplify the Cmr gene to knock it out in ML-6
	R: CATCGACGAGTTACGCCCCGCCCTGCCACTCAT
mocA2	F: CGGGGCGTAACTCGTCGATGATCACCGGAA	Used to amplify mocA to knock it out in ML-6
	R: CAGGTCGACTCTAGAGGATCAGCAGCGCCCCGTCGAAGAA
16S rRNA	F: GACGGTAACCGGAGAAGAAG	Used as an internal standard for qRT-PCR
	R: CGTGGACTACCAGGGTATCT
ML6_01047	F: CCTTCGTGGGGATGTGGGTCTG	Used to amplify ML6_01047 for qRT-PCR
	R: CCATGTCGATCAGCAGCCAGTC
ML6_01048	F: GGCGAACAAGTTCCACCAGCAC	Used to amplify ML6_01048 for qRT-PCR
	R: GTCCGTAATCGCACAGCACCTC
ML6_01049	F: ACGCACGAGAAGCACAACATCC	Used to amplify ML6_01049 for qRT-PCR
	R: TCGCGCAGGTAGAGGTAGAACG
ML6_01050	F: GGGCGTCTTCAACCTCGTCAAC	Used to amplify ML6_01050 for qRT-PCR
	R: GCGTTGCCGAAGATGATCTGTCC
ML6_01052	F: ACCTCTCGCAGCACTTCACTCTC	Used to amplify ML6_01052 for qRT-PCR
	R: CCACCTCGACCTCGACGACATC
ML6_01055	F: CGTGTACCAGCAGCCGTTCG	Used to amplify ML6_01055 for qRT-PCR
	R: AGATGTCGGTCTCGGTCCACTTG

Prediction of the interaction between E1 and the MocA protein in M. oxydans ML-6.

The protein structure of MocA was processed with Maestro 11.9 (Schrodinger, USA). Energy minimization and geometric structure optimization of the MocA protein were performed using the Schrödinger Protein Preparation Wizard (62). The method adopts grid-based ligand docking and uses a hierarchical filtering sequence to generate the ligand conformation of the protein receptor in the active region (63). All chemical compounds were organized following the default settings of the LigPrep module. For screening in the Glide module, the prepared receptors were transferred, and their active sites were predicted for the active site located in the generated receptor grid generation (64). Finally, molecular docking and screening of MocA were performed by the SP process. The pattern of the binding of compounds to proteins was visualized using PyMOL 2.1 software.

Gene expression and purification of recombinant MocA in M. oxydans ML-6.

Expression of the mocA gene and purification of recombinant MocA protein were performed according to the method of Hou et al. (65) with slight modification, and the primers used in this study are shown in Table 3. The mocA gene was amplified and cloned into the vector pET-28a(+), and then, the recombinant vector pET-mocA, which was cleaved by ApaI, was transferred into the receptor E. coli BL21(DE3). IPTG (0.4 mM) was utilized to induce recombinant MocA expression when the OD₆₀₀ of strain DE3 was approximately 0.5 to 0.6. The molecular mass of the MocA protein was determined by SDS-PAGE (66), and the result was further confirmed by Western blotting (67). And the detailed biochemical characterization of the recombinant protein MocA are shown in the supplemental material.

To verify the function of the purified protein, 1 mL of the MocA protein (1.01 mg/mL) was first added to 9 mL of phosphate-buffered saline (PBS; 50 mM PBS, pH 7.2: 68.5 mM NaCl, 1.35 mM KCl, 2.15 mM Na₂HPO₄, 0.7 mM KH₂PO₄) containing 30 mg/L E1 to react at 30°C for 4 h. After the reaction was completed, 10 mL of methanol was added to homogenize E1 as described by Li et al. (17). The sample was analyzed via HPLC and LC-MS/MS (55).

mocA gene knockout and complementation in M. oxydans ML-6.

The knockout strain was constructed by homologous recombination using the linearized pUC19 vector in the In-Fusion Snap Assembly cloning kit (TaKaRa Biotechnology, Dalian, China) (68). First, amplification of the mocA gene was performed by PCR using the primers mocA-F/-R with ML-6 genomic DNA as the template. Second, the Cm^r gene was amplified by PCR using the pET-RFP plasmid as the template. Then, the mocA1-Cm^r-mocA2 fragments were obtained and ligated to pUC19-Linearized-Control-Vector using the In-Fusion Snap Assembly cloning kit. The plasmid was selected by resistance to 70 mg/L ampicillin. Finally, the linear fragments of mocA1-Cm^r-mocA2 were cloned from the pUC-mocA1-Cm^r-mocA2 plasmid with the primers mocA-F/-R and introduced into ML-6 competent cells by an electroporator (Gene Pulser Xcell; Bio-Rad, USA) at 1.8 kV, 400 Ω, and 25 μF. The knockout strain was selected through chloramphenicol resistance at a concentration of 50 mg/L, and the fragments of the Cm^r gene were verified by the primers CmR-F/R with knockout strain genomic DNA as the template. Detailed primer information is provided in Table 3. The complementation strain ML-6ΔmocA(pNit-mocA) was obtained by ligating the PCR-amplified mocA gene into HindIII-digested pNit-QC2 (69). The abilities of the mutant strain ML-6ΔmocA and complementation strain ML-6ΔmocA(pNit-mocA) to catabolize E1 were determined by inspecting cell growth and substrate degradation and validated by the method used for the wild-type strain ML-6.

Phytotoxicity assessment.

The phytotoxicity test of E1 solutions (30 mg/L) treated with strain ML-6 was carried out using the wheat seed germination method to evaluate the detoxification effect (48). This experiment was conducted in the dark for 7 days, and parameters such as seed growth and root length were measured. The GI was calculated by the method described by Ademakinwa (48). We set up 3 groups of tests with 20 seeds in each group: distilled water (CK), E1 solutions treated with strain ML-6 (T1), and E1-only solutions (T2).

Statistical analysis.

All experiments, including those for the control group, were carried out in triplicate, and the results are expressed as means and standard deviations (SD). One-way analysis of variance (ANOVA) was used to compare data between different groups with significance at P values of <0.05 or <0.001 (70). The relevant data in this study were analyzed by SPSS software 22.0 (71).

Data availability.

The whole-genome (accession number CP092891.1) and the 16S rRNA gene (accession number MW007985.1) of strain ML-6 have been deposited in GenBank. The data supporting transcriptomics are available in SRA under accession numbers SRR23194572, SRR23194573, and SRR23194574 for E1-cultured cells at 8 h, SRR23194575, SRR23194576, and SRR23194577 for glucose-cultured cells at 8 h, SRR23194578, SRR23194579, and SRR23194580 for E1-cultured cells at 12 h, and SRR23194581, SRR23194582, and SRR23194583 for glucose-cultured cells at 12 h.