Environmental estrogens (EEs) are one of the most important environmental contaminants with great disorder effect and sex influence, and can cause serious environment problems^[1-2]. Typical natural estrogens include 17β-estradiol (E2), estrone (E1), and estriol (E3); 17β-estradiol exhibits the highest estrogenic activities^[3]. Bioremediation using microorganisms are considered as the efficient way to eliminate the EEs pollution because of its high efficiency, low cost and little secondary pollution^[4-5]. Microorganisms can utilize estrogens as carbon source and degrade them into non-estrogenic compounds^[5]. Series of estrogen- degrading strains have been isolated from activated sludge and farm composting, such as Sphingomonas spp., Rhodococcus zopfii, Rhodococcus equi, Nitrosomonas europaea and Pseudomonas spp.^[6-8]. Some fungi and algae were also found with the estrogen-degrading capabilities^[9-11].

In bacteria, 17β-estradiol oxidization was considered as the first step for E2 biodegradation; therefore, enzyme catalyzing this reaction is the key enzyme for E2 transformation. In human, hydroxysteroid dehydrogenases (HSDs) belonging to the short-chain dehydrogenase/reductase (SDR) family, catalyze the conversion of inactive/active forms of estrogens by NAD(P)H-linked oxidization/ reduction of steroid molecules^[12]. 3, 17β-HSD and 17β-HSD in Comamonas testosterone have been proved to catalyze the oxidoreduction at the C₃ site of testosteroneand at C₁₇ site of 17β-estradiol^{[8, 13-14]}. Several regulators (Rep1 and Rep2, HsdR, PhaR, TetR, LuxR, and BRP protein) have been found to participate in the degradation of testosterone in C. testosteroni by regulating the expression of 3α-hsd or 3, 17β-hsd genes^[15-19]. However, up to now, only a few enzymes involved in the bacterial degradation pathway of 17β-estradiol were identified; the enzymatic characteristics and their regulation factors were not very clear.

Pseudomonas putida SJTE-1 (CGMCC No. 6585) has been confirmed with efficient estrogen-degrading capability; its whole genome sequence and comparative proteomics have been obtained and analyzed^{[6, 20]}. In 17β-estradiol environment, significant expression changes of many proteins involved in carbon catabolism, cellular transportation and transcriptional regulation have been observed^[20]. In this work, a novel 17β-HSD (17β-HSD 2) was characterized responsible for E2 oxidization in P. putida SJTE-1, and one AraC regulator was identified to participate in the regulation of E2 degradation in this strain.

1 Materials and methods 1.1 Strains, chemicals and cultures

All strains and plasmids used in this work were listed in Table 1. E. coli strains were cultured at 37 ℃ and P. putida SJTE-1 was cultured at 30 ℃. Luria-Bertani (LB) medium (tryptone 10.0 g, yeast extract 5.0 g, NaCl 10.0 g/L, pH 7.0) and minimal medium (MM) (K₂HPO₄ 3.815 g, KH₂PO₄ 0.5 g, (NH₄)₂HPO₄ 0.825 g, KNO₃ 1.2625 g, Na₂SO₄ 0.2 g, CaCl₂ 0.02 g, FeCl₃ 0.002 g, MgCl₂ 0.02 g/L, pH 7.0) were used for strains culture. 17β-Estradiol, estrone, estriol, 17α-ethinyl estradiol, testosterone, phenanthrene (of HPLC grade, > 98.5%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). All other reagents were of analytical reagent grade. Estrogenic compounds were dissolved in dimethyl sulfoxide (DMSO) to form 10 mg/mL solution as DMSO was not utilized by strains.

1.2 Standard DNA manipulation

Oligonucleotides used for PCR amplification, reverse transcription (RT), quantitative PCR (q-PCR), homologous recombination, plasmid construction, electrophoretic mobility shift assay (EMSA) and DNase I footprinting were synthesized at Invitrogen Ltd. (Shanghai, China) and listed in Table 2. Restriction endonucleases and DNA-modifying enzymes were purchased from TaKaRa Biocompany (Dalian, China). The PCR products were recovered with the QIAquick gel extraction kit (Qiagen, Shanghai, China). Plasmid DNAs were isolated using the QIAprep Mini-spin kit (Qiagen, Shanghai, China), and genomic DNA was obtained using the QIAamp DNA minikit (Qiagen, Shanghai, China). All the constructed plasmids were confirmed by DNA sequencing (Invitrogen, Shanghai, China). Other general techniques for restriction endonuclease manipulation, agarose gel electrophoresis, SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and native PAGE were carried out with standard protocols.

1.3 Multiple sequences alignment (MSA) and homology modelling

Multiple sequences alignment of 17β-HSD proteins and AraC proteins from P. putida SJTE-1 and other organisms were performed with Clustal W software. The structure conformation of 17β-HSD2 protein was constructed by homology-modelling method achieved in SWISS-MODEL, with the short-chain dehydrogenase/reductase (5jla.1.A) from Homo sapiens as template.

1.4 Reverse transcription and quantitative PCR (RT-qPCR)

The transcription profiles of target genes (17β- hsd2 and araC) in P. putida SJTE-1 cultured with different carbon sources were detected with RT-qPCR. Strain SJTE-1 was cultured in the minimal medium with 0.2% glucose, 10 mg/L or 20 mg/L different steroid compounds (17β-estradiol, estrone, estriol, 17α-ethinyl estradiol, testosterone) as sole carbon source to the mid-exponential phase; and the total RNA was extracted using the Total RNA Extraction Reagents (Vazyme, Nanjing, China) according to the protocol. The yield of RNA was estimated using a Nanodrop UV spectrometer (Thermo Scientific, DE, USA). Reverse transcription was achieved with 1 μg RNA and 20 ng random primers using the PrimeScript Reverse Transcriptase Kit (TaKaRa, Dalian, China). The quantitative PCR was performed using the Premix Ex Taq and the gene-specific primers (Table 2) in the IQTM 5 Multicolor Real-time PCR Detection System (Bio-Rad, CA, USA). At the meanwhile, the transcription profiles of target genes (17β-hsd2 and araC) in P. putida SJTE-1 transformed with plasmids pBS-araC cultured with different carbon sources were also detected. The strain was cultured in MM meidium with 0.2% glucose or 20 mg/L E2 to the mid-exponential phase, and 0.1 mmol/L isopropyl β-D-1-thiogalactopyranoside (IPTG) was supplied. The RNA extraction and RT-qPCR detection were performed as above. The relative fold change in mRNA quantity was calculated using the DDCt method. Five independent experiments were performed for each RNA sample and the average values with the standard errors were calculated.

1.5 Heterogeneous expression and affinity purification of recombinant proteins

The recombinant 17β-HSD2/AraC proteins were expressed and purified as described before^[22]. Plasmids pET28a-hsd2/pET28a-araC were transformed into the competent E. coli BL21 (DE3) cells. Single colony was inoculated in LB media with 50 μg/mL kanamycin and cultured at 37 ℃ overnight. The culture were transferred into fresh LB with 50 μg/mL kanamycin and cultured to OD₆₀₀=0.5. 0.1 mmol/L IPTG was added and cells were induced at 37 ℃ for 3 h. Cells were harvested and resuspended with ice-cold lysis buffer [20 mmol/L Tris-HCl, 300 mmol/L NaCl, 5 mmol/L imidazole, 5 mmol/L β-mercaptoethanol (β-ME), 1 mmol/L phenylmethanesulfonyl fluoride (PMSF), 10% glycerol, pH 7.9]. After sonication (300 W, 10 s/ 10 s, 20 min), the supernatant was obtained and loaded into affinity chromatography of Ni-NTA resin (Bio-Rad, CA, USA) at 4 ℃. After washed with washing buffer (20 mmol/L Tris-HCl, 300 mmol/L NaCl, 5 mmol/L β-ME, 10% glycerol, 1 mmol/L PMSF, and 50 mmol/L imidazole pH 7.9), the recombinant protein was eluted from the column using elution buffer (20 mmol/L Tris-HCl, 300 mmol/L NaCl, 5 mmol/L β-ME, 10 % glycerol, 1 mmol/L PMSF, 200 mmol/L imidazole, pH 7.9). All eluted solutions were analyzed by SDS-PAGE. The eluted recombinant proteins were dialyzed and stored in the storage buffer (20 mmol/L Tris-HCl, 50 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L DTT, 50% glycerol, pH 8.0) or 1× PBS buffer (Na₂HPO₄ 1.42 g/L, KH₂PO₄ 0.27 g/L KCl 0.2 g/L, NaCl 8.0 g/L, pH 7.4) at –80 ℃. The concentration of proteins was determined using a BCA protein assay kit (TaKaRa, Dalian, China).

1.6 Enzymatic assay of 17β-HSD2

The enzymatic activity of recombinant 17β-HSD2 protein was measured by detecting the yield of NADH at 355 nm (excitation) /460 nm (emission)^[17]. The reaction system (20 μL) contained 20 μmol/L NAD⁺, and different concentrations of 17β-HSD2 protein and steroids (0–0.5 μmol/L of 17β-HSD2, 0-0.5 μmol/L of 17β-estradiol, estrone, estriol, 17α-ethinyl estradiol, testosterone, phenanthrene) in reaction buffer (20 mmol/L Tris-HCl, 25 mmol/L NaCl). The samples without 17β-HSD2 or steroids were used as blank controls. The enzymatic parameters of 17β-HSD2 (K_m, V_max, optimal pH and temperature) were determined. Effect of metal ions (Mg²⁺, Mn²⁺, Cu²⁺, Ca²⁺, Zn²⁺, and Ni²⁺) was analyzed. One unit of enzyme activity is defined as the amount of enzyme catalyzing the oxidation of 1 μmol of E2 per min under the above-indicated conditions at 25 ℃. All the experiments were repeated five times, and results were the average values with standard errors.

1.7 High Performance Liquid Chromatography (HPLC) detection of enzymatic products

The solutions of enzymatic reaction were extracted with ethyl acetate, dried in the nitrogen blowing apparatus, dissolved with acetonitrile, and detected with HPLC system (Agilent 1260 Infinity LC, US). The Agilent exclipse plus C18 column (3.5 μm, 4.6 mm×150 mm, Agilent, US) and Diode Array Detector (DAD) detector were used with the mobile phase of acetonitrile and water (1:1 V/V) at a flow rate of 1 mL/min at 30 ℃, and the UV wavelengths of 280 nm and 315 nm. The quantities of estrogens were calculated from their respective peak areas by using a standard curve of individual standards; the R² values for the standard curves were > 0.99. Five independent experiments were performed and results were calculated as average values with standard errors.

1.8 Green fluorescent protein (GFP) fluorescence assay

Plasmid pBS-P_hsd2-eGFP (Table 1) was transformed into E. coli BL21(DE3) strain and the recombinant strain was cultured in the minimal medium with 0.2% glucose or 10 mg/L 17β-estradiol. The cells were collected at different time points, and the fluorescence values of eGFP were measured with excitation at 485 nm and emission at 527 nm. Five independent experiments were performed, and the average values with the standard errors were calculated.

1.9 Electrophoretic mobility shift assay (EMSA)

The DNA fragments were amplified from the upstream region of 17β-hsd2 gene with 5'-FAM labeled primers (Table 2). The recombinant AraC protein and DNA fragments were mixed in different molecular ratios, and incubated in the binding buffer (20 mmol/L Tris-HCl, 50 mmol/L NaCl, 1 mmol/L DTT, 0.1 mmol/L EDTA, pH 7.5) at 37 ℃ for 30 min. All the input DNA amounts were 4 pmol in 20 μL binding system. The reaction mixture was analyzed with native PAGE, and the gel was photographed in the Bio-Rad Imaging System (Bio-Rad, CA, USA).

1.10 Detection of estrogens effect on protein-DNA interaction

The binding of AraC protein to the upstream fragments of 17β-hsd2 gene was performed as above. 17β-estradiol and estrone (50 pmol or 250 pmol) were supplied into the binding system and DMSO was used as negative control. After 15 min incubation at 37 ℃, the mixture was analyzed with native PAGE and the gel was photographed in BioRad Imaging System (Bio-Rad, CA, US).

1.11 DNase I foot-printing assay

The 1.87 μg of DNA (the upstream fragments of 17β-hsd2 gene) was bound with AraC protein at 25 ℃ for 30 min performed as above. Then 1.33 U of DNase I was added into the binding mixture and incubated at 25 ℃ for 1 min. The DNA fragment without AraC protein addition was used as control. The reaction products were extracted with phenol to remove proteins and ethanol was added to precipitate DNA. The mixture were analyzed with native PAGE and displayed in the BioRad Imaging System (Bio-Rad, CA, US).

2 Results 2.1 Sequence alignments and structure conformation showed that 17β-HSD2 contained the conserved structure of SDR

Based on the analysis of P. putida SJTE-1 genome sequence (CP015876.1), one gene (A210_13000, 2861306..2862133) was predicted as hsd gene and its encoding protein was probably an HSD protein (ANI03512.1). Multiple sequences alignment showed that its secondary structure contained the conserved Rossmann-fold of SDRs, a central twisted parallel β-sheets consisting of 6–7 β-strands flanked by 3–4 α-helices from each side^[23]. The amino acid sequences and the secondary structure of this protein were similar to those of 17β-HSDs from P. aeruginosa PAO1, Rhodococcus sp. P14, C. testosteroni ATCC 11966, and E. coli MG1655 (Figure 1-A). They all contained the two conserved motifs of SDR members, the N-terminal Rossmann-fold Gly-X-X-X-Gly-X-Gly motif for cofactor binding and the Tyr-X-X-X-Lys motif for proton acceptance. The conserved triad motif of HSD, Ser-Tyr-Lys (residues 152, 165, 169), was also found in this protein. Serine at residue 152 probably functioned as the active site, and the highly conserved tyrosine (residue 165) may act as a proton acceptor^[24]. Structure modelling showed that this protein exhibited 39.43% sequence similiarity to the template, the SDR from Homo sapiens (5jla.1.A), and its coverage was 89%. Their structures were quite similar as the local quality estimate of this HSD protein to the template was about 0.8 (Figure 1-B). Therefore, this protein was a 17β-HSD.

2.2 Protein ANI03512.1 could be induced by 17β-estradiol and was a 17β-HSD

The transcription levels of 17β-hsd2 gene in P. putida SJTE-1 cultured with different steroid compounds as sole carbon sources were measured. Results showed when 10 μg/mL or 20 μg/mL 17β-estradiol were used, the transcription of 17β-hsd2 increased about 3.6 or 4.8 folds compared to that in the culture with glucose. Tetosterone could also enhance the transcription of 17β-hsd2 gene about 2.5 folds, while estrone and estriol only had a little induction (Figure 2-A). 17α-ethinyl estradiol and phenanthrene also exhibited a little inductive activity (data not shown). The eGFP fluorescence results showed that the expression of egfp under the 17β-hsd2 promoter showed an increase of about 1.7 folds after 3 h induction with E2, and an increase of about 2.8 folds after 9 h induction (Figure 2-B). This demonstrated that protein ANI03512.1 could be induced by steroidic compounds, especially by 17β-estradiol.

Then the enzymatic properties of protein ANI03512.1 were studied. Its encoding gene was cloned and expressed in E. coli BL21 (DE3), and the purified protein exhibited 29.0 kDa with over 96% purity (Figure 3-A). The yield of recombinant protein was over 11 mg/L and the specific activity of the purified protein for E2 transformation was 25.28 U/mg. Enzymatic assay showed that the recombinant protein used NAD⁺ but not NADP⁺ as its reaction cofactor; 17β-Estradiol, testosterone and estradiol could be oxidized effciently in vitro. Obvious substrate preference for protein ANI03512.1 to 17β-estradiol was observed with efficient transformation efficiency (Table 3). The optimal reaction temperature was 37 ℃ and the optimal reaction pH was 9.0, similar to other HSDs (Figure 3-B, 3-C)^[25]. Mg²⁺ and Mn²⁺ could enhance the reaction efficiency, while Zn²⁺ and Cu²⁺ repressed the enzyme activity greatly (Figure 3-D). The K_m value of protein ANI03512.1 for E2 was 0.0802±0.004 mmol/L and its V_max value was 56.26±0.02 μmol/min·mg; its specificity constant (k_cat/K_m) for E2 was 28.267/s·mmol·L^–1 (Table 4). Further HPLC analysis showed that estrone was the oxidation product of 17β-estradiol catalyzed by protein ANI03512.1 and the transformation efficiency was above 97.4% in five minutes (Figure 4). Therefore, protein ANI03512.1 could be induced by 17β-estradiol and could oxidize 17β-estradiol into estrone in vitro efficiently. It was named as 17β-HSD2.

2.3 AraC could bind to the speicific sites in the promoter region of 17β-hsd2 gene and repress the expression of 17β-HSD2

Genome analysis of P. putida SJTE-1 showed that one araC gene (A210_12995, 2860257. 2861129) encoding a potential regulator AraC (ANI06277.1) located in the upstream of 17β-hsd2 gene. RT-qPCR detection showed that this araC gene could also be induced by 17β-estradiol, dependent on the concentration of E2 (Figure 5-A). To analyze the potential role of AraC to 17β-HSD2, we detected the transcription levels of hsd2 gene in P. putida SJTE-1 with plasmids pBS-araC. IPTG induced the overexpression of AraC greatly, and then the transcription of hsd2 gene was repressed significantly. When E2 was used, the transcription of hsd2 gene was induced and increased; while with IPTG added, overexpressed AraC caused transcriptional repression of hsd2 gene (Figure 5-B). The repression extent of AraC to the transcription of hsd2 gene decreased significantly in E2 environment compared to that in glucose environment, even the expression of AraC induced greatly with IPTG; this implied that E2 could release the binding of AraC and enhance the transcription of hsd2 gene. Multiple sequences alignment showed that the secondary structure of AraC protein in P. putida SJTE-1 was similar to that of other AraC regulators, with the conserved N-terminal HTH (helix-turn-helix) structure and the variable C-terminal (Figure 5-C). These indicated this AraC protein could be induced by 17β-estradiol and may participate in the 17β-estradiol degradation via regulating the transcription of 17β-hsd2.

The binding ability of AraC protein to the upstream fragment of 17β-hsd2 gene was determined. The recombinant AraC protein (30.5 kDa) was obtained by hetergenous expression and affinity purification (Figure 6-A). EMSA detection showed this AraC protein could directly bind to the 176 bp upstream fragment of 17β-hsd2 gene at low protein/DNA ratio, and the binding was in a concentration dependent mode (Figure 6-B). Addition of high-concentration 17β-estradiol or estrone could release the binding of AraC protein from the upstream fragment of 17β-hsd2 gene to some extent; 17β-estradiol of low-concentration had little influence on the AraC binding to DNA (Figure 6-B). Further DNase I foot printing assay showed when AraC protein added, the regions close to its transcription start site in the promoter of 17β-hsd2 gene were covered, probably acting as the specific sites for AraC protein binding (Figure 6-C). This region was around 40 bp upstream of 17β-hsd2 with the sequence ATATGCXXGTTXCCCTGAXX, similar to the conserved binding sites of AraC regulator^[26]. These results demonstrated that AraC protein could directly bind to the specific sites in the promoter region of 17β-hsd2 gene and repress its expression; the substrate (E2) and the product (E1) of 17β-HSD2 could release the binding of AraC and facilitate the transcription of 17β-hsd2 gene.

Biodegradation with microorganisms is considered as efficient strategy to remediate the pollution of environmental estrogens. HSDs were considered as the key enzyme for 17β-estradiol transformation in bacteria^{[12-14, 27]}. Although several 3, 17β-HSDs were found to participate in the oxidization at C₁₇-site of steroid compounds, the key catalyzing enzyme of 17β-estradiol was not really clear. The reason was that the reported 3, 17β-HSDs could not be induced by 17β-estradiol, even the oxidization of 17β-estradiol could be achieved in vitro^{[13, 15, 28]}. Previous work showed that P. putida SJTE-1 could degrade 17β-estradiol efficiently and several hydroxysteroid dehydrogenases were induced in the culture with 17β-estradiol as carbon source^{[6, 20]}. In this work, 17β-HSD2 in P. putida SJTE-1 could oxidize 17β-estradiol into estrone efficiently, whose enzymatic activity was higher than or similar to those of the reported HSDs^{[13, 27]}. The 17β-HSD2 could be sigificantly induced by 17β-estradiol and testosterone, and catalyzed the transformation of two compounds. It implied 17β-HSD2 was probably the key enzyme for estrogen transformation in P. putida SJTE-1 and important for the estradiol degradation of this strain.

At the meanwhile, the regulators involved in the biodegradation of 17β-estradiol in bacteria and their regulation mechnisms were also unclear. PhaR, TetR, LuxR, and BRP protein have been found to regulate the 3, 17β-HSD and influence testosterone utilization in C. testosteroni^[16-19]. However, these regulation was not obvious when 17β-estradiol used; not only because 17β-estradiol was not utilized in C. testosteroni as good as testosterone, but also because this 3, 17β-hsd gene could not be efficiently induced by 17β-estradiol^{[13, 29]}. In this work, AraC regulator was proved to able to bind to the specific sites in the promoter region of 17β-hsd2 gene directly and repress the transcription of 17β-hsd2 gene. Estrogens could induce the expression of AraC, and overexpressed AraC could repress the transcription of 17β-hsd2 gene. The binding of AraC could be released by high concentration of E2 and its repression effect would be reduced to generate more 17β-HSD2 and facilitate the E2 biodegradation. The potential binding region was found in the araC-17β-hsd2 intergenic region, close to the transcription start sites of 17β-hsd2. The sequences of the binding site were ATATGCXXGTTXCCCTG AXX, similar to the conserved binding sites (ATATG CXXGAAAXXXTCCTTA) of reported AraC regulator^[26]. Therefore, AraC protein regulated the expression of 17β-hsd2 gene in a negative mode in P. putida SJTE-1. In a word, this work could promote the estrogen degradation mechanism study and the regulatory network analysis in bacteria.