Identification and Characterization of the LysR-Type Transcriptional Regulator HsdR for Steroid-Inducible Expression of the 3α-Hydroxysteroid Dehydrogenase/Carbonyl Reductase Gene in Comamonas testosteroni

Abstract

3α-Hydroxysteroid dehydrogenase/carbonyl reductase (3α-HSD/CR) from Comamonas testosteroni is a key enzyme in steroid degradation in soil and water. 3α-HSD/CR gene (hsdA) expression can be induced by steroids like testosterone and progesterone. Previously, we have shown that the induction of hsdA expression by steroids is a derepression where steroidal inducers bind to two repressors, RepA and RepB, thereby preventing the blocking of hsdA transcription and translation, respectively (G. Xiong and E. Maser, J. Biol. Chem. 276:9961-9970, 2001; G. Xiong, H. J. Martin, and E. Maser, J. Biol. Chem. 278:47400–47407, 2003). In the present study, a new LysR-type transcriptional factor, HsdR, for 3α-HSD/CR expression in C. testosteroni has been identified. The hsdR gene is located 2.58 kb downstream from hsdA on the C. testosteroni ATCC 11996 chromosome with an orientation opposite that of hsdA. The hsdR gene was cloned and recombinant HsdR protein was produced, as was anti-HsdR polyclonal antibodies. While heterologous transformation systems revealed that HsdR activates the expression of the hsdA gene, electrophoresis mobility shift assays showed that HsdR specifically binds to the hsdA promoter region. Interestingly, the activity of HsdR is dependent on decreased repression by RepA. Furthermore, in vitro binding assays indicated that HsdR can come into contact with RNA polymerase. As expected, an hsdR knockout mutant expressed low levels of 3α-HSD/CR compared to that of wild-type C. testosteroni after testosterone induction. In conclusion, HsdR is a positive transcription factor for the hsdA gene and promotes the induction of 3α-HSD/CR expression in C. testosteroni.

INTRODUCTION

Comamonas testosteroni is a Gram-negative bacterium that belongs to the Betaproteobacteria (34). These strictly aerobic, nonfermentative, chemoorganotrophic bacteria rarely attack sugars but grow well on organic acids and amino acids (2). Moreover, C. testosteroni strains are able to use steroids as the sole carbon source and may be an attractive means for the removal of these stable compounds from the environment. Interestingly, the catabolic enzymes for steroid degradation usually are not constitutively expressed but are induced by their respective steroid substrates (18, 20, 24). Hence, steroids play a particularly important role in certain prokaryotes, as they may simultaneously serve both as signal molecules and as the carbon source.

Since the pioneering work of Talalay and coworkers (18, 33), it is well known that 3α-hydroxysteroid dehydrogenase/carbonyl reductase (3α-HSD/CR) is one of the first enzymes of the steroid-catabolic pathway and therefore plays an important role in steroid metabolism. In previous investigations, 3α-HSD/CR was identified to catalyze the oxidoreduction at position 3 of the steroid nucleus of a variety of C_19-27 steroids (22, 24). Surprisingly, this enzyme also was capable of catalyzing the carbonyl reduction of a variety of nonsteroidal xenobiotic aldehydes and ketones (24). It has been demonstrated that this substrate pluripotency not only enhances the metabolic capacity of insecticide degradation but also increases the resistance of C. testosteroni toward the steroid antibiotic fusidic acid (22, 23).

3α-HSD/CR from C. testosteroni is one of the enzymes whose expression is induced by steroids such as testosterone and progesterone (18, 20, 23, 24), and this is why we were interested in the mode of the molecular regulation of its gene (hsdA). In previous investigations, we identified two genes, repA and repB, that are involved in hsdA regulation, and we reported a two-repressor model to control hsdA gene expression. RepA was identified to specifically bind to both operators Op1 and Op2 and to force the DNA between them to form a loop structure. The two palindromic 10-bp motifs Op1 (TCAAAGCCCA) and Op2 (TGGGCTTTGA), working as cis-acting operator elements for hsdA regulation, were localized at 0.935 and 2.568 kb upstream of hsdA, respectively, while Op2 overlaps the −10 binding site (TTTGAT) of the σ70 RNA polymerase by 5 bp. Thus, the resulting DNA-loop-RepA complex strongly blocks the transcription of the hsdA gene. In the presence of appropriate steroids, however, they bind to RepA, thereby reducing its ability to bind to the operator region (38). Upon the dissociation of RepA from the operators, RNA polymerase may bind to the promoter, and then the transcription of 3α-HSD/CR mRNA is initiated. RepB was demonstrated to bind to the mRNA of 3α-HSD/CR and to interfere with 3α-HSD/CR translation (39). The teiR gene, encoding a positive regulator of steroid-degrading enzymes, including 3α-HSD/CR, was identified to mediate steroid sensing and signaling in C. testosteroni ATCC 11996 via a kinase mechanism (6).

In the present study, a novel regulator (HsdR; for 3α-hydroxysteroid dehydrogenase/carbonyl reductase regulator) for 3α-HSD/CR expression in C. testosteroni has been identified which was recognized as a member of the LysR-type transcriptional regulator family. The LysR-type transcriptional regulator (LTTR) family, formally documented by Henikoff et al. (9), is a well-characterized group of transcriptional regulators. LTTRs are dual-function regulators acting as both autorepressors and activators of target promoters, frequently of genes colocated with the LTTRs in the chromosome (9, 29, 32). The common features of this family comprise sequence lengths of around 300 residues, high sequence similarity at the N-terminal winged helix-turn-helix (wHTH) motif for DNA binding, and a less conserved C-terminal inducer binding domain. Also, with few known exceptions (29), LTTRs act as homotetramers (32). LTTRs regulate the expression of a wide variety of genes, including operons involved in amino acid metabolism, oxidative stress, bacterial virulence, and the degradation of aromatic compounds (29, 32).

Here, the hsdR gene was found to be a positive transcriptional regulator for hsdA expression and locates 2.58 kb downstream of the hsdA gene on the C. testosteroni chromosome with an orientation opposite that of hsdA. Studies with wild-type and hsdR knockout mutant strains confirmed that HsdR is necessary for the induced expression of the hsdA gene. Electrophoretic mobility shift assays (EMSAs) showed that HsdR and RepA can simultaneously bind to different sites of the hsdA promoter region. In addition, HsdR interacts with RNA polymerase, as revealed by HsdR-RNA polymerase binding. The expression of HsdR itself in C. testosteroni is not induced by testosterone. From these results we conclude that HsdR is a positive transcription factor for induced hsdA expression.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

Host strains Escherichia coli HB101 (Promega) and C. testosteroni ATCC 11996 (Deutsche Sammlung für Mikroorganismen) were used for cloning and gene expression. The subcloning of fragments was carried out in plasmids pUC18 (containing the ampicillin resistance gene; obtained from Invitrogen) and plasmid pK18 (containing the kanamycin resistance gene; a gift from Ciba Pharmaceuticals, Inc., Department of Biotechnology, Basel, Switzerland). The plasmid copy numbers determined were 80 copies of pK18 and pUC18 per cell in E. coli. For the overexpression and purification of HsdR, E. coli strain BL21(DE3)pLysS together with plasmid pET15b from Novagen was used. The tac promoter (274 bp) was obtained by BamHI digestion from plasmid pHA10, which was a gift from H. Arai (1). Plasmid pCR2.1-TOPO (Invitrogen) served for the PCR cloning of hsdR fragments and sequencing. Bacterial cells were grown in a shaker (180 rpm) in standard I nutrient broth medium (SIN) (Merck) or LB medium at 37 (E. coli) or 27°C (C. testosteroni). Growth media contained 100 μg/ml ampicillin and/or 30 μg/ml kanamycin when required.

DNA manipulations, sequencing, and reagents.

Recombinant DNA work was carried out by following standard techniques according to Sambrook and Russel (30). All of the primers were prepared by MWG (Ebersberg, Germany). Before further cloning, fragments prepared by PCR were cloned into pCR2.1-TOPO and then checked for correct sequence by MWG. Restriction enzymes, T4 ligase, and shrimp alkaline phosphatase were obtained from Roche Applied Science, New England BioLabs, MBI, Promega, and Amersham Biosciences and were used according to the manufacturers' instructions. Sodium lauroyl sarcosinate was from Fluka. The steroid compounds were supplied by Sigma. Ampicillin and kanamycin were from AppliChem and Calbiochem, respectively. Incomplete Freund's adjuvant was from MP Biomedicals.

Plasmid construction.

The hsdR gene was cloned from C. testosteroni ATCC 11996 chromosomal DNA by using a forward primer containing an NdeI site (5′-CATATGGATTTCAATGCGC-3′) and a reverse primer containing a BamHI site (5′-GGATCCAAGAGCGGTCATGC-3′). The full-length hsdR gene then was cloned into pCR2.1-TOPO to yield plasmid pTOPOHsdR, which, after sequence confirmation (MWG), was used as the template for further PCRs (Fig. 1). To generate hsdR gene constructs that are controlled by the lac or the tac promoter, respectively, plasmid pTOPOHsdR was digested with KpnI and XbaI, and the resulting KpnI-XbaI fragment was ligated into pK18 downstream from the lac promoter to yield pKHsdR3 or downstream from the tac promoter to yield pKtacHsdR1. In addition, pTOPOHsdR was digested by BamHI and NdeI and subcloned into pET15b downstream from the N-terminal His tag coding sequence to yield pETHsdR2, which was used for recombinant HsdR protein production.

Fig 1.

Genetic organization of the hsdA gene and its regulatory elements together with the schematic illustration of the various constructs of the hsdR gene. (A) Two negative regulator genes, repA and repB, are located in the vicinity of the hsdA gene (38, 39), and a positive regulator gene, teiR (7), is located upstream of the hsdA gene. A novel transcriptional regulator, HsdR, encoded by hsdR, of hsdA expression was identified 2.58 kb downstream of the hsdA gene. Open reading frames are indicated as arrows, and restriction sites are abbreviated as one- or two-letter codes (A, AvrII; B, BamHI; Bg, BglII; Bs, BssHI; C, ClaI; E, EcoRI; H, HindIII; K, KpnI; M, MluI; N, NdeI; No, NotI; P, PstI; X, XamI). Op, operator; ksi, gene encoding ketosteroid isomerase. (B) The hsdR gene and fragments thereof are drawn as horizontal lines and were subcloned into vector pK18 (pK), pET15b (pET), or pCR2.1-TOPO (pTOPO). The first four plasmids were used for E. coli transformation, and the plasmid in the bottom line was used for the preparation of an hsdR knockout mutant in C. testosteroni.

To elucidate if HsdR regulates the transcription of the hsdA gene, a series of plasmids constructed in our laboratory were used. As described previously, a 5.257-kb EcoRI fragment of C. testosteroni chromosomal DNA was cloned into pUC18 to yield p6 (38). Plasmid p6 contains the 3α-HSD/CR gene, hsdA, together with its regulatory region and the two repressor genes repA and repB (38, 39). AvrII and XbaI were used for the double digestion of p6 and ligation to yield pAX1 (38). With pAX1 as the template, a shorter derivative, pDel13n, was constructed in which 13 upstream bases critical for HsdR action had been deleted (Fig. 2). To vary the distance between Op1 and Op2, restriction enzymes AvrII and MluI together with PCRs were used to generate mutant plasmids p67, p67-3, p67-5, and p67-7 from p6, which comprised a spacing of 67, 64, 62, and 60 bp between Op1 and Op2 (40), respectively (Fig. 3). The expression of hsdA served as a detection system for HsdR transcriptional regulation in cotransformation experiments with the plasmids described above.

Fig 2.

HsdR activates the expression of 3α-HSD/CR. 3α-HSD/CR (in μg/mg protein) was assayed by ELISA after the cotransformation of E. coli with pKHsdR3 and plasmids containing hsdA. Plasmid p6 contained the entire 5.257-kb EcoRI fragment of C. testosteroni chromosomal DNA, including hsdA and its regulatory elements (38). In plasmid pAX1, operator Op1 was deleted. Finally, in plasmid pDel13n an additional 13 bases upstream of the hsdA promoter had been deleted. The induced expression of 3α-HSD/CR is observed with plasmid pAX1 but not with p6 (control) or pDel13n. Bars represent the averages and standard deviations from at least three independent measurements (*, P < 0.05; ***, P < 0.0001; both by t test). A, AvrII; E, EcoRI.

Fig 3.

HsdR activity is associated with the function of RepA. Under normal conditions, active RepA binds to two operators, Op1 and Op2, upstream of the hsdA gene and forces the DNA between these two palindromic sequences to form a loop structure (38, 40). Plasmids p67, p67-3, p67-5, and p67-7, with various distances between Op1 and Op2 (67, 64, 62, and 60 bp, respectively), upstream of the hsdA gene were cotransformed into E. coli HB101 with plasmid pK18 (control) or pKtacHsdR1 (containing hsdR). 3α-HSD/CR expression in E. coli, as determined by ELISA, was sensitive to the DNA-loop topology (for details see the text). Bars represent the averages and standard deviations from at least three independent measurements (***, P < 0.0001 by t test). E, EcoRI.

Construction of an hsdR-disrupted mutant of C. testosteroni.

An hsdR-disrupted mutant of C. testosteroni was prepared by homologous integration. A DNA fragment ranging from 251 to 650 bp of the hsdR gene was generated by PCR using forward primer 5′-CTGCCGCCAGCGGGC-3′ and reverse primer 5′-CAGTGCATGGGCTGC-3′ and plasmid pTOPOHsdR as the template. The 400-bp fragment was cloned into pCR2.1-TOPO containing the kanamycin resistance gene to yield pTOPOHsdR3-1. Since plasmid pTOPOHsdR3-1 cannot replicate in C. testosteroni, and because of the sensitivity of wild-type C. testosteroni to kanamycin, only mutants of C. testosteroni harboring the kanamycin resistance gene of plasmid pTOPOHsdR3-1 integrated within the chromosomal DNA can grow in medium containing kanamycin. Accordingly, C. testosteroni was transformed with 10 μg of pTOPOHsdR3-1 by electroporation (1.8 kV; 1-cm cuvette; Bio-Rad), which contains hsdR sequences homologous to C. testosteroni chromosomal DNA. The cells were spread on 30 μg/ml kanamycin SIN agar plates and cultured in a 27°C incubator overnight. The colonies were proven by PCR for homologous integration.

Preparation and purification of recombinant HsdR protein.

The overexpression of HsdR was performed in E. coli strain BL21(DE3)pLysS with plasmid pETHsdR2, and the recombinant protein was purified by its His tag sequence. In brief, cells transformed with plasmid pETHsdR2 (Fig. 1) were grown in SIN medium with 100 μg/ml ampicillin at 37°C in a shaker (180 rpm). One hundred microliters of the overnight culture was used to inoculate 3 ml of fresh medium. When the bacteria had grown to an optical density of 0.4 to 0.6 at 595 nm, target protein expression was induced by the addition of isopropyl-β-d-thiogalactoside to a final concentration of 1 mM. After induction for 4 h at 37°C or overnight at room temperature, cells were harvested by centrifugation. The cell pellet was either stored at −80°C for further usage or directly suspended in 200 μl of lysis buffer (50 mM NaH₂PO₄, 300 mM NaCl, 10 mM imidazole, pH 8.0) (Qiagen) containing different concentrations of sodium lauroyl sarcosinate (Fluka Chemie AG, Buchs, Switzerland). Cells were lysed by freezing (at −20°C for 30 min) and thawing (at room temperature for 30 min) 3 times, and the resulting mixture was centrifuged at 10,000 × g for 20 min. The supernatant was applied to a mini-nickel-nitrilotriacetic acid (Ni-NTA) metal affinity chromatography column (Qiagen). After washing 2 times with 600 μl of washing buffer (50 mM NaH₂PO₄, 300 mM NaCl, 75 mM imidazole, and different concentrations of sodium lauroyl sarcosinate, pH 8.0) (Qiagen), HsdR was eluted from the column by applying 100 μl of elution buffer 4 times (50 mM NaH₂PO₄, 300 mM NaCl, 250 mM imidazole, and different concentrations of sodium lauroyl sarcosinate, pH 8.0). Samples containing pure and soluble HsdR protein were assessed by SDS-polyacrylamide gel electrophoresis. The purification of RepA protein was performed as described previously (38). The concentration of purified protein was determined by the method of Bradford with Roti-Quant solution (Roth) using bovine serum albumin as the standard (3). Protein analysis by SDS-polyacrylamide gel electrophoresis was carried out according to Laemmli (12).

Immunization and preparation of antisera against HsdR.

On the first day, rabbits were injected subcutaneously with an emulsion of 0.5 ml water and 0.5 ml incomplete Freund's adjuvant. For immunization, 1 μg of purified HsdR protein was dissolved in an emulsion of 0.5 ml water and 0.5 ml incomplete Freund's adjuvant, and rabbits were immunized on days 7, 37, and 67. The antiserum was collected at day 74, and antibody titer determination in the rabbit serum was performed by enzyme-linked immunosorbent assay (ELISA).

ELISA of 3α-HSD/CR and HsdR.

Proteins for 3α-HSD/CR or HsdR ELISA detection were prepared from 3 ml of bacterial cell culture and subsequent centrifugation at 10,000 × g for 10 s. The pellet was washed 3 times with 1 ml of phosphate-buffered saline (PBS) and resuspended in 200 μl of PBS with 100 μg/ml lysozyme. The suspension was frozen at −20°C and thawed at room temperature 3 times. Finally, the samples were centrifuged again at 10,000 × g for 20 min. The supernatant was diluted into 1 mg/ml protein and used for 3α-HSD/CR or HsdR ELISA detections.

To quantify 3α-HSD/CR protein expression an ELISA was established, and respective antibodies directed against 3α-HSD/CR from C. testosteroni were prepared in rabbits (19). For HsdR detection, antibodies against HsdR were prepared in rabbits as described above. ELISA plates were coated with protein containing 3α-HSD/CR or HsdR diluted in coating buffer. After washing, antibodies against 3α-HSD/CR or HsdR were added in a 1:1,000 dilution. As the secondary antibody, peroxidase-conjugated swine anti-rabbit immunoglobulin (Dako, Denmark) was used in a 1:1,000 dilution. The further procedure corresponded to that of the chloramphenicol acetyltransferase ELISA kit from Roth.

HsdR-RNA polymerase interaction.

Three hundred ng RNA polymerase (a generous gift from Ruth Schmitz-Streit, Kiel University) was used to coat each well (the coating buffer contained 1.59 g Na₂CO₃, 2.93 g NaHCO₃, and 0.2 g NaN₃, pH 9.6, in 1 liter). The plate was incubated at 37°C for 30 min. After washing 3 times (the washing buffer contained 8 g NaCl, 0.2 g KH₂PO₄, 1.5 g Na₂HPO₄, 0.2 g KCl, and 0.5 ml Tween 20, pH 7.4, in 1 liter), different amounts of purified HsdR diluted with washing buffer were added. Following an incubation at 37°C for 30 min, rabbit antibodies against HsdR diluted with washing buffer (1:1,000) were added to the wells. After incubation for another 30 min at 37°C and washing, peroxidase-conjugated swine anti-rabbit immunoglobulin (Dako, Denmark) diluted 1:1,000 with washing buffer was added. Following a further incubation for 30 min at 37°C and washing, 100 μl ABTS solution [2′-azino-di-(3-ethylbenzthiazoline-6-sulfonate) with H₂O₂ in glycine-citric acid buffer; Roche, Mannheim, Germany] was added into each well and incubated at 37°C for 30 min. Finally, the samples were assayed in an ELISA reader at 415 nm with reference at 490 nm (Bio-Rad, Hercules, CA).

EMSA.

For gel mobility shift assays, digoxigenin-11-dUTP-labeled 203-bp DNA fragments containing the hsdA promoter region were generated by PCR with forward primer p7-1 (5′-AGGGGAATTACATCGTCCGTTTGCATGTAGCC-3′), reverse primer pCla (5′-CCGCATCGCGTATATCG-3′), and plasmid p6 as the template. Incubation mixtures (20 μl) contained 10 ng of labeled DNA fragment, 100 ng of herring sperm DNA, and different concentrations of purified HsdR and RepA protein in binding buffer consisting of 50 mM Tris-HCl (pH 8.0), 100 mM KCl, 12.5 mM MgCl₂, 1 mM EDTA, 5% glycerol, 100 μg/ml bovine serum albumin, and 1 mM dithiothreitol (DTT). After incubation at 27°C for 30 min, samples were separated on 5% polyacrylamide nondenaturing gels in 0.05 M Tris borate-EDTA buffer (pH 8.3) for 2 h at 10 V/cm and blotted onto nitrocellulose membranes. Labeled DNA was visualized by chemiluminescence using the digoxigenin luminescence detection kit (Roth).

Nucleotide sequence accession number.

The nucleotide sequence of the hsdR gene reported in this study has been submitted to the GenBank database with the accession number JF747025.

RESULTS

Cloning and sequence analysis of HsdR from C. testosteroni.

A new transcriptional factor, HsdR, named from 3α-hydroxysteroid dehydrogenase/carbonyl reductase regulator, was identified in C. testosteroni. The hsdR gene locates 2.58 kb downstream of hsdA on the C. testosteroni ATCC 11996 chromosome with an orientation opposite that of hsdA (Fig. 1). The open reading frame of hsdR consists of 912 bp and translates into a protein of 303 amino acids. BLAST analysis showed that HsdR is a member of the LysR-type transcriptional regulator family (LTTRs).

Activation of hsdA expression by HsdR.

For the identification of HsdR as an activator of hsdA expression, plasmid pKHsdR3, containing the hsdR gene, was cotransformed into E. coli HB101 with plasmid p6, pAX1, or pDel13n, respectively. After cotransformation with pKHsdR3, the expression level of 3α-HSD/CR with pAX1 increased compared to that of the control vector pK18, but not with p6 (Fig. 2).

According to our previous findings, there are two operators (Op1 and Op2) present in plasmid p6 which form a DNA-loop structure upon the binding of both with repressor RepA (38). In addition, Op2 overlaps the −10 region of the hsdA promoter region (38). Both the loop structure and the occupation of the hsdA promoter by RepA lead to a strong repression of hsdA gene expression (38). Since in pAX1 only operator Op2 is present, the DNA-loop structure cannot be formed, as hsdA expression has already increased with empty control vector pK18. However, after cotransformation with plasmid pKHsdR3, the increased expression of hsdA could be observed. This indicates that HsdR drives hsdA gene expression as a transcriptional activator (Fig. 2).

The cotransformation of pDel13n, in which 13 bases were deleted upstream of the AvrII-EcoRI fragment, with pK18 also leads to a slight increase in 3α-HSD/CR expression compared to that with p6 (Fig. 2). The DNA-loop structure also could not be formed due to the lack of bases important for RepA binding (38). Interestingly, the cotransformation of the hsdR gene (pKHsdR3) did not significantly enhance 3α-HSD/CR expression with pDel13n, a fact which leads us conclude that the deleted nucleotides may be necessary for HsdR action (Fig. 2).

To demonstrate that the ability of HsdR to activate hsdA expression is affected by RepA binding, a series of plasmids, p67, p67-3, p67-5, and p67-7, in which various numbers of bases were present between the two operators Op1 and Op2, were employed. With these plasmids, we have previously shown that a critical distance between Op1 and Op2 together with additional deletions of nucleotides 3, 5, and 7 result in DNA rotations that lead to altered orientations of both operators to each other (40). As a consequence, RepA binding and subsequent hsdA repression was influenced (40).

E. coli strain HB101 was transformed with these plasmids and plasmid pKtacHsdR1 (Fig. 3). 3α-HSD/CR expression was the lowest with plasmid p67-3, probably because the 64-bp distance and relative orientation between Op1 and Op2 is suitable for strong RepA binding and hsdA repression (38, 40). In this conformation, RepA is able to prevent HsdR from activating the hsdA promoter. With plasmid p67-7, which harbors a 60-bp spacing between Op1 and Op2, the cotransformation of HsdR leads to a significant increase in 3α-HSD/CR expression compared to that of the empty control vector pK18. The altered DNA conformation, in which repression by RepA is not as strong as that in p67-3, allows HsdR to perform its action on hsdA expression. This effect becomes most clear with plasmids p67 and p67-5, in which the positioning between Op1 and Op2 was turned by at least 72° along the DNA axis compared to the orientations of p67-3 and p67-7, such that RepA-Op1 and RepA-Op2 binding became sterically unlikely (40). In the absence of Op1-RepA-Op2, HsdR could bind to the hsdA promoter domain and activate 3α-HSD/CR expression (Fig. 3). This result clearly shows that HsdR activity is dependent on decreased repression by RepA.

Overexpression and purification of HsdR.

To produce purified HsdR protein, E. coli BL21(DE3)pLysS cells were transformed into plasmid pETHsdR2 and induced by IPTG at 37°C. However, the recombinant HsdR protein was present in the form of inclusion bodies, even after attempting to dissolve it with different detergents, such as sodium lauroyl sarcosinate (SLS). To solve this problem, the protein then was induced at room temperature overnight and dissolved with SLS. Surprisingly, the solubility of HsdR induced at room temperature was highly increased compared to that at 37°C, especially in the presence of SLS. The low temperature probably slowed down the speed of protein production and gave the protein enough time to fold properly (37). The dissolved HsdR protein then was purified with an Ni-NTA chromatography column under native conditions. The molecular mass of the recombinant protein (33.4 kDa) plus the His tag sequence (2.2 kDa) as seen on the SDS-polyacrylamide gel (35.6 kDa) was identical to that predicted from the amino acid sequence. The purified protein was used for binding assays to DNA and RNA polymerase, as well as for the preparation of polyclonal antibodies.

Binding of HsdR to the promoter region of the hsdA gene.

The specific interaction between HsdR and the promoter of the hsdA gene was demonstrated by gel mobility shift assays. A DNA fragment from −65 to +137 bp relative to the transcriptional start site of the hsdR gene was labeled with digoxigenin and incubated with purified HsdR protein. After electrophoresis, the formation of the HsdR-DNA complex was seen as shifted bands in Fig. 4A (lanes 2 and 3). Two shifted bands were observed when 100 pmoles HsdR protein was present. It seems that there are several binding sites at the hsdA promoter, including high-affinity binding sites and low-affinity binding sites, the former being occupied after the addition of small amounts of HsdR protein which resulted in the formation of the fast-migrating band. When large amounts of HsdR protein are present, both high-affinity and low-affinity binding sites are bound, and the slower migrating band is formed. The extent of the shifted bands became weaker upon reducing the amount of HsdR protein, as shown in Fig. 4A (lane 3). This result indicates that HsdR can bind to the hsdA promoter.

Fig 4.

HsdR binds to the hsdA promoter region. A DNA fragment of 203 bp containing the promoter domain of the hsdA gene was labeled with DIG-11-dUTP upon amplification by PCR. The labeled DNA fragment was mixed with different concentrations of HsdR and/or RepA, respectively, and herring sperm DNA was used as the competitor DNA. The reaction mixtures were subjected to 5% native PAGE. (A) Ten ng of DNA probe was incubated with different amounts of HsdR protein. Compared to free DNA (lane 1), HsdR-DNA binding leads to a shift of the corresponding bands (lanes 2 and 3). (B) Different concentrations of purified RepA were added to the HsdR-DNA reaction mixtures. Compared to the control (lane 5), RepA and HsdR were shown to bind to different sites of the hsdA promoter (lanes 1 to 4) (for details see the text).

To figure out if RepA competes with HsdR to bind to the hsdA promoter, the DNA fragment described above was used for the competition binding of RepA with HsdR. This fragment contained one high-affinity determinant (Op2) of RepA (9). The shifted bands formed by RepA (Fig. 4B, lane 1) and HsdR (Fig. 4B, lane 4) indicated that HsdR and RepA can independently bind to the same DNA probe. It is proposed that if they compete for the same binding determinant, the formation of the HsdR-DNA complex will be inhibited by RepA. However, it turned out that two shifted bands occurred after the addition of 6 pmoles RepA to the HsdR reaction mixtures (Fig. 4B, lane 3): one shifted band is similar to that formed by HsdR, and the other one, with slower electrophoretic mobility, obviously represents an HsdR-DNA-RepA complex. Furthermore, the shifted HsdR-DNA-RepA complex became very strong after the addition of more RepA protein (12 pmoles) to the HsdR reaction mixtures (Fig. 4B, lane 2). These findings show that HsdR and RepA can simultaneously bind to different sites of the hsdA promoter.

Interaction between HsdR and RNA polymerase.

To determine HsdR as an activator on the transcription level for the target gene, the in vitro binding activities of HsdR and RNA polymerase were measured (Fig. 5). In this ELISA experiment, two distinct negative controls were set up, one with no RNA polymerase coated on the wells and no HsdR protein added and the other with only HsdR protein (20 ng) added. As shown in Fig. 5, the optical density with 20 ng HsdR and 300 ng RNA polymerase was approximately 3-fold higher than that with 0.625 ng HsdR. At the same time the value gradually decreased with decreasing amounts of HsdR protein, as shown in Fig. 5. This indicates that HsdR binds to RNA polymerase, thereby potentially increasing the concentration of RNA polymerase in the promoter domain of the target gene and enhancing its expression.

Fig 5.

HsdR interacts with RNA polymerase (RNAP). RNAP first was coated onto the ELISA plates, followed by the addition of various concentrations of HsdR protein. After incubation with primary antibodies against HsdR, peroxidase-conjugated swine anti-rabbit immunoglobulin (anti-rabbit) was added as a secondary antibody. Finally, the samples were assayed in an ELISA reader (Bio-Rad). A clear and concentration-dependent binding of RNAP to HsdR is seen. Bars represent the averages and standard deviations from at least three independent measurements.

HsdR is essential for the induced expression of hsdA.

To demonstrate if HsdR is involved in the induction of hsdA expression, an hsdR gene knockout mutant of C. testosteroni (CT-HsdR-Ko) was prepared by homologous integration. Wild-type C. testosteroni and the hsdR knockout mutant strain CT-HsdR-Ko were induced overnight with 0.5 mM testosterone, and ELISA was used to measure hsdA expression. As shown in Fig. 6, the expression level of 3α-HSD/CR in wild-type C. testosteroni highly increased after incubation with testosterone, which was not the case in the hsdR knockout mutant. In the absence of the inducer testosterone, 3α-HSD/CR expression in both the wild-type strain and the hsdR knockout mutant was at the same basal level. It seems that testosterone binding to RepA decreases its affinity to the operators Op1 and Op2 such that the loop unfolds (38). As a consequence, the hsdA promoter becomes accessible for the transcription factor HsdR, which now can perform its function to enhance 3α-HSD/CR expression. According to this result, HsdR is a critical factor for hsdA gene regulation.

Fig 6.

HsdR is necessary for the induced expression of 3α-HSD/CR. Wild-type cells (C.T.) and hsdR knockout mutants of C. testosteroni (CT-HsdR-Ko) were cultured overnight in SIN medium in the presence or absence of 0.5 mM testosterone. ELISA revealed that after testosterone induction, 3α-HSD/CR expression increased considerably in wild-type C. testosteroni compared to that observed in the hsdR knockout mutants. In the absence of the steroidal inducer testosterone, 3α-HSD/CR expression in the hsdR knockout mutants and in wild-type cells occurred at the same basal level. Bars represent the averages and standard deviations from at least three independent measurements (***, P < 0.0001 by t test).

HsdR expression is not induced by testosterone.

To determine whether the expression of HsdR itself is sensitive to testosterone induction, the amount of HsdR in C. testosteroni was measured by ELISA with respective primary antibodies. After the addition of testosterone 3α-HSD/CR expression increased, whereas HsdR expression did not change (Fig. 7). This reveals that HsdR expression is not induced by testosterone in C. testosteroni.

Fig 7.

3α-HSD/CR but not HsdR expression is induced by testosterone. C. testosteroni was cultured overnight at 27°C in the presence or absence of 0.5 mM testosterone. Total protein was extracted and adjusted to 1 mg/ml. 3α-HSD/CR and HsdR expression levels were determined by ELISA. Whereas 3α-HSD/CR expression was induced by testosterone (A), HsdR expression did not change upon testosterone induction (B). Therefore, HsdR itself is not induced by testosterone but is necessary for steroid signaling in C. testosteroni. Bars represent the averages and standard deviations from at least three independent measurements (***, P < 0.0001 by t test).

DISCUSSION

Microorganisms capable of utilizing various naturally occurring steroids as carbon and energy sources are relatively widespread in nature (16, 17, 31, 41). The complete assimilation of these substrates is achieved through an adaptive complex metabolic pathway involving many enzymatic steps of oxidation that are responsible for the breakdown of the steroid nucleus (8, 13). Catabolic enzymes for steroid degradation are not constitutively expressed but are induced by their respective steroid substrates (18, 20, 24). There is considerable interest in the mechanism of the induction of steroid-catabolic enzyme expression and the basis of their transcriptional regulation. In previous investigations, 3α-HSD/CR from C. testosteroni ATCC 11996 was identified to be induced by steroids such as testosterone, and the expression of this enzyme was controlled by several regulators, including the repressors RepA and RepB and positive regulator TeiR (6, 38, 39).

In the present investigation, the novel transcription factor HsdR was found to serve as an activator of hsdA expression. To search for further cis- or trans-acting elements for 3α-HSD/CR expression which might locate upstream or downstream of the hsdA gene, a 5.257-kb EcoRI genomic fragment containing hsdA was extended leftward or rightward. Sequence analysis revealed that an LTTR gene (which we named hsdR) is located 2.58 kb downstream of the hsdA gene. Extensive studies revealed that LTTRs have a highly conserved N-terminal DNA binding domain and a less conserved C-terminal coinducer recognition domain, and they often act as homotetramers (29, 32). The HELIXTURNHELIX program (http://www.pasteur.fr) predicted that HsdR contained a winged helix-turn-helix DNA binding motif in the N-terminal region. Moreover, at the C terminus a LysR-type substrate binding domain was identified, as revealed by a BLAST search. In the course of our further investigation, HsdR was shown to bind to the promoter region of the hsdA gene. In general, LysR-type regulators act as tetramers or dimers to perform their function and to recruit RNA polymerase for subsequent gene transcription. The recruitment of RNA polymerase also was shown in this study for HsdR-induced hsdA expression.

LTTRs constitute the largest family of prokaryotic regulatory proteins identified so far (15). The genes regulated by LTTRs have diverse functions, including the degradation of organic compounds. In this study, HsdR has been identified to regulate the expression of 3α-HSD/CR, which is one of the first enzymes in the steroid-catabolic pathway, and it also catalyzes the carbonyl reduction of nonsteroidal aldehydes and ketones (24). Apart from this enzyme, many LTTRs are associated with degradation pathways of aromatic compounds. A large group of LTTRs regulates a single target operon only, such as CatR controlling catBCA expression for catechol metabolism in Pseudomonas putida (28). Operons such as the clcABDE operon of plasmid pAC27 (4), the tcbCDEF operon in plasmid pP51 (36), and the cbnABDE operon from Ralstonia eutropha (21), which are involved in chlorocatechol metabolism, were controlled by ClcR, TcbR, and CbnR, respectively (4, 21, 36). Two paralogous LTTRs, BenM and CatM from Acinetobacter sp. strain ADP1, controlling the expression of several operons involved in benzoate degradation, were extensively studied (5, 27).

In general, LTTRs have been described as transcriptional activators of a single divergently transcribed gene that exhibited negative autoregulation (14, 25, 32). Extensive research has led to them being regarded as global transcriptional regulators, acting as either activators or repressors of single or operonic genes; they often are divergently transcribed but can be located elsewhere on the bacterial chromosome (10, 11). In this study, hsdR was located 2.58 kb downstream of its target gene hsdA on the C. testosteroni chromosome. In addition, a further short-chain dehydrogenase/reductase gene, SDRx, was found to be divergently transcribed from the hsdR gene. SDRx seemingly is related to 7α-HSDs, as revealed by phylogenetic analysis. 7α-HSD is an NADP(H)-dependent oxidoreductase belonging to the short-chain dehydrogenase/reductase (SDR) superfamily (35). 7α-HSDs are widespread among bacteroides and clostridia and occur in E. coli and Ruminococcus species as well. They catalyze the dehydrogenation of a hydroxyl group at position 7 of the steroid skeleton of bile acids (26). Interestingly, the degradation velocities of the steroids cholic acid, testosterone, and estradiol decreased in SDRx knockout mutants of C. testosteroni (7).

Based on the results obtained in the present study, a model for the regulation of hsdA expression by HsdR is proposed. In the absence of inducing steroids, the transcriptional repressor RepA binds to Op1 and Op2 to form a loop structure which contains the hsdA promoter domain. Due to the resulting DNA configuration, other transcriptional regulators, such as HsdR, cannot perform their function at the hsdA promoter even if they are already bound to the promoter (Fig. 8A). Therefore, 3α-HSD/CR expression is only at basal levels in both the wild-type strain and the hsdR knockout mutant (Fig. 6). In the presence of an inducer such as testosterone, RepA is released from the operators and the loop structure is disrupted, such that HsdR can activate the hsdA promoter and increase the concentration of RNA polymerase at the promoter domain (Fig. 8B). Finally, hsdA expression increases. In conclusion, HsdR is a positive transcription factor for the hsdA gene and promotes the induction of 3α-HSD/CR expression in C. testosteroni. Furthermore, HsdR activity is dependent on decreased repression by RepA.

Fig 8.

Model for the regulation of hsdA expression by HsdR in C. testosteroni. (A) In the absence of the inducing steroid testosterone, the RepA protein binds to operators Op1 and Op2 and blocks hsdA transcription. The loop structure formed by RepA binding also affects other transcription regulators, such as HsdR, to activate the hsdA promoter. (B) In the presence of testosterone, RepA is released from the operators and the loop structure is disrupted. Thus, HsdR now can recruit RNA polymerase (RNAP) to activate hsdA transcription.