The desA and desB genes from Clostridium scindens ATCC 35704 encode steroid-17,20-desmolase

Abstract

Clostridium scindens is a gut microbe capable of removing the side-chain of cortisol, forming 11β-hydro­xyandrostenedione. A cortisol-inducible operon (desABCD) was previously identified in C. scindens ATCC 35704 by RNA-Seq. The desC gene was shown to encode a cortisol 20α-hydroxysteroid dehydrogenase (20α-HSDH). The desD encodes a protein annotated as a member of the major facilitator family, predicted to function as a cortisol transporter. The desA and desB genes are annotated as N-terminal and C-terminal transketolases, respectively. We hypothesized that the DesAB forms a complex and has steroid-17,20-desmolase activity. We cloned the desA and desB genes from C. scindens ATCC 35704 in pETDuet for overexpression in Escherichia coli. The purified recombinant DesAB was determined to be a 142 ± 5.4 kDa heterotetramer. We developed an enzyme-linked continuous spectrophotometric assay to quantify steroid-17,20-desmolase. This was achieved by coupling DesAB-dependent formation of 11β-hydroxyandrostenedione with the NADPH-dependent reduction of the steroid 17-keto group by a recombinant 17β-HSDH from the filamentous fungus, Cochliobolus lunatus. The pH optimum for the coupled assay was 7.0 and kinetic constants using cortisol as substrate were K_m of 4.96 ± 0.57 µM and k_cat of 0.87 ± 0.076 min⁻¹. Substrate-specificity studies revealed that rDesAB recognized substrates regardless of 11β-hydroxylation, but had an absolute requirement for 17,21-dihydroxy 20-ketosteroids.

Clinical studies in the 1950s reported urinary excretion of 17-ketosteroids following rectal infusion of cortisol for the treatment of ulcerative colitis. Concomitant neomycin treatment precluded urinary 17-ketosteroid excretion, implicating gut bacteria in the conversion of a C₂₁ glucocorticoid to C₁₉ androstane/androgenic metabolites (1, 2). Later, it was shown that rat and human fecal bacteria were capable of this cortisol side-chain cleavage reaction, which became known as steroid-17,20-desmolase (3, 4). Serial dilutions of human and rat fecal samples revealed that ∼10⁶ bacterial gram⁻¹ of wet-weight express steroid-17,20-desmolase activity (5). Subsequently, an organism was isolated in pure culture that expressed steroid-17,20-desmolase and 20α-hydroxysteroid dehydrogenase (20α-HSDH) activities, and was named Clostridium scindens, whose species epithet means “to cut,” a reference to separation of the side-chain from the steroid D-ring (6).

Cortisol-inducible steroid-17,20-desmolase was partially purified, and 20α-HSDH was purified to apparent electrophoretic homogeneity (SDS-PAGE) by traditional chromatographic separation from cell extracts of C. scindens ATCC 35704 (7, 8). The application of genome-wide transcriptomics (RNA-Seq) identified a polycistronic cortisol-inducible operon (desABCD), which included a gene encoding 20α-HSDH (desC) (9). The desA and desB genes are annotated as N-terminal and C-terminal transketolases, respectively. Transketolases catalyze the thiamine pyrophosphate (TPP)-dependent transfer of glycol aldehyde to an aldose acceptor. Comparison of the transketolation reaction to that of steroid-17,20-desmolase suggests an analogous reaction, and that desAB encodes a novel steroid transketolase (9). Clostridium cadavaris and Butyricicoccus desmolans were shown previously to generate 20β-dihydrocortisol as well as 11β-hydroxyandrostenedione from cortisol (10). When the genomes of C. cadavaris and B. desmolans were sequenced and reported, the desAB genes were located via BLAST search, and were found to be flanked by a gene (DesE) encoding a short-chain dehydrogenase/reductase (SDR) family protein. We overexpressed the SDR gene from B. desmolans ATCC 43058 in Escherichia coli and showed that the enzyme is a pyridine nucleotide-dependent 20β-HSDH (11). The clustering of genes encoding cortisol 20α-HSDH or 20β-HSDH with desAB in different gut microbial species suggests that desAB encodes steroid-17,20-desmolase; however, this function has not yet been demonstrated. Here, we report that the desAB genes encode steroid-17,20-desmolase. In addition, we developed a novel, enzyme-coupled continuous spectrophotometric assay to quantify steroid-17,20-desmolase activity.

MATERIALS AND METHODS

Bacterial strains and chemicals

E. coli DH5α (turbo) competent cells from New England Biolabs (Ipswich, MA) were used for cloning, and E. coli BL21-CodonPlus (DE3) RIPL was purchased from Stratagene (La Jolla, CA) and used for protein overexpression. The pETDuet and pET-51b(+) vectors were obtained from Novagen (San Diego, CA). Restriction enzymes were purchased from New England Biolabs and a QIAprep Spin Miniprep kit was obtained from Qiagen (Valencia, CA). Isopropyl β-D-1-thiogalactopyranoside was purchased from Gold Biotechnology (St. Louis, MO). Streptactin resin was purchased from IBA GmbH, (Goettingen, Germany). Steroids were purchased from Steraloids (Newport, RI). Amicon Ultra-15 centrifugal filter units with 30 kDa and 50 kDa molecular mass cutoffs were obtained from Millipore (Billerica, MA). All other reagents were of the highest possible purity and were purchased from Fisher Scientific (Pittsburgh, PA).

Cloning of desA and desB from C. scindens and 17β-HSDH from C. lunatus

Genomic DNA was extracted from C. scindens ATCC 35704 using Fast DNA isolation kit from Mo-Bio (Carlsbad, CA) according to the manufacturer’s protocol. The nucleotide sequence of the coding region of a cDNA encoding 17β-hydroxysteroid dehydrogenase (17β-HSDH) from C. lunatus was entered into the Codon Optimization Tool from Integrated DNA Technologies (Coralville, IA). The resulting sequence, whose deduced amino acid sequence was identical to the wild-type 17β-HSDH sequence was synthesized as double-strand DNA by gBlocks(R) gene fragments service from Integrated DNA Technologies. C. scindens ATCC 35704 genomic DNA and gBlocks 17β-HSDH double-strand DNA were used as templates to amplify the desAB and 17β-HSDH genes using oligonucleotide primers reported in Table 1. Amplification of desAB and 17β-HSDH inserts was performed on an Applied Biosystems ProFlex PCR System (Foster City, CA) using Phusion High-Fidelity DNA Polymerase (Stratagene, La Jolla, CA). Inserts were gel-purified from agarose and cloned into the pETDuet and pET-51b(+) vectors, respectively, using appropriate restriction enzymes to digest insert and vector and ligated with DNA ligase. Recombinant plasmids were transformed into chemically competent E. coli DH5α cells via the heat-shock method, plated, and grown for 16 h at 37°C on lysogeny broth (LB) agar plates supplemented with ampicillin (100 µg/ml). A single colony from each transformation was inoculated into LB medium (5 ml) containing ampicillin (100 µg/ml) and grown to saturation. The cells were subsequently centrifuged (3,220 g, 15 min, 4°C) and plasmids were extracted from the resulting cell pellet using the QIAprep Spin Miniprep kit (Qiagen, Valencia, CA). The sequences of the inserts were determined via Sanger sequencing (W. M. Keck Center for Comparative and Functional Genomics at the University of Illinois at Urbana-Champaign).

TABLE 1.

Oligonucleotides and synthetic gene sequences used in this study

	5′-Forward Primer-3′
	5′-Reverse Primer-3′
desA	ATATATGAATTCGATGGCTAAAGTATGTATCAATGATGTGC
	ATATATAAGCTTTTACTTCACCTTCGCCAGTTTCTG
desB	ATATATCATATGCATCACCATCATCACCACAGCATGGCAAATATGGGCGGTTTCAC
	ATATATCTCGAGCTATTTCTTTAACTGCTCCAGTGCAACT
17β-HSDH (CL17HSDH)	ATATATGGATCCGATGCCACATGTAGAAAACGCCTC
	ATATATAAGCTTTTAAGCCGCCCCGCCGTC
CL17HSDH GenBlock (E. coli Codon usage)	ATGCCACATGTAGAAAACGCCTCGGAGACGTATATCCCAGGACGTCTGGACGGCAAGGTAGCACTGGTCACAGGTTCCGGGCGCGGAATTGGGGCCGCCGTTGCCGTTCACTTAGGGCGCTTAGGAGCCAAGGTCGTGGTTAACTATGCCAATAGTACGAAGGACGCGGAAAAAGTTGTGTCAGAAATCAAGGCCCTTGGTTCAGATGCCATCGCCATTAAGGCAGATATCCGTCAGGTTCCAGAAATCGTGAAATTATTTGACCAGGCGGTTGCGCATTTCGGACATCTTGACATTGCAGTCTCCAACAGCGGGGTTGTTTCTTTTGGGCACTTAAAAGACGTTACCGAAGAAGAGTTTGACCGTGTATTCAGCTTAAACACGCGTGGACAGTTCTTCGTTGCTCGTGAAGCATATCGTCACCTGACGGAGGGAGGGCGTATTGTTTTGACTTCATCGAACACATCTAAGGATTTTTCGGTGCCCAAGCACTCTCTTTACTCCGGGTCTAAAGGGGCGGTAGATTCTTTTGTTCGCATTTTCTCCAAAGATTGTGGGGATAAAAAGATTACAGTCAACGCGGTTGCTCCTGGAGGGACAGTGACAGATATGTTCCACGAAGTTTCTCATCATTACATTCCGAATGGTACAAGTTACACAGCCGAACAGCGTCAACAGATGGCTGCACATGCCTCTCCATTACACCGTAATGGATGGCCGCAGGATGTCGCAAATGTGGTTGGTTTTCTGGTGTCTAAGGAGGGTGAGTGGGTTAATGGCAAAGTTCTTACGTTGGACGGCGGGGCGGCTTAA

Restriction sites are in bold and italic.

Gene expression and purification of recombinant steroid-17,20-desmolase and 17β-HSDH

For protein expression, plasmids extracted containing correct insert from the E. coli DH5α cells were transformed into E. coli BL-21 CodonPlus (DE3) RIPL chemically competent cells via the heat-shock method (42°C for 30 s) and grown overnight at 37°C on LB agar plates supplemented with ampicillin (100 µg/ml) and chloramphenicol (50 µg/ml). After 16 h, five isolated colonies were used to inoculate 10 ml of fresh LB medium supplemented with antibiotics and grown at 37°C for 6 h with vigorous aeration. The precultures were then added to fresh LB medium (1 L) supplemented with the same antibiotics at the same concentrations and grown with vigorous aeration at 37°C. At OD₆₀₀ of 0.3, isopropyl β-D-1-thiogalactopyranoside was added to each culture at a final concentration of 0.1 mM and the temperature was decreased to 16°C. Following 16 h of culturing, cells were pelleted by centrifugation (4,000 g, 30 min, 4°C) and resuspended in 30 ml of binding buffer (20 mM Tris-HCl, 150 mM NaCl, 20% glycerol, and 10 mM 2-mercaptoethanol at pH 7.9). The cell suspension was subjected to four passages through an EmulsiFlex C-3 cell homogenizer (Avestin, Ottawa, Canada), and the cell lysate was clarified by centrifugation at 20,000 g for 30 min at 4°C.

The recombinant steroid-17,20-desmolase was then purified using TALON® Metal Affinity Resin (Clontech Laboratories, Mountain View, CA) per the manufacturer’s protocol. The recombinant protein was eluted using an elution buffer composed of 20 mM Tris-HCl, 150 mM NaCl, 10% glycerol, 10 mM 2-mercaptoethanol, and 250 mM imidazole at pH 7.9. The recombinant 17β-HSDH was then purified using Strep-Tactin® resin (IBA GmbH, Goettingen, Germany) per manufacturer’s protocol. The recombinant protein was eluted using an elution buffer composed of 20 mM Tris-HCl, 150 mM NaCl, 10% glycerol, 10 mM 2-mercaptoethanol, and 2.5 mM desthiobiotin at pH 7.0. The purity of the proteins was assessed by SDS-PAGE, and protein bands were visualized by staining with Coomassie Brilliant Blue G-250 Dye. The protein concentrations were calculated based on the computed molecular mass and extinction coefficient. The observed subunit mass for each protein was calculated by migration distance of purified protein to standard proteins using ImageJ (https://imagej.nih.gov/ij/docs/faqs.html).

Gel filtration chromatography

Gel filtration chromatography was carried out on a Superose-6 10/300 analytical column (GE Healthcare, Piscataway, NJ) attached to an ÄKTAxpress chromatography system (GE Healthcare, Piscataway, NJ) at 4°C. The eluted recombinant steroid-17,20-desmolase from the metal affinity resin was concentrated and loaded on to the column with a buffer composed of 50 mM Tris-Cl, 150 mM NaCl, 10% glycerol, and 10 mM 2-mercaptoethanol at pH 7.0. The native molecular mass of the protein was calculated by peak retention volume of protein to retention volume of standard proteins.

LC/MS analysis

Steroid-17,20-desmolase activity was determined by incubating 1 μM recombinant enzyme with 50 μM cortisol in the presence of 1 µM MnCl₂ and 10 µM TPP in 50 mM glycine-glycine buffer (pH 7.0) at room temperature overnight. After incubation, the products were extracted by vortexing the reaction with 2 vols of ethyl acetate for 1 to 2 min and then recovering the organic phase. The organic phase was then evaporated under nitrogen gas. The residue was dissolved in 50 μl methanol and analyzed using a Shimadzu UHPLC system with DAD detector coupled with a Shimadzu LCMS-IT-TOF System (Shimadzu Corporation, Kyoto, Japan). The LC operating conditions were as follows: LC column, C-18 analytical column (Capcell Pak C18, Shiseido, Japan), 250 mm × 2 mm i.d., particle size, 3 μm; mobile phase, water containing 0.1% formic acid (A), and acetonitrile containing 0.1% formic acid (B); total flow rate of mobile phase, 0.2 ml/min; total run time including equilibration, 41 min. The initial mobile phase composition was 70% mobile phase A and 30% mobile phase B. The percentage of mobile phase B was changed linearly over the next 5 min until 35%. Over the next 25 min, the percentage was increased to 98% linearly. Then the percentage was maintained for 5 min, the mobile phase composition was allowed to return to the initial conditions, and allowed to equilibrate for 5 min. The injection volume was 10 μl. The DAD detector was used at a wavelength of 254 nm. Peak retention times and peak areas of samples were compared with standard steroid molecules.

The mass spectrometer (LCMS-IT-TOF) was operated with an ESI source in positive mode. The nebulizer gas pressure was set at 150 kPa with the source temperature of 200°C and the gas flow at 1.5 lmin⁻¹. The detector voltage was 1.65 kV. High-purity nitrogen gas was used as collision cell gas. The raw chromatogram and mass spectrogram data were processed with the LC solution Workstation software (Shimadzu).

Continuous enzyme-coupled spectrophotometric assay

Steroid-17,20-desmolase activity was measured aerobically at 25°C by monitoring the conversion of NADPH to NADP⁺ 340 nM (ε=6,220 M⁻¹·cm⁻¹) reflecting DesAB-catalyzed conversion of cortisol to 11β-hydroxyandrostenedione followed by NADPH-dependent 17β-HSDH conversion of 11β-hydroxyandrostenedione to a 11β-testosterone. Control reactions were run leaving out one component at a time. Linearity with respect to time and enzyme concentration were determined. The standard reaction mixture contained 50 mM MOPS buffer, pH 7, 50 μM of cortisol, 1 µM MnCl₂, 10 µM TPP, 200 μM NADPH, 1 μM each enzyme, and buffer to a final volume of 0.5 ml. The reaction was initiated by the addition of the steroid-17,20-desmolase. The initial velocity of enzyme catalyzed reactions was plotted against the substrate concentrations, and the kinetic parameters were estimated by fitting the data to the Michaelis-Menten equation by nonlinear regression method using the enzyme kinetics module in GraphPad Prism (GraphPad Software, La Jolla, CA). An alternative methodology was utilized for substrates lacking a 17-hydroxyl, or those substrates whose 17-keto products had low specificity for recombinant C. lunatus 17β-HSDH (rCL17HSDH), was to run the basic assay without rCL17HSDH and NADPH. Reactions were terminated with 1N HCl and reaction products extracted and separated by LC/MS as described above. A standard curve of peak area vs. concentration of the reaction products was generated to quantify reaction rates.

Determination of optimal pH

The buffers used to study the pH profiling of the coupled enzymatic assay contained 50 mM of buffering agent, 1 µM MnCl₂, 10 µM TPP, and 150 mM NaCl. Buffering agents used were as follows: sodium citrate (pH 6.0), MOPS (pH 6.5–7.0), glycine-glycine (pH 6.5–8.5), Tris-Cl (pH 9), and glycine-NaOH (pH 9.5–10). To determine the optimal pH, purified recombinant steroid-17,20-desmolase and 17β-HSDH at final concentrations of 1 μM each were added to 50 μM cortisol and 200 μM NADPH in 500 μl of buffer. The oxidation of NADPH was measured continuously at 340 nm for 5 min by spectrophotometry and initial velocities were calculated.

RESULTS

Cloning, overexpression of desAB, and purification of recombinant DesAB

The desA gene encodes a 296 amino acid protein (CLOSCI_00899) with predicted TPP binding site (H74, G154, N183, K253), metal-binding site (D153, E155) and active-site (H37, H268) based on multiple sequence alignment with characterized transketolases and organisms expressing steroid-17,20-desmolase activity, or encoding a predicted desABE operon (Fig. 1). The desB encodes a 327 amino acid protein (CLOSCI_00900) comprising the C-terminal portion of a typical full-length transketolase, and includes conserved residues with transketolases as diverse as Lactobacillus salivarius UCC118 (YP_536833) (12) and Homo sapien (P29401.3) (13) including TPP-binding (F95), and active-site residues (E70). Also included in this alignment are DesA and DesB sequences from known steroid-17,20-desmolase expressing strains such as gut bacteria B. desmolans (WP_031476415.1, WP_031476417.1), C. cadaveris (WP_027640053.1, WP_027640052.1), urinary isolate Propionimicrobium lymphophilum ACS-093-V-SCH5 (14, 15) (WP_016455355.1, WP_016455356.1), and inferred by homology and desABE operon structure from urinary tract isolate Arcanobacterium urinimassiliense (16) (WP_084789932.1, WP_073996552.1).

Fig. 1.

Steroid-17,20-desmolase (DesAB) shares conserved catalytic and cofactor binding residues with transketolases. Sequences in partial CLUSTAL-Omega alignment include H. sapien transketolase (TKT_Hsap), transketolase from L. salivarius (TKT_Lsa1), DesAB from C. scindens ATCC 35704 (desAB_Csci), DesAB from B. desmolans (desAB_Bdes), DesAB from C. cadavaris (desAB_Ccad), DesAB from P. lymphophilum (desAB_Plym), and DesAB from A. urinimassiliense (desAB_Aaur). Conserved residues are highlighted based on functional role depicted by the color key.

Because key conserved transketolase catalytic and cofactor binding residues are distributed across both the desA and desB genes, we chose the pETDuet expression vector, designed for overexpression of two target genes in E. coli BL21(DE3)RIL, as His-tagged recombinant proteins. SDS-PAGE analysis of the elution fraction following affinity chromatography revealed two bands at 32.7 kDa and 38.4 kDa, corresponding to the predicted molecular mass of rDesA and rDesB, respectively (Fig. 2A). Gel filtration chromatography determined the native molecular mass of rDesAB to be 142 ± 5.4 kDa, suggesting that the rDesAB is a heterotetramer (Fig. 2C).

Fig. 2.

SDS-PAGE of affinity purified rDesA and rDesB and native molecular mass estimate of rDesAB. A: Gene organization and domain architecture of steroid-17,20-desmolase. B: SDS-PAGE of pETDuet coexpressed rDesA and rDesB (Lane DesAB), protein molecular mass marker (M). C: Native molecular size analysis of rDesAB by size-exclusion chromatography.

Development of a coupled, continuous spectrophotometric assay to measure steroid-17,20-desmolase activity

Next, we incubated recombinant DesAB overnight in the presence of cortisol, TPP, and Mn²⁺ in glycine-glycine buffer, pH 7.0. Extraction of the reaction buffer with ethyl acetate and separation by reverse-phase HPLC revealed net conversion of cortisol (R_T = 17.5 min) to a single product, with retention time identical to 11β-hydroxyandro­stenedione standard (R_T =23.5 min) (Fig. 3). The control cortisol peak was subjected to ESI-TOF-MS and resulted in a major mass ion in positive mode of 363.17 m/z, consistent with the molecular weight of cortisol (362.46 amu). The major mass ion of the recombinant DesAB (rDesAB) product was 303.15 m/z, consistent with side-chain cleavage and formation of 11β-hydroxyandrostenedione (302.414 amu). These results demonstrate that the DesAB is the steroid-17,20-desmolase.

Fig. 3.

LC/MS of rDesAB reaction products. Standards include authentic cortisol and 11β-hydroxyandrostenedione. Reaction assay conditions include 1 μM recombinant enzyme with 50 μM cortisol in the presence of 1 µM MnCl₂ and 10 µM TPP in 50 mM glycine-glycine buffer (pH 7.0) at room temperature overnight. Control reactions omitted rDesAB. Mass spectra corresponding to HPLC peaks depicted in inset.

The steroid-17,20-desmolase assay previously utilized is discontinuous, time-consuming, and expensive. We therefore sought another approach to quantify steroid-17,20-desmolase activity using a continuous enzyme-coupled spectrophotometric assay. The steroid-17,20-desmolase reaction yields a 17-ketosteroid product which, if coupled to an NAD(P)H-dependent 17α- or 17β-HSDH, could be used to measure reaction velocity by monitoring NAD(P)H oxidation at 340 nm (Fig. 4). Consultation of the relevant literature revealed a well characterized NADP(H)-dependent 17β-HSDH from the filamentous fungus C. lunatus, whose 1.044 kB cDNA sequence is available at NCBI (Accession number AAD12052.1) (17). We trimmed the 5′- and 3′- untranslated region sequences and optimized the resulting 813 bp coding sequence for codon-usage in E. coli with the Codon Optimizer tool from Integrated DNA Technologies and the gene insert was synthesized using the GenBlock synthetic biology service from the same company. The insert was PCR-amplified using primers designed for bi-directional cloning into pET51b(+) vector, allowing for expression of rCL17HSDH as a 32.2 kDa N-terminal strep­tavidin affinity-tagged protein (Fig. 4).

Fig. 4.

SDS-PAGE of purified recombinant 17β-HSDH from C. lunatus (rCL17HSDH) and schematic representation of enzyme-coupled continuous spectrophotometric steroid-17,20-desmolase activity assay. A: SDS-PAGE of rCL17HSDH. Protein molecular weight marker (M). B: Schematic of conversion of cortisol to 11β-hydroxyandrostenedione, catalyzed by rDesAB, followed by NADPH-dependent conversion of 11β-hydroxyandrostenedione to 11β-hydroxytes­tosterone by rCL17HSDH.

rCL17HSDH showed NADPH-dependent activity toward 11β-hydroxyandrostenedione, but not cortisol in 50 mM glycine-glycine buffer, pH 7.0. Oxidation of NADPH was observed at 340 nm only in the presence of rDesAB, cortisol, and rCL17HSDH. Previous characterization of rCL17HSDH showed pH optimum between 7.0 and 8.0 (18). We determined that DesAB was most active between pH 7.0 and pH 7.5 by observing conversion of cortisol to 11β-hydroxyandrostenedione between pH 6.0 to 8.5 (data not shown). Therefore, the pH optima of both rDesAB complex and rCL17HSDH overlap. After optimization of the coupled assay with respect to enzyme concentration, we determined that pH optimum for the coupled reaction was pH 7.0 in MOPS buffer. There was a precipitous decline in activity from pH 6.5 to pH 6.0, nearly equivalent to the loss of activity from pH 7.0 to 8.0. At pH 10.0, the coupled assay retained ∼5% of its maximum activity (Fig. 5).

Fig. 5.

Biochemical characterization of rDesAB by enzyme-coupled continuous spectrophotometric assay. Effect of pH on steroid-17,20-desmolase activity (top panel). See Materials and Methods for reaction conditions. Michaelis-Menten and Lineweaver-Burk plot of rDesAB activity with increasing cortisol (bottom panel).

Determination of kinetic constants and substrate-specificity

We varied the concentration of cortisol in the presence of 200 µM NADPH, 1 µM MnCl₂, 10 µM TPP in 50 mM MOPs buffer pH 7.0. The substrate-saturation curve and Lineweaver-Burke plot are shown in Fig. 5. The K_m for rDesAB was 4.96 ± 0.57 µM and V_max of 12.76 ± 1.12 nmol·min⁻¹·mg⁻¹. The turnover number (k_cat) was determined to be 0.87 ± 0.076 min⁻¹ and catalytic efficiency (k_cat/K_m) was 0.17 ± 0.029 min⁻¹·μM⁻¹. Suspecting that an aldose-acceptor may be utilized in the reaction, glyceraldehyde-3-phosphate and ribose-5-phosphate were tested at 50 µM, 100 µM, and 500 µM; however, we did not observe increased enzymatic activity.

We tested substrate-specificity using a combination of methods, depending on whether steroids were 17-hydroxylated. The enzyme-coupled continuous spectrophotometric assay was relegated to 17-hydroxylated substrates. Those reaction products from substrates lacking 17-hydroxy were characterized by LC/MS. Furthermore, we tested NADPH-dependent rCL17HSDH activity against the expected C-19 steroid-17,20-desmolase products to determine whether we could make a direct comparison between substrates whose products differed structurally. We determined that Ring-A reduced products (50 µM concentration), such as etiocholanolone and 11β-hydroxyetiocholanolone, are not substrates for rCL17HSDH. These substrates yielded only 8.06 ± 2.65% and 4.73 ± 0.68% of the maximal activity, respectively, relative to 11β-hydroxyandrostenedione (three technical replicates). Therefore, ring-A reduced substrates were tested for activity using rDesAB with separation of steroids by LC/MS.

We determined that 11β-hydroxylation is not necessary for steroid-17,20-desmolase (Table 2). 20α-Hydroxy or 20β-hydroxy derivatives of cortisol were not substrates. Those substrates lacking 17-hydroxyl group, and/or 21-hydroxyl group, such as corticosterone and progesterone metabolites were not side-chain cleaved (Fig. 6, supplemental Fig. S1). Interestingly, tetrahydrocortisol and tetrahydrodeoxycortisol were also not recognized by rDesAB, indicating 3α-hydroxy, 5β-reduced metabolites are not substrates.

TABLE 2.

Substrate specificity for rDesAB from C. scindens ATCC 35704

Steroid	Trivial Name	Relative Activity %
4-PREGNEN-11β, 17, 21-TRIOL-3, 20-DIONE	CORTISOL	100.00 ± 2.78a
4-PREGNEN-17, 21-DIOL-3, 20-DIONE	11-DEOXYCORTISOL	186.73 ± 3.60
4-PREGNEN-11α, 21-DIOL-3, 20-DIONE	11-EPICORTICOSTERONE	NAb,c
4-PREGNEN-11β, 21-DIOL-3, 20-DIONE	CORTICOSTERONE	NAc
4-PREGNEN-11β-OL-3, 20-DIONE	11β-HYDROXYPROGESTERONE	NAc
4-PREGNEN-11β, 17-DIOL-3, 20-DIONE	21-DESOXYCORTISOL	NA
4-PREGNEN-11β, 17, 20α, 21-TETROL-3-ONE	20α-DIHYDROCORTISOL	NA
4-PREGNEN-11β, 17, 20β, 21-TETROL-3-ONE	20β-DIHYDROCORTISOL	NA
4-PREGNEN-17, 20α, 21-TRIOL-3-ONE	20α-DIHYDROCORTEXONE	NA
4-PREGNEN-17, 20β, 21-TRIOL-3-ONE	20β-DIHYDROCORTEXONE	NA
5β-PREGNAN-3α,11β,17, 21-TETROL-20-ONE	TETRAHYDROCORTISOL	NAc
5β-PREGNAN-3α,17, 21-TETROL-20-ONE	TETRAHYDRODEOXYCORTISOL	NAc

^aValues represent mean ± SD from three technical replicates.

^bNA = No activity.

^cDetermined by LC/MS (See Fig. 6)

Fig. 6.

Determination of steroid-17,20-desmolase activity against substrates lacking a 17-hydroxy, and tetrahydro(deoxy)cortisol by LC/MS. A: HPLC chromatographs with control (left panel) and rDesAB catalyzed (right panel). Cortisol and 11-deoxycortisol are included as positive controls. HPLC peaks are labeled with corresponding major mass ions (m/z). Mass spectra are presented in supplemental Fig. S1. B: Overlay of control (red) and rDesAB catalyzed (blue) HPLC chromatographs with tetrahydrocortisol (top panel) and tetrahydrodeoxycortisol (lower panel). MS data are included as insets and correspond to major peaks. Steroid metabolites were not detected in any minor peaks. C: Functional groups important for rDesAB specificity highlighted in orange.

DISCUSSION

In the current study, we report that the desA and desB genes from C. scindens ATCC 35704 encode steroid-17,20-desmolase. Recombinant DesAB displayed substrate-specificity consistent with native steroid-17,20-desmolase; the enzyme recognizes 20-keto-17, 21-dihydroxy steroids (7, 8). C. scindens ATCC 35704 expresses NAD(H)-dependent 20α-HSDH (9), whereas B. desmolans, C. cadavaris, and P. lymphophilum express 20β-HSDH (11) in addition to steroid-17,20-desmolase. Oxidation-reduction of the C20 oxygen is predicted to function as a regulatory “switch,” interconverting cortisol between DesAB substrate and nonsubstrate forms. The desD gene is coexpressed with desABC (9) and encodes a predicted 51.2 kDa transport protein in the major facilitator superfamily. DesD is predicted to transport cortisol from the gut lumen into the bacterial cytoplasm. RNA-Seq analysis also identified a cortisol-inducible ABC-type multidrug transport system predicted to export 11β-hydroxyandrostenedione (9).

The conversion of cortisol to 11β-hydroxyandrostenedione by C. scindens is an important step in a multi-species gut microbial endocrine pathway that results in the formation of 11β-hydroxyandrogens (9). Previous studies identified a number of cortisol metabolites following incubation with mixed human or rat fecal bacteria, including 3α,11β,17β-trihydroxy-5β-androstane and 3α,11β,17β-trihydroxy-5α-androstane, demonstrating the ability of gut microbes to form 5α-, 17β-reduced metabolites (3, 4), which are potent androgens. Some of the taxa responsible for these biotransformations have been identified. Clostridium paraputrificum is known to express 3α-HSDH and 5β-reductase (19), Clostridium innocuum is reported to encode 3βHSDH and 5β-reductase (20), Clostridium sp. Strain J-1 encodes 3α-HSDH and 5α-reductase (19), capable of reducing the 3-oxo-Δ⁴ bonds in ring-A. 17α-HSDH has been reported in C. scindens VPI 12708 (21), while 17β-HSDH has been detected in Eggerthella. lenta DSM 2243 and C592 (22). The genes encoding these reductases have yet to be identified, but are likely to affect steroid-17,20-desmolase activity in the gut. E. lenta also expresses 21-dehydroxylase (23), and 21-dehydroxylated metabolites are detected following incubation of cortisol in fecal suspensions (3, 4). Due to the absolute requirement for the C21-hydroxy group, steroid-17,20-desmolase and cortisol 21-dehydroxylation by E. lenta are competing reactions in the gut (23).

Extrapolations from radiometric studies in nonhuman primates (24) and data from humans (25) suggest that 15% of the 15 mg of cortisol synthesized daily is detected in the gut. Based on these studies, the concentration of cortisol metabolites in the ∼0.4L human colon (26) is ∼15 μM. 11β-hydroxyandrostenedione is now viewed as a pro-androgen, whose downstream 5α-, 17β-reduced metabolites are now thought to play an important role in activation of androgen receptor (27, 28). The field of microbial endocrinology has emerged and recognizes the need to characterize androgen production by C. scindens (29–34). Many host immune cells express androgen-receptor (35, 36), and it is possible that C. scindens affects immune function through generation of androgens. Androgens directly affect bacterial growth and virulence (37–39). Intriguingly, steroids have been shown to influence Clostridium difficile germination (40). 11β-hydroxy-androgens are also inhibitors of the host enzyme 11β-HSD2 (41, 42), but in kidney and vascular smooth muscle may be pro-hypertensive (42). Inhibition of 11β-HSD2 in colonocytes may be protective against colorectal cancer (43). Prostate tissue contains 11β-hydroxyandrogens despite inhibition of host enzymes involved in synthesizing androgens (chemical castration) (27, 28). The role of 11β-hydroxyandrostenedione formation by the gut microbiota in androgen accumulation in the prostate is unknown; however, we discovered that bacteria isolated from urine also express the bacterial steroid-17,20-desmolase pathway (11), and the majority of cortisol excretion from the body is via the urine (24, 25). Of note, we recently reported identification of the steroid-17,20-desmolase gene cluster in the genomes of urinary bacterial isolates and confirmed steroid-17,20-desmolase activity in an isolate of Propionimicrobium lymphophylum (11). Verification that desA and desB genes encode steroid-17,20-desmolase is an important step toward the application of nucleic acid-based approaches, such as quantitative polymerase chain and metagenomic sequencing, in determining whether correlations exist between levels/expression of desAB and host phenotypes and disease states.

As gut metagenomic sequencing becomes less costly and more widespread, assigning function to particular nucleic acid and deduced amino acid sequences will be important for hypothesis generation and testing. The desA and desB genes are a clear case in point. The desABCD operon is annotated as participating in carbohydrate metabolism, due largely to the desA and desB gene product homology to N-terminal and C-terminal TPP-dependent transketolases, respectively (Fig. 1). Early culture-based studies established that steroid-17,20-desmolase activity is induced by cortisol (6–8). Utilization of RNA-Seq to identify differentially expressed genes, coupled with biochemical characterization of the recombinant gene products, was necessary to determine the nucleic acid sequences encoding this bacterial steroid biotransforming pathway (9).

A recent report described side-chain cleavage of cholesterol by Mycobacterium tuberculosis and revealed a thiolase-derived enzyme, a steroid-aldolase, involved in microbial steroid-degradation (44) (Fig. 7). It is therefore possible that additional side-chain cleaving gut microbial members will be identified that are capable of aldolase or transketolase cleavage of host and dietary steroids in the gut. This is in contrast to eukaryotic cells, which typically utilize oxygen-dependent P450 monooxygenases (45).

Fig. 7.

Schematic representation of side-chain cleavage reactions for steroid-17,20-desmolase from C. scindens ATCC 35704, steroid-transaldolase from M. tuberculosis, and Cytochrome P450 monooxygenase (CYP11A1) in Eukaryotes.

The low turnover number of DesAB is within the range of previously reported prokaryotic enzymes involved in secondary metabolism (46), and may indicate that selection pressure may be less stringent for optimization of enzyme rate for functions that may not contribute greatly to organismal fitness. Studies are now ongoing to understand important amino acid residues involved in substrate-binding to steroid-17,20-desmolase and catalysis, as well as the identity and fate of the side-chain. Development of a novel enzyme-coupled continuous spectrophotometric assay to quantify steroid-17,20-desmolase activity against cortisol is expected to hasten results from these efforts. Current limitations of the assay with respect to androgen specificity of rCL17HSDH may be addressed in the future, either through protein engineering or by identifying additional naturally occurring 17α- or 17β-HSDHs with broader substrate recognition. In addition, future studies will be required to determine the identity of the side-chain product and whether an aldose acceptor is utilized in vivo. A detailed understanding of the steroid-17,20-desmolase will be important in uncovering how anaerobic bacteria coopted pentose-phosphate pathway enzymes for steroid side-chain cleavage.