The mouse genome project has made available the sequence of a cluster containing eight members of the aldo-keto reductase family located in the chromosome 13 20. Using specific primers, we have cloned a cDNA fragment containing the entire coding region of the gene identified as AKR1C21 without notification of any activity. Sequence analysis of the cDNA fragment shows that it encodes a putative protein of 323 amino acids having a calculated MW of 36745 Daltons. We have deposited the sequence in GenBank under the accession number AY742217. Comparison of the deduced amino acid sequence with that of other aldo-keto reductase members (Figure 1) shows that mouse 17α-HSD shares 77, 70 and 72 % amino acid sequence identity with mouse type 5 17β-HSD, 3α-HSD and 20α-HSD respectively. It also shares 70, 69, 71, 72, 72, 73, 74 % with rat 3α-HSD and 20α-HSD, rabbit 20α-HSD as well as human 3α-HSD1, 17β-HSD5, 3α-HSD3 and 20α-HSD. The genomic structure as derived from public data bank indicates that mouse 17α-HSD gene is 12.5 kb long covering 9 exons and is transcribed into a 1.2 kb mRNA.

Preliminary data using [¹⁴C]4-dione as substrate showed that the product resulting from the transformation of 4-dione by the enzyme overexpressed in HEK-293 cells is different from T, the metabolite produced by a 17β-HSD activity. To verify the identity of this metabolite, we used HPLC to analyze extracts of culture medium of HEK-293 cells over expressing 17α-HSD in presence of [¹⁴C]4-dione (Figure 2B). We also analyzed metabolites obtained after [¹⁴C]4-dione incubation with the pure enzyme (Figure 2C). Comparison of the HPLC elution and TLC migration profile of the product with a commercial epiT standard clearly indicates that the product is epitestosterone, the 17α-epimer of testosterone. Using TLC separation, we were also able to verify the identity of epi3α-diol, which is produced from ADT, in addition of epitestosterone In order to confirm the 17α-HSD nature of the activity, comparison of the products obtained from incubation of 4-dione (Figure 3A) and ADT (Figure 3B) with 17β-HSD type 5 and 17α-HSD was accomplished.

Using 17α-HSD stably expressed in HEK-293 cells, we have characterized the substrate specificity of this enzyme in cultured cells by comparison with non transfected cells. As illustrated in Figure 4 and Table 1, mouse 17α-HSD possesses two activities. The strongest activity is 17α-reductase activity responsible for the transformation of 4-dione into epiT as well as the transformation of ADT, 5α-dione and DHEA into epi3α-diol, epiDHT and epi5-diol, respectively. The second activity is 3α-reductase, responsible for the transformation of DHT to 3α-diol, 5α-dione to ADT and DHP to allopregnanolone. Moreover, we found that purified enzyme displays the same catalytic activity as the overexpressed enzyme in intact cells.

Using Q_RT-PCR to quantify mRNA expression levels of 17α-HSD in male and female (Table 2) mouse tissues, we found that the expression of this enzyme is highly specific for the mouse kidney, while type 5 17β-HSD, the enzyme catalyzing the formation of T, is markedly expressed in the liver. Both enzymes are expressed at very high levels, more than 15 millions copies/μg total RNA. Since the estimated amount of total RNA in a single liver cell is about 50 pg 21, we consider that an expression level of 20,000 copies/μg total RNA corresponds to approximately 1 copy/cell in an homogenous cell population. Therefore, 15 millions copies/μg total RNA should correspond to 750 copies/cell.

A cDNA fragment of a coding region of mouse 17α-HSD (AKR1C21) was amplified by polymerase chain reaction (PCR) from a mouse spleen cDNA and the oligoprimer pair (5'-ggg-gtc-gac-ttt-gaa-gag-gga-cac-ata-atg-a-3' and 5'-ggg-ggt-acc-acc-cat-agg-ctt-ttc-agg-aga-3') derived from the DNA sequence NM_029901 from GenBank database. Mouse spleen cDNA was obtained by reverse transcription of 20 μg of mouse spleen total RNA using 400 U SuperScript II reverse transcriptase (Invitrogen, Burlington, Ontario, Canada) and oligo-d(T)24 as primer in a reaction buffer containing 50 mM Tris-HCl pH 8.3, 75 mM KCl, 3 mM MgCl₂, 10 mM DTT and 0.5 mM dNTPs. The resulting cDNA fragments were subcloned into a pCMVneo expression vector (pCMVneo-m17α-HSD) which was subsequently used to produce a stably transfected HEK-293 cell line. Plasmid DNA was prepared using the Qiagen Mega Kit (Qiagen, Chatsworth, CA, USA). Sequence of the pCMVneo-m17α-HSD was determined using an ABI 3730/XL automatic sequencer, to verify the identity of the amplified sequence.

Stable transfection of pCMVneo-m17α-HSD into HEK-293 cells was performed as described previously 29. Briefly, HEK-293 cells were cultured in 6-well falcon flasks to approximately 3 × 10⁵ cells/well in Minimum Essential Medium (MEM) (Invitrogen) supplemented with 10% (vol/vol) FCS (Wisent, Saint-Bruno, Québec, Canada) at 37°C under a 95% air- 5% CO₂ humidified atmosphere. Five μg of pCMVneo-m17α-HSD was transfected using Exgen 500 reagent (Fermentas, Burlington, Ontario, Canada). After 6 h incubation at 37°C, the transfection medium was removed and 2 ml of MEM were added. Cells were further cultured for 48 h, then transferred into 10 cm Petri dishes and cultured in MEM containing 700 μg/ml of G-418 (Invitrogen) in order to inhibit the growth of non-transfected cells. The medium containing G-418 was changed every two days until resistant colonies were observed.

The cDNA encoding mouse 17α-HSD was subcloned into a pGEX vector (Amhersham Biosciences, Baie d'Urfé, Québec, Canada) and expressed in Escherichia coli BL21(DE3) pLysS as a fusion protein with glutathione-S-transferase (GST). The fusion protein was isolated using a Glutathione-Sepharose 4B column, as described by the manufacturer. The purified 17α-HSD enzyme was separated from GST by digestion with thrombin. 17α-HSD is not adsorbed on the DEAE column and is recuperated in the flow-through fraction, while GST and fusion protein remain on the column. With this method we obtain about 10 mg of a purified enzyme preparation per 100 ml of cell culture. Analysis of samples obtained during the purification process was performed using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), as described before 30. The broad range molecular weight standards was purchased from Bio-Rad (Missisauga, Ontario, Canada).

Enzymatic activities of 17α-HSD were determined using both purified enzyme and the cultured HEK-293 cells stably transfected with pCMVneo-m17α-HSD, as previously described 29. Briefly, 3 μg of purified 17α-HSD were incubated with 10 mM NADPH, 0.1 μM of 4-dione in phosphate saline buffer, 50 mM, pH 7.3, for 20 minutes. For the intact cells, 0.1 μM of the [¹⁴C]-labeled steroid (PerkinElmer, Boston, Massachussetts, USA) was added to freshly changed culture medium in a 6-well culture plate. Non-transfected HEK-293 cells were used as control of the background. After incubation, the steroids were extracted with 2 ml of ether. The organic phases were pooled and evaporated to dryness. The steroids were then solubilized in 50 μl of dichloromethane, applied to Silica gel 60 thin layer chromatography (TLC) plates (Merck, Darmstad, Germany). To obtain a better separation and identification of metabolites, different solvent systems were used. Metabolites of substrates 4-dione, DHEA and 5α-dione were separated in chloroform : ether (9:1), while DHP, DHT and ADT products were separated in the toluene : acetone (4:1) solvent system. Substrates and metabolites were identified by comparison with reference steroids, revealed by autoradiography and quantified using the PhosphorImager System (Molecular Dynamics, Sunnyvale, California, USA). The enzymatic reaction was carried out using the condition in which the activity varies linearly with the enzyme concentration and incubation time, indicating that the cofactor concentration produced by the cells is in excess; the reverse reaction was consequently prevented. In our conditions, this linearity was observed at even more than 60% transformation. Determination of the kinetic parameters was done by Lineweaver-Burk graph analysis using Enzfitter software.

¹⁴C-Labeled steroids were analyzed using Waters NovaPak reverse-phase C18 HPLC column (3.9 × 150 mm, 4 μm). The mobile phase was MeOH/H₂O/THF (26 : 56 : 18 v/v), with a flow rate of 0.7 ml.min^-1. Radioactivity was monitored in the eluent using Beckman 171 HPLC Radioactivity Monitoring System. Unlabelled steroids (4-dione, DHEA, 5α-dione, ADT, T, 5-diol, DHT, 3α-diol, epiT, epi5-diol, epiDHT and epi3α-diol) were obtained from Steraloids (Newport, Rhode Island, USA) and used as standards.

Total RNA of indicated tissues was isolated using Trizol Reagent (Invitrogen, Burlington, Ontario) as described by the manufacturer. Twenty μg of total RNA was converted to cDNA by incubation at 42°C for 1 h with 400 U SuperScript II reverse transcriptase (Invitrogen), using oligo-d(T)24 as primer in a reaction buffer containing 50 mM Tris-HCl pH 8.3, 75 mM KCl, 3 mM MgCl₂, 10 mM DTT and 0.5 mM dNTPs. The tissues were collected in C57BL6 mice at 12-15 weeks of age obtained from Charles River, Inc. (Saint-Constant, Québec, Canada). The mice were housed individually in vinyl cages. The photoperiod was 12 h of light and 12 h of darkness (lights on at 07:15 h). Certified rodent food (Lab Rodent Diet) and tap water were provided ad libitum. The experiment was conducted in an animal facility approved by the Canadian Council on Animal Care (CCAC) and the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC). The study was performed in accordance with the CCAC Guide for Care and Use of Experimental Animals. The collected organs were rapidly trimmed, snap-frozen in liquid nitrogen and stored at -80°C until RNA extraction.

Total RNA from pituitary gland, adrenal, liver, kidney, spleen, thymus, stomach, heart, lung, ovary, uterus, clitoral gland, mammary gland, testis, prostate and preputial gland, prepared as described above, were analyzed for the expression of 17α-HSD mRNA using quantitative RealTime RT-PCR (Q_RTPCR). cDNA corresponding to 20 pg of the initial total RNA was used to perform fluorescent-based Realtime PCR quantification using the LightCycler Realtime PCR apparatus (Roche Inc. Nutley, NJ). Reagents were obtained from the same company and were used as described by the manufacturer. The conditions for the PCR reactions were: denaturation at 95°C for 10 sec, annealing at 62°C for 5 sec and elongation at 72°C for 8 sec. Oligoprimer pairs (5'-ttg-att-gcc-ctt-cgc-tac-cag-3', 5'-aaa-tgg-cag-cag-gta-tgt-atc-gc-3') allowed the amplification of approximately 170 bp of the mouse 17α-HSD sequence. Data calculation and normalization was performed using second derivative and double correction method as previously described 31. 17α-HSD mRNA expression levels are expressed as number of copies/μg total RNA using a standard curve of Cp versus logarithm of the quantity. The standard curve was established using known cDNA amounts of 0, 10², 10³, 10⁴, 10⁵ and 10⁶ copies of cDNA and a LightCycler 3.5 program provided by the manufacturer (Roche Inc).