INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

The Ke 6 gene, encoded within the major histocompatibility complex, has been intimately linked to the development of cysts in the kidney and liver of mice (1-6). The expression of the Ke 6 gene is severely down-regulated in all murine models of PKD¹ that have been examined to date: cpk, jck (1, 3) and pcy mice (2). The inhibition of the Ke 6 gene expression using antisense deoxyoligonucleotides in embryonic kidneys in organ cultures gives rise to numerous cysts against a background of normal nephrogenesis (5, 6). Importantly, the human Ke 6 gene is located on chromosome 6p21 (7) where the human autosomal recessive PKD mutation has been mapped (8) and therefore there is a possibility that the Ke 6 gene is the primary mutation in autosomal recessive PKD. The structure of the mouse Ke 6 gene (2), cDNA (1), and protein (4) has been extensively characterized in our laboratory.

Our earlier data base searches with the Ke 6 sequence showed that it belongs to the short-chain dehydrogenase/reductase family (1). At that time, Ke 6 was most similar to bacterial oxidoreductases with no close similarity to any mammalian oxidoreductase in the data base, leaving the exact function of Ke 6 unknown. Recently, we searched the data base again with Ke 6 and found it is similar to 17HSD4 (9). This prompted us to examine Ke 6 for 17HSD activity with various androgen and estrogen substrates.

As reported here, we find that Ke 6 protein is an NAD-dependent 17HSD, making it the seventh member of this class of enzyme. It efficiently catalyzes the oxidation of estradiol, testosterone, and dihydrotestosterone and also the reduction of estrone to form biologically active estradiol. The identification of the substrate of Ke 6 allows us to postulate that regulated sex steroid metabolism plays a crucial role in the development of the mammalian kidney, and aberrations in this regulation leads to developmental problems. We propose that the developing kidney is affected by abnormal estrogen/androgen metabolism or by abnormal glucocorticoid metabolism, which could occur as a consequence of elevated sex steroid levels. Estrogen levels have been implicated in the inhibition of 11HSD enzyme in a study by Low et al. (10), where it was determined that gonadectomy of the normal female mouse stimulated the expression of the 11HSD1 gene and estradiol replacement inhibited its activity. Elevated levels of estrogens and androgens in the cpk/cpk kidney could result in a defect in glucocorticoid metabolism by the inhibition of 11HSD1.

The involvement of a steroid dehydrogenase in renal cyst formation is particularly intriguing since previously several laboratories had demonstrated that the rodent kidney is extremely vulnerable to cyst formation in response to steroid administration at a critical time in development (11-17). Moreover, the cystogenic potential of glucocorticoids has been directly demonstrated in metanephric kidneys in organ culture by us (5) and others (18). Recessive PKD in humans and mice is primarily a developmental disorder and is clinically, phenotypically, and genetically distinct from the autosomal dominant form of PKD, which manifests in adults (19). Autosomal recessive PKD in humans occurs in neonates or in early childhood and is usually fatal due to the severity of kidney maldevelopment (19).

In this study we have identified the substrates of Ke 6 and have characterized its kinetic properties and phylogeny. We have used Northern and Western analyses to show that Ke 6 is expressed in the ovary and testis, which are targets for estrogen and androgens, as is the kidney. We have also localized the expression of Ke 6 within the rat ovary. These findings support a function of Ke 6 as a 17HSD, a regulator of sex steroid concentrations in these tissues.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Phylogenetic Analysis-- The Feng-Dolittle algorithm (20) was used to construct the phylogenetic tree of mouse Ke 6 and its homologs. The method of Fitch and Margoliash (21) was then used to obtain the branch order for the sequences. Branch lengths are calculated by a linear regression analysis of the best fit of the pairwise distances and the branching order. The lengths of the branches are proportional to the relative distance between the sequences.

Recombinant GST-Ke 6 Fusion Protein-- The full-length mouse Ke 6a cDNA (13.3.1) (1) was cloned into the NotI-SalI sites of pGEX vector, and the reading frame of Ke 6 protein was checked by sequencing of the vector-Ke 6 juncture. Transformed bacteria were grown in LB containing 50 µg/ml ampicillin at 37 °C for 5-6 h. 0.1 mM isopropyl-1-thio--D-galactopyraonoside was added for induction of protein synthesis when the cell density reached a A₆₀₀ = 0.8-1.0. The cells were incubated for 2-3 h at 25-30 °C. Cells were harvested, ultrasonicated on ice, and incubated with 0.1 mg/ml lysozyme, 1% Triton X-100 at 4 °C for 15 min. After centrifugation, the supernatant was incubated with preswollen 50% glutathione-bonded Sepharose (Amersham Pharmacia Biotech). After binding for about 1 h at room temperature, the Sepharose was washed with PBS and the GST-Ke 6 protein was eluted with 10 mM reduced glutathione at room temperature. The purity and molecular weight of uncleaved and thrombin-cleaved protein was checked by SDS-PAGE and Western analysis.

Ke 6 Antibody-- The polyclonal Ke 6 antibody used in this study was generated by the injection of two rabbits (R3664 and R3691) with a 13-residue Ke 6 peptide from amino acids 54-65 synthesized by a commercial facility. The peptide was conjugated to diphtheria toxin, and immunization was carried out over a 5-month period, with intervals of 3 weeks between boosts. The antiserum from R3664 was used for Western analysis and the antiserum from R3691 was used for immunohistochemistry. The generation and characterization of these antibodies have been described previously (4).

SDS-PAGE and Western Blot Analysis-- 2 µg of affinity-purified Ke 6-GST protein and thrombin cleaved Ke 6-GST protein were separated on a 80 × 60 × 1-mm 12% polyacrylamide gel with a 4% stacking gel according to the protocol described by Laemmli (22). The electrophoresis was carried out in 25 mM Tris, pH 8.3, 200 mM glycine, 0.1% SDS at 80 V. The gel was stained with Coomassie Blue stain, destained, and photographed. For Western blot analysis, 25 ng of protein were loaded onto 12% SDS-PAGE gels and run in conditions described above. After electrophoresis, the gel was electrotransferred at 25 V in transfer buffer (200 mM glycine, 250 mM Tris base, and 20% methanol) onto Immobilon-P membranes, and Western analysis was performed as described previously (4). The primary antibody (1:250) was applied to the membrane in 1% bovine serum albumin, PBS, 0.1% Tween 20 for 1 h and then washed in PBS and incubated with the secondary antibody (peroxidase-conjugated goat antibody to rabbit IgG) in PBS, 0.1% Tween 20, washed in PBS, and developed using the ECL kit from Amersham Pharmacia Biotech. The membrane was stripped in 100 mM -mercaptoethanol, 63.5 mM Tris, pH 6.7, 2% SDS at 70 °C for 30 min. After washing the membrane was antibody-treated the same way as described above. For peptide blocking experiment, the same Western membrane was stripped and treated with the same concentration of Ke 6 antisera (1:250) preblocked overnight with 2 mg/ml peptide in PBS at 4 °C.

Enzyme Assays-- 3 µg of Ke 6 protein was used in steroid assays. The GST protein was used as a negative control and showed no steroid metabolic activity. The reaction was carried out with ¹⁴C-labeled steroids in 50 mM Hepes buffer and 1 mM NAD/NADP for oxidation reactions or potassium phosphate buffer and 1 mM NADH/NADPH for reduction reactions. The reaction mixture contained 20% glycerol, 1 mM MgCl₂, 1 mM EDTA, and 1 mM dithiothreitol as described previously (23). The ethanol concentration was constant in all reactions with different amounts of steroid. For concentration dependence curves, steroids were added to the reaction mixture at the 10 nM to 5 µM range. The reaction time and protein concentrations were within the linear period of the reaction, and all experiments were repeated at least five times. The steroids were extracted with an equal volume of methylene dichloride, dried in a Speed Vac, and applied to TLC on silica gel plates (Whatman) in 30 µl of methylene dichloride and then separated in benzene-acetone (4:1) solvent system. The plates were exposed to Kodak BioMax MS autoradiography films for all ¹⁴C-labeled compounds. For the two ³H-labeled substrates (retinol and androstenediol), the plates were exposed to the film in Transcreen filters. The signals were quantitated using the Collage software. The conversion of a specific substrate to its 17HSD product was determined by running radiolabeled standards of both the substrate and product on the TLC and identifying the conversion of steroids by their migration with the standards. To determine the conversion of dihydrotestosterone, unlabeled androstanedione was run as a standard on the TLC, and its migration was detected by UV light. Steady state reaction velocities were obtained from plots of steroid production and analyzed using the GraphPad software (GraphPad, San Diego, CA).

RNA Preparation and Northern Analysis-- Poly(A)⁺ RNA was extracted from kidney, spleen, testis, and ovaries of 6-week-old C57Bl/J6 mice as described previously (1-4). RNA was separated on a 1% agarose-formaldehyde gel followed by blotting onto nylon membrane (Magnagraph, Micron Separations). Hybridization and washing of the blots was carried out as described previously (1-4).

Immunochemistry-- The ovaries from 23-day-old Sprague-Dawley rats (Taconic Technical Service, Germantown, NY) were infiltered with O.C.T. compound (Miles, Inc., Elkhart, IN) and frozen in liquid nitrogen. Ten-micron sections of the ovaries were fixed in acetone at 20 °C. After blocking and permeabilization in 1% horse serum in 0.2% Triton X-100, the sections were incubated with rabbit anti-mouse antibody to Ke 6 protein at 1:100 dilution in PBS overnight at 4 °C. The immunostaining was done according to the procedure described previously (24). After thorough washing in PBS, the sections were incubated with goat anti-rabbit IgG conjugated to fluorescein isothiocynate for 1 h. Ovary sections were examined for the localization of Ke 6 protein on a Zeiss LSM 410 confocal microscope, and images were recorded with Adobe Photoshop software.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Phylogenetic Analysis Shows Ke 6 Is Close to Mammalian 17HSD4-- When Ke 6 was first cloned and sequenced, its homology with short-chain dehydrogenase/reductase was noted (1). The proteins closest to Ke 6 were bacterial enzymes; there was no striking sequence similarity to a mammalian enzyme. However, our recent data base search uncovered an intriguing similarity between Ke 6 and human 17HSD4, which are 35% identical. The sequence similarity extends over the entire length of the protein. The level of homology among mammalian steroid dehydrogenases is not particularly high, and there is only about 22-25% identity among other members of the 17HSD family and between 11HSD1 and 2 (25). Therefore, a 35% identity with 17HSD4 was considered significant since proteins with this degree of sequence similarity may have similar function (26, 27).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Phylogenetic relationship of Ke 6 to members of the short-chain alcohol dehydrogenase family. A phylogenetic tree of Ke 6 with human 17HSD and two bacterial proteins was constructed with the Feng-Dolittle algorithm (18). The branch distances can be used to calculate the relative distances of each protein from each other. Thus, Ke 6 is 124.2 (41.3 + 6.1 + 13.8 + 63) units from 17HSD4 and 116.9 units form 7HSD, which metabolizes bile acids. Ke 6 is 149.5 units from 17HSD2, 157.6 units from 17HSD3, and 180.8 units from 17HSD1.

Identification of the Androgenic/Estrogenic Ke 6 Substrates and Determination of the Kinetic Properties of Ke 6 Protein-- The full-length mouse Ke 6 cDNA was cloned into pGEX expression vector to produce a glutathione transferase-Ke 6 fusion protein. The purity of the recombinant protein produced in bacteria was determined by SDS-PAGE (Fig. 2A). The 61-kDa GST-Ke 6 fusion protein is cleaved by thrombin to 34 kDa which is the native molecular mass for Ke 6 protein. We determined that the Ke 6-GST fusion protein had the correct reading frame by sequencing of the vector-cDNA junction and also by Western blot analysis (Fig. 2B). Antisera raised against a 14-amino acid synthetic Ke 6 peptide recognizes the Ke 6-GST fusion protein, indicating that the correct reading frame of the Ke 6 protein is maintained in the recombinantly produced protein. For the initial identification of the substrates of Ke 6, we used radiolabeled androstenediol, estradiol, testosterone, dihydrotestosterone in oxidation reactions, and radiolabeled estrone, androstenedione, and dehydroepiandrosterone in reduction reactions with the appropriate cofactors. We found that Ke 6 uses NAD/NADH as the cofactor. Ke 6 can oxidize estradiol, testosterone, and dihydrotestosterone (Fig. 3) but not androstenediol (not shown). In the reductive mode, Ke 6 can convert estrone into estradiol, but cannot convert androstenedione (Fig. 3) or DHEA (not shown). The reduction of androstanedione by the Ke 6 enzyme has not yet been examined.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Analysis of the Ke 6-GST fusion protein. A, SDS-PAGE of Ke 6-GST fusion protein. Sigma broad range molecular mass markers (1), 2 µg of Ke 6-GST protein (2), and 2 µg of thrombin cleaved Ke 6-GST protein (3) were run on a 80 × 60 × 1-mm 12% SDS-PAGE. The electrophoresis buffer was 25 mM Tris, pH 8.3, 200 mM glycine, 0.1% SDS, and the gel was run at 80 V for 1 h. The arrows indicate the 61 kDa Ke 6 fusion protein in lane 2 and the 34-kDa Ke 6 protein released by thrombin cleavage in lane 3. The 27-kDa band in lane 3 is the GST protein. B, Western analysis of recombinant Ke 6 protein. 25 ng of thrombin cleaved Ke 6-GST fusion protein (1) and uncleaved Ke 6-GST protein (2) were run on a SDS-PAGE, 12% gel, as described above, and the proteins were electroblotted onto an Immobilon-P membrane. The transfer was carried out for 12-14 h at 25 V in transfer buffer (200 mM glycine, 250 mM Tris base, and 20% methanol). The application of the antibody and membrane washing conditions are described under "Materials and Methods." The migration of molecular mass standards are indicated on the left side of the Western blot. The 61-kDa fusion protein is cleaved to the expected 34-kDa size for the Ke 6 protein.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Identification of substrate and cofactor usage of Ke 6. 3 µg of Ke 6-GST fusion protein or GST protein were used in steroid assays as described under "Materials and Methods." The 14C-labeled steroids and cofactors for each reaction are indicated. In panel A, [14C]androstenedione was loaded in the last lane as a marker. The arrow indicates the migration of the solvent run. The positions of the substrates and products are marked on the sides of the TLC autoradiography.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   The enzyme concentration dependence of Ke 6 for estradiol, testosterone, dihydrotestosterone, and estrone. Experiments were performed using affinity-purified Ke 6 recombinant protein as described under "Materials and Methods." The activities were obtained from plots of estrone production in estradiol (black squares) oxidation reactions; androstenedione production in testosterone (open circles) oxidation reactions; androstanedione production in dihydrotestosterone (open diamonds) oxidation reactions, and estradiol production in estrone (black triangles) reduction reactions. The error bars indicate the S.E. of independent experiments and are seen when they exceed the size of the symbols.

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Lineweaver-Burk plots of the conversion of (A), estradiol (E2), (B), testosterone (T), (C), dihydrotestosterone (DHT), and estrone (E1) (D) by the Ke 6 enzyme. The results of four to six independent experiments were used for linear regression calculation (R > 0.9). Experiments were performed using affinity-purified Ke 6 recombinant protein as described under "Materials and Methods." Raw data were analyzed with the GraphPad program to calculate Km and Vmax values.

Ke 6 Does Not Have Any 7HSD, 11HSD, or Retinol Dehydrogenase Activity-- Because of the close homology of Ke 6 with bacterial 7HSD, we examined the Ke 6 protein for 7HSD activity using a spectrophotometric assay which detects changes in A_340 nm readings during the oxidation/reduction of the cofactors. Cholic acid and chenodeoxycholic acid were tested as substrates for 7HSD oxidation in the presence of NADP or NAD. There was no change in the spectophotometric readings in the Ke 6 reactions or the GST protein (negative control) reactions with cholic acid or chenodeoxycholic acid as substrates with either cofactors. In addition, we also examined if cholic acid would be able to inhibit the oxidation of estradiol. Steroid conversion assays with 100 nM of radiolabeled estradiol as the substrate and varying amount of cholic acid (10 nM, 100 nM, 200 nM, 1 µM, and 10 µM) as the inhibitor did not show any difference in estradiol oxidation. Therefore, cholic acid is not a substrate for the Ke 6 enzyme as it cannot act as an inhibitor at even very high concentrations.

Expression of the Ke 6 Gene in the Gonads-- We examined the level of expression of the Ke 6 gene within the ovary and testis because of the enzymatic activity of Ke 6 on sex steroids. Northern blot analyses were performed to compare the level of expression of the Ke 6 mRNA within the kidney, spleen, ovary, and testis (Fig. 6A). The level of the Ke 6 mRNA in the gonads is slightly less than in the kidney. Interestingly, in the ovary and the testis the length of the Ke 6 mRNA appears to be slightly longer than that in the kidney, which could be due to differences in poly(A) tail lengths. The gonadal Ke 6 mRNA does not represent Ke 6b, the spleen-specific transcript, which is about 1.2 kilobase pairs and distinctly larger than the Ke 6a mRNA. The cloning and sequencing of the Ke 6b transcript has determined the presence of an unspliced intron which is responsible for the difference in length between the Ke 6a and Ke 6b mRNAs (2). The Ke 6 gene has been fully sequenced, and there are no unknown splice acceptor-donor sites (2). Therefore, it is unlikely that the slightly longer transcript seen in the gonads is due to an alternately spliced transcript of the Ke 6 gene.

View larger version (65K): [in this window] [in a new window]   Fig. 6.   Northern and Western blot analysis of the expression of Ke 6 in kidney, ovary, testis, and spleen. In A, in the left panel, 1.7 µg of poly(A)+ RNA from normal mouse ovary (O), spleen (S), and kidney (K) and in the right panel, 3 µg of poly(A)+ RNA from normal mouse kidney (K) and testis (T) were run on 1% agarose-formaldehyde gels. Blotting, hybridization, and washing were done as described under "Materials and Methods." In B, 40 µg of protein from normal (+/+) and cpk/cpk mouse kidney (K), 40 µg of protein from normal mouse testis (T) and ovary (O), Ke 6a cDNA transfected Chinese hamster ovary cells (C), and 80 µg of protein from normal mouse spleen (S) were run on a 10% SDS-PAGE gel. The immunoblotting using Ke 6 antisera at 1:250 dilution was done as described under "Materials and Methods." In the bottom panel, the Western membrane was stripped and then treated with the same concentration of Ke 6 antibody preblocked with a 13-amino acid peptide to which the Ke 6 antisera were raised. The details of the stripping and antibody blocking protocol are described under "Matrials and Methods." The arrowhead indicates the absence of Ke 6 protein signal.

Localization of Ke 6 Protein in the Ovary-- Cryostat sections of rat ovaries were treated with polyclonal Ke 6 antisera. Intense Ke 6-specific immunostaining was visible within the cumulus cells immediately surrounding the oocyte (Fig. 7). The mural granulosa cells and theca cells were nonimmunoreactive. The 17HSD activity of Ke 6 suggests its role in regulating estrogens and androgens levels in the ovary, and consequently its role in follicle maturation and differentiation. It is very intriguing that Ke 6 is present in a key location within the ovary, the cumulus cells, which remain attached to the oocyte even when the oocyte is expelled from the follicle in ovulation. The strategic location of Ke 6 indicates that it plays a very important role in steroid regulation, which either protects the egg from its microenvironment or provides the egg with a critical steroid for its ongoing differentiation. Moreover, the Ke 6 protein could serve as a useful marker of cumulus cells because it is the only known marker of this cell type. The Ke 6 protein is also present in endothelial cells of the blood vessels within the ovary (Fig. 7C). Previously, Ke 6 staining was also seen in the glomerulus in the kidney (4). The presence of Ke 6 within endothelial cells, in general, may serve to regulate the steroid levels of the blood that enters the target tissues, and in this regard, Ke 6 could control the amount of biologically potent steroid that enters all tissues. However, if Ke 6 protein is found only in blood vessels of particular organs, for example, the kidney and ovary, then its 17HSD activity could play an important role in controlling the hormone level in blood that infiltrates that organ and reflect the local needs of that tissue.

View larger version (160K): [in this window] [in a new window]   Fig. 7.   Localization of Ke 6 protein by indirect immunofluorescence in rat ovary. Ke 6 antiserum was used for indirect staining of Ke 6 protein on cryostat sections of rat ovaries. Positive staining is seen in the cumulus cells (CC) in A-D. Ke 6 positive staining is also seen in endothelial cells (EC) of blood vessels in B. Cumulus cells and endothelial cells are indicated in panels A and B. The negative staining region represents granulosa cells and theca cells. Bar in A-C is 250 µm and in D is 1,000 µm.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The enzyme 17HSD catalyzes a key reaction in the biosynthesis and metabolism of sex steroids in both gonadal and nongonadal tissues (29). Oxidoreduction at carbon 17 of estrogens and androgens, carried out by 17HSDs, modulates the levels of biologically active estrogens and androgens within tissues. The levels of estrogens and androgens in turn control a variety of important physiological functions, such as growth, reproduction, and differentiation. To date, six isozymes of the 17HSD family have been identified and characterized, and in addition, it has been determined that retinol dehydrogenase also possesses 17HSD activity (29). The enzymes belong to the short-chain alcohol dehydrogenase family and use nicotinamide adenine dinucleotide or its phosphate as cofactors. The seven 17HSDs (including retinol dehydrogenase) have different substrate and cofactor specificities, tissue distributions, and subcellular localizations, which suggests each of these enzymes has a distinct physiological role in endocrine (glandular) and intracrine (local) metabolism of sex steroids (30).

In Table I, we have compared Ke 6 to other 17HSDs (28, 29, 31-35). Ke 6 is similar to 17HSD2, 17HSD4, and 17HSD6 in preference for NAD⁺ as a cofactor and for oxidative activity. However, Ke 6 also has reductive activity for estrone, suggesting a broader role in regulating the sex steroid concentration. These enzymatic differences, which distinguish Ke 6 from the other 17HSDs, are in agreement with the phylogenetic analysis (Fig. 1), which shows that Ke 6 is on a branch that is clearly distinct form the other 17HSDs, and that the closest enzymes to Ke 6 are an E. coli reductase and two steroid dehydrogenases: 7HSD and 17HSD4, which oxidize their substrate.

View this table: [in this window] [in a new window]   Table I Properties of the Ke 6 enzyme and other 17HSDs

An important function of the ovary is the synthesis and secretion of 17-estradiol, and this involves two cell populations (36). DHEA, produced by the adrenals, is released into the circulation and taken up by the well vascularized theca cells where it is converted into androstenediol by 3HSD (37). Androstenedione is then aromatized into estrone by P450 aromatase in the granulosa cells (37). Estrone can then be converted into the biologically active estradiol by the reductive activity of 17HSDs. Alternatively, androstenedione may be converted to testosterone by an androgenic 17HSD to testosterone, which is then aromatized to estradiol (34). The localization of Ke 6 protein within the cumulus cells confirms the presence of more than one 17HSD within the ovary. 17HSD1, an NADPH-dependent enzyme that is expressed at high levels in the ovary, is found in the granulosa cells of primary and antral follicles (38). Type 2, 3, 4, and 6 17HSDs are not present in the ovary (39). The physical separation of the two 17HSDs, Ke 6, and the 17HSD1 within the ovary, achieved by the cellular organization of granulosa cells and cumulus cells within the follicle, is interesting. The location of 17HSD1 within the granulosa cells and Ke 6 within the cumulus cells may allow the ovary to achieve a fine tuned balance of the optimal sex steroid levels. 17HSD1 has strong reductive activity and synthesizes estradiol (40). Whereas Ke 6 has both oxidative and reductive activities, inactivating and synthesizing estradiol. The smaller Michaelis constant of Ke 6 and its greater activity with estradiol over estrone suggest that it is primarily an oxidative enzyme that can switch to a reductive mode determined in the appropriate physiologic milieu. Because the Ke 6 enzyme can inactivate androgenic steroids, which 17HSD1 cannot do, the strategic location of Ke 6 within the cumulus cells could have a critical role in protecting the egg from exposure to androgen, which is known to be produced by the ovary. It is also possible that after ovulation, when the oocyte is within the oviduct, it is supplied estradiol that is synthesized within the cumulus cells by the reductive activity of the Ke 6 enzyme.

The kidney is also a highly steroid-sensitive organ equipped with several steroid dehydrogenases, which inactivate glucocorticoids and sex steroids. The particular location of specific 17HSDs and 11HSD enzymes along the nephrons may serve specific needs. Besides Ke 6, the kidney expresses 17HSD2, 17HSD4, and 17HSD6, which are also oxidative enzymes and use the same cofactor as the Ke 6 enzyme. The Ke 6 protein is located within the S3 segment of the proximal tubules and within the intercalated cells of the collecting ducts (4). 11HSD1 is present in the proximal tubules (41); 11HSD2 is present in the collecting ducts (42). The location of 17HSD2, 17HSD4, and 17HSD6 within the kidney has not been determined to our knowledge. The presence of four 17HSDs and two 11HSDs within the kidney indicates the importance of maintaining an optimal level of sex steroids and glucocorticoids which may be particularly critical during development.

Previously, we had determined that the Ke 6 gene is severely down-regulated in the kidneys of three different recessive murine models of polycystic kidney disease, the cpk, pcy, and jck mouse (1-4). The cpk/cpk mouse was initially described and characterized as a model for the human recessive polycystic kidney disease because of the gross enlargement of the kidneys due to numerous cysts (43). The disease closely resembled human autosomal recessive PKD in several aspects, recessive inheritance of the mutation, earliest cysts in the proximal tubules and later predominance of cysts in the collecting tubules, and aggressive nature of the disease with death occurring at infancy (43). Recently, Woo and Greenbaum (44), reported that the ovary of the cpk/cpk mouse is markedly underdeveloped, determined by morphological, histological, and functional studies. The cpk/cpk females failed to superovulate in response to human chorionic gonadotropin treatment. Histological analyses revealed the presence of atresic follicles, and the sizes of the ovaries and uteri were much smaller than normal (44). Recently the same laboratory also showed abnormal testicular development in the cpk homozygote males with fewer and immature seminiferous tubules.² It is possible that abnormal sex steroid metabolism is responsible for the underdevelopment of the gonads in the cpk/cpk mice. We expect to find a reduction in the expression of the Ke 6 gene in the gonads of the cpk homozygotes since we had previously found that Ke 6 mRNA levels is reduced in all tissues of this recessive murine PKD model (1). Another recessive murine model of polycystic kidney disease, the kat (kidney, anemia, and testes) mouse has been described with testicular abnormalities (45), which further indicates the tight linkage between kidney and gonadal differentiation.

The 17HSD ability of Ke 6 to promote either estradiol synthesis (reduction) or estradiol, testosterone, and dihydrotestosterone inactivation (oxidation) could be important in the development of both the kidney and the gonads of the cpk homozygotes by maintaining optimal levels of sex steroids within these organs. The development of the urinary and genital systems are closely associated in the earliest stages. The urogenital ridge gives rise to both the nephric and gonadal structures, and androgens and estrogens are likely to regulate differentiation and organization of both organ systems. Regulation of the intracellular levels of these steroids by Ke 6 and other 17HSDs would be expected to be important in the developing kidney and gonads.