| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Biochemistry, Vanderbilt University School of Medicine, Nashville, Tennessee 37232
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
estradiol, female reproductive tract, mechanisms of hormone action, uterus
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
It was established almost 50 years ago that providing retinoic acid (RA) in place of a source of retinol in the diet prevented these changes in the epithelium [4, 5]. Dietary RA can restore normal uterine epithelium and maintain fertility, although retinol itself is required for successful parturition [3, 6, 7]. We now know that the biological effects of vitamin A are mediated through the action of nuclear receptors of the RAR and RXR families and that RA is the active hormonal form that binds to the receptors to control gene transcription [8]. This is presumably the mechanism by which activated stem cells are directed to an appropriate rate of cell division and subsequent terminal differentiation.
Any stimulation of stem cell division hastens or induces the keratinized phenotype in sensitive epithelia of the vitamin A-deficient animal. Mechanical abrasion of the trachea in the vitamin A-deficient rat accelerates epithelial keratinization during the repair process [9, 10]. Another example is that the epithelium of the uterus undergoes keratinization in rats deprived of retinol [1]. However, if the rats have been ovariectomized, making the uterus essentially quiescent, no keratinization is observed. If the ovariectomized animals are put under constant, exogenous estrogen stimulation to stimulate uterine growth, keratinization does occur, as in the normal, vitamin A-deprived rat that has continued to maintain the estrous cycle [11]. Interestingly, a recent report suggests that the mouse responds differently, with ovariectomy accelerating the metaplasia in the uterus of the vitamin A-deficient animal [12].
We have reported that administration of estrogen to the prepubertal rat will, within 12 h, induce in the uterine lining epithelium the ability to synthesize RA from retinol [13]. This is accompanied by the appearance of cellular RA-binding protein, type II (CRABP[II]), in the epithelial cells. In the mature rat, CRABP(II) is expressed in the uterine lining epithelium only during estrus, the state that is preceded by an increase in circulating estrogen [14]. Whether the cells are also capable of RA synthesis at this time has not yet been established. However, we have observed that the glandular epithelium of the human breast expresses CRABP(II) and is capable of RA synthesis [15]. The same may be true for the uterine epithelium of the mature rat during estrus.
The production of RA from retinol is by two sequential oxidations, first producing retinal that is then converted to RA. We sought to identify the enzyme catalyzing the conversion of retinol to retinal in the rat NMU mammary epithelial cell line capable of synthesizing RA from retinol and to determine if it also was present in CRABP(II)-expressing cells of the rat uterus at estrus.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
These studies were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, with the oversight of medical center veterinarians and approval by our local Institutional Animal Care and Use Committee.
Retinoids
All-trans-retinol (Sigma, St. Louis, MO) was dissolved in dimethyl sulfoxide (DMSO; 20 mM) and stored at -70°C. Immediately before use, 0.5 ml was added to 6 ml of ethanol containing NaBH4 (
25 mg) to reduce any contaminating retinal. After 5 min of incubation at room temperature, an equal volume of 4 M NaCl + 0.15 M KCl was added and the retinol extracted with hexane. The hexane extract was dried under a stream of N2, and the retinol was redissolved in DMSO and examined spectrophotometrically by dilution of an aliquot into ethanol and monitoring the spectrum. The concentration was calculated using a molar extinction coefficient for retinol of 52 480 at A325. Retinal and RA standards used for high-performance liquid chromatography (HPLC) were from Sigma.
Cell Culture
NMU cells, a rat mammary carcinoma cell line, and Cos-7 cells were obtained from ATCC (Manassas, VA) and cultured in 50:50 (v/v) Dulbecco modified Eagle medium/Ham F12 with 10% (v/v) fetal bovine serum (normal media).
Transient transfection of expression vectors for estrous uterus epithelial retinol dehydrogenase (eRolDH) and aldehyde dehydrogenase (ALDH) 6 into Cos-7 cells was performed using SuperFect from Qiagen (Valencia, CA) according to the manufacturer's protocol with slight modification. Cells were seeded 12 h before transfection to reach approximately 70–80% confluence. The DNA (5 µg) and SuperFect (30 µl) were complexed for 10 min before application to the cells in 1 ml of normal media for 2.5 h. Cells were rinsed with PBS, normal media replaced, and transfection assayed after approximately 36–40 h. Empty vector was transfected as a control.
Reverse Transcription-Polymerase Chain Reaction for Short-Chain Alcohol Dehydrogenases
Examination of expression of known short-chain alcohol dehydrogenases (SCADs) and screening for possible novel SCADs was performed by reverse transcription-polymerase chain reaction (RT-PCR) of oligo-dT-primed cDNA using RNA from NMU cells or rat liver as template. The RNA was isolated using the RNeasy Mini kit (Qiagen) according to the manufacturer's protocol. The cDNA was synthesized with Superscript II from Invitrogen (Carlsbad, CA). Primers for PCR to highly homologous regions of SCAD family members were as follows: oligo "A" (forward), 5'-CTGTGACTCGGGCTTTGG-3'; oligo "S" (forward), 5'-CTGGTNAAYAATGCTGG-3'; oligo "S" (reverse), 5'-CCAGCATTRTTNACCAG-3'; oligo "B" (forward), 5'-GGYTACTGCRTCTCCAAG-3'; oligo "B" (reverse), 5'-AWACTTGGAGAYGCAGTA-3'; and oligo "C" (reverse), 5'-CTTGGCATCCCARCCWGG-3'. To screen for SCAD expression, PCR with these primers in several combinations was performed for 33 cycles, with denaturation for 45 sec at 94°C, annealing for 45 sec at 45°C, and elongation for 1 min at 72°C. Candidate bands were excised from agarose gels, gel-purified with the QIAQuick gel extraction kit (Qiagen), and cloned with the pGEM-T Easy kit (Promega, Madison, WI). Sequencing of cloned PCR products was performed by Vanderbilt's DNA Sequencing Shared Resource, a part of the Vanderbilt-Ingram Cancer Center.
Complementary DNA Library Screening
The ZAP Express cDNA library used was prepared by Stratagene (La Jolla, CA) from mRNA obtained from the rat uterus at estrus. Approximately 2 x 105 plaques were screened with the 233-base pair (bp) "AS" PCR fragment as probe and labeled with the Redi-Prime kit (Amersham, Piscataway, NJ), with hybridization overnight at 65°C. Filters were washed for 20 min at low stringency (room temperature in 40 mM sodium phosphate buffer [pH 7.2], 1 mM Na2EDTA, 5% SDS, and 0.5% BSA), then at high stringency for 20 min (65°C in 40 mM sodium phosphate buffer [pH 7.2], 1 mM Na2EDTA, and 1% SDS). Five independent clones were purified through two more rounds of screening using the same conditions. The plasmids were excised from phages according to the Stratagene protocol and then sequenced. Three of the five clones were confirmed as positive, with overlapping sequences containing the "AS" fragment sequence. These overlapping sequences were assembled to give 1636 bp of sequence containing an open reading frame of 960 bp beginning at position 176. The sequence has been deposited in GenBank with accession number AF337953 and, for purposes of this work, is termed eRolDH.
Construction of Mammalian Expression Vectors
To construct a mammalian expression vector, the open reading frame of eRolDH was obtained by PCR from the estrous uterus cDNA library using the forward primer 5'-CCGGAATTCATTATGCTGCTTTGGGTGTTG-3', containing an EcoRI restriction site, and the reverse primer 5'-CGCGGATCCTCACACAGCTTGGGGATTG-3', containing a BamHI site. The PCR conditions were as described above, except for an annealing temperature of 55°C. The gel-purified PCR product was ligated into pcDNA3.1(-) for sense orientation or into pcDNA3.1(+) for antisense orientation used for probe transcription (see below). The proper sequence of the expression vector was confirmed by sequencing. The expression vector for human ALDH6 was as described elsewhere [16].
Assay for Retinal and RA Production in Cultured Cells
NMU or transiently transfected Cos cells in 60-mm dishes were incubated with 2 µM retinol in normal media containing 10 µM BSA as carrier for 4 h. Cells were then scraped free and cells and media extracted together as described below. For quantitation, cells from duplicate plates were trypsinized, pelleted, and resuspended in homogenization buffer (20 mM Hepes [pH 7.4], 150 mM KCl, 2 mM Na2EDTA, and 1 mM ßME with protease-inhibitor cocktail from Roche [Indianapolis, IN]). Cells were lysed by sonication and the lysate used for protein assay with the BCA kit (Pierce, Rockford, IL). Retinoids in culture media were extracted and analyzed as described below.
Retinoid Extraction and HPLC
Extraction and analysis of retinal and RA were as described previously [13]. Briefly, samples to be analyzed were mixed with an equal volume of 4.25 M NaCl in 0.25 M KOH, and 2x volume of 100% ethanol containing 250 µg/ml of butylated hydroxytoluene. This solution was extracted twice with a 2x volume of hexane to remove neutral retinoids (retinol and retinal). To recover RA, the aqueous fraction was acidified with 0.03x volume of 6 N HCl and extracted once with hexane. Extracted fractions were dried under nitrogen and resuspended in mobile phase (92:8:0.1 hexane:dioxane:acetic acid) for injection onto the HPLC column. Separation by HPLC was accomplished on a Whatman Partisil 5 silica column (4.6 x 250 mm) (Clifton, NJ) at a flow rate of 2 ml/min. A Waters 996 Photodiode Array detector (Milford, MA) in conjunction with Millenium32 software (Waters) was used to detect retinoids, with monitoring at 370 nm for retinal and 352 nm for RA. The amount of retinoid present was quantitated by relating the peak area to peak areas from injections of known quantities of retinoid standards.
Northern Blot Analysis of eRolDH Expression
Uteri were harvested from normal female Sprague-Dawley rats at estrus or diestrus as determined by vaginal cytology or from prepubertal animals treated with eCG as described previously [17]. The RNA from cultured cells and uterine or liver tissue was obtained using Trizol Reagent (Gibco, Rockville, MD) or the RNeasy kit (Qiagen). An 
-32P-labeled uradine triphosphate (UTP) probe for eRolDH was prepared with the MAXIscript T7 in vitro transcription kit from Ambion using a template prepared by cloning the eRolDH coding sequence antisense into the pcDNA3.1(+) vector. Plasmid was linearized with AccI and gel-purified for use in the transcription reaction. This gave a template consisting of 222 bp of the 3' terminal coding sequence. Template for a rat cyclophilin internal control was purchased as linearized plasmid from Ambion (Austin, TX).
Twenty micrograms total RNA were separated on 1% formaldehyde/agarose gel and transferred to a Nytran membrane using the TurboBlotter system (Schleicher & Schuell, Keene, NH) by downward capillary action. The blot was prehybridized in 5x SSC (1x SSC: 0.15 M sodium chloride and 0.015 M sodium citrate), 5x Denhardt solution, 50% formamide, and 1% SDS for at least 3 h, then approximately 5 x 106 cpm/ml of -32P-labeled UTP antisense RNA probe for eRolDH, labeled as described above, were added and hybridized overnight at 65°C. The blot was washed twice for 15 min at room temperature in 2x SSC and 0.1% SDS, then 15 min at 42°C in 0.2x SSC and 0.1% SDS, then 15 min at 65°C in 0.2x SSC and 0.1% SDS. The blot was exposed to BioMax MS film (Kodak, Rochester, NY) overnight at -70°C with two intensifying screens. The blot was stripped by incubating at 80°C in 40 mM Tris-HCl (pH 7.5), 0.1x SSC, and 1% SDS and then hybridized with -32P-labeled UTP cyclophilin probe as described above.
Ribonuclease protection assays were performed using the RPA III kit from Ambion, with 10 µg of total RNA from multiple tissues, approximately 1 x 105 cpm of labeled eRolDH probe, and 1 x 104 cpm of cyclophilin probe per reaction according to the manufacturer's instructions with slight modification. Before addition of hybridization buffer, the precipitated RNA and probes were dissolved in 1 µl of water. After hybridization overnight at 42°C and RNase A/T1 digestion, products were resolved on 6% polyacrylamide/urea/Tris-borate-EDTA sequencing gels and visualized by autoradiography.
Antibodies
Antibodies to eRolDH were generated by the Protein and Immunology Core of the Clinical Nutrition Research Unit (directed by D.O.). A region of the eRolDH sequence with little homology to other SCAD family members, YIEKSLHRLKSSTSC, was chosen for generation of rabbit polyclonal antibodies. The peptide was synthesized by PeptidoGenic (Livermore, CA). For immunization, 2 mg of peptide were conjugated to Pierce's Imject Activated KLH and injected intradermally with Hunter's TiterMax Gold (CytRx Corp., Norcrosse, GA) as adjuvant. Rabbits were boosted i.m. after 5 wk, and serum was obtained 10 days after boosting. The immunoglobulin (Ig) G fraction was purified from serum using Protein A Sepharose from Pierce, and the antibody populations were affinity purified using the immunizing peptide immobilized on a Sulfolink column (Pierce). Rabbit polyclonal antibodies specific for rat cellular retinol-binding protein (CRBP), rat CRABP(II), and human ALDH6 (the homologue of mouse RALDH3) have been described previously [14, 16].
Western Blot Analysis and Immunohistochemistry
Tissue to be analyzed was processed by Polytron (Brinkman, Westbury, NY) in homogenization buffer twice for 30 sec on ice at 12 000 rpm. Cultured cells were trypsinized and pelleted, then homogenized in buffer by sonication for fifteen 1-sec pulses on ice. Cell and tissue lysates were centrifuged for 10 min at 10 000 x g. The supernatant liquid was centrifuged at 170 000 x g to pellet microsomes. The microsomal pellet was washed by resuspension in buffer by sonication and recentrifugation. Protein content of subcellular fractions was quantitated with the BCA protein assay kit.
Samples containing 100 µg of microsomal protein from tissues, 25 µg of microsomal protein from transfected cells, or 50 µg cytosolic protein were separated by SDS-PAGE and transferred to Immobilon polyvinylidene fluoride (Millipore, Bedford, MA). After blocking in 5% milk for 10 min, blots were incubated with affinity-purified IgG to eRolDH or ALDH6 overnight at 4°C. Secondary antibody incubation was with horseradish peroxidase-conjugated goat anti-rabbit antibody at 1:5000, followed by detection with enhanced chemiluminescence reagents (both from Amersham).
Animals at estrus and diestrus were killed in a CO2 atmosphere, and uteri to be analyzed by immunohistochemistry were dissected and fixed overnight, embedded, and sectioned as previously described [14]. After deparaffinization, rehydration, and blocking with 3% BSA, tissues were incubated overnight at 4°C with affinity-purified IgG to eRolDH, CRBP, or CRABP(II). Secondary biotin-conjugated goat anti-rabbit and tertiary alkaline phosphatase-conjugated mouse anti-biotin antibodies (Jackson Immunoresearch) were used at 1:1000 dilution and each incubated for 1 h at room temperature. Antibody staining was visualized by incubation for 15–30 min in BCIP/NPT/INT substrate from DAKO (Oxford, U.K.). After counterstaining in hematoxylin, slides were mounted with aqueous mountant from Serotec.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The CRABP(II) and, presumably, also the retinol dehydrogenase (RolDH) enzyme for which we were searching are expressed in only a single cell layer in the uterus at estrus, a period of 12 h in a 4- to 5-day cycle. This limited the amount of material available for study. During this time, however, we observed that several human breast epithelial cell lines could synthesize RA from retinol [15]. Normal human breast epithelium and several of those cell lines also express CRABP(II). We hypothesized that this observation might extend to the rat and, therefore, screened several rat breast cell lines available from ATCC. One, the NMU cell line, was capable of RA synthesis and also expressed CRABP(II). We then initiated our search for a candidate RolDH with this cell line, hypothesizing that the enzyme might also be involved in RA synthesis in the uterine epithelium.
Employing an RT-PCR strategy with primers to conserved regions of SCADs, we investigated whether mRNAs of known RolDHs were present. The primers were designed to amplify the various sequences of each of the three known rat RolDHs previously identified as present in liver, with RolDHI (GenBank accession no. RNU18762) used as the prototype (Fig. 1A). Reactions with these primers gave only two of five PCR products of the expected sizes when NMU RNA served as the template, whereas liver RNA gave products of the expected sizes for each primer pair (Fig. 1B). This suggested that the NMU cell line did not contain those known RolDHs and that the first-step enzyme for RA synthesis by these cells would be novel.
|
The two PCR products that were produced ("AB" and "AS") were cloned and sequenced. The sequence of the "AB" product was identical to ß-hydroxybutyrate dehydrogenase, a known SCAD with no activity for retinoids. This established the ability of these primers to screen for expression of other members of the SCAD family. Sequencing of the 233-bp "AS" product revealed a novel sequence, with no exact matches to any other sequence in the GenBank database, but BLAST results showed partial nucleotide similarity with a number of other SCADs, many with activity for retinoids. Using RT-PCR of RNA from other sites with either known RA synthesis or CRABP(II) expression, including the uterus of the prepubertal rat after hormone treatment, in vitro-decidualized cells from the uterus, and a cDNA library prepared from mRNA isolated from the rat uterus at estrus, we found the "AS" band but not the full pattern of SCAD products (data not shown).
To confirm the presence of this novel sequence in the rat uterus and to obtain the full-length coding sequence, we used the 233-bp fragment as a probe to screen the estrous uterine library. Of 1 x 105 plaques initially plated, three rounds of screening produced three clones that hybridized to the "AS" fragment. Sequencing showed that all three had partial overlap with each other. The sequence resulting from alignment of these clones contained a 960-bp open reading frame in addition to some 5' and 3' untranslated regions (Fig. 2). The theoretical translation of this sequence yielded a protein sequence with several motifs, confirming that it was a member of the SCAD family (shaded motifs in Fig. 2). Two motifs, GXXXGXG at residues 36–42 and YXXXK at residues 176–180, are found at the cofactor-binding site and the active site, respectively. Another highly conserved sequence with unknown function, LXNNAG, is present at residues 109–114. We have observed a fourth region of homology in many of the SCADs with activity for retinoids, HPRTRYSAGWDAK (residues 276–288 of RolDHI from rat). This sequence (partially corresponding to the location of the "C" PCR primer; see above) was not conserved in this novel sequence. The sequence will be termed eRolDH throughout to indicate that it encodes an RolDH expressed in the epithelium of the estrous uterus.
|
As with the 233-bp fragment, a BLAST search with the entire coding sequence revealed no exact match but considerable similarity to other SCAD family members with known RolDH activity (Table 1). A dendrogram showing the relationships among the SCAD members with homology to eRolDH suggests that the sequence lies on a separate branch from other RolDHs (Fig. 3). The closest match, with 87% identity, was with a human SCAD described as both a novel 3-hydroxysteroid dehydrogenase and a putative RolDH [18, 19]. The two sequences reported by these groups (GenBank accession nos. AF343729 and AY017349) are identical in the coding sequence except for a single nucleotide difference. Work by one group suggests that the enzyme is much more efficient as a hydroxysteroid dehydrogenase and is not active with retinol, whereas the other group found increased metabolism of retinol in cells induced to express the enzyme or in Cos cells transfected with the coding sequence. However, no expression of the message was found by Northern blot analysis in human uterus or mammary gland [18]. Thus, it appears doubtful that this sequence is the human homologue of rat eRolDH.
|
|
Primers designed for the open reading frame of eRolDH were used in RT-PCR with NMU RNA or the estrous uterus cDNA library and produced the expected 960-bp product of identical sequence from both templates (data not shown).
Expression of the Novel SCAD and Demonstration of RolDH Activity
To test that the novel sequence did indeed encode a functional enzyme able to oxidize retinol to retinal, a mammalian expression vector containing the eRolDH open reading frame was constructed from the 960-bp PCR product from the estrous uterus library. When the eRolDH sequence was transfected into Cos cells that were then provided with 2 µM retinol in culture, the cells synthesized significantly greater amounts of retinal than the cells transfected with empty vector (Fig. 4A). The average yield of retinal from eRolDH-transfected cells was 18.9 ± 5.3 pmol (mean ± SD, n = 16) per 60-mm dish compared to 3.9 ± 1.6 pmol (n = 14) for similar cultures of vector-transfected cells. This oxidation of retinol seen in vector-transfected cells may be nonenzymatic, because nontransfected Cos cells fixed with paraformaldehyde could also produce a small amount of retinal when incubated with retinol (data not shown). We also tested the ability of the enzyme to oxidize retinol bound to CRBP by cotransfection of increasing amounts of a CRBP expression vector. Expression of CRBP was confirmed by Western blot analysis, but no inhibition of the production of retinal was observed (data not shown).
|
To our knowledge, all SCADs identified to date with activity with retinoids also have some activity with steroids, including the 11-cis-RolDH of the retina, an enzyme whose accepted physiological role is as an RolDH [20]. We did not examine this enzyme for possible steroid dehydrogenase ability.
We prepared polyclonal antibodies against a peptide in the C-terminus of eRolDH (underlined in Fig. 2) that was considerably different in sequence from other SCADs. Western blot analysis of microsomal protein from transfected cells probed with these antibodies showed expression of an approximately 32-kDa protein only in eRolDH-transfected cells, correlating with RolDH activity (Fig. 4B). No immunoreactivity was observed in microsomes of vector-transfected cells or in cytosol from either cells, confirming microsomal localization as observed for most other SCAD family members.
Cotransfection of eRolDH and an Aldehyde Dehydrogenase Confers RA Biosynthetic Ability
To investigate further the ability of eRolDH to function as an RolDH, we tested whether the enzyme could function in an RA biosynthetic pathway. The known retinaldehyde dehydrogenase RALDH3 (the homologue of human ALDH6) is present at sites of RA synthesis in the embryo and retina [21–23], and we have also found it expressed in human breast epithelial cells that make RA, but not in a cancer cell line that is impaired in RA synthetic ability [16]. We cotransfected plasmids with the eRolDH and human ALDH6 sequences into Cos cells and analyzed RA production from retinol provided in culture. When both enzymes were present, the Cos cells showed a potent synthesis of RA (Fig. 5A). When ALDH6 was transfected without eRolDH, considerably less RA was produced. No RA was recovered from cells transfected with eRolDH only or with empty vector. Interestingly, the level of activity from cotransfected Cos cells was significantly higher, by approximately sevenfold, than the RA synthetic ability of a similar culture of NMU cells (data not shown). Thus, eRolDH can function as an RolDH both independently and in a pathway of RA biosynthesis. Again, we confirmed appropriate expression of transiently expressed proteins by Western blot analysis of cytosol for ALDH6 or microsomes for eRolDH (Fig. 5B).
|
Expression of eRolDH mRNA
As previously noted, CRABP(II) expression in the uterus occurs only at estrus. Consequently, we probed a Northern blot of total RNA (20 µg) from the estrous and diestrous uterus as well as RNA from NMU cells and liver using an antisense RNA probe transcribed from the last 222 nucleotides of the eRolDH coding sequence. A major transcript of approximately 3.9 kilobases (kb) was observed for both NMU cells and the estrous uterus (Fig. 6A, lanes 1 and 2). Significantly, no hybridization was observed with diestrous uterus or liver RNA (lanes 3 and 4). Thus, in the uterus, the mRNA for eRolDH appears to follow the same regulation pattern as CRABP(II).
|
The epithelium of the uterus of the eCG-treated prepubertal rat is another system in which hormonal treatments induce the expression of CRABP(II) and synthesis of RA [13]. To examine expression in this system and to verify the absence of this mRNA in the diestrous uterus and liver, we used the same eRolDH probe in the more sensitive ribonuclease protection assay. As expected, protection of a 222-bp fragment was observed with RNA from NMU cells, the estrous uterus, and the uterus of the eCG-treated prepubertal rat (Fig. 6B, lanes 1, 2, and 5). No protection was seen with RNA from the uterus at diestrus or from the liver (lanes 3 and 4). This confirmed the presence of eRolDH mRNA as seen by Northern blot analysis and extended expression of eRolDH to another place at which RA synthesis in culture, concurrent with CRABP(II) expression, has been observed.
Western Blot Analysis of eRolDH Expression
To confirm that eRolDH protein is expressed in NMU cells and to compare its apparent size to that of eRolDH produced by transfection, we probed a Western blot of microsomes prepared from eRolDH-transfected Cos cells and from NMU cells with anti-eRolDH antibodies (described above). The only immunoreactivity observed was a band of approximately 32 kDa for both sources (Fig. 7A). Microsomes obtained from the estrous and diestrous uterus were tested in the same procedure, revealing the 32-kDa band in the estrous but not in the diestrous sample, which is consistent with mRNA expression (Fig. 7B).
|
Immunolocalization of eRolDH in the Uterus
To determine sites of expression of eRolDH, immunohistochemical analyses of sections of rat uterus at estrus and diestrus were performed with the anti-eRolDH antibodies described above. Staining with anti-eRolDH antibodies localized to the luminal but not to the glandular epithelium of the uterus at estrus (Fig. 8A, left). No specific staining of any other cell type was observed. The apparent staining of macrophages that can be seen was also present in sections incubated without primary antibody. At diestrus, epithelial staining was not present, nor was staining observed in any other cell type (Fig. 8A, right).
|
We then compared the cellular expression of CRABP(II) and CRBP to that of eRolDH by immunohistochemistry. Expression of these proteins paralleled that of eRolDH in the luminal epithelium: staining at estrus (Fig. 8B, i–iii) and lack of staining at diestrus (Fig. 8B, iv–vi). As we observed previously, staining with anti-CRBP antibodies at diestrus localizes to the stromal cell layer just underlying the epithelium but then appears in the epithelium at estrus, which is consistent with a change in the retinol metabolism of those cells. To ensure the specificity of the eRolDH antibodies, we stained sections of liver, where other SCADs but not eRolDH are expressed, and observed no staining in these sections (data not shown). Thus, we observed coordinate regulation of several parts of a putative RA biosynthetic pathway in epithelial cells of the uterus at estrus but not at diestrus.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although a human cDNA with high sequence identity to this sequence has been isolated [18, 19], several lines of evidence suggest that it is not the homologue of rat eRolDH. First, no expression of this human sequence was observed in human mammary gland or uterus [18], whereas eRolDH expression at both the mRNA and protein levels was found in a rat mammary cell line and in rat uterus in the present study. Second, we could not detect expression of this human sequence by ribonuclease protection assay (data not shown) in human mammary cell lines that we have previously reported to synthesize RA from retinol [15].
Consistent with our proposal that this is the enzyme involved in RA biosynthesis in the uterus is the appearance of eRolDH in response to administration of eCG to the prepubertal rat, which was previously demonstrated to induce RA biosynthesis capability and CRABP(II) expression in the epithelium [13]. Expression of eRolDH in the uterus of the mature animal was limited to the epithelium at estrus, but not during diestrus. Because of technical limitations, it is difficult to obtain primary cultures of epithelial cells from normal cycling adult uterus at estrus, so we cannot demonstrate that this epithelium also synthesizes RA. However, the fundamental changes induced in the uterus by eCG treatment of a prepubertal rat (i.e., induction of the proliferative phase of the uterus, presumably under control of estrogen secretion from the ovary in response to the FSH-like activity of eCG) is thought to mimic the normal proliferative phase brought about by increased estrogen levels culminating at the estrous phase of the cycle.
The factors that produce the estrous state in the normal cycle of the mature animal induce the expression not only of this novel RolDH in the epithelium but also the expression of CRBP and CRABP(II). In addition, preliminary results suggest that one of the retinaldehyde dehydrogenases, RALDH2, is expressed in the uterine epithelial lining at estrus but not at diestrus (unpublished results). Clearly, this is an important and, perhaps, the first demonstration of the coordinate regulation of proteins in an RA biosynthetic pathway in the normal physiology of the adult animal.
Previous work in ovariectomized animals showed that the epithelium of the relatively quiescent uterus had little, if any, need for vitamin A, because no metaplastic phenotype was observed in these animals on a vitamin A-deficient diet. However, constant stimulation by estrogen of vitamin A-deficient animals resulted in a dramatic induction of the vitamin A-deficient phenotype seen in normal, deficient animals [11]. During the normal estrous cycle, no proliferation of uterine epithelium is required unless the animal becomes pregnant, and in fact, the purpose of the cycle is simply to prepare the uterus for that event. Thus, it would appear that the components of a pathway for synthesizing RA are induced by estrogen and are present in the epithelium at the estrous phase of the cycle in preparation for the expected proliferation required in pregnancy. Whether RA is actually synthesized at estrus or synthesis is initiated after mating to control the differentiation of stem cells during the subsequent proliferation remains to be investigated. Of perhaps more importance is that this system of RA biosynthesis is not specific for the uterus, because it is present in the rat mammary NMU cell line. However, this is a line derived from tumor tissue. Consequently, the question of whether it is present or can be induced in other vitamin A-dependent epithelia at the time of stem cell proliferation for repair or growth will be an important area of future study.
|
FOOTNOTES |
|---|
2 Correspondence: David E. Ong, Department of Biochemistry, Vanderbilt University School of Medicine, 610 MRB, 23rd Ave. at Pierce, Nashville, TN 37232-0146. FAX: 615 343 7347; david.e.ong{at}vanderbilt.edu
Received: 3 May 2002.
First decision: 29 May 2002.
Accepted: 12 June 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-hydroxysteroid dehydrogenase: unique tissue distribution and catalytic properties. J Biol Chem 2001 276:22278-22286-hydroxysteroid/retinoid short-chain dehydrogenase/reductase. Endocrinology 1999 140:5275-5284-hydroxysterol short-chain dehydrogenase isozyme. J Biol Chem 1998 273:17910-17916-hydroxysterol short-chain dehydrogenase isozyme. Gene 2000 251:175-186[CrossRef][Medline]-hydroxysteroids. J Biol Chem 1998 273:19778-19785-hydroxysteroid dehydrogenases from rat and human prostate. J Biol Chem 1997 272:15959-15966
ß)-hydroxysteroid epimerase. J Biol Chem 2000 275:29452-29457-hydroxysterol short-chain dehydrogenase. J Biol Chem 1997 272:33125-33131This article has been cited by other articles:
|
|
 |
|
  M. Zhang, P. Hu, C. R. Krois, M. A. Kane, and J. L. Napoli Altered vitamin A homeostasis and increased size and adiposity in the rdh1-null mouse FASEB J, September 1, 2007; 21(11): 2886 - 2896. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Arslan, L. I. Gold, K. Mittal, T.-C. Suen, I. Belitskaya-Levy, M.-S. Tang, and P. Toniolo Gene expression studies provide clues to the pathogenesis of uterine leiomyoma: new evidence and a systematic review Hum. Reprod., April 1, 2005; 20(4): 852 - 863. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X.-H. Li, B. Kakkad, and D. E. Ong Estrogen Directly Induces Expression of Retinoic Acid Biosynthetic Enzymes, Compartmentalized between the Epithelium and Underlying Stromal Cells in Rat Uterus Endocrinology, October 1, 2004; 145(10): 4756 - 4762. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
