| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a
b Laboratoire de Physiologie Animale, INRA, 78352 Jouy-en Josas, France
c UNCEIA, 94703 Maisons-Alfort, France
d UMR INRA-CNRS 6073, 37380 Nouzilly, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
epididymis, fertilization
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This protein was first described under the name ß-Trace by Clausen [6] as a major protein present in the cerebrospinal fluid. In 1993, Hoffmann et al. [7] showed that ß-Trace has the same sequence as the lipocalin-type PGDS characterized by Shimizu et al. [8] in the rat brain. The expression of PGDS in the central nervous system is limited to particular structures, such as the leptomeninges, choroid plexus, astrocytes, and oligodendrocytes of white brain cells [1, 9, 10]. Lipocalin PGDS was recently shown to be expressed in other organs as well, such as the liver, kidneys, and heart [11].
PGDS belongs to the lipocalin superfamily. Lipocalins are small, secretory proteins in various extracellular fluids of the body (tears, nasal secretion, luminal uterine fluid, urine). These proteins possess characteristic three-dimensional folding, delimiting an internal hydrophobic pocket involved in the binding of small, lipophilic molecules such as retinoic acid (RA) (for review [12, 13]).
Lipocalin-type PGDS is, so far, the only lipocalin reported to have an enzymatic activity (EC 5.3.99.2) [14]. The enzymatic activity is particularly well documented in the brain, where, by its prostaglandin (PG) H2 D-isomerase activity, PGDS catalyzes the transformation of PGH2 produced from arachidonic acid by cyclooxygenase (COX1 and COX2) activity into PGD2.
A major primary PG in the central nervous system [10], PGD2 is involved in sleep induction, olfactory function, and body temperature regulation [15, 16]. It is also present in other organs, such as the smooth muscles [15], and could be related to muscle relaxation/contraction and regulation of the secretion of chloride ions [17]. The biologic effects of PGD2 are mediated by a specific 7-transmembrane domain receptor coupled to the adenylate cyclase pathway [18]. This receptor is present in leptomeninges [19] and in other organs, including the epididymis in the human [20].
PGDS is considered to be a dual-function protein in the brain: an enzyme producing PGD2, and a lipophilic ligand-binding protein considered to be a carrier with high affinity for several molecules, such as retinaldehyde, RA, biliverdin, and bilirubin [21].
The biochemical and physiological functions of PGDS are uncertain in the male genital tract, although PGDS activity has been previously described in vitro in rat epididymal tissues [22] and in bull semen [4]. To our knowledge, no evidence exists for an enzymatic activity of PGDS in vivo in the epididymis. Also, no information is available concerning the hypothetical lipocalin function of PGDS in the genital tract. Furthermore, at least two other lipocalins are also present in the epididymal fluid: epididymal-RA-binding protein (E-RABP) in different species [23–25], and protein 24p3 in the mouse [26].
In the present study, which was performed on various mammalian species, we investigated the physiological functions of PGDS in the epididymis by developing two main approaches: establishment of a putative correlation between the quantity of PGDS in the ejaculate and male fertility, and ex vivo studies of its enzymatic activity and RA- and testosterone-binding abilities.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The study was performed mostly in adult rams (Ile de France). The cDNA cloning of PGDS was also performed in the stallion (Pony). Rats (Wistar) were used for the pharmacological inhibition of the PG pathway. Epididymal regions and collection of epididymal fluids by microperfusion have been described previously [27]. Epididymal tissues were rapidly frozen at -80°C for RNA and protein extraction.
The quantification of PGDS was performed in rams and bulls (Holstein, Montbéliard, Normand). No differences were observed between races for the bull. The semen of 133 bulls and 65 rams of known fertility were collected through artificial vaginas. The fertility of the bull was determined according to "indicators" given by the French national genetic evaluation program for fertility traits carried out twice a year for all dairy bulls [28]. These values are expressed as a rate of success after artificial insemination and in terms of deviation from the mean of the population. These values are estimated using all available artificial insemination records and are corrected for effects such as month of insemination and inseminator. The fertility of rams was estimated from the results of at least 250 artificial inseminations for each ram.
The seminal plasma for each species was obtained after centrifugation (10 000 x g, 10 min, 4°C) and immediately frozen and stored at -20°C.
Reagents and Chemicals
Complete Dulbecco modified Eagle medium (DMEM), X-OMAT Blue XB1 autoradiography films, and goat anti-rabbit immunoglobulin G coupled to horseradish peroxidase were purchased from Sigma Chemical Co. (St. Louis, MO). [3H]Arachidonic acid and [3H]RA were purchased from NEN Life Science Products (Les Ulis, France), acrylamide (30% [w/v] acrylamide, 0.8% [w/v] N,N-methylenebisacrylamide) and nitrocellulose membrane from Millipore (St. Quentin, France), electrophoresis calibration kit (standard proteins) and [3H]testosterone from Amersham Pharmacia Biotech (Paris, France), and Bradford assay from Bio-Rad (Paris, France). All other chemicals were molecular-biology grade from Sigma.
Quantification of PGDS by Immunoblotting in the Semen
PGDS was quantified in the semen by immunoblotting using our specific antiserum against PGDS and as described in a previous study [3]. For each experiment and each animal, the same volume of seminal plasma was loaded on the gel. After electrophoresis, transfer onto a nitrocellulose membrane, and incubation with the antibodies, revelation was performed with a chemiluminescent system (Renaissance; NEN Life Science Products, Boston, MA). The films were analyzed and signals quantified with the Imagequant software (Molecular Dynamics, Amersham). The results are expressed as the intensity signal in arbitrary units and correspond to the average of three immunoblots after normalization allowed by the presence of an internal standard in each blot. The Pearson coefficient of correlation between fertility and PGDS concentration was calculated using the CORR procedure in SAS software (SAS Institute, Cary, NC) after logarithmic transformation of the PGDS concentrations to normalize their distribution. Distribution of Log PGDS was evaluated with the UNIVARIATE procedure in SAS software.
Complementary DNA of PGDS in Ovine and Equine Species
Total RNA was prepared from 200 mg of frozen tissue according to the isothiocynate guanidinium technique described by Chomczynski and Sacchi [29]. PGDS cDNA for both species was obtained by reverse transcription-polymerase chain reaction (RT-PCR) using 5 µg of total RNA from the caput epididymis, in which PGDS expression is high [3]. The RT was performed by Moloney murine leukemia virus reverse transcriptase (Gibco-BRL, Gaithersburg, MD) with oligo(dT) (Pharmacia, Saclay, France). The specific cDNA was then amplified by PCR with the following primers: for the equine species, human primer as forward primer (5'-CACACCACTGGCACCAGGCC) and porcine primer as reverse primer (5'-GTCTCAGGTCTCGGGGTGTTGGA-3'); and for the ovine species, bovine primer as forward primer (5'-TGTGCCAGCCCGGTC-3') and porcine primer as reverse primer (5'-GTCTCAGGTCTCGGGGTGTTGGA-3'). These primers enabled us to obtain directly the full-length cDNA of PGDS. The PCR was performed for 30 cycles (94°C, 30 sec; 55°C, 30 sec; and 72°C, 30 sec). The cDNA obtained was purified of agarose gel (kit Jetsorb; Bioprobes Systems, Montreuil, France) and sequenced by the "Dye-Terminator" technique using a Perkin-Elmer sequencer (ABI 377; Paris, France).
Immunodetection of Epididymal COX1 and COX2 Cyclooxygenase Protein and Localization of COX2 mRNA by RT-PCR
Immunodetection of COX1 and COX2 in cytosolic extracts of various epididymal regions was performed using immunoblots. Epididymal tissue proteins were extracted in 50 mM Tris-HCl (pH 7.2) with a cocktail of protease inhibitors (0.5 mM EDTA; 2 mM para-aminobenzamidine; 10 µg/ml each of antipain, leuptine, bestatine, pepstatine A, and 0.5 mM PMSF); after centrifugation (100 000 x g, 4°C, 30 min), the supernatants were kept at -20°C. COX1 and COX2 were immunodetected with specific monoclonal antibodies as described by Charpigny et al. [30, 31] after electrophoresis and transfer on the nitrocellulose membrane. The mRNA expression of COX2 was studied by RT-PCR (as described above for PGDS) using specific primers [32]. Total RNAs of ovine 14-day-old embryos were used as positive control of the RT-PCR reaction, and a band was observed at approximately 0.5 kilobases.
Analysis of the PG Pathway in the Epididymis
Production of PGs in the epididymis was evaluated from arachidonic acid as described by Charpigny et al. [31] using epididymal tubules of the anterior (caput = E2) and median (corpus = E6) regions. Briefly, isolated tubules were ligatured at both ends and incubated for 5 h in the presence of 10 µCi [3H]arachidonic acid (specific activity, 6845 GBq/mmol) in DMEM medium with 20 mM Hepes and 2 mM carnitine. After incubation, the fluids, tissues, and supernatants were kept at -80°C. After extraction with methanol/acetic acid (85%/0.01%), cyclooxygenase molecules (PGs and thromboxanes) were separated and quantified by high-performance liquid chromatography and scintillation detection.
Study of the enzymatic activity of luminal PGDS from PGH2 was performed on epididymal fluids (caput and cauda) and seminal plasma. The method used was refined from various methods described previously [33–35]. The samples (1 and 10 µg for epididymal fluids and 50 and 100 µg for seminal plasma) were incubated in the presence of 5 nmol PGH2 (Cayman, Ann Arbor, MI) in 0.1 M Tris-HCl (pH 9) and 1 mM dithiothreitol (DTT) at 24°C for 10, 30, 60, and 120 sec. Quantification of PGD2 was performed using a specific assay kit (MOX-PGD2; Spi-Bio, Ann Arbor, MI) in which PGD2 is stabilized. The sensitivity of the assay is 3.24 pg/ml and is in the range of previously published results obtained for bull seminal plasma [4].
Treatment of Male Rats with Nonsteroidal Anti-Inflammatory Drugs: Effects on Fertility and Mobility Characteristics of Spermatozoa
Two cyclooxygenase inhibitors were used in this experiment: flurbiprofen (COX1 inhibitor), and indomethacin (COX1 and COX2 inhibitor) (both from Sigma). The efficiency of these drugs to inhibit PG synthesis in the male genital tract was investigated as follows: Three rats were treated daily for 7 days with 3.3 mg/kg of flurbiprofen and 1.7 mg/kg of indomethacin. Three other rats were treated with the diluent only (150 mM Na2CO3, pH 7.3). All were treated twice a day by i.p. injection. After treatment, PGE2 and PGF2 were evaluated in seminal vesicle secretions by radioimmunoassay, as previously described by Charpigny et al. [30, 31].
To study the effect of COX inhibition on rat fertility and caudal spermatozoa mobility, five adult male rats were treated with the drugs and five control rats with the diluent only for 60 days. The fertility of each male was evaluated before and after treatments by breeding with four females (per male). The caudal spermatozoa concentration and mobility were quantified after dilution in the milieu described by Jeulin et al. [36] using HTM-IVOS (Hamilton Thorne Research, Beverly, MA). The motility parameters measured included percentage of motile cells, mean track velocity, mean progressive velocity, mean linearity, mean straightness, and mean lateral head displacement. These parameters were quantified for 350 spermatozoa in each sample.
PGDS RA- and Testosterone-Binding Ability
Epididymal PGDS in the ram was purified using electroelution of spots obtained after two-dimensional (2D) electrophoretic separation of caput epididymal fluid. Two hundred micrograms of purified PGDS or total protein from caput epididymal fluid were incubated with 10 µCi (0.3 nmol) all-trans [3H]RA (NEN Life Science Products, Les Ulis, France) with or without 150 nmol nonradioactive all-trans RA (Sigma) for 1 h at 0°C in 20 mM Tris-HCl (pH 7.2) and 150 mM NaCl. RA bound by proteins was separated from unbound RA by gel filtration column at 4°C (Superdex 75, 16/60; Pharmacia, Paris, France) in the incubation buffer. Radioactivity and absorbance at 280 nm were measured in each eluted fraction. The same incubation and separation protocols were used to study testosterone binding with caput epididymal protein. Two hundred micrograms of proteins were incubated for 2 h at 4°C in the presence of 50 µCi (0.52 µmol) 1,2,6,7-[3H]testosterone (Amersham) with or without 140 µM cold testosterone (Sigma).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We estimated the quantity of PGDS in the semen of rams and bulls of known fertility using quantitative immunoblotting with a specific antibody against PGDS [3] as described in Materials and Methods. The amount of PGDS for a same volume of ejaculate was expressed as intensity of chemiluminescence (Fig. 1). A high variation in intensity was observed between animals, with a range from 0 to 100 000 arbitrary units. To investigate a putative intraindividual variation in PGDS concentration in ejaculates, PGDS was quantified in samples from 12 representative bulls collected several times (3 to 5) at an interval of several weeks. The intensity obtained in immunoblotting for the same volume of ejaculate in each animal did not present a significant variation between collected samples, suggesting that PGDS concentration is relatively constant in the semen (Fig. 2). In the ram and bull, the overall intensity of PGDS was not sampled as a normal distribution (Fig. 1, A and C). Low and zero intensities were found in bulls of all ranges of fertility. However, high intensities were observed in animals with normal (0) or high (>0) fertility, but no statistical correlation was found by analysis of log-transformed values. Distribution of the values in rams had no statistical relationship with the fertility index, although the lowest intensities were found in animals with the lowest fertility (<0) and the highest intensity in animals with normal or high fertility.
|
|
Complementary DNA and Protein Sequences of PGDS in Horses and Sheep
The protein sequences of PGDS in horses and sheep were deduced from the cDNA sequences obtained by RT-PCR using orthologous primers and epididymal RNA as described in Materials and Methods. The size of the cDNA fragments amplified and sequenced was 658 and 676 base pairs in the horse and ram, respectively (Fig. 3). According to the deduced protein sequences, PGDS possessed 194 and 191 amino acids in the horse and ram, respectively, corresponding to a theoretical molecular mass of 21.6 and 21.1 kDa, respectively. The following were found in these sequences: the characteristic lipocalin sequences GRW (41–43), TDY (123–125), and two cysteines (positions 89 and 189 for the horse and 89 and 186 for the sheep) [12]; and a cysteine at position 65 involved in enzyme activity. Sequence homology was 76% between these species and ranged from 70% to 90% with other previously sequenced species (Fig. 4). The theoretical peptide signal in the horse PGDS and the putative N- and O-glycosylations in both species have already been deduced [37]. The N-terminal sequence of the secreted protein in the ram was determined by N-terminal microsequencing in our previous study [3]. The putative sites of N-glycosylation are asparagines (N) at positions 51 and 78 in the stallion and positions 33, 51, and 78 in the ram. Serine (S) 109 and 113 and threonine (T) 152 are putative sites of O-glycosylation in the ram.
|
|
Because ovine and equine epididymal PGDS possesses the cysteine necessary for the enzyme function and the characteristic lipocalin sequences, we investigated both functions in the epididymis.
Metabolic PG Pathway in the Epididymis
To investigate the putative enzymatic function of PGDS, we performed an overall study of the metabolic PG pathway in the epididymis. The aim of this approach was to determine if this pathway is functional in the epididymis by detecting the cyclooxygenases (COX1 and COX2) in the ram epididymis, analyzing in vitro PG synthesis in isolated epididymal tubules using exogenous radioactive arachidonic acid, and evaluating the effect of the pharmacological inhibition of this pathway on fertility in the rat.
Detection of cyclooxygenases (COX1 and COX2) in the epididymis COX1 was detected by immunoblotting using a specific antiserum in tissue extracts of the ram epididymis [31, 32] (Fig. 5). The protein was detected as a single 72-kDa band from the caput to the cauda. COX1 was also detected in testicular tissue extracts, but it was not detected in epididymal fluid from various epididymal regions or in spermatozoa membrane extracts (data not shown). The presence of COX2 protein was investigated by immunodetection using a specific antibody [30, 31] in epididymal tissues, fluids, and spermatozoa membrane extracts, and its mRNA presence was investigated by RT-PCR using specific primers and ovine trophoblast RNA as a positive control [31]. Neither the protein nor its mRNA was detected in the epididymis using these two approaches. These results suggest that only COX1 is expressed in the ram epididymis.
|
Metabolic PG pathway in the epididymis The presence of COX1 in the ram epididymis indicated that the PG pathway is functional in this organ. To investigate this pathway, we used an in vitro incubation technique with closed-end epididymal tubules as previously described [38] in the presence of [3H]arachidonic acid. The radioactive PGs produced in the luminal fluids, tissues, and supernatants during incubation were detected and quantified as described in Materials and Methods.
In the first experiment (two animals), the PGs synthesized inside and outside the tubules of the caput (E2) and corpus (E6) were analyzed (Fig. 6). The neosynthesized PG profiles were similar for both epididymal regions without any significant qualitative or quantitative difference. For both regions, the same products were detected inside and outside the tubule in relatively similar amounts. The most abundant of the different PGs detected was 6-keto-PGF1
(prostacyclin degradation product). The other major PGs detected were PGE2, PGF2 and its degradation product (13,14-dihydro-15-keto PGF2), thromboxane B2 (derived from thromboxane A2), and PGD2. In the epididymal caput and corpus (lumen and supernatant), PGD2 represented 7–10% of the total PGs detected, without any difference outside and inside the tubule.
|
In the second experiment using the same protocol, we analyzed the neosynthesized PGs in the corpus region in the absence of the endogenous luminal PGDS. For this purpose, region E6 tubules, which do not normally synthesize PGDS, were incubated intact (control) or after removing the luminal content by flushing (washing) the lumen with PBS, so that PGDS was no longer present inside the tubule. The percentages of each PG detected in epididymal fluids, tissues, and supernatants are shown in Table 1. The percentages of neosynthesized PGD2 were approximately 11% in the fluid, 8% in the tissue, and 10% in the supernatant, without any significant difference regarding the incubation conditions (control/washing).
|
To summarize our investigation of this pathway, PGD2 production in the epididymis under our experimental conditions was quantitatively irrespective of the presence of PGDS in the lumen. These results strongly suggest that luminal PGDS is not involved in the enzyme production of PGD2. These results were corroborated by the absence of PGD2 production when PGH2 was directly added to the epididymal fluid and its absence in seminal plasma even in the presence of PGDS cofactors such as DTT, which is reported to be involved in enzyme activity in the brain [35] (data not shown).
Effect of PG pathway inhibition on fertility and spermatozoa mobility using nonsteroidal anti-inflammatory drugs This study was performed in adult rats. Seven days of daily i.p. treatment with flurbiprofen and indomethacin were sufficient to inhibit synthesis of PGE2 and PGF2 by seminal vesicles (data not shown), suggesting that the doses used were sufficient to inhibit the whole PG pathway in the male genital tract. After 60 days of this treatment, fertility of the males was estimated by breeding with normal females. No differences were observed between treated animals and controls (Table 2). The motility of caudal spermatozoa was unchanged after treatment, except for progressive mobility. These results suggest that inhibition of the PG pathway does not adversely affect male reproduction in the rat.
|
Binding Ability of PGDS to Hydrophobic Compounds in the Ram Epididymis
PGDS RA-binding ability Because recombinant PGDS has been previously shown to bind RA [39] and retinoids are necessary to maintain male reproductive function [40], we investigated the ability of epididymal PGDS to bind RA. PGDS purified by electroelution from the spots obtained in 2D electrophoresis of the caput fluid, as shown previously [3], was incubated with 10 µCi (0.3 nmol) [3H]RA. The eluted fraction containing PGDS after gel filtration also contained a large amount of [3H]RA, showing that PGDS is able to bind RA (Fig. 7A). An excess of 150 nmol nonradioactive RA did not affect the binding of radioactive RA (data not shown). Several peaks of radioactivity were detected (Fig. 7B) when a similar binding experiment was performed with total proteins from the epididymal caput fluid in the presence of [3H]RA, including one corresponding to PGDS size in the gel filtration separation. In the presence of total protein in caput epididymal fluid, binding of RA in the fractions where PGDS was eluted represents approximately 6% of the total RA binding.
|
PGDS testosterone-binding ability When [3H]testosterone was used as a potential ligand for PGDS, significant binding was observed in the fractions that correspond to PGDS elution (Fig. 7C). In the presence of a large amount of cold testosterone (140 mmol), most of the other binding disappeared, except for the PGDS fractions, for which significant binding of [3H]testosterone was still observed. The testosterone binding on the PGDS fractions represents 40% of the total binding to protein in the epididymal fluid.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
High PGDS concentrations observed in bull ejaculates presenting normal (0) and high (>0) fertility suggest that PGDS is correlated with male fertility and, thus, is important for male reproduction. Previously, PGDS has been reported to be a fertility-associated protein in bull seminal plasma [1, 41] and to be a biochemical marker of sperm quality in humans [42]. However, because low PGDS concentrations were also observed in semen of rams and bulls with high fertility, a high amount of PGDS in the genital tract likely is not essential for male reproductive function, or other proteins can assume its function when the concentration of this protein is low.
To investigate the function of epididymal PGDS, we first cloned cDNA in the ram and stallion to predict the primary sequence in these species. The protein sequence of ovine and equine PGDS showed high homology with that of PGDS in other mammals (e.g., mouse, human, bull), demonstrating high conservation during evolution. In both species, the sequences involved in the lipocalin structure (Fig. 3) and the cysteine at position 65 necessary for enzyme activity are present, suggesting that PGDS could have both activities in the epididymal fluid.
Several putative sites of glycosylation (N and O) were observed in the protein sequences for both species. We have previously described PGDS isoforms using 2D electrophoresis in the epididymal fluid of rams and stallions [3]. The difference of 8–10 kDa between the theoretical and the apparent molecular mass of PGDS and the presence of at least six spots with a pI range of 4–7 are probably the result of posttransductional modifications at these putative sites. In the cerebrospinal fluid of the human and rat, PGDS is also glycosylated [10, 43], as is PGDS in the human epididymal fluid [5].
The enzyme substrate of PGDS is PGH2, which results from the conversion of arachidonic acid by COX activity (COX1 and COX2). COX2 is not expressed in the ram epididymis, although COX2 is present in the rat epididymis [44], suggesting variation between species. COX1 was detected only in tissue extracts in the ram epididymis and not in the lumen content. COX1 has been shown to be localized on spermatozoa membranes in the bull [45], but this was not found in the ram. Because COX1 localization is exclusively intracellular in the ram epididymis, the availability of PGH2 (a very labile molecule) for PGDS in the epididymal extracellular fluid is highly unlikely, although PGH2 has been shown to be exchanged between cells [46].
The COX pathway is functional in epididymal tubules, as shown by our ex vivo experimental conditions, and several PGs were detected in the epididymal tissues and fluid as well as outside the tubule. PGD2 is produced at a relatively low level (
8%), which is surprising in view of the high amount of PGDS in the epididymis. This paradox has already been pointed out by Hoffamn et al. [1] for cerebral PGDS: The high concentration of PGDS in the cerebrospinal fluid is not related to a high amount of PGD2 [47]. Moreover, we have shown that the amount of neosynthesized PGD2 is irrespective of the presence of PGDS in the lumen. The low PGD2 production in the tubule in the presence of a high concentration of PGDS in the lumen may be related to rapid transformation of PGD2 into PGJ2 by a nonenzymatic reaction, and then into 
12-PGJ2 or 13,14-dihydro-15-keto-PGD2 [48]. However, none of these compounds were detected among all the PGs observed under our conditions.
Taken together, these results strongly suggest that the enzymatic activity of PGDS is very low (or even absent) in the epididymal lumen, with our experimental conditions being close to in vivo conditions. Previous studies demonstrating the existence of enzymatic activity of epididymal PGDS were performed in drastic conditions (pH 9) that seem incompatible with the epididymal milieu [4, 49]. Moreover, the enzymatic function of PGDS occurs only when the molecule is in its reduced form [50], but the protein is secreted in vivo with its disulfide bonds [1].
The secretion of PGD2 in the lumen of the tubule might be related to other luminal pathways, such as the activity of glutathione S-transferase or albumin [51]. Because we observed PGD2 inside and outside the tubule, a more likely hypothesis is that PGD2 might originate from intracellular enzyme activity and then be secreted into the lumen or outside the tubule. The biological effect of PGD2 in the brain is mediated by its specific receptor. This receptor has been identified in the rat epididymis [19]. However, its regionalization is restricted to the caput, whereas PGDS is more abundant in the cauda (unpublished results). This receptor is not expressed in the human epididymis or testis, although PGDS is present [20, 49]. Obviously, no colocalization of luminal PGDS and expression of PGD2 receptor is observed, confirming the hypothesis that the function of luminal PGDS is not the production of PGD2. The biological effects of PGD2 in the epididymis are unknown. It could be involved in smooth muscle contraction [15]. Whatever the function, it is not essential for male reproduction, because 2 mo of COX pathway inhibition by nonsteroidal anti-inflammatory drugs inducing overall reduction in PG synthesis does not affect fertility in the rat. Previous studies with short-term treatment involving nonsteroidal anti-inflammatory drugs showed the same results [52, 53]. Furthermore, inhibition of PG synthesis by disruption of COX1 and COX2 genes had no drastic consequences on reproduction in mice [54–56].
Because enzyme activity of PGDS in the epididymal fluid is unlikely, we investigated its function as a putative androgen and RA transporter. RA was a good candidate, because 1) it is present in the epididymis [57], 2) it is important for maintenance of the epididymal epithelium [40], and 3) Tanaka et al. [39] previously demonstrated that recombinant human PGDS is able to bind RA. Our results showed clearly that PGDS binds RA in vitro after purification and in the epididymal fluid. However, saturation of the binding sites was not obtained in the presence of excess ligand, suggesting that the binding is not specific, as it is for several other luminal proteins able to bind RA in vivo.
PGDS showed a high capacity to bind to testosterone. This high capacity to bind and, potentially, transport steroid suggests that PGDS could be involved in the regulation and maintenance of functions of the epididymal epithelium. In several species, a specific testicular androgen-binding protein has been previously described, but most of this protein is absorbed by the efferent ducts and proximal epididymis [58, 59]. In the more distal epididymis, PGDS could play an important role in transporting androgens such as dihydrotestosterone, which is actively formed in the proximal region.
The most likely function of PGDS is, therefore, the binding of lipophilic molecules. The nature of the potential lipophilic ligands in vivo is unknown, and further investigations are necessary to characterize these in vivo ligands and their binding characteristics (affinity).
Other lipocalins are also secreted by the epididymis (e.g., E-RABP in several mammals [23–25]). This suggests that lipocalins have an important function in sperm maturation and/or storage. Because these lipocalins have different regionalizations in the epididymal fluid, they could function in concert for the binding and trafficking of a common endogenous ligand in different epididymal regions; each could also have their own ligand. Specific endogenous ligands need to be identified for all these lipocalins, including PGDS.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by a grant from l'Institut National de la Recherche Agronomique (INRA, France) and from Région Centre (France). The nucleotide sequence reported in this paper has been submitted to EMBL Data Bank with accession numbers AJ133469 and AJ133642. 
2 Correspondence. FAX: 33 2 47 42 77 43; jdacheux{at}tours.inra.fr
Accepted: September 19, 2001.
Received: March 21, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
receptor in bovine preovulatory follicles. Endocrinology 1996; 137:3348-3355[Abstract]This article has been cited by other articles:
|
|
 |
|
  K. Ravi Ram and M. F. Wolfner Seminal influences: Drosophila Acps and the molecular interplay between males and females during reproduction Integr. Comp. Biol., September 1, 2007; 47(3): 427 - 445. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Moura, D. A. Chapman, H. Koc, and G. J. Killian Proteins of the Cauda Epididymal Fluid Associated With Fertility of Mature Dairy Bulls J Androl, July 1, 2006; 27(4): 534 - 541. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kim, P. Yang, M. Suraokar, A. L. Sabichi, N. D. Llansa, G. Mendoza, V. Subbarayan, C. J. Logothetis, R. A. Newman, S. M. Lippman, et al. Suppression of Prostate Tumor Cell Growth by Stromal Cell Prostaglandin D Synthase-Derived Products Cancer Res., July 15, 2005; 65(14): 6189 - 6198. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Mueller, D. R. Ripoll, C. F. Aquadro, and M. F. Wolfner Comparative structural modeling and inference of conserved protein classes in Drosophila seminal fluid PNAS, September 14, 2004; 101(37): 13542 - 13547. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Fouchecourt, J.-J. Lareyre, P. Chaurand, B. B. DaGue, K. Suzuki, D. E. Ong, G. E. Olson, R. J. Matusik, R. M. Caprioli, and M.-C. Orgebin-Crist Identification, Immunolocalization, Regulation, and Postnatal Development of the Lipocalin EP17 (Epididymal Protein of 17 Kilodaltons) in the Mouse and Rat Epididymis Endocrinology, March 1, 2003; 144(3): 887 - 900. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
