| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Anatomy and Cell Biology, Martin Luther University Halle-Wittenberg, Faculty of Medicine, D-06097 Halle (Saale), Germany
b Large Animal Clinic for Theriogenology and Ambulatory Services, Faculty of Veterinary Medicine, University of Leipzig, D-04103 Leipzig, Germany
c Institute of Veterinary Anatomy, Justus-Liebig-University Giessen, D-35392 Giessen, Germany
ABSTRACT
Employing postpubertal testicular tissue, we determined the cDNA coding sequence of a truncated canine relaxin-like factor (RLF) consisting of a signal peptide of 28 amino acids (aa), a B-domain of 23 aa, a truncated C-domain of 34 aa, and an A domain of 26 aa, respectively. Within the B-domain of canine RLF, the putative relaxin receptor binding motif contained a single substitution with the C-terminal arginine replaced by a serine residue, and the putative RLF receptor binding motif was truncated. Leydig cells specifically expressed RLF in the normal postpubertal and cryptochid testis as well as in testicular Leydig cell adenoma. The epididymis was an additional source of RLF in the dog. In the female reproductive tract, expression of immunoreactive RLF and relaxin were compared. Within the ovary, RLF, but not relaxin, was detected in follicular theca interna and granulosa cells and the corpus luteum. In the nonpregnant uterus, luminal and glandular epithelium coexpressed RLF and relaxin. Uteroplacental tissue at early stages of gestation revealed RLF expression in the proliferative fetal villous cytotrophoblast and in maternal uterine cells. In the mature canine placenta, the trophoblast surrounding the maternal blood vessels and the hemophagous cytotrophoblast of the paraplacental zone expressed RLF. Canine relaxin was absent in the paraplacental areas. Western analysis of placental tissue extracts revealed the presence of specific immunoreactive bands likely resembling unprocessed and enzymatically cleaved RLF. Differential expression of RLF and relaxin appears to reflect distinct autocrine and paracrine functions of RLF in canine reproductive tissues.
corpus luteum, epididymis, follicle, implantation/early development, Leydig cells, ovary, placenta, pregnancy, relaxin, testes, trophoblast, uterus
INTRODUCTION
Relaxin-like factor (RLF), also known as Leydig cell-derived insulin-like factor (INSL-3) [1] and the structurally closely related heterodimeric polypeptide hormone relaxin constitute the relaxin family within the superfamily of insulin-like molecules.
Relaxin-like factor is a specific product of testicular Leydig cells [1–4] and cells of the ovarian theca interna layer [5–7]. In rodents, RLF gene expression is developmentally regulated [8], and mice with a deletion of the RLF gene display a failure of gubernaculum testis formation with impaired testicular descent and disturbances in ovarian folliculogenesis [9, 10].
The sequence of canine relaxin has been determined [11, 12], and the syncytiotrophoblast has been identified as the placental source of relaxin in the pregnant bitch [12–14]. Reaching by far the highest relaxin serum levels of all species with up to 10 µg/ml [15], and given the cross-reactivity of RLF and relaxin for their corresponding receptors [16, 17], relaxin may act as a substitute for RLF during gestation in this species. With the exception of the recently cloned relaxin in the dromedary [18], RLF is regarded as a substitute for relaxin in domestic ruminants that do not possess a functional relaxin gene [19, 20] but contain functional binding sites for relaxin [21].
In the present study we have identified a functional and unique canine RLF gene product and reveal its differential expression pattern within canine reproductive tissues.
MATERIALS AND METHODS
Collection of Tissues and RNA Isolation
Uteroplacental tissues from six pregnant domestic dogs (Canis domesticus) at approximately Days 25 (n = 2), 30 (n = 2), and 50 (n = 2) of gestation as determined by measurement of the crown-rump length of the fetus, uterine tissues (n = 5), ovarian tissues (n = 8), normal postpubertal testes (n = 5), cryptorchid testes (n = 3), and testicular Leydig cell adenoma (n = 4) were collected during the course of medically indicated treatments. Canine kidney was cryoconserved at a local clinic from a male welp killed in a road accident and served as control tissue in the Northern analysis. No animal was operated upon or killed specifically for this study. The ovarian, testicular, and uteroplacental tissues at about Days 20 and 50 of gestation were fixed in Bouin solution and embedded in paraffin. Uteroplacental tissue at Days 30 and 50 of gestation, kidney, and normal postpubertal testicular tissue were snap frozen in liquid nitrogen and used for isolation of total RNA with Trizol reagent (Life Technologies, Eggenstein, Germany). For the cloning of canine RLF, mRNA was isolated from 75 µg of total testicular RNA using oligo(dT)-coated magnetic beads (Dynal, Hamburg, Germany). The amount of mRNA isolated was determined spectrophotometrically at 260 and 280 nm [22].
Cloning of Canine RLF
Rapid amplification of 5'- and 3'-cDNA ends (5'- and 3'-RACE)-polymerase chain reaction (PCR) and reverse transcriptase (RT)-PCR were performed on mRNA of normal postpubertal testicular tissue at Day 30 of gestation to determine the cDNA coding sequence of canine RLF. All PCR primers employed in this study (Fig. 2 and Table 1) flanked the putative single intron at the N-terminus of the C-domain of RLF to preclude any genomic DNA amplification (Figs. 1 and 2). Using approximately 600 ng of mRNA, first-strand cDNA-synthesis was done using the Superscript reverse transcriptase kit (Life Technologies). The PCR reactions were carried out in a 50-µl solution containing 1 µl of cDNA, 5 µl of 10x Advantage cDNA polymerase mix buffer, 100 µM of dNTP, 10 pmol of each primer, and 2.5 U Advantage cDNA mix polymerase (Clontech, Heidelberg, Germany). For the initial amplification of a cDNA fragment of canine RLF, a degenerate oligonucleotide primer pair designed according to relatively conserved nucleic acid sequences within the signal peptide and the A-domain of human and ruminant RLF molecules (primers 1 and 2; Figs. 1 and 2 and Table 1) was employed [2, 7]. The PCR cycles consisted of an initial denaturation for 3 min at 95°C, followed by 40 cycles of denaturation at 95°C and annealing at 56°C, both for 1 min each, an elongation step for 2 min at 72°C, and a final extension cycle for 10 min at 72°C. Employing gene-specific primers (GSP; Table 1) in combination with a universal primer (primer 10; Figs. 1 and 2 and Table 1), the 5'- and 3'-RACE-PCR were performed as described previously [12]. The complete cDNA coding sequence of canine RLF was amplified by RT-PCR for 40 cycles at an annealing temperature of 68°C and a specific primer pair located at both ends of the coding sequence (primers 7 and 8; Figs. 1 and 2 and Table 1). All PCR products were separated on a 1% low-melting point agarose gel, purified by Magic column extraction, and cloned into the pGEM-T vector (both Promega, Heidelberg, Germany). Sequence analysis of the PCR-amplified cDNA clones was performed employing the Prism dye Deoxy Terminator cycle sequencing kit (Perkin Elmer, Foster City, CA) and T7 or SP6 sequencing primers.
|
|
|
Northern Analysis
Twenty micrograms of total RNA extracted from postpubertal canine testicular tissue and kidney were run on a 1% formaldehyde agarose gel containing ethidium bromide to assess equal loading by comparison of the rRNA bands. The RNA molecular weight markers (RNA ladder 0.44–1.77 kilobases [kb]; Life Technologies) were included with the gel. After blotting for 1.5 h [23] and prehybridization for 30 min at 62°C with Easy-Hyb hybridization buffer (Boehringer Mannheim, Mannheim, Germany), nylon membranes (Hybond N+; Amersham, Braunschweig, Germany) were incubated overnight at 62°C in 15 ml of the same solution containing approximately 200 ng/ml of digoxygenin (DIG)-labeled canine RLF cRNA probe. Membranes were washed 2x 20 min in 2x saline-sodium citrate (SSC, 1x SSC is 0.15 M NaCl and 0.015 M sodium citrate)/0.1% SDS at room temperature, followed by 2x 20 min in 0.5x SSC/0.1% SDS at 58°C, and specific hybridization signals were visualized by chromogenic membrane staining with nitroblue tetrazolium/5-bromo-4-chloro-3-indolylphosphate toluidinium (Sigma, Deisenhofen, Germany) with the DIG luminescence detection kit (Boehringer Mannheim).
Digoxygenin-Labeling of cRNA and In Situ Hybridization
For cRNA synthesis, 5 µg of the pGEM-T plasmid clone containing the insert for canine RLF were digested with the restriction enzymes NcoI (antisense cRNA) and SacI (sense cRNA; both New England Biolabs, Frankfurt, Germany), and phenol-extracted digested plasmid at 1 µg was used as template for cRNA synthesis employing a cRNA synthesis kit (AMS Biotechnology, Wiesbaden, Germany) and a 10x DIG-RNA labeling mix (Boehringer Mannheim). After precipitation of the DIG-labeled cRNA, the pellet was dissolved in 70 µl of diethyl pyrocarbonate-water, and the quantity of cRNA was determined spectrophotometrically at 260 and 280 nm and by dot-blot analysis of serial dilutions of DIG-labeled cRNA [24]. Nonradioactive in situ hybridization was performed according to the procedure by Lewis and Wells [25] with modifications as described previously [12].
Immunohistochemistry
Immunohistochemical staining was performed on 6-µm-thick paraffin sections of canine testicular, ovarian, and uteroplacental tissues at about Days 20 and 50 of gestation mounted on Aptex-coated slides. After dewaxing and rehydration, all staining procedures were performed with PBS buffer containing 0.1% Tween (PBS-T). Negative controls were performed in all cases by replacing the primary antibody by normal serum of the same species as the primary antibody. After dewaxing and rehydration, ovarian tissue sections were boiled for 25 min in citrate buffer. Ovarian and testicular tissue sections were treated with 3% H2O2 in methanol to inactivate endogenous peroxidase. Unspecific binding sites were blocked with PBS-T containing 3% BSA and 20% normal swine serum for 1 h at room temperature. For the specific detection of RLF, a primary rabbit polyclonal antiserum to RLF, generated against the peptide within the B-domain of RLF -E31KLCGHHFVRALVRV45- (single letter code; numbering according to human RLF; BioGenes, Berlin, Germany), was diluted at 1:250 and 1:500 for the ovarian and testicular sections, respectively, in PBS-T containing 3% BSA plus 3% nonimmune goat serum (Dianova, Hamburg, Germany) and incubated overnight at 4°C. Ovarian and testicular tissue sections were washed 3x 5 min in PBS-T and incubated with a peroxidase-conjugated goat anti-rabbit secondary antibody (Life Technologies) diluted at 1:500 in PBS-T containing 1% BSA for 1 h at room temperature. Specific staining was visualized employing the peroxidase substrate diaminobenzidine (Pierce, Rockford, IL). Uteroplacental tissue sections were incubated with proteinase K (Boehringer Mannheim) at 30 µg/ml for 30 min at 37°C, and endogenous alkaline phosphatase was inactivated with 20% acetic acid. Tissue sections were blocked in PBS-T containing 3% BSA and 20% normal swine serum for 1 h at room temperature to saturate nonspecific binding sites. For the detection of RLF epitopes, the rabbit polyclonal RLF antiserum was diluted 1:250 in PBS-T containing 3% BSA plus 3% nonimmune goat serum (Dianova) and incubated overnight at 4°C. Following 3x 10 min washing in PBS-T, the tissue sections were incubated for 1 h at room temperature with an alkaline phosphatase-conjugated goat anti-rabbit secondary antibody, diluted 1:250 in PBS-T with 3% BSA. Immunohistochemical staining of ovarian and uterine tissues with a rabbit polyclonal serum against canine relaxin (no. 78513; generously provided by Professor B.G. Steinetz) were performed as described previously [12]. For the detection of cytokeratin, the tissue sections were incubated overnight at 4°C with the primary mouse monoclonal antibody to pan-cytokeratin, diluted 1:250 (MNF 116; Dako, Hamburg, Germany) in PBS-T with 3% BSA and 3% normal rabbit serum. After washing 3x 5 min in PBS-T, the alkaline phosphatase-anti-alkaline phosphatase (APAAP) technique was employed using a rabbit anti-mouse bridging antibody at 1:50 and mouse-APAAP complexes at 1:100 (both Dako). Incubations with the secondary antibody and the APAAP complexes were repeated and specific immunostaining was visualized with the HistoMark Red AP substrate (KPL, Gaithersburg, MD). Tissue sections were counterstained with hematoxylin and mounted in glycerol gel.
Western Analysis
Snap-frozen tissue samples from dog testes and canine placental tissue at Day 50 of gestation were used for protein isolation with a lysis-buffer containing 63 mM Tris, 2% SDS, and 10% saccharose. After boiling for 5 min and subsequent centrifugation, the amount of protein in the supernatant was determined with the protein assay kit (BioRad, Munich, Germany), and the supernatant was stored at -20°C until used for Western analysis. The extracted proteins were separated on a 15% SDS-polyacrylamide gel and blotted onto a Hybond nylon membrane (Amersham). For the immunodetection of RLF, blots were incubated for 1 h at room temperature with the above-described rabbit polyclonal RLF antiserum, diluted at 1:2000 in PBS-T containing 5% milk powder and 1% BSA. After washing with PBS-T, specific binding was detected with a peroxidase-conjugated goat anti-rabbit Ig (Life Technologies) diluted at 1:5000 in PBS-T containing 5% milk powder and was visualized with the ECL Western blotting detection reagent (Amersham). Blots incubated with a rabbit nonimmune serum (Dako) diluted in the same way as the primary antiserum were used as negative controls. To characterize further the specificity of the rabbit polyclonal RLF antiserum, the cDNA for human RLF was expressed in the prokaryotic expression vector pGEX-5 (Pharmacia, Freiburg, Germany) and the RLF fused to glutathione-S-transferase (GST) and GST alone as a negative control were used in Western analysis. Loading of equal amounts of RLF-GST fusion protein and GST was confirmed by incubation with an antiserum directed against GST (Pharmacia).
RESULTS
Employing RT-PCR and RACE-PCR on mRNA isolated from postpubertal testicular tissue, we have determined the cDNA coding sequence for canine RLF consisting of 336 base pairs (bp) encoding 111 amino acids (aa). According to the predicted cleavage sites of other RLF molecules [3, 7, 26], canine RLF encoded a signal peptide of 28 aa, a B-domain of 23 aa, a truncated C-domain of 34 aa, and an A-domain of 26 aa, respectively (Figs. 1 and 2). Control experiments using either mRNA or water as templates in the PCR reactions yielded no amplification products (data not shown).
The signal peptide of canine RLF contained two additional residues (Leu23 and Gly24; Fig. 3). The B-domain of canine RLF was unique among RLF molecules. The canonical putative relaxin receptor motif -RALVR- conserved within the B-domain of all RLF molecules, with the exception of the mouse RLF [8], was altered to -R18ALVS22- with a single substitution of the C-terminal arginine to serine (Figs. 2 and 3). The lack of C-terminal six amino acids reduced the classical putative RLF receptor binding motif -GGPRW- to the truncated motif -SG–W- (Figs. 2 and 3). The locations and positions of the cysteine residues within the A- and B-domain were conserved, with an additional Cys B19 residue present immediately downstream of the putative relaxin receptor binding motif within the B-domain (Figs. 2 and 3). The truncated C-domain of canine RLF was devoid of the N-terminal 16 amino acids (Fig. 3) representing the shortest C-domain of all known RLF molecules [27]. In contrast to the B-domain, the deduced peptide sequence of the A-domain of canine RLF was much more conserved (Fig. 3), displaying highest amino acid homology with porcine RLF (88.5% with the A-domain; overall amino acid homology 61%; Table 2) [26; 28]. When equal amounts of total RNA from postpubertal testicular tissue and dog kidney were processed for Northern analysis, a single specific hybridization signal at 0.9 kb was exclusively detected in RNA of testicular origin (Fig. 4).
|
|
|
Tissue localization of RLF mRNA was performed by nonradioactive in situ hybridization and immunoreactive RLF was detected by immunohistochemistry employing a rabbit RLF antiserum generated against an RLF-specific synthetic peptide. The specificity of the rabbit RLF antiserum was demonstrated by Western analysis on extracts of canine testicular tissue revealing two bands at approximately 12 kDa and 5 kDa, suggesting an uncleaved and processed canine RLF (Fig. 5A). The same RLF antiserum specifically detected the human RLF-GST fusion protein but not GST alone (Fig. 5B). Control experiments included Western blots with an antiserum against GST to verify equal loading of RLF-GST and GST alone and incubation with a rabbit nonimmune serum replacing the RLF antiserum that did not result in specific staining (data not shown).
|
In agreement with the detection of specific RLF hybridization signals, immunoreactive RLF was detected in interstitial and extraparenchymal Leydig cells of normal postpubertal testicular tissues (Fig. 6A), cryptorchid testis (data not shown) and in the epididymis, with the epithelial cell layer displaying cytoplasmic and nuclear localization of immunoreactive RLF (Fig. 6C). Leydig cells of the testicular Leydig cell adenoma displayed weak and homogeneous expression of RLF (Fig. 6D). When the primary antiserum was replaced with nonimmune serum of the same species, sections were devoid of staining (Fig. 6B). When treated with the DIG-labeled sense RLF cRNA, germ cells and Sertoli cells within the seminiferous epithelium as well as all other sections investigated in this study were devoid of hybridization signals (Fig. 6, F and J).
|
In the female reproductive tract, paraffin-embedded ovarian and uterine tissues of nonpregnant bitches and uteroplacental at Days 25 and 50 of gestation, including paraplacental tissue sections at Day 50 of gestation, were investigated. In ovarian tissues, specific RLF transcripts (Fig. 6E) and immunoreactive RLF (Fig. 6G) were colocalized in the follicular granulosa cell layer, the theca interna, and in cells of the corpus luteum. Sections treated with the sense RLF cRNA probe were devoid of staining (Fig. 6F). Employing the polyclonal rabbit antiserum against canine relaxin (no. 78513), we were unable to detect immunoreactive relaxin in canine ovarian tissues.
Uterine sections from the nonpregnant bitch displayed expression of RLF transcripts (Fig. 6H) and immunoreactive RLF (data not shown) exclusively within the luminal epithelium and the endometrial glands. Uterine epithelium coexpressed immunoreactive canine relaxin (data not shown). In the pregnant uterus, the uterine glands exclusively expressed RLF (Fig. 6L).
Within the placental girdle at Day 25 of gestation, RLF mRNA was detected in the fetal cytotrophoblast cells of the chorionic villi and the maternal glandular epithelium (Fig. 6I). The syncytiotrophoblast surrounding the tip of the protruding fetal chorionic villus was devoid of RLF transcripts (Fig. 6I). In uteroplacental tissue at Day 50 of gestation, RLF mRNA (Fig. 6, K and M) and immunoreactive RLF (Fig. 6O) were expressed in cytokeratin-positive fetal trophoblast cells of the zonary placental girdle (Fig. 6K) and in the paraplacental regions (Fig. 6, O and P). Sections treated with the DIG-labeled sense cRNA did not shown hybridization signals (Fig. 6J). Control experiments replacing primary antibodies with species-specific nonimmune serum were devoid of immunostaining (Fig. 6N).
DISCUSSION
Although composed of the classical N'B-C-AC' domain structure preceded by a signal peptide sequence, canine RLF contained an extended truncation at the junction between the B- and C-domain and unique structural changes within the B-domain that are likely to affect the interaction of this smallest of all RLF molecules with binding sites for both relaxin and RLF. The RLF has been shown to interact with specific RLF receptors [17] and to cross-react with the relaxin receptor at a 100-fold lower binding affinity than relaxin [16]. Located within the relaxin receptor binding motif of relaxin molecules, including canine relaxin [12], the two flanking arginine residues (ArgB13 and ArgB17; H2-relaxin numbering) have been shown to be essential for high affinity interaction of relaxin with its receptor [29]. An H2 relaxin derivative with ArgB17 replaced by alanine was biologically inactive [29]. In canine RLF, the mutation of the C-terminal arginine to serine within the putative relaxin receptor binding motif [30] will likely result in a reduced affinity to relaxin binding sites. In the pregnant bitch the highest relaxin serum levels of all species have been determined reaching up to 10 µg/ml [15]. Given the cross-reactivity of RLF and relaxin with their corresponding receptors [16], low affinity interaction of RLF with relaxin binding sites or even the partial blockage of the binding of relaxin to its cognate receptor may be a potential function of RLF in the pregnant bitch. Conserved among RLF molecules and located separate from the relaxin binding domain at the C-terminus of the B-chain, the pentameric motif -GGPRW- has been demonstrated to be conserved among RLF molecules, facilitating binding of RLF to specific RLF receptors [30]. In human RLF, both the proline B25 (ProB25) and the tryptophan B27 (TrpB27) residue have been shown to be important for the interaction with specific RLF binding sites. Replacement of L-ProB25 with the D-enantiomer resulted in a significant decrease in affinity, indicating L-ProB25 to be important for the proper positioning of the TrpB27 indole ring [30]. In contrast to TrpB27, the ProB25 residue did not appear to be essential for RLF receptor binding itself because the affinity of an RLF-AlaB25 derivative was comparable to the human RLF [30]. In canine RLF, the putative RLF receptor binding motif was reduced to a unique -SGW- sequence but contained the essential tryptophan residue preceded by a glycine residue that may, by analogy to the AlaB25-human RLF derivative [30], confer molecular flexibility onto the indole ring that is important for bioactivity. Unlike the marmoset [31], canine postpubertal testicular and placental tissues were devoid of RLF splice variants and similar to RLF transcripts in the human [1], marmoset [31], mouse [3, 8], rat [32], and in ruminant species [5, 7, 18, 26], we detected a single population of RLF transcripts at approximately 0.9 kb in the postpubertal canine testis, suggesting canine RLF to be a single gene product as has been shown for the human [28], the pig [28], and the marmoset [31]. For the first time, immunoreactive canine RLF peptides corresponding in size to both unprocessed and enzymatically cleaved RLF could be detected in normal postpubertal testicular tissue, indicating translation of testicular RLF transcripts and enzymatic processing of the truncated canine RLF prohormone to form a heterodimeric hormone consisting of an A- and B-domain. Human RLF has recently been shown to be a secreted hormone [33], suggesting that, in addition to its potential systemic effects, canine RLF possesses autocrine and paracrine functions within the reproductive tract. Assuming, however, that canine RLF is stabilized by insulin- and relaxin-like disulfide bonds, the presence of the seventh cysteine residue CysB19 is likely to confer instability onto the RLF dimer, as the additional free SH group of CysB19 will engage in a thiol disulfide exchange reaction eventually resulting in the separation of the A- and B-chain of the RLF heterodimer and the loss of RLF bioactivity over time.
As shown in the human testis [4], cryptorchidism did not affect the expression of RLF in canine testicular Leydig cells. Expression of RLF was also observed in extraparenchymal Leydig cells located in the tunica albuginea of the canine testis. Albugineal Leydig cells have also been demonstrated in the postpubertal human testis displaying impaired synthesis of immunoreactive testosterone when compared with the parenchymal Leydig cell compartment [34]. Whether human extraparenchymal Leydig cells express RLF is unknown. As in the mouse Leydig tumor cell line MA-10 [3, 35], weak and homogeneous RLF expression was observed in canine testicular Leydig cell adenoma. This is in contrast to our previous results in human testicular Leydig cell adenoma that displayed decreasing RLF expression from the periphery to the center of the tumor [36].
The ovary, the nonpregnant uterus, and uteroplacental tissue have all been identified as sources of RLF within the female reproductive tract [5, 7, 37, 38]. Within the ovary of the nonpregnant bitch, RLF was expressed in both the theca interna cell layer and granulosa cells of follicular tissues as well as the corpus luteum. Although the theca interna is regarded the major source of RLF in various species [5, 7, 37, 38], granulosa cells within the bovine follicle have also been reported to express RLF, albeit at much lower levels than the theca interna cell layer [5]. In long-term primary cultures of bovine thecal cells the expression of RLF has been shown to be insulin dependent [39]. In the mouse, ovarian RLF expression is correlated with the stage of follicular development and stage of pregnancy [8, 35] and disturbances in folliculogenesis seen in RLF knockout mice [10] indicate a paracrine role for RLF within the ovary. Employing a rabbit polyclonal antiserum against canine relaxin (no. 78513), we detected relaxin epitopes within the canine nonpregnant uterus and the placenta [12] but not in the canine ovary suggesting that, within the ovary of the bitch, a bioactive canine RLF may replace relaxin and mediate the effects on folliculogenesis and luteinization. Given the peptide sequences of the relaxin and RLF receptor binding sites within the B-domain, canine RLF would most likely interact with specific ovarian RLF receptors rather than relaxin binding sites.
In the fallow deer [7] but not the marmoset [31], the uterine epithelium has been identified as a source of RLF. Expression of both RLF and relaxin in the glandular and luminal uterine epithelium of the nonpregnant uterus of the bitch implicated both members of the relaxin family to participate in a local physiology within the nonpregnant uterus as has been proposed for marmoset relaxin [40]. Although the role of RLF in the uterus and its potential interactions with relaxin are currently unknown, relaxin has been implicated in neovascularization of the endometrial lining for its ability to stimulate the expression of vascular endothelial growth factor in human endometrial stromal cells [41]. Maternal uterine epithelium and fetal trophoblast cells within the uteroplacental tissues of the pregnant bitch displayed differential expression of RLF and relaxin, suggesting distinct functions for both members of the relaxin family within the canine placenta. Similar to members of the placental lactogen family [42], fetal canine trophoblast cells entering specific differentiation pathways revealed differences in the control of RLF and relaxin gene expression. In canine uteroplacental tissue obtained as early as 2 days postimplantation at Day 20 of gestation [43, 44], RLF was expressed in highly proliferative fetal villous cytotrophoblast cells of the early placental girdle protruding into the uterine luminal gland openings. Within the mature labyrinthine endotheliochorial girdle placenta of the bitch [45], the large trophoblast population surrounding the maternal capillaries expressed RLF. Regarded as an early secretion product and a marker of pregnancy in the bitch [13, 15], canine relaxin has been shown to be expressed exclusively in the terminally differentiated fetal syncytiotrophoblast of the placental girdle but was absent in the paraplacental zones of the carnivore placenta [12, 46]. In contrast, RLF was a product of hemophagous columnar cytotrophoblast in the paraplacental extravasate zone known to phagocytose erythrocytes and provide iron and possibly other nutrients required by the fetus [47–49]. We have previously demonstrated differential RLF gene expression in populations of binucleate trophoblast cells derived from the deer placenta suggesting RLF to be an early trophoblast differentiation marker within the ruminant placenta [7]. A 0.7-kb RLF gene fragment immediately upstream of the RLF initiation codon promotes RLF transcription in mouse Leydig cells [50], and steroidogenic factor-1 has recently been identified as a key regulator of RLF promoter activity [51]. The characterization of trophoblast-specific regulatory mechanisms affecting RLF gene expression will be an essential contribution to our understanding of the role of RLF during trophoblast differentiation.
ACKNOWLEDGMENTS
We thank Mrs. Christine Fröhlich, Elke Bernhard, and Elisabeth Schlüter for excellent technical assistance. The authors are grateful to the members of the Small Animal Clinic, Faculty of Veterinary Medicine, University of Leipzig, for excellent cooperation. We are also grateful to Professor B.G. Steinetz, Nelson Institute of Environmental Medicine, New York University Medical Center, Old Forge Road, Tuxedo, NY 10987, for providing the rabbit anti-dog relaxin antiserum no. 78513.
FOOTNOTES
First decision: 2 August 2000.
1 Correspondence: T. Klonisch, Department of Anatomy and Cell Biology, Martin Luther University Halle-Wittenberg, Faculty of Medicine, Grosse Steinstrasse 52, D-06097 Halle (Saale), Germany. FAX: 0049 345 557 1700;thomas.klonisch{at}medizin.uni-halle.de 
Accepted: September 11, 2000.
Received: June 30, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  R. Ivell and R. Anand-Ivell Biology of insulin-like factor 3 in human reproduction Hum. Reprod. Update, July 1, 2009; 15(4): 463 - 476. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Truong, N. V. Bogatcheva, C. Schelling, G. Dolf, and A. I. Agoulnik Isolation and Expression Analysis of the Canine Insulin-Like Factor 3 Gene Biol Reprod, November 1, 2003; 69(5): 1658 - 1664. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Klonisch, K. Steger, A. Kehlen, W. R. Allen, C. Froehlich, J. Kauffold, M. Bergmann, and S. Hombach-Klonisch INSL3 Ligand-Receptor System in the Equine Testis Biol Reprod, June 1, 2003; 68(6): 1975 - 1981. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Ivell and R. A.D. Bathgate Reproductive Biology of the Relaxin-Like Factor (RLF/INSL3) Biol Reprod, September 1, 2002; 67(3): 699 - 705. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. F. Irving-Rodgers, R. A.D. Bathgate, R. Ivell, R. Domagalski, and R. J. Rodgers Dynamic Changes in the Expression of Relaxin-Like Factor (Insl3), Cholesterol Side-Chain Cleavage Cytochrome P450, and 3{beta}-Hydroxysteroid Dehydrogenase in Bovine Ovarian Follicles During Growth and Atresia Biol Reprod, April 1, 2002; 66(4): 934 - 943. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Hombach-Klonisch, S. Seeger, G. Tscheudschilsuren, J. Buchmann, B. Huppertz, G. Seliger, B. Fischer, and T. Klonisch Cellular localization of human relaxin-like factor in the cyclic endometrium and placenta Mol. Hum. Reprod., April 1, 2001; 7(4): 349 - 356. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
