| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Female Reproductive Tract |
Department of Obstetrics and Gynecology,2 Prince of Wales Hospital, Shatin, New Territories, Hong Kong, China
Department of Obstetrics and Gynecology,3 Chinese University of Hong Kong, Shatin, New Territories, Hong Kong, China
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
fallopian tubes, growth factors, oviduct
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Vascular endothelial growth factor (VEGF) is a glycoprotein consisting of six isoforms with 121, 145, 165, 183, 189, and 206 amino acids, respectively (VEGF121, VEGF145, VEGF165, VEGF183, VEGF189, and VEGF206, respectively) [4–6]. These isoforms have different heparin-binding properties that affect their solubility [4]. Two VEGF receptor tyrosine kinases have been identified, flt-1 and flk-1/KDR [7, 8]. VEGF is a potent mitogen for vascular endothelium [9], and it stimulates vascular permeability [10]. These two biological properties are important in the cascade of events leading to angiogenesis in, for example, the developing follicle and corpus luteum [11–13]. However, VEGF is widely distributed in normal human tissues, including some luminal organs such as those of the digestive system that are not undergoing active angiogenesis [14, 15]. The role of VEGF in the digestive system may be to increase vascular permeability and promote secretory activity in the epithelial cells of luminal organs. VEGF has been localized in human oviduct [16]. The presence of VEGF, a known permeability-enhancing factor, within the oviductal epithelium suggests that VEGF may be involved in the regulation of vascular permeability within the oviduct and it may therefore be a controller of oviductal secretion.
To support the early reproductive events that occur predominantly in the ampulla during the periovulatory period, oviductal fluid secretion undergoes both temporal and regional modulation. The amount of oviductal fluid is greatest in the preovulatory period [17], which implies that the expression of VEGF in oviduct may also be hormone dependent. This hypothesis is supported by the observation that VEGF expression in bovine oviducts was highest in the preovulatory phase when both gonadotropins and estrogen levels are high [18]. However, to date, the relationships among gonadotropins, ovarian hormone production, and VEGF expression in the oviduct have not been investigated. More fluid is produced by cells in the ampulla, where the early reproductive events occur, than by cells in the isthmus [19]. However, the differences of VEGF expression in the various regions of the human oviduct have not yet been examined.
We examined the localization of VEGF in the human oviduct using immunohistochemistry, the differences of VEGF mRNA expression in various regions of the oviduct, and the modulation of VEGF mRNA expression in oviduct mucosal cells derived from fertile women throughout the ovulatory cycle. The understanding of the role of VEGF in the regulation of oviductal fluid secretions may be important in the development of mechanisms to control gamete transport and embryo growth.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Participants
Twenty-two normally cycling parous women who underwent laparoscopic tubal sterilization or hysterectomy for benign gynecological conditions were recruited into the study. All women had regular monthly menstrual cycles and were not taking any exogenous hormone supplements. The cycle was divided into five stages determined by the date of commencement of the last menstrual period (LMP) and were confirmed by appropriate serum concentrations of LH, FSH, estradiol (E2), and progesterone (P). Five women were in stage 1 (early follicular phase, Days 1–5), four were in stage 2 (midfollicular phase, Days 6–12), five were in stage 3 (periovulatory phase, Days 13–17), five were in stage 4 (early or midluteal phase, Days 18–23), and three were in stage 5 (late luteal phase, Days 24–28). No woman was included whose LMP did not correspond with stage-appropriate serum LH, FSH, E2, and P concentrations. Informed consent was obtained from each participant.
Oviduct Tissue Dissection and Preparation
Oviduct tissue was excised during laparoscopic tubal sterilization or hysterectomy for benign gynecological conditions. Excised oviduct tissue was immediately placed in 20 ml of Hepes-buffered Quinn Human Tubal Fluid (Irvine Scientific, Santa Ana, CA). This medium was used for all tissue manipulation. The oviduct was rinsed to minimize blood cell contamination. Only two of the investigators performed the tissue dissection and preparation, and the technique was closely monitored to ensure accurate replication.
Isthmic, ampullary, and infundibular parts of oviduct were collected according to their distance from the proximal uterine connection or the distal fimbrial opening. The 2-cm proximal and distal portions of the oviduct were taken as isthmic and infundibular parts respectively, and the ampullary parts were those 4–6 cm from the proximal ends. Oviduct tissue was fixed in formalin and embedded in paraffin blocks. Five-micrometer tissue sections were cut and fixed on glass slides. The accuracy of the classification of the three regions of oviduct was confirmed by histological examination of the tissue sections. The landmarks used for the infundibulum and the ampulla were the numerous fingerlike projections (the fimbriae) and the extensive longitudinal mucosal folds, respectively, and the isthmus was typically marked by its thick smooth muscle layer with only a few mucosal folds.
The lumen of the oviduct was exposed by incising along the antimesenteric border, and only the mucosal layer was dissected off macroscopically. The mucosal strips were cut into 1-mm3 pieces and digested by trypsinization (0.05% trypsin and 0.53 mM EDTA; Gibco BRL, Carlsbad, CA). Immunohistochemistry confirmed that >95% of the cells stained positively for the specific marker cytokeratin, confirming the presence of epithelial cells. An aliquot of the cells was placed into an Eppendorf tube and stored at -70°C for subsequent mRNA isolation.
Immunohistochemistry
Thirty-three tissue sections of the isthmic, ampullary, and infundibular regions of 11 oviducts were analyzed for the presence of VEGF by specific immunohistochemical staining. Tissue sections fixed on glass slides were dewaxed in graded xylenes, rehydrated gradually through graded alcohols, and washed in PBS.
Sections were stained for VEGF with an antibody directed against amino acids 1–20 of human VEGF, which therefore recognizes all VEGF isoforms. Sections were first incubated with 10% horse serum in PBS for 20 min to suppress nonspecific binding of IgG. Primary antibody, polyclonal goat anti-human VEGF (sc-152-G; Santa Cruz Biotechnology, Santa Cruz, CA) diluted to 4 µg/ml in PBS with 1.5% horse serum, was then applied, and the sections were incubated for 120 min at room temperature. Biotinylated secondary antibody, donkey anti-goat IgG (Santa Cruz Biotechnology) diluted to 1.0 µg/ml in PBS with 1.5% horse serum, was then added, and sections were incubated for 30 min at room temperature. Sections were then incubated with horseradish peroxidase (HRP)-streptavidin complex for 30 min, color was developed with HRP substrate mixture for 1 min, and sections were counterstained with Mayer haematoxylin solution for 1 min. After gradual dehydration through alcohols and xylenes, the sections were permanently mounted in aqueous mounting medium. Staining was detected under light microscopy by the same investigator.
As negative controls, normal goat IgG was substituted for the primary antibodies to exclude nonspecific binding of the secondary antibodies, and a five-fold excess of peptide competitor (sc-152-P; Santa Cruz Biotechnology) was added to the primary antibodies to exclude binding of primary antibodies with cross-reacting materials.
Semiquantitative Reverse Transcription Polymerase Chain Reaction
Messenger RNA was extracted from the epithelial cells of isthmic, ampullary, and infundibular regions of oviduct using the Oligotex direct mRNA kit according to the manufacturer's instructions (Qiagen, Hilden, Germany). Integrity of RNA was checked by electrophoresis of samples in a 2% agarose gel stained with ethidium bromide. All mRNA samples were treated with 1 U of RNase-free DNase (Boehringer Mannheim/Hoffman-La Roche, Basel, Switzerland) before reverse transcription (RT) to remove contaminating genomic DNA. DNase was then inactivated by incubation for 5 min at 90°C, and 100 ng of mRNA was used for cDNA synthesis with Multiscribe reverse transcriptase (PE Biosystems). All resultant cDNA was then used for polymerase chain reaction (PCR) with Amplitaq Gold DNA polymerase (PE Biosystems, Foster City, CA). Each PCR cycle consisted of denaturation at 95°C for 45 sec, annealing at 55°C for 45 sec, and extension at 72°C for 1 min. Thirty cycles were performed, followed by a final extension at 72°C for 5 min.
Forward and reverse primers specific to VEGF were derived from the published primer sequences [20]. The oligonucleotide primers were forward GGGCAGAATCATCACGA and reverse TGGTCTGCATTCACATTTG (Mwgag Biotech, Ebersberg, Germany). To ensure that the detected product was amplified cDNA rather than contaminating genomic DNA, primers were designed to cross intron/exon boundaries. VEGF165 and to a lesser extent VEGF121 are the predominant isoforms of VEGF in bovine oviduct [18]. However, each of the VEGF isoforms showed the same pattern of expression during the different phases of the ovulatory cycle [18]. Therefore, primers were specifically designed to detect all VEGF isoforms for increased sensitivity. Primers were tested with cDNA obtained from luteal phase human endometrium, a known source of VEGF mRNA, and the identity of the PCR products was confirmed by sequence analysis [20]. ß-Actin was coamplified with VEGF to provide a semiquantitative internal control for RNA quantity and PCR efficiency. ß-Actin is commonly used as a standard when comparing samples under different hormonal conditions [11, 21, 22] because it is constitutively expressed [23]. It has also been specifically used with human oviduct tissue, and its expression is not affected by gonadotropins or sex hormones [24]. Forward and reverse primers specific to ß-actin were also derived from the published primer sequences [23]: forward ATCGTGGGGCGCCCCAGGCAC and reverse CTCCTTAATGTCACGCACGATTTC (Mwgag Biotech).
The RT-PCR assay was validated. The amount of coamplified product for the VEGF (experimental) and ß-actin (internal standard) primer sets was linear and parallel with increasing amounts of cDNA. The cycle number and primer concentrations were optimized in the exponentially increasing phase of detectable product. Thirty cycles of PCR were performed for each cDNA sample with 100 pmol VEGF primer and 5 pmol ß-actin primer added. To estimate within-assay variability, RNA from five oviduct tissue samples collected from women in various phases of the ovulatory cycle were combined and reverse transcribed to form a pool that was amplified as four replicates during the PCR. Within-assay variability typically ranged from <1% to 12%.
Twenty percent of each PCR product was separated by gel electrophoresis on a 2% agarose gel with 0.5 µg/ml ethidium bromide in Tris-borate-EDTA buffer. The separated PCR products were visualized under ultraviolet (UV) light. A video camera sent the UV-illuminated gel image to a computer, where the software package Gel Doc (Model GS-1000; Bio-Rad Laboratories, Hercules, CA) enabled an image of the gel to be recorded. The integrated optical density (IOD) was determined for each PCR product by the Molecular Analyst Software (Bio-Rad). The IOD ratio between the PCR-amplified VEGF product (297 base pairs) and its simultaneously amplified control ß-actin (543 base pairs), the VEGF:ß-actin ratio, was obtained for each sample. The ratio was normalized to the lowest value for subsequent analysis.
Hormone Assays
Both serum LH and serum FSH were quantified using a two-site sandwich immunoassay and direct chemiluminescence technology (Bayer, Tarrytown, NY) [25]. Two mouse anti-LH (or anti-FSH) ß subunit antibodies, one labeled with acridinium ester and the other bound to paramagenetic particles in the solid phase reagent, were used. The chemiluminescence reaction allows the detection of the bound anti-mouse LH (or FSH) ß antibodies, calculated as relative light units (RLUs). Therefore the patient sample LH (or FSH) concentration can be determined by the direct relationship between the amount of LH (or FSH) in the sample and the RLUs detected by the system. Interassay coefficients of variation (CVs) of the LH assay were 6.2%, 5.2%, and 6.2% at concentrations of 7.2, 36.5, and 96.5 IU/ml, respectively. Interassay CVs of the FSH assay were 6.7%, 5.6%, and 5.0% at concentrations of 5.4, 43.2, and 102.5 IU/ml, respectively.
Serum E2 and P were both quantified using a competitive immunoassay with direct chemiluminescence technology (Bayer) [26]. E2 (or P) in the patient sample binds to a mouse anti-E2 (or anti-P) antibody that is labeled with acridinium ester. Unbound antibody binds to an E2 (or P) derivative covalently coupled to paramagnetic particles in the solid phase reagent. The chemiluminescence reaction allows the detection of the antibodies bound to the E2 (or P) derivative added in the solid phase reaction, calculated as RLUs. Therefore, the patient sample E2 (or P) concentration is determined by the inverse relationship between RLUs detected by the system and the total RLUs added in the start reaction. Interassay CVs of the E2 assay were 9.8%, 4.2%, and 8.7% at concentrations of 250, 859, and 2830 pmol/L, respectively. Interassay CVs of the P assay were 12.0%, 4.6%, and 3.8% at concentrations of 3.8, 24.0, and 71.1 nmol/L, respectively.
As an external quality assurance, the accuracy of the assays over the study period was considered satisfactory with <8% bias from consensus values in a United Kingdom External Quality Control Assessment Scheme.
Statistics
One-way ANOVA was used to evaluate the differences in serum hormone concentrations at various menstrual stages. Two-way ANOVA was used to evaluate the differences in VEGF:ß-actin ratio in various oviduct regions and at various menstrual stages. The Tukey method was used to determine the least significant difference (LSD), the minimum difference between the two groups that can be considered significant. The Pearson correlation analysis was used to determine whether significant correlations were present between serum gonadotropin and sex hormone concentrations and VEGF:ß-actin ratio. Statistical significance was accepted at P 
0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The serum concentrations of LH, FSH, E2, and P at each stage of the cycle at which oviductal tissues were collected are summarized in Table 1. The serum concentrations of LH and E2 were significantly higher during the periovulatory stage than during the other menstrual stages. The serum concentration of P was significantly higher during the early or midluteal stage than during the other menstrual stages. However, the differences in the serum concentration of FSH among various menstrual stages were not significant.
|
Localization of VEGF in the Human Oviduct by Immunohistochemistry
Strong and specific cytoplasmic VEGF immunohistochemical staining was observed in the luminal epithelium, smooth muscle cells, and blood vessels within the oviduct. Cytoplasmic VEGF staining was observed in all three oviduct regions and in all five stages of an ovulatory cycle. Immunolocalization of VEGF in various regions of human oviduct is shown in Figure 1. No immunostaining was observed in negative controls.
|
Differences in VEGF mRNA Expression in Various Oviduct Regions and Various Stages of the Ovulatory Cycle
Sixty-six tissue sections, 22 sections from each oviduct region, were used to evaluate the modulation of VEGF expression in various oviduct regions and during various stages of the ovulatory cycle. By two-way ANOVA, the VEGF:ß-actin ratio differed significantly among the three oviduct regions (P < 0.01; LSD = 0.7), and the differences during the various stages of an ovulatory cycle approached significance (P = 0.06; LSD = 2.2). The VEGF:ß-actin ratio was significantly higher (P < 0.05) in the infundibular region than in the isthmus throughout the ovulatory cycle except during the early follicular stage. The VEGF:ß-actin ratio was significantly higher (P < 0.05) in the ampullary region than in the isthmus throughout the ovulatory cycle except during the midfollicular stage. The VEGF:ß-actin ratio during the periovulatory stage was significantly higher (P < 0.05) than that during the midfollicular stage in the ampulla and that during the early follicular stage in the infundibulum (Fig. 2). One representative result for RT-PCR products of VEGF and ß-actin in the three oviduct regions is shown in Figure 3.
|
|
Correlations between VEGF mRNA Expression with Serum LH, FSH, E2, and P Concentrations
VEGF:ß-actin ratio in the ampullary region was used for the subsequent analysis of its correlations with serum hormone concentrations because the ampulla is the predominant site of oviductal fluid secretion. There were significant positive correlations between the normalized VEGF:ß-actin ratio and both serum LH (correlation coefficient: r = 0.69, P < 0.01) and FSH concentrations (r = 0.78, P <0.01) (Fig. 4). VEGF mRNA expression relative to the expression of the ß-actin control increased with increasing serum LH and FSH concentrations. However, there were no significant correlations between the normalized VEGF:ß-actin ratio and either serum E2 concentrations (r = 0.29, P > 0.05) or P concentrations (r = -0.10, P > 0.05) (Fig. 4).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As previously demonstrated by Gordon et al. [16], VEGF is immunolocalized in both the endothelial cells and epithelial cells of human oviduct. This observation suggests a possible mechanism of transport of the plasma-derived components into the oviduct lumen by increasing both the vascular permeability and epithelial permeability. Similar to findings in the bovine oviduct [18], VEGF is expressed along the whole human oviduct throughout the ovulatory cycle. However, in contrast to observations of no remarkable differences in VEGF immunostaining in various regions of the bovine oviduct [18], VEGF mRNA expression in human oviduct is more prominent in the ampullary and infundibular regions. Although the different findings in bovine and human oviducts may be a genuine species variation, they are more likely to be related to the immunostaining technique used in the bovine study, which was not quantitative [18]. In addition, the stronger expression of VEGF in the ampulla, where the surface area of oviductal epithelium is wider [29], than in the isthmus is consistent with the ampulla as the main site of oviductal secretion [19], where oviduct fluid supports the early reproductive events [1].
The female reproductive system undergoes a number of programmed cyclical processes during the course of the ovulatory cycle. VEGF, under the influence of gonadotropins and/or ovarian hormones, plays a crucial regulatory role in these processes. The roles of VEGF in ovary and endometrium have been well studied, and these roles and the modulation of expression of VEGF differ in the various reproductive tissues. In the endometrium, where the main role of VEGF is to induce angiogenesis to form a receptive endometrium for embryo implantation [30], the expression of VEGF is promoted by estrogen and progesterone [31, 32]. This finding is compatible with the observation that VEGF expression in the endometrium increases during the midproliferative and midsecretory phases, when estrogen and progesterone levels, respectively, are high [32, 33]. However, in ovarian granulosa cells, VEGF expression is not affected by estrogen or progesterone [34]. Instead, these cells are stimulated by hCG, and this relationship holds even under steroid-ablated conditions [34]. The different regulatory mechanism of VEGF expression may be related to its roles in ovary, where it does more than induce angiogenesis. VEGF both promotes angiogenesis during follicular growth and corpus luteum formation and increases vascular permeability, thus promoting transudation of plasma for the accumulation of antral fluid in growing follicles [35]. The increase in vascular permeability may facilitate the delivery of large precursors, such as lipids, that are used for ovarian hormone synthesis.
Our study is the first to describe the expression and modulation of VEGF mRNA in human oviduct epithelial cells during the course of the ovulatory cycle. The observation of the highest VEGF expression in the human oviduct during the periovulatory stage is consistent with the regulatory role of VEGF in oviductal secretion, which supports the early reproductive events. Because of the small sample size for each stage (three to five samples), the differences were significant only for the comparison of the periovulatory stage with the early and midfollicular stage in the infundibulum and ampulla, respectively. This observation is supported by the significant positive correlation between VEGF mRNA expression and serum LH and FSH concentrations, which reached the highest levels during the periovulatory stage.
VEGF mRNA expression in the human oviduct is positively correlated with serum LH and FSH concentrations but not with E2 or P concentrations. The absence of significant correlations between VEGF mRNA expression in oviduct and either serum E2 or P concentrations suggests that the VEGF expression in the human oviduct is steroid independent, and its correlation with gonadotropins is likely to be direct rather than an action mediated by ovarian steroids. These findings suggest that the regulation of VEGF expression in the oviduct may be similar to that observed in the ovary rather than that in the endometrium. The lack of correlation of VEGF expression in the oviduct with E2, a known angiogenic factor, supports the notion that the main role of VEGF in oviduct is permeability promotion rather than angiogenesis. This hypothesis would also explain why VEGF regulation in the oviduct is different from that in the endometrium, where the predominant role of VEGF is angiogenesis.
Although a positive correlation in human oviduct between VEGF mRNA expression and serum gonadotropins levels was demonstrated, their causal relationship has not been established. Therefore, we intend to evaluate the in vitro responses of cultured oviductal cells to hCG or estradiol treatment and their effect on regulating VEGF mRNA expression. The results of this study may also help to exclude the correlation of VEGF expression with estradiol, as suggested by the current observations. Moreover, there is evidence that the two VEGF receptors, KDR and flt-1, may mediate different functions of VEGF. We also intend to determine whether the localization and regulation of the two VEGF receptors follow a similar pattern or a pattern similar to that of VEGF itself. We will use the same methodology as used in the present study to answer these questions.
Our results show that VEGF expression in the human oviduct may increase vascular permeability and epithelial cell secretion, which may contribute to the support of fertilization and early embryo development.
|
FOOTNOTES |
|---|
Received: 23 October 2002.
First decision: 31 October 2002.
Accepted: 12 December 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Wanggren, P.G. Lalitkumar, A. Stavreus-Evers, B. Stabi, and K. Gemzell-Danielsson Prostaglandin E2 and F2{alpha} receptors in the human Fallopian tube before and after mifepristone treatment Mol. Hum. Reprod., September 1, 2006; 12(9): 577 - 585. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Nasu, H. Itoh, A. Yuge, Y. Kawano, and H. Narahara Insulin-like Growth Factor-I Regulates Vascular Endothelial Growth Factor Secretion by Human Oviductal Epithelial Cells and Stromal Fibroblasts Reproductive Sciences, July 1, 2006; 13(5): 368 - 371. [Abstract] [PDF]
|
|
|
|
 |
|
 P. M. Lam, C. Briton-Jones, C. K. Cheung, L. S. Po, L. P. Cheung, and C. Haines Increased mRNA expression of vascular endothelial growth factor and its receptor (flt-1) in the hydrosalpinx Hum. Reprod., November 1, 2003; 18(11): 2264 - 2269. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
