Introduction
It is well known that ovarian activities are regulated not only by endocrine hormones but also by autocrine and paracrine local growth factors (Eppig, Reference Eppig2001; Fortune, Reference Fortune2003; Van den Hurk & Zhao, Reference Van den Hurk and Zhao2005). Among these factors, epidermal growth factor (EGF) seems to be an important regulator of ovarian physiology. EGF is a protein belonging to the EGF family, which consists of at least eight members (Riese & Stern, Reference Riese and Stern1998). EGF protein has been demonstrated in oocyte and granulosa cells of early and late staged follicles (human: Maruo et al., Reference Maruo, Ladines-Llave, Samoto, Matsuo, Manalo, Ito and Mochizuki1993; Bennett et al., Reference Bennett, Osathanondh and Yeh1996, hamster: Roy & Greenwald, Reference Roy and Greenwald1990; pig: Singh et al., Reference Singh, Rutledge and Armstrong1995a) while the EGF mRNA has been described only in oocyte and granulosa cells from pig antral follicles (Singh et al., Reference Singh, Rutledge and Armstrong1995a). Both protein and mRNA for EGF were also found in rat (Tekpetey et al., Reference Tekpetey, Singh, Barbe and Armstrong1995) and porcine corpora lutea (Kennedy et al., Reference Kennedy, Brown and Vaughan1993; Singh et al., Reference Singh, Kennedy, Tekpetey and Armstrong1995b).
The action of EGF in both follicles and luteal cells is mediated by a membrane receptor, ErbB1, which belongs to the ErbB superfamily (Riese & Stern, Reference Riese and Stern1998). This EGF receptor (EGF-R) is a glycoprotein transmembrane receptor with an intrinsic tyrosinase-kinase domain in the cytoplasmic portion of the protein (Carpenter, Reference Carpenter1999), and binds to at least six different EGF family members: EGF itself, transforming growth factor-α, heparin binding EGF-like growth factor, amphiregulin, betacellulin and epiregulin (Riese & Stern, Reference Riese and Stern1998). EGF-R mRNA and protein have been identified in oocyte and granulosa cells of early- and late-stage follicles (mouse: Hill et al., Reference Hill, Hammar, Smith and Gross1999; rat: Chabot et al., Reference Chabot, St-Arnaud, Walker and Pelletier1986; Feng et al., Reference Feng, Knecht and Catt1987; hamster: Garnett et al., Reference Garnett, Wang and Roy2002; cattle, Lonergan et al., Reference Lonergan, Carolan, Van Langendonckt, Donnay, Khatir and Mermillod1996; pig: Singh et al., Reference Singh, Rutledge and Armstrong1995a; human: Maruo et al., Reference Maruo, Ladines-Llave, Samoto, Matsuo, Manalo, Ito and Mochizuki1993; Bennett et al., Reference Bennett, Osathanondh and Yeh1996; Qu et al., Reference Qu, Godin, Nisolle and Donnez2000), and also in luteal cells of pig (Kennedy et al., Reference Kennedy, Brown and Vaughan1993; Singh et al., Reference Singh, Kennedy, Tekpetey and Armstrong1995b) and rat (Tekpetey et al., Reference Tekpetey, Singh, Barbe and Armstrong1995).
All this evidence suggests that EGF plays a pivotal role in controlling ovarian activity in mammals. Indeed, in vitro, we demonstrated a beneficial effect of EGF on oocyte growth in goat primary follicles (Silva et al., Reference Silva, Van den Hurk, Matos, Santos, Pessoa, Moraes and Figueiredo2004a). EGF also promotes in vitro proliferation of porcine granulosa cells obtained from preantral follicles (Morbeck et al., Reference Morbeck, Flowers and Britt1993) and growth of early follicles in cow (Gutierrez et al., Reference Gutierrez, Ralph, Telfer, Wilmut and Webb2000), hamster (Roy, Reference Roy1993), mouse (Boland & Gosden, Reference Boland and Gosden1994) and human (Roy & Kole, Reference Roy and Kole1998). In antral follicles, EGF has been shown to stimulate in vitro oocyte maturation (mouse: Smitz et al., Reference Smitz, Cortvrindt and Hu1998; De La Fuente et al., Reference De La Fuente, O'Brien and Eppig1999; sheep: Guler et al., Reference Guler, Poulin, Mermillod, Terqui and Cognie2000; cattle: Lonergan et al., Reference Lonergan, Carolan, Van Langendonckt, Donnay, Khatir and Mermillod1996; human: Goud et al., Reference Goud, Goud, Qian, Laverge, Van der Elst, De Sutter and Dhont1998; pig: Prochazka et al., Reference Prochazka, Kalab and Nagyova2003; Li et al., Reference Li, Liu, Jiao and Wang2002), cumulus cells expansion (mouse: O'Donnell et al., Reference O'Donnell, Hill and Gross2004), granulosa cell proliferation (pig: May et al., Reference May, Bridge, Gotcher and Gangrade1992) and estrogen production (human: Misajon et al., Reference Misajon, Hutchinson, Lolatgis, Trounson and Almahbobi1999).
The goat is an ideal model for the transgenic production of therapeutic recombinant proteins in the milk because of the high yield of purified product and relatively short generation interval (Reggio et al., Reference Reggio, James, Green, Gavin, Behboodi, Echelard and Godke2001). Thus, it is very important to understand the mechanisms that control folliculogenesis in this species to produce a large number of in vitro matured oocytes either to provide cytoplasts for cloning of transgenic goats or to produce large number of zygotes from valuable animals. With regard to EGF and its receptor, thus far, no information is available on their mRNA expression and protein localization in goat early follicles. Data about expression of both protein and mRNA for EGF in goat antral follicles are also lacking, while expression of EGF-R mRNA and protein has been studied only in oocyte and cumulus cells of goat antral follicles (Gall et al., Reference Gall, Chene, Dahirel, Ruffini and Boulesteix2004).
The aim of the present study was to examine the expression of EGF and EGF-R mRNA and protein in goat ovaries obtained from slaughterhouses, with special attention to early and late-stage follicles as possible sources of an EGF/EGF-R system. To this end, mRNA expression was detected by reverse transcriptase polymerase chain reaction (RT-PCR) and protein distribution was evaluated using immuno-histochemistry.
Materials and methods
Ovaries
During the breeding season, ovaries (n = 50) with large antral follicles and/or corpora lutea from slaughtered adult mixed-breed goats were recovered and transported to the laboratory in a Thermos flask, within 1 h. Ten of the ovaries were fixed overnight at room temperature in 4% (w/v) paraformaldehyde in phosphate-buffered saline (PBS, pH 7.4), and subsequently dehydrated and embedded in paraffin wax (Histoplast, Shandon Scientific, Pittsburgh, USA) in preparation for immunohistochemical studies. The remaining 40 ovaries were used to recover cells and tissues for RT-PCR.
Immunohistochemistry
Immunohistochemical study for EGF and EGF-R was performed on serial 5μm sections cut from 10 ovaries of five different goats. These sections were mounted on poly-l-lysine coated slides, dried overnight at 37 °C, deparaffinized in xylene and rehydrated in a graded ethanol series. Endogenous peroxidase was blocked by incubating the deparaffinized sections in 3% hydrogen peroxide in methanol for 10 min. The sections were then washed with PBS (pH 7.4) and the epitopes activated by microwaving the sections for 7 min at 900 W in 0.01 M citrate buffer (pH 6.0). Following microwave treatment, the sections were washed in PBS/0.05% Tween (PBS-T, Merck, Darmstadt, Germany) before being incubated for 30 min with 5% normal goat serum in PBS to minimize non-specific binding. The primary antibodies used were: (1) rabbit polyclonal anti-EGF (Z-12, Santa Cruz Biotechnology, Santa Cruz, CA) and (2) rabbit polyclonal anti-EGF receptor (SC-03, Santa Cruz Biotechnology), both diluted 1:100 in PBS containing 5% normal goat serum. The sections were incubated overnight at 4 °C in appropriate dilutions of the antibodies. All other incubations and washes were performed at room temperature. After incubation with an antibody, sections were washed three times with PBS-T and incubated for 45 min with goat anti-rabbit IgG (Vector Laboratories, Burlingame, CA), diluted 1:200 in PBS containing 5% normal goat serum. Next, the sections were washed three times in PBS-T before being incubated for 45 min with an avidin―biotin complex (1:600, Vectastain Elite ABC kits; Vector Laboratories, Burlingame, CA). The sections were then washed three times in PBS and stained with diaminobenzidine (DAB; 0.05% DAB in Tris/HCl pH 7.6, 0.03% H2O2 ― Sigma tablets, St Louis, MO) until a precipitate formed or for a maximum of 20 min. The stained sections were rinsed in PBS and water, and counterstained for 10 s in Mayer's haematoxylin. Finally, the sections were washed for 10 min in running tap water, dehydrated in a graded ethanol series and then xylene, and mounted in Depex. The staining intensity for both EGF and EGF-R immunoreactive protein expression was scored as follows: absent (−), occasionally found (−/+), weak (+), moderate (++) or strong (+++). Sections were analysed in this way by two independent researchers.
Controls for non-specific staining were performed by: (1) replacing the primary antibody with IgGs from the same species in which the specific antibody was raised, at the same concentration; (2) incubation with DAB reagent alone to exclude the possibility of non-suppressed endogenous peroxidase activity; and (3) preabsorbing the antibody (EGF-R) overnight at 4 °C with its blocking peptide at 20-fold excess (Santa Cruz Biotechnology).
Classification and measurement of follicles
Early follicles were classified as (1) primordial (one layer of flattened/cuboidal granulosa cells), (2) primary (a single layer of cuboidal granulosa cells) and (3) secondary (two or more layers of cuboidal granulosa cells). These follicles were considered either healthy, when a morphologically normal oocyte was surrounded by granulosa cells organized in discrete layers, or atretic, when pyknosis was present in oocyte and/or granulosa cells. Secondary follicles with irregular spaces between the layers of granulosa cells, but without pyknotic granulosa cells or degenerating oocytes were classified as healthy, since these irregular spaces are considered to be early signs of antrum formation (Hirshfield, Reference Hirshfield1983). Antral follicles were classified into two groups: (1) small antral follicles (<3 mm in diameter; with multiple granulosa cells enclosing an antrum) and (2) large antral follicles (3–6 mm). Among the healthy large antral follicles, it was not possible to distinguish between subordinate and dominant follicles. Antral follicles were classified as healthy when pyknotic cells were absent or occasionally present (less than 5% per follicle), while those follicles having more than 5% of pyknotic granulosa cells were considered atretic. The diameter of follicles was calculated according to the method described by Van den Hurk et al. (Reference Van den Hurk, Dijkstra, Hulshof and Vos1994).
Collection of cells and tissues for RT-PCR
The recovered ovaries were rinsed in saline (0.9% NaCl) containing antibiotics (100 IU/ml penicillin and 100μg/ml streptomycin). Then, 10 ovaries were allocated for isolation of early-staged follicles and the others were used to provide antral follicles, oocytes, cumulus cells, mural granulosa cells and samples of corpora lutea and ovarian surface.
Early-stage follicles, i.e. primordial, primary and secondary, were isolated using the mechanical procedure described previously (Lucci et al., Reference Lucci, Amorim, Bao, Figueiredo, Rodrigues, Silva and Gonçalves1999). Briefly, ovaries were cut individually into small fragments using a tissue chopper (Mickle Laboratory Engineering, Gomshal, Surrey, UK) adjusted to 75μm. The fragments were then placed in PBS containing 5% bovine serum albumin (Sigma) at room temperature and aspirated 40 times, using a large Pasteur pipette (diameter ∼1600μm), and 40 times with a smaller pipette (diameter ∼600μm). The suspension was then filtered successively through 500 and 100μm nylon mesh filters. After repeated washing to completely remove the stromal cells, follicles from each of the three different categories (primordial, primary and secondary) were placed in separate Eppendorf tubes in groups of 15. All samples were stored at −80 °C until the RNA was extracted. Previously, we had performed histological analysis to confirm the classification of goat preantral follicles after isolation (Lucci et al., Reference Lucci, Amorim, Bao, Figueiredo, Rodrigues, Silva and Gonçalves1999).
From a second group of ovaries (n = 20), cumulus–oocyte complexes (COCs) were aspirated from small (1–3 mm) and large (3–6 mm) antral follicles using an 18-gauge needle attached to a tube in line with a vacuum pump. From the follicle content thus collected, compact COCs were selected as described by Van Tol & Bevers (Reference Van Tol and Bevers1998). Thereafter, the cumulus was separated from the oocyte by a combination of vortexing and aspiration via a narrow-bore Pasteur pipette. After removal of denuded oocytes, the remaining cumulus cells were collected separately and washed four times in PBS. Groups of either 10 denuded oocytes or cumulus cells from 10 COCs were packed in tubes and stored at −80 °C until RNA extraction.
To collect theca cells, small (n = 10) and large antral follicles (n = 10) were isolated from goat ovaries (n = 5) and dissected free of stromal tissue using forceps, as described previously for bovine ovaries (Van Tol & Bevers, Reference Van Tol and Bevers1998). Those follicles that had denuded oocytes and signs of atresia were discarded. The follicles were then bisected and the mural granulosa cells scraped off using a scalpel blade, and then washed and stored until RNA extraction. Next, the theca cell layers were vortexed for 1 min in 1 ml HEPES-buffered M199 (Gibco BRL, Paisley, UK) supplemented with penicillin (100μg/ml) and streptomycin (100μg/ml) to remove contaminating mural granulosa cells, transferred to a fresh 1 ml of buffer, vortexed for another minute, washed twice in 2 ml HEPES-buffered M199, collected and stored at −80 °C. From another group of ovaries (n = 5), small pieces of corpus luteum and surface epithelium were collected and stored at −80 °C until RNA extraction. Since ovaries collected from slaughterhouses were used it was not possible to determine the exact stage of luteal phase. Three samples of each tissue sample were collected and analysed. Appropriate tests to ensure the purity of samples (oocytes, cumulus cells, mural granulosa cells or theca cells) had previously been performed using gene differential expression. In these previous studies, GDF-9 was demonstrated in granulosa cells but not in the theca, while Kit Ligand was absent in oocytes and present in granulosa cells (Silva et al., Reference Silva, Van den Hurk, van Tol, Roelen and Figueiredo2004b, Reference Silva, Van den Hurk and Figueiredoc).
Extraction of total RNA and reverse transcription
Isolation of total RNA combined with on-column DNase digestion was performed using the RNeasy mini kit and the RNase-free DNase set (Qiagen, Valencia, USA). Following the manufacturer's instructions, 350μl lysis buffer was added to each frozen sample and the lysate aspirated through a 20-gauge needle before being centrifuged at 10000 g for 3 min at room temperature. The lysates of theca cells, corpus luteum and ovarian surface samples were then subjected to proteinase K treatment (6.7 mAU/ml, Qiagen, Valencia, USA) at 55 °C for 10 min. Thereafter, all lysates were diluted 1:1 with 70% ethanol and introduced to a mini-column. After binding of the RNA to the column, DNA digestion was performed using RNase-free DNase (340 Kunitz units/ml) for 15 min at room temperature. After washing the column three times, the RNA was eluted with 30μl RNase-free water.
Prior to the reverse transcription reaction, the eluted RNA samples were incubated for 5 min at 70 °C, and chilled on ice. Reverse transcription was then performed in a total volume of 20μl made up of 10μl of sample RNA, 4μl 5× reverse transcriptase buffer (Gibco BRL, Breda, The Netherlands), 8 units RNAsin, 150 units Superscript II reverse transcriptase (BRL), 0.036 U random primers (Life Technologies, Leiden, The Netherlands) and containing 10 mM dithiothreitol (DTT) and 0.5 mM of each dNTP. The mixture was successively incubated for 1 hr at 42 °C and 5 min at 80 °C, and then stored at −20 °C. Minus RT blanks were prepared under the same conditions, but without inclusion of reverse transcriptase.
Amplification of cDNA by PCR
PCR reactions were carried out in 200μl tubes (Biozym, Landgraaf, The Netherlands), using 1μl cDNA as template in 25μl of a mixture containing 2 mM MgCl2, 200μM of each dNTP, and 0.5μM each of primers and 0.625 units Taq DNA polymerase (HotStarTaq, Qiagen, Valencia, USA) in 1× PCR buffer. The primers for EGF-R and EGF used for amplification are presented in Table 1.
Oligonucleotide primers used for PCR analysis of goat cells and tissues
s, sense; a, antisense.
The thermal cycling profile for EGF and EGF-R during amplification was: initial denaturation and activation of the polymerase for 15 min at 94 °C, followed by 40 cycles of 15 s at 94 °C, 30 s at 50 °C and 45 s at 72 °C. Final extension was for 10 min at 72 °C. During the amplification of EGF cDNA, heminesting was used to increase the specificity and sensitivity. For heminesting, 1μl of the first round product was transferred to another 200μl tube containing 24μl amplification mixture, and amplified for 25 cycles using the same thermal cycling profile. All reactions were performed in a 24-well thermocycler (Perkin-Elmer, Gouda, The Netherlands). Finally, 10μl of the product was resolved by electrophoresis in 1% agarose gel containing ethidium bromide. A 100 base pair (bp) DNA ladder (Gibco BRL) was included as a reference for fragment size and image of each gel was recorded using a digital camera (Olympus C-4040, New York, USA).
A standard sequencing procedure (ABI PRISM 310 Genetic Analyzer, Applied Biosystems) was used to verify the specificity of the PCR products.
Results
Immunohistochemistry
The immunolocalization of EGF and EGF-R proteins in healthy goat follicles is illustrated in Fig. 1 and Table 2. In primordial, primary and secondary follicles, a moderate EGF immunoreaction was generally observed in oocyte and granulosa cells, with the exception of the oocyte of secondary the follicle that had a weak detection (Fig. 1A–C). In the secondary follicles, a weak reaction was also present in early theca cells (Fig. 1C). In small antral follicles, strong EGF immunostaining was localized in oocyte, cumulus cells and mural granulosa cells, but theca cells had weak staining (Fig. 1D). In large antral follicles, oocyte and theca cells showed weak EGF immunostaining while cumulus cells and mural granulosa cells had moderate staining (Fig. 1E). In addition, moderate to strong and strong EGF immunoreactivity is generally observed in corpora lutea (Fig. 1F) and ovarian surface epithelium (Fig. 1G), respectively. Both oocyte and granulosa cells from atretic follicles were irregularly stained for EGF (not shown). No specific immunoreaction was observed when control stainings were carried out (Fig. 1H).
EGF immunoreactivity in healthy ovarian follicles, corpus luteum and ovarian surface epithelium. (A) Primordial follicle, (B) primary follicle, (C) secondary follicle, (D) small antral follicle, (E) large antral follicle, (F) corpus luteum, (G) ovarian surface epithelium, (H) negative control. O, oocyte; G, granulosa cells; MGC, mural granulosa cells; CC, cumulus cells; T, theca cells; CL, corpus luteum; S, ovarian surface epithelium. Scale bars represent 25μm (A–C, F–H), 50μm (D) and 100μm (E).
Localization of mRNA and relative intensity of immunohistochemical staining for EGF and EGF-R in the ovaries of goats
aWhole follicles.
−, absent; −/+, occasionally found; +, weak; ++, moderate; +++, strong immunoreaction.
The EGF-R protein was immunohistochemically demonstrated in healthy oocytes of primordial, primary and secondary follicles (Fig. 2A–C). The protein was distributed throughout the cytoplasm of the oocytes, while the staining intensity in primordial follicles was stronger than in primary and secondary follicles (Table 2). Occasionally, granulosa cells of primordial follicles were weakly stained for EGF-R (Fig. 2A), but a moderate immunoreaction was observed in granulosa cells of primary and secondary follicles (Fig. 2B, C, Table 2). No EGF-R staining was detected in the theca of secondary follicles (Fig. 2C). Oocytes of small antral follicles showed strong EGF-R immunoreactivity, but cumulus cells, mural granulosa cells, and occasionally theca cells were weakly stained (Fig. 2D). In large antral follicles, moderate EGF-R immunostaining was present in all follicular compartments, except for the oocytes, which were weakly stained (Fig. 2E). Follicles with signs of atresia were irregularly stained for EGF-R (not shown). Apart from follicles, corpora lutea and ovarian surface epithelium immunoreacted with the EGF-R antibody, moderately to strongly (Fig. 2F) and strongly (Fig. 2G), respectively. No specific immunoreaction was observed when control stainings were carried out (Fig. 2H).
EGF-R immunoreactivity in healthy ovarian follicles, corpus luteum and ovarian surface epithelium. (A) Primordial follicle, (B) primary follicle, (C) secondary follicle, (D) small antral follicle, (E) large antral follicle, (F) corpus luteum, (G) ovarian surface epithelium, (H) negative control. O, oocyte; G, granulosa cells; MGC, mural granulosa cells; CC, cumulus cells; T, theca cells; CL, corpus luteum; S, ovarian surface epithelium. Scale bars represent 25μm (A–C, F–H), 50μm (D) and 100μm (E).
Expression of mRNA for EGF and EGF-R in goat ovaries
Using primers for EGF, the first round of amplification yielded abundant products only in samples of cDNA prepared from oocytes collected from small and large antral follicles. After heminesting, however, transcripts for EGF were observed in cDNA from primordial, primary and secondary follicles as well as from cumulus, mural granulosa or theca cells collected from small or large antral follicles. EGF expression was also detected in ovarian surface epithelium, but no EGF mRNA was detected in corpus luteum (Fig. 3). Amplification of –RT blanks or water controls yielded no specific products in any of the reactions.
Expression of EGF and EGF-R mRNA in different follicle and cell types in goat ovaries. Follicle and cell types are indicated at the top. One-hundred base pair ladders are included as markers for fragment size.
With the use of specific primers for EGF-R, amplification of cDNA from primordial, primary and secondary follicles and from oocytes, cumulus cells, mural granulosa and theca cells from small or large antral follicles in all cases resulted in an abundant product after one round of amplification. Expression for EGF-R was also detected in corpus luteum and ovarian surface (Fig. 3). Amplification of –RT blanks or water controls yielded no specific products in any of the reactions.
Figure 4 shows the sequence of the amplified product for EGF and EGF-R. At the nucleotide level, EGF product showed 93% identity with EGF cDNA from ovine species. Similarly, the EGF-R product displayed 97% homology with EGF-R cDNAs from bovine. These results confirmed the specificity of EGF and EGF-R products. A truncated EGF-R has been described in human species, but to our knowledge there is no information about this truncated form in ruminants.
Nucleotide sequence of the amplified product of (A) goat EGF (gEGF) and (B) EGF-R (gEGF-R) and their alignment with the corresponding ovine EGF (oEGF) and bovine EGF-R (bEGF-R).
Discussion
The present study examined the distribution of EGF and EGF-R mRNA and protein in goat ovaries, to determine whether EGF signalling may play a role in ovarian functioning in this species, and particularly in folliculogenesis. With regard to EGF, we demonstrated the presence of protein in oocyte and granulosa cells of primordial, primary and secondary follicles. A similar distribution of EGF was described previously for pig (Singh et al., Reference Singh, Rutledge and Armstrong1995a) and human (Bennett et al., Reference Bennett, Osathanondh and Yeh1996) early-staged follicles. Using RT-PCR, the current study demonstrated the expression of mRNA for EGF in caprine primordial, primary and secondary follicles and, to our knowledge, is the first study to describe EGF mRNA expression in mammalian early follicles. The mRNA and the protein for EGF-R were both detected in primordial, primary and secondary goat follicles. This supports our previous conclusion that EGF is involved in oocyte growth of goat primary follicles in vitro (Silva et al., Reference Silva, Van den Hurk, Matos, Santos, Pessoa, Moraes and Figueiredo2004a). In other species, both protein and mRNA for EGF-R are demonstrated in early follicles (pig: Singh et al., Reference Singh, Rutledge and Armstrong1995a; human: Maruo et al., Reference Maruo, Ladines-Llave, Samoto, Matsuo, Manalo, Ito and Mochizuki1993; Bennett et al., Reference Bennett, Osathanondh and Yeh1996; Qu et al., Reference Qu, Godin, Nisolle and Donnez2000; hamster: Garnett et al., Reference Garnett, Wang and Roy2002). In vitro studies with preantral follicles showed that EGF promotes growth (cow: Wandji et al., Reference Wandji, Eppig and Fortune1996; Saha et al., Reference Saha, Shimizu, Geshi and Izaike2000, Gutierrez et al., Reference Gutierrez, Ralph, Telfer, Wilmut and Webb2000; sheep: Hemamalini et al., Reference Hemamalini, Rao, Tamilmani, Amarnath, Vagdevi, Naidu, Reddy and Rao2003; buffalo: Gupta et al., Reference Gupta, Nandi, Ravindranatha and Sarma2002) proliferation of their granulosa cells (pig: Morbeck et al., Reference Maruo, Ladines-Llave, Samoto, Matsuo, Manalo, Ito and Mochizuki1993), and that it reduces the rate of atresia (cow: Wandji et al., Reference Wandji, Eppig and Fortune1996). Additionally, EGF has been shown to regulate the expression of TGFß receptors (hamster: Yang & Roy, Reference Yang and Roy2001) and connexin 43 (pig: Bolamba et al., Reference Bolamba, Floyd, McGlone and Lee2002; rabbit: Kennedy et al., Reference Kennedy, Floyd, Clarkson and Lee2003) in preantral follicles. Connexin 43 is essential for gap junction formation in granulosa cells and plays a critical role in early follicle growth at least in rodents, since the absence of its gene disrupts progression of follicles beyond primary stages in transgenic mouse ovaries (Juneja et al., Reference Juneja, Barr, Enders and Kidder1999).
In both small and large caprine antral follicles, mRNA and protein for EGF and its receptor were present in oocyte, cumulus cells, mural granulosa cells and theca cells. This expression pattern argues for autocrine and paracrine roles for EGF in directing the development of goat antral follicles. In goat antral follicles, EGF-R expression was recently described to be present in the oocyte and in cumulus cells (Gall et al., Reference Gall, Chene, Dahirel, Ruffini and Boulesteix2004), while in agreement with our present findings, both protein and mRNA have been found in all antral follicle compartments of other species (pig: Singh et al., Reference Singh, Rutledge and Armstrong1995a; hamster: Garnett et al., Reference Garnett, Wang and Roy2002; human: Maruo et al., Reference Maruo, Ladines-Llave, Samoto, Matsuo, Manalo, Ito and Mochizuki1993). In vitro studies with antral follicles have demonstrated that EGF stimulates oocyte maturation (mouse: Smitz et al., Reference Smitz, Cortvrindt and Hu1998; Merriman et al., Reference Merriman, Whittingham and Carroll1998; De La Fuente et al., Reference De La Fuente, O'Brien and Eppig1999; sheep: Guler et al., Reference Guler, Poulin, Mermillod, Terqui and Cognie2000; cattle: Lonergan et al., Reference Lonergan, Carolan, Van Langendonckt, Donnay, Khatir and Mermillod1996; Sakaguchi et al., Reference Sakaguchi, Dominko, Yamauchi, Leibfried-Rutledge, Nagai and First2002; human: Goud et al., Reference Goud, Goud, Qian, Laverge, Van der Elst, De Sutter and Dhont1998; pig: Singh et al., Reference Singh, Meng, Rutledge and Armstrong1997, Prochazka et al., Reference Prochazka, Srsen, Nagyova, Miyano and Flechon2000, Reference Prochazka, Kalab and Nagyova2003). In cumulus cells, EGF promotes Ca2+ efflux and improves their expansion during maturation (Hill et al., Reference Hill, Hammar, Smith and Gross1999; O'Donnell et al., Reference O'Donnell, Hill and Gross2004). It was furthermore reported that EGF stimulates granulosa cells proliferation in vitro (pig: May et al., Reference May, Bridge, Gotcher and Gangrade1992) and estrogen production through aromatase activation (human: Misajon et al., Reference Misajon, Hutchinson, Lolatgis, Trounson and Almahbobi1999; goat: Behl & Pandey, Reference Behl and Pandey2001). There is evidence for EGF regulating the in vitro cellular activities of granulosa cells either by inhibiting LH receptor production (Hattori et al., Reference Hattori, Yoshino, Shinohara, Horiuchi and Kojima1995) or inhibin secretion (Serta & Seibel, Reference Serta and Seibel1993) or by stimulating FSH receptor expression (Luciano et al., Reference Luciano, Pappalardo, Ray and Peluso1994) and binding affinity (May et al., Reference May, Buck and Schomberg1987). Very recently, Park et al. (Reference Park, Su, Ariga, Law, Jin and Conti2004) demonstrated that LH stimulation induces transient expression of the EGF family members amphiregulin, epiregulin and beta-cellulin in antral follicles. In vitro, these growth factors promote the morphological events triggered by LH, including cumulus expansion and oocyte maturation, which supports the idea that EGF-related growth factors are paracrine mediators that propagate the LH signal throughout the follicle, via EGF-R (Park et al., Reference Park, Su, Ariga, Law, Jin and Conti2004).
Apart from follicles, EGF, EGF-R protein and EGF-R mRNA were detected in goat corpora lutea, suggesting a possible role of EGF in luteal activity. The presence of EGF protein but not EGF mRNA in goat corpora lutea points to an EGF origin different from luteal cells. This is in contrast with data from other species, in which EGF mRNA and protein have both been demonstrated in luteal cells (rat: Tekpetey et al., Reference Tekpetey, Singh, Barbe and Armstrong1995; pig: Kennedy et al. Reference Kennedy, Brown and Vaughan1993, Singh et al., Reference Singh, Kennedy, Tekpetey and Armstrong1995b). The importance of EGF/EGF-R for luteal activity is supported by Tekpetey et al. (Reference Tekpetey, Singh, Barbe and Armstrong1995), who demonstrated EGF-induced progesterone production by rat luteal cells that had been cultured in vitro.
The strong expression of both mRNA and proteins for EGF and EGF-R in goat ovarian surface epithelium (OSE) also suggests a role for EGF/EGF-R at this site. A paracrine effect of EGF from ovarian surface on neighbouring follicles cannot be excluded. It is well established that the OSE must proliferate to repair the ovulatory defects in the ovarian surface and that approximately 90% of the ovarian cancers arise in the OSE (Auersperg et al., Reference Auersperg, Wong, Choi, Kang and Leung2001). Studies on expression of EGF and its receptor in healthy OSE are very limited but EGF-R has been localized in the malignant OSE (van der Burg et al., Reference Van der Burg, Henzen-Logmans, Foekens, Berns, Rodenburg, van Putten and Klijn1993). Moreover, EGF and EGF-R-like peptides are expressed in about one-half of all ovarian tumours; yet their role in tumorigenesis is unclear (van Haaften-Day et al., Reference Van Haaften-Day, Russell, Boyer, Kerns, Wiener, Jensen, Bast and Hacker1996). McClellan et al. (Reference McClellan, Kievit, Auersperg and Rodland1999) demonstrated that EGF regulates proliferation and apoptosis in human ovarian surface epithelial cells. Aberration in the expression of the EGF/EGF-R system in OSE thus may evoke a deviation of the normal proliferation and programmed death of its cells, which may lead to tumour formation.
In conclusion, the present study demonstrates that EGF and its receptor are expressed in goat ovarian follicles at all stages of follicle development, in corpora lutea, and in ovarian surface epithelium. This widespread distribution of EGF and its receptor shows that, in goat ovaries, it may play an important role in various processes, including early and advanced folliculogenesis, luteal activity and surface epithelium behaviour. The demonstrated presence of an EGF/EGF-R system in early-staged follicles supports previous findings of in vitro studies, in which we showed a stimulating effect of EGF on oocyte growth in goat primary follicles.
Acknowledgements
This study was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – CAPES, Brazil. The authors thank the members of the Department of Image Processing and Design (Faculty of Biology, Utrecht, The Netherlands) for help with photography and H.T.A. van Tol (Faculty of Veterinary, University of Utrecht) for help with PCR studies.
