| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Institute of Molecular Animal Breeding and Biotechnology3 and Laboratory for Functional Genome Analysis (LAFUGA),4 Gene Center of the Ludwig-Maximilians University, 81377 Munich, Germany
Institute of Veterinary Anatomy,5 Ludwig-Maximilians University, 80539 Munich, Germany
ABSTRACT
The epidermal growth factor receptor (EGFR) and its ligands are emerging as key molecules in regulating female reproduction. Here, we used a transgenic mouse model to evaluate whether and at which level of the reproduction cascade higher-than-normal levels of the EGFR ligand betacellulin (BTC) in the reproductive organs affect fertility. Western blots and immunohistochemistry revealed increased BTC levels in uterus and ovaries from transgenic females, particularly evident in granulosa cells of antral follicles. Onset of puberty, estrous cyclicity, and the anatomy and histology of reproductive organs at puberty were not altered as compared to control females. Fertility tests revealed a reduction (
50%) in litter size as the major reproductive deficit of transgenic females. Embryo implantation was delayed in transgenic females, but this was not the reason for the reduced litter size. Transgenic females produced a normal number of oocytes after natural ovulation. The in vivo fertilization rate was significantly reduced in untreated transgenic females but returned to normal levels after superovulation. Impaired oocyte fertilization in the absence of superovulation treatment was associated with MAPK3/MAPK1 hyperactivation in BTC transgenic ovaries, whereas similar levels of MAPK3/MAPK1 activation were detected in transgenic and control ovaries after superovulation treatment. Thus, tight regulation of MAPK3/MAPK1 activity appears to be essential for appropriate granulosa cell function during oocyte maturation. Our study identified hitherto unknown effects of BTC overabundance in reproduction and suggests BTC as a novel candidate protein for the modulation of fertility.
betacellulin, EGFR, female reproductive tract, fertilization, growth factors, implantation, MAPK3/MAPK1, transgenic mice, uterus
Peptide growth factors are omnipresent molecules that coordinate every conceivable aspect of mammalian development, growth, physiology, and pathology. Reproduction is not an exception, and its success depends on the appropriate expression and activity of growth factors belonging to several different families. Among them, the epidermal growth factor (EGF)-like peptides are emerging as major players in regulating different aspects of female reproduction. The EGFR system comprises the seven EGF-like growth factors amphiregulin (AREG), betacellulin (BTC), epidermal growth factor (EGF), epigen, epiregulin (EREG), heparin-binding EGF-like growth factor (HBEGF), and transforming growth factor alpha (TGFA) and four tyrosine kinase receptors called ERBB1 (corresponding to the EGFR), ERBB2 (neu), ERBB3, and ERBB4 [1–3]. The ERBBs consist of an extracellular domain where ligand binding takes place, a short transmembrane domain, and a cytoplasmic region containing the catalytic protein tyrosine kinase. Ligand binding causes receptor dimerization and the phosphorylation of cytoplasmic domains, which serve in their turn as docking sites for a variety of molecules whose recruitment initiates a cascade of intracellular signaling events. Subsequently, the ligand-receptor complex is internalized and degraded within lysosomes [2, 4]. The EGFR ligands are synthesized as transmembrane type I proteins that are cleaved to release the mature growth factor. Members of the ADAM (a disintegrin and metalloprotease) family of enzymes have been identified as their main sheddases [5]. Interestingly, both the precursor, membrane-bound form and the mature, circulating form of the EGFR ligands are believed to be actively engaged in signaling (juxtacrine and paracrine or endocrine actions, respectively). Concerning reproduction, EGFR ligands have been shown to be involved in the process of sexual maturation [6], oocyte maturation and ovulation [7–12], preimplantation embryonic development [13], and implantation [14–18]. These actions have been demonstrated to be dependent on a functional EGFR.
BTC, a rather poorly characterized EGFR ligand initially isolated from the conditioned medium of a mouse pancreatic β-cell carcinoma cell line [19], belongs to this group of growth factors. BTC was identified as one of the EGFR ligands expressed in the mouse uterus exclusively at the sites of blastocyst apposition at the time of attachment reaction (Day 4) and through the early phase of implantation (Day 5) [16]. In addition, it was identified as a mediator of luteinizing hormone (LH) [7, 8, 10], prostaglandin (PG), and progesterone receptor (PGR) [9] actions in the ovulatory follicle. Finally, BTC was recently identified as a possible ovarian mediator of bone morphogenetic protein 15 actions, an oocyte-specific growth factor that plays a major role in determining ovulation quota in mammals [20].
Although, in many aspects, BTC reproduces the actions of other EGFR ligands, it has some unique structural and functional properties (reviewed in [21]). Knockout mice lacking BTC are healthy and fertile [22], suggesting a high degree of functional redundancy within the family of EGFR ligands. Hypothesizing that the opposite approach, overexpression of BTC, would be more informative, we recently generated transgenic mice overexpressing this growth factor [23]. Since elevated levels of BTC are present in female reproductive tissues, we took advantage of this model to clarify 1) whether the physiological timely controlled expression of BTC in the uterus is a prerequisite for blastocyst apposition and implantation; 2) whether overexpression of BTC affects the expression of other tightly regulated EGFR ligands, such as AREG, in the uterus; 3) whether abnormal EGFR expression in the uterus impairs reproductive performance of female mice; and 4) whether overexpression of BTC in the ovary affects oocyte developmental capacity. To clarify these points, we performed a systematic study of BTC transgenic female mice and nontransgenic littermate controls, covering all steps of the reproduction cascade.
The generation and genotyping of transgenic mice overexpressing BTC has been described in detail elsewhere [23]. Briefly, ubiquitous expression of a cDNA sequence coding for full-length BTC was obtained by placing it under the control of the cytomegalovirus immediate-early enhancer fused to the chicken beta actin gene promoter. The relevant sequences were released from the vector and used for pronuclear microinjection into fertilized oocytes (FVB/N). Transgenic and nontransgenic control mice from the FVB/N strain were maintained in a specific pathogen-free facility with controlled photoperiod (14L:10D) and temperature (22°C–25°C) and received a standard rodent diet (V1534; Ssniff, Soest, Germany) and water ad libitum. Mice used for expression studies and for phenotype analysis were weaned at an age of 3 wk, marked by ear piercing, and housed in cages separated by sex. At the time of weaning, tail tips were clipped, frozen on dry ice, and stored at –80°C for genotype analysis. For the determination of organ weights, the structures were dissected free of adjacent tissues and weighed to the nearest 0.1 mg. The female fertility testing was performed with animals from lines 2 and 4 [23]; for all other studies, animals from line 2 were employed. All experimental procedures involving animals were approved by the author's institutional committee on animal care and carried out with permission from the responsible veterinary authority.
Western Blot and Immunohistochemistry of BTC
Detection of BTC and actin in ovaries by Western blotting was performed as described previously [23]. For immunohistochemistry, the ovaries, oviducts, and uterus were fixed in 4% paraformaldehyde (in phosphate-buffered saline [PBS]) at 4°C overnight and routinely processed for paraffin embedding. Serial 5-µm sections were either stained with hematoxylin and eosin and examined by light microscopy or further processed for immunohistochemistry. For this purpose, sections were deparaffinized and rehydrated in PBS for 20 min. Blocking of endogenous peroxidase activity was achieved by incubation in 3% H2O2 in PBS for 10 min at 37°C. The sections were incubated with the primary antibody, a goat anti-mouse BTC antibody (R&D Systems, Wiesbaden, Germany) for 1 h at 37°C. Then the sections were rewashed in Tris-buffered saline two times for 5 min and incubated with the secondary antibody, a biotinylated rabbit anti-goat Ig (Dako, Glostrup, Denmark). Slides were incubated with avidin-biotin complex for 30 min at room temperature, stained with diaminobenzidine (DAB), counterstained with hematoxylin, mounted, and examined under a bright-field microscope.
Onset of Puberty and Estrous Cycle Tracking
Nontransgenic and BTC transgenic females were observed daily from the time of weaning (21 days) for the appearance of vaginal opening. Vaginal opening is widely used as a marker for female sexual maturation in mice because it is regulated by estrogen and coupled to the initiation of ovarian activity. Once vaginal opening occurred, daily cytological analysis of vaginal smears was performed to unambiguously ascertain the occurrence of puberty. The cycle's length was defined as the interval between proestrus smears, and the onset of estrous cyclicity was defined as the first day of the first cycle less than or equal to 6 days in length [24].
The breeding performance was systematically analyzed by mating BTC transgenic females and control females with nontransgenic FVB males of proven fertility. All animals were 2 mo old at the beginning of the experiment, and each mating pair was housed in a separate cage for exactly 3 mo. After birth, the pups remained in the cages for 3 wk. The interval to the first litter, the number of litters, and litter size were recorded. After 3 mo, the females were housed individually and monitored for further 3 wk.
Adult BTC transgenic and control females were mated with fertile nontransgenic males, and the morning of finding a vaginal plug was considered Day 1 of pregnancy. By the end of Day 4 (2400 h) or in the morning (0900 h) of Day 5 of pregnancy, the mice were injected intravenously (200 µl/mouse) with a solution of Evans Blue dye (1% in PBS) and killed 5 min after injection to examine the attachment reaction and progression of the implantation process. Blue bands along the uterus indicated the sites of blastocyst apposition [25]. After the observation of the implantation sites, the uterus was flushed to recover embryos that were not implanted. Animals without any blue sites or flushed embryos were not included in the study.
Uteri from control and transgenic females were collected at the indicated time points and homogenized in TriZol reagent (Invitrogen, Karslruhe, Germany), and total RNA was isolated according to the manufacturer's instructions. Fifteen micrograms of total RNA were separated by electrophoresis in a formaldehyde gel and blotted on positively charged nylon membranes (Pall Corporation, Pensacola, FL). 32P-labeled probes were generated with the Rediprime labeling system (GE Healthcare, Buckinghamshire, UK) and hybridized in Rapid-hyb buffer (GE Healthcare). Signals were visualized on Phosphorimager-Storm (GE Healthcare). The Btc [23] and the Gapdh [26] cDNA probes have been described previously. The Areg cDNA probe was generated by PCR using the primers AREG1 (forward, 5' TGC TGC TGG TCT TAG GCT CA 3') and AREG2 (reverse, 5' CGT TTC CAA AGG TGC ACT GTG 3').
Natural Ovulation, Superovulation, and Fertilization
For evaluating the natural ovulation rate, adult wild-type and BTC transgenic females were mated with nontransgenic males and examined every morning and evening for the presence of a copulatory plug (= Day 0). Positive animals were killed 12 h later, the genital tract was excised, and the oocyte/cumulus masses were collected from the oviducts by flushing. The total number of oocytes and the number of fertilized oocytes were recorded after enzymatic dissociation from the surrounding cumulus with 0.3% hyaluronidase (Sigma, Deisenhofen, Germany). Superovulation was induced in females (8 wk old) with a single intraperitoneal injection of 5 units of equine chorionic gonadotropin (eCG, Intergonan; Intervet, Unterschleissheim, Germany). After 48 h, the females received 5 units of human chorionic gonadotropin (hCG, Ovogest; Intervet) intraperitoneally and were mated to nontransgenic males. The total number of oocytes and the number of fertilized oocytes were recorded as described previously.
In Vitro Maturation and In Vitro Fertilization
Induction of follicle growth to the preovulatory stage with eCG was performed as described previously, and the animals were killed 44 h later. The ovaries were removed and stored in Opti-MEM medium (Gibco, Grand Island, NY) with 10% fetal calf serum (FCS; Biochrom AG, Berlin, Germany). Large antral follicles were punctured with 25-gauge needles (BD Medical Systems, Drogheda, Ireland) to release cumulus-oocyte complexes (COCs). Only COCs that consisted of an oocyte surrounded by at least three complete layers of cumulus cells were selected for further culture. COCs were taken through three washes of Opti-MEM medium plus 10 IU/ml eCG and 10% FCS before being cultured in groups of 20 in 80-µl drops of pre-equilibrated medium and overlaid with silicone oil (Sigma) in an atmosphere of 5% CO2 in humidified air at 37°C. After 16 h of culture, COCs were removed from the drop, and the oocytes were denuded of cumulus cells by repeated gentle pipetting in medium containing 0.1% hyaluronidase. Morphological changes of the cultured oocytes were evaluated and classified. Immature oocytes were classified as GV when the germinal vesicle (GV) was present, as metaphase I (MI) when GV was broken down, as metaphase II (MII) when the first polar body was extruded, and as degenerated when oocytes were dark, granulated, or fragmented. The degree of cumulus expansion was assessed according to a subjective scoring system using a scale of 0 (no expansion) to +4 (maximal expansion) [20].
For in vitro fertilization, sperm were collected from the caudae epididymidum and vasa deferentia of fertile nontransgenic male mice in IVF medium (Vitrolife, Kungsbacka, Sweden). The percentages of motile and progressive spermatozoa and the sperm concentration were recorded. Oocytes obtained from eCG-stimulated transgenic or control females and subjected to in vitro maturation as described previously were washed three times in the fertilization medium (IVF, Vitrolife) and placed in fertilization drops, where the spermatozoa were added with a concentration of 1 x 106/ml. Six hours after fertilization, oocytes were denuded from the remaining cumulus cells by repeated gentle pipetting, evaluated for successful fertilization (appearance of two pronuclei), and cultured for detection of cleavage after further 24 h.
Time Course of Spontaneous Oocyte Maturation
Follicle growth to the preovulatory stage was induced with eCG, the animals were killed 44 h later, and the ovaries were collected as described previously. COCs were punctured from large antral follicles. Structures consisting of an oocyte surrounded by at least three complete layers of cumulus cells were selected for further culture. Oocytes were denuded in Opti-MEM complete medium with 1 µg/ml hyaluronidase by repetitive pipetting and cultured in groups of 20 in 80-µl drops of pre-equilibrated medium and overlaid with silicone oil (Sigma) in an atmosphere of 5% CO2 in humidified air at 37°C. They were regularly examined under a stereomicroscope, and the time course of spontaneous oocyte maturation in vitro was determined by detecting germinal vesicle breakdown (GVBD), a hallmark of oocyte meiotic resumption, and the extrusion of the first polar body, indicating nuclear maturation. The time oocytes were released from the follicles was considered time 0. The percentage of oocytes in a well that had undergone GVBD or extrusion of the first polar body was defined as (number of oocytes in GVBD or metaphase II stage/total number of oocytes) x 100.
MAPK3/MAPK1 Activity in the Ovaries
Ovaries were collected from control and transgenic females either at the proestrus stage or 3 h after hCG application to eCG-stimulated animals (as described previously for the superovulation). This protocol has been chosen because almost 100% of GVBD is attained at 3–4 h after hCG administration to eCG-primed mice [27]. Ovaries were homogenized in lysis buffer (Cell Signaling, Danvers, MA), and 25 µg of total protein were separated by SDS-PAGE and transferred to PVDF membranes by electroblotting. Loading of equal amounts of protein for each sample was verified with Ponceau staining. An antibody that reacts specifically with the active, phosphorylated form of MAPK3/MAPK1 (Cell Signaling; #4376) and a secondary goat anti-rabbit-HRP-conjugated antibody (Cell Signaling) were employed. Bound antibodies were detected using an enhanced chemiluminescence detection reagent (ECL Advance Western Blotting Detection Kit, GE Healthcare) and appropriate x-ray films (GE Healthcare). After detection, membranes were stripped and incubated with a second antibody recognizing total MAPK3/MAPK1 (Cell Signaling; #9102). Band intensities were quantified using the ImageQuant software package (GE Healthcare).
The data presented correspond to means ± SD. Student t-tests were performed to compare the data derived from transgenic and control mice. Differences were considered significant when the value of P was < 0.05.
Organ Weights, Gross Anatomy, and Histology
In a preliminary characterization of the mouse lines employed in this study [23], we reported expression of transgene-derived BTC in the gonads and an increase in the relative weight of the ovaries of transgenic animals (see also Fig. 1A). Further studies employing Western blot analyses revealed the presence of increased levels of the growth factor in the uterus as well (data not shown). Thus, we extended our studies of organ weight to include the weight of the uterus. As shown in Table 1, while the increased relative weight of the ovary in transgenic females was confirmed, no difference in the weight of the uterus could be detected. Gross macroscopical evaluation and histological examination of tissue sections failed to reveal any obvious structural alteration in the reproductive organs of BTC transgenic females (data not shown). Follicles at the different growth stages as well as corpora lutea could be observed in ovaries from animals of both genotypes (data not shown), strengthening the impression of absence of major pathological alterations.
|
|
Spatial Localization of Transgene Expression in the Reproductive Tissues
Since the localization of BTC overexpression at the cellular level is crucial for interpreting the reproductive deficits in transgenic females, we performed immunohistochemical studies. While the uterus of control females was negative (Fig. 1B), BTC-positive cells were readily detected in the uterine stroma of transgenic animals (Fig. 1C). BTC was not detectable in the oviducts of animals from both genotypes (Fig. 1, D and E). While control ovaries were negative for BTC (Fig. 1, F and G), high amounts of BTC were obvious in transgenic ovaries (Fig. 1H), consistent with the results of the Western blot analysis (all females were at the proestrus stage). Although increased BTC levels were visible in the interstitial compartment and in corpora lutea, particularly high levels were detected in mural granulosa cells of antral follicles (Fig. 1I). Interestingly, the intensity of the signal became progressively weaker in granulosa cells localized close to the antrum, and cumulus cells were only sporadically positive.
Vaginal opening, the first observable consequence of the rise in circulating estradiol that accompanies the onset of puberty in rodents, occurred essentially at the same age in transgenic and nontransgenic females (between Days 24 and 26). Similarly, the interval between vaginal opening and the first estrus was undistinguishable between both genotypes (data not shown). The interval between vaginal opening and the onset of estrous cyclicity, which takes several days in normal mice, was also unchanged in the transgenic females as compared to control littermates (Fig. 2A). Figure 2B shows the duration of the first three cycles, demonstrating that, although the first cycle was significantly longer in the transgenic group, the difference disappeared after the second cycle.
|
To further evaluate the reproductive performance of BTC transgenic females, we conducted a continuous mating study using sexually mature (2 mo old) female mice and known fertile nontransgenic male mice. Breeding data derived from mating pairs kept together for a defined period of 3 mo revealed that, except for a single transgenic female from line 4 (which was excluded from the study), every transgenic and control female produced at least one litter. Furthermore, the interval to the first litter was not different between both groups (Table 2). However, a significant decrease in the average number of litters per animal in transgenic females was present in line 4 and reached borderline statistical significance in line 2. More important, the mean litter size was reduced by approximately 50% in transgenic females as compared to control animals (Table 2).
|
Although most of the initial routine mouse breeding for expanding the BTC transgenic lines involved mating of transgenic males to control females, a few matings involving transgenic females from line 2 paired to control males were performed sporadically. Retrospective analysis of these matings revealed that the mean litter size of transgenic females (5.7 pups/litter, n = 13) was significantly reduced as compared to nontransgenic age-matched animals (9.9 pups/litter, n = 11), confirming the results of the controlled fertility study.
Since a role in blastocyst attachment has been proposed for several EGFR ligands, including BTC, we decided to evaluate whether a perturbation of this process was the cause for the reduced litter size of BTC transgenic females. The initiation of the attachment reaction, as detected by the blue dye method, was observed in all of the control mice at 2400 h on Day 4 of pregnancy (Table 3). In contrast, none of the plug-positive transgenic females showed blue bands at this time. After evaluation of blue sites, the uterus was flushed to recover unattached structures. While only a few blastocysts could be recovered from control uteri (reflecting the ongoing attachment process), a large number of embryos was recovered from transgenic females at this stage. Degenerated or unfertilized oocytes were obtained from transgenic females but not from control animals. Interestingly, the mean number of total viable embryos either attached to the endometrium or recovered from the uterus of transgenic females at Day 4 was markedly reduced when compared to the number of embryos present in the uterus of control females at this time point. In the morning (0900 h) of Day 5, all transgenic females exhibited blue bands, indicating a delay but not a block in the implantation process (Table 3). Again, a few (2) nonviable embryos were constantly flushed from the uterus, and the amount of viable embryos either attached to the endometrium or recovered from the uterus was markedly reduced when compared to the number of embryos present in the uterus of control females at Day 4.
|
To clarify whether the delayed blastocyst implantation observed in BTC transgenic females was associated with deferred expression of AREG in the uterus, as described previously in transgenic mice overexpressing TGFA [17], we assessed the expression of Areg and Btc mRNA in the uterus during the early pregnancy (Days 1–7). Btc mRNA was detected in the uterus at every day in transgenic but not in control females (Fig. 3). Interestingly, the typical Areg peak at Day 4 was present in the uterus from both control and transgenic females (Fig. 3), indicating that the dynamics of Areg expression in the pregnant uterus is not perturbed by the excess of BTC.
|
Ovulation and Fertilization Rates
To evaluate whether ovulation or fertilization rates were reduced in transgenic animals and could be responsible for the reduction in litter size, we mated transgenic and nontransgenic females with fertile nontransgenic males. As shown in Figure 4A, the number of oocytes obtained by natural ovulation did not differ between the two genotypes (10.7 ± 3.0 vs. 10.4 ± 2.3 oocytes in transgenic and control females, respectively). To further assess the ovarian function of BTC transgenic females, we induced superovulation by treating animals with exogenous gonadotropins. The results show that hormonal treatment was successful in animals from both genotypes, even if transgenic females showed fewer (P = 0.02) oocytes (24.0 ± 5.3) as compared to control females (30.3 ± 5.9) (Fig. 4A). There were no differences in cumulus thickness or in the nuclear maturation status (evaluated after cumulus removal with hyaluronidase) between oocytes from transgenic or control females after either natural ovulation or superovulation (data not shown). Evaluation of fertilization rate revealed a statistically significant (P = 0.01) reduction in the percentage of fertilized oocytes collected after natural ovulation in transgenic (54.7% ± 8.9%) as compared to control (81.7% ± 5.3%) females (Fig. 4B). Interestingly, this was not the case for the oocytes obtained after treating females with the exogenous gonadotropins eCG and hCG: oocytes obtained from superovulated transgenic females showed approximately the same fertilization rate (92.2% ± 4.5%) as those from control females (95.1% ± 4.8%).
|
In Vitro Maturation and Fertilization
To confirm the in vivo observation of impaired fertilization in a defined in vitro system, in vitro maturation and fertilization of oocytes was performed. Oocytes were isolated from control (n = 11) and BTC transgenic females (n = 15) and matured in vitro. After in vitro maturation, oocytes were classified, under a light microscope, as germinal vesicle (GV), metaphase I (MI), or metaphase II (MII). As shown in Table 4, the percentage of oocytes matured to MII was significantly lower in the transgenic group as compared to control animals. Conversely, a higher proportion of oocytes from transgenic females remained at MI, indicating a block or delay in the maturation process. Cumulus expansion was attained to the same degree in both groups (Table 4).
|
Analysis of the in vitro fertilization of oocytes matured in vitro, as shown by the appearance of two pronuclei after 6 h of fertilization, revealed a reduction in the percentage of fertilized oocytes of the transgenic animals as compared with control animals (Table 5). The cleavage rate, evaluated 24 h after the fertilization, was also reduced in oocytes from transgenic mice (Table 5).
|
Since data from the in vitro maturation experiment provided information concerning only the endpoint (after 16 h) of nuclear maturation, possible effects related to accelerated or retarded maturation (e.g., oocyte aging) could not be excluded. To clarify this point, we evaluated the time course of nuclear maturation in an in vitro experiment. Oocytes isolated from preovulatory follicles of eCG-primed BTC transgenic ovaries underwent in vitro spontaneous maturation, as evaluated by GVBD (Fig. 5A), and achievement of metaphase II (as shown by extrusion of the first polar body; Fig. 5B) at a rate similar to that observed with oocytes from control mice, indicating that the time course of oocyte nuclear maturation is not altered by the increased levels of BTC in granulosa cells.
|
MAPK3/MAPK1 Activity in the Ovary of BTC Transgenic Mice
As a first attempt to characterize the molecular mechanisms underlying the impaired maturation and fertilization of naturally ovulated oocytes from BTC transgenic females, we studied the activation level of MAPK3/MAPK1 (formerly known as Erk1 and Erk2). Immunohistochemistry using a phospho MAPK3/MAPK1-specific antibody revealed that the oocytes and several granulosa cells were positive for phosphorylated MAPK3/MAPK1 in animals from both genotypes (Fig. 6, A–D). To obtain quantitative information, we performed Western blots with ovarian extracts obtained from females in proestrus or after treatment with eCG and hCG. As shown in Figure 6E, the levels of activated MAPK3/MAPK1 during proestrus were significantly increased in the ovaries of transgenic females as compared to control animals. No difference was detected between BTC transgenic and control mice when ovaries were collected after treatment with eCG and hCG.
|
In the present study, transgenic mice were employed to study the effects of increased levels of BTC in female reproduction. The expression pattern of BTC in transgenic tissues, affecting only specific cell populations instead of the whole organ, may appear unexpected considering the promoter employed (the potentially ubiquitous beta actin). However, deviations from the anticipated expression pattern are common in transgenic mouse models, mostly because of positional effects [28]. In spite of higher-than-normal BTC levels, the uterus and ovaries of BTC transgenic females developed normally and did not show gross histological abnormalities when sexual maturity was attained. BTC transgenic females behaved quite similar to control females during the first weeks of puberty (with the exception of a longer first estrous cycle). Importantly, starting with the third cycle, the mean cycle length in both groups was shorter than 6 days in duration, the maximal cycle length considered normal for different mouse strains [24]. Since body weight plays (at least partially) a role in triggering puberty, the longer first estrous cycle may be related to the fact that transgenic females have a lower body weight at the time of puberty as compared to control animals [23]. Controlled fertility studies, however, revealed that BTC transgenic females exhibit a statistically significant decrease of litter size.
BTC Transgenic Females Exhibit Delayed Implantation but Unaltered AREG Expression in the Uterus
Attachment of the blastocyst to the uterine epithelium initiates the implantation process and is accompanied by locally increased vascular permeability, which can be visualized in the mouse uterus as distinct blue bands after an intravenous injection of a macromolecular blue dye solution [25]. Evaluation of the attachment reaction revealed that blastocyst implantation is delayed in BTC transgenic mice. However, since the total number of embryos either attached (visualized as a blue band) or recovered from the uterus of transgenic females at Day 4 or Day 5 was almost identical with the observed litter size, the reduction in the number of viable embryos has already taken place at this stage. Thus, the reason for the reduction in litter size must be found in processes taking place before implantation. The constant presence of unfertilized/degenerated structures in flushes from transgenic uteri (but not from uteri of control mice) indicates that oocyte maturation or fertilization may be affected. Several studies strongly suggest an important role for EGFR and its ligands in blastocyst attachment. TGFA [29, 30], HBEGF [31], AREG [14], EREG, and BTC [16] are present in the mouse uterus at sites and time points relevant to the implantation process. Even more convincingly is the report of delayed blastocyst attachment reaction in transgenic mice with timely inappropriate expression of TGFA in the uterus [17]. In the previously mentioned study, delayed implantation was associated with deferred expression of AREG, which is probably the EGFR ligand most relevant for uterine receptivity [14, 25]. In our animals, BTC overexpression is being directed by a heterologous promoter that does not follow the endogenous expression pattern of the growth factor (as was the case for the TGFA transgenic mice). In BTC transgenic females, however, the typical AREG peak in the uterus at Day 4 of pregnancy was present in the uterus of both control and transgenic females, indicating that there is no interference between BTC overexpression and the characteristic peak of AREG expression in the uterus around implantation. At this point it is not clear whether the delay in implantation is a direct consequence of increased BTC expression in the uterine tissue or an indirect effect. Further comparative studies will be necessary to assess the transcriptional status of other implantation-relevant molecules, such as leukemia inhibitory factor [32] or calcitonin [33]. While our findings stress the importance of EGFR ligands for the implantation process, our results also underline the need of cautious interpretation of implantation defects with regard to their real relevance for overall reproductive performance.
Impaired Fertilization Is Associated with MAPK3/MAPK1 Hyperactivation
Mating and in vitro fertilization experiments revealed that the fertilization rate was severely reduced in transgenic females as compared to control littermates. In order to obtain insight into the molecular mechanisms behind the maturation/fertilization defect, we studied the activity of MAPK3/MAPK1 in the ovaries of BTC transgenic females. MAPK3/MAPK1 are activated by a wide variety of extracellular signals and play a critical role in regulating cell growth and differentiation [34]. The reason why we have chosen to study specifically these proteins is threefold: 1) MAPK3/MAPK1 phosphorylation is induced by LH receptor activation in granulosa cells and mediates oocyte maturation [35, 36]; 2) MAPK3/MAPK1 phosphorylation is a general effect of EGFR activation [37] and has been specifically observed in granulosa cells [38]; and 3) compared to other EGFR ligands, BTC has been shown to be a particularly potent activator of the MAPK3/MAPK1 pathway [39]. Thus, ligand binding to the EGFR and the LHR converge in the activation of MAPK3/MAPK1. Our results show that the phosphorylation level of MAPK3/MAPK1 was significantly increased in the ovaries from BTC transgenic females in proestrus. In contrast, its activation was similar in transgenic and control mice after treatment with exogenous gonadotropins. This remarkable finding correlates well with the rescue of the in vivo fertilization rate after superovulation of transgenic females and demonstrates a close association between the reproductive defect (reduced fertilization rate) and MAPK3/MAPK1 hyperactivation. On treatment with the exogenous gonadotropins eCG and hCG, the negative influence of BTC must be overpowered by circulating hormones or local molecules that are not present (at least not in concentrations high enough) in the in vitro situation. For instance, hCG, administrated to mouse females as part of the superovulation protocol, has been demonstrated to exert a positive influence on oocyte developmental capacity when allowed to act for a prolonged period of time before oocyte retrieval in human patients [40, 41]. Along the same lines, estradiol secretion increases drastically on superovulation as a consequence of a higher number of follicles secreting the hormone, but this will not occur in vitro. Active somatic cell steroidogenesis prior to ovulation and an appropriate steroid milieu are essential prerequisites for successfully generating competent oocytes [42]. More intensive studies will be needed to evaluate whether these or other candidate molecules are involved in the rescue of fertilization rate after superovulation.
Recent studies have implicated selected EGFR ligands (including BTC) as paracrine mediators of luteinizing hormone (LH) actions in the ovulatory follicle [43, 44]. Specifically, LH was shown to stimulate the expression of AREG, EREG, and BTC in rodent preovulatory follicles, and these growth factors triggered meiosis and cumulus expansion (and the expression of genes associated with this process) in cultured follicles in an EGFR-dependent manner [7, 8]. These findings help to explain the puzzling observation that, in spite of complex actions on the oocyte, granulosa, and cumulus cells, LH receptor expression is restricted to mural granulosa cells [45]. Particularly enlightening are recent experiments by Hsieh et al. [12] employing mice deficient for AREG or EREG or carrying a hypomorphic EGFR allele. The authors show that progressive genetic disruption of the EGFR network in the periovulatory follicle prevents LH-mediated oocyte meiotic reentry and ovulation.
It is appealing to postulate that hyperactivation of MAPK in granulosa cells via the EGFR by higher-than-normal BTC levels leads to a dysfunction of granulosa or cumulus cells during the maturation process, compromising the oocyte developmental competence and impairing fertilization rates. The in vitro maturation of transgenic oocytes was only slightly retarded, and, as shown by the analyses of spontaneous maturation kinetics, meiotic competence is also preserved. However, since the ability of the oocyte to undergo meiotic maturation is a poor marker for further developmental capacity [46], maturation defects cannot be completely excluded. Supporting our findings, spontaneous maturation of oocytes isolated from preovulatory follicles of eCG-primed, AREG-deficient mice carrying the hypomorphic Egfr allele wa-2 was shown to occur in vitro at a rate similar to that seen in control oocytes [12]. This indicates that nuclear maturation of the oocytes is a robust process that is not readily impaired by reduction [12] or increase (our findings) in the levels of EGFR ligand activity. In addition, transgenic mice overexpressing TGFA in the ovary (with particularly high levels in granulosa cells) show an inhibitory effect on follicular development and gonadotropin-stimulated steroidogenesis [47]. Although experiments on the nuclear maturation capacity of these oocytes were not reported, precluding a comparison with BTC transgenic mice, both studies indicate that an excess of EGFR ligands in granulosa cells is associated with impaired oocyte developmental capacity.
Although we concentrated our studies on BTC effects in the reproductive organs, additional effects of the growth factor via the central nervous system cannot be fully excluded in our model. Members of the EGFR system, particularly TGFA and the EGFR, are involved in sexual maturation [6, 48], and mice carrying a defective EGFR [49] or expressing a dominant-negative ERBB4 receptor in astrocytes [50] showed impaired adult reproductive function. Importantly, these animals also displayed delayed vaginal opening and puberty. Since the excess of BTC probably causes a continuous activation of the EGFR, one would expect the opposite effect, namely, precocious puberty as observed in transgenic mice overexpressing TGFA [47]. However, our studies revealed that BTC transgenic females exhibit normal sexual maturation, arguing against effects in the hypothalamic-pituitary axis.
In summary, overexpression of BTC in the uterus and ovary of transgenic females resulted in delayed implantation, a significant reduction of litter size, and a reduced response of the ovaries to gonadotropins. The reason for the smaller litter size is impaired fertilization, and this effect could be rescued by treatment with exogenous gonadotropins. In vitro maturation and fertilization rates of transgenic oocytes were also significantly reduced. The fertilization defect was correlated with increased MAPK3/MAPK1 activity, the convergence spot of EGFR and LHR signaling. The discovery that BTC excess negatively affects the fertilization rate reveals a new action of this EGFR ligand in reproduction. Further studies on this network of peptide growth factors and receptors may help to develop new strategies for the treatment of infertile patients, for the improvement of in vitro culture conditions, or the development of new contraception strategies.
ACKNOWLEDGMENTS
We thank Dr. Ingrid Renner-Müller for veterinary care, Dr. Rebecca Kenngott for help with histological analyses, Petra Renner and Tanja Mittmann for excellent animal care, and Steffen Schiller and Joseph Millauer for genotyping the mice.
FOOTNOTES
1Supported in part by the Deutsche Forschungsgemeinschaft (FOR 478, GRK 1029). A.A.G. is a recipient of a fellowship from the CAPES, Brazil. 
Correspondence: 2Marlon R. Schneider, Institute of Molecular Animal Breeding and Biotechnology, Gene Center, Ludwig-Maximilians University, Feodor-Lynen-Str. 25, 81377 Munich, Germany. FAX: 49 89 218076849; e-mail: schnder{at}lmb.uni-muenchen.de
Received: 3 May 2007.
First decision: 31 May 2007.
Accepted: 13 September 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
