| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Obstetrics and Gynaecology, University of Adelaide, Adelaide 5005, Australia
b University of Adelaide Reproductive Medicine Unit, The Queen Elizabeth Hospital, Woodville 5011, Australia
c Fertilitetscentrum AB, Göteborg 402 29, Sweden
d Pest Animal Control Cooperative Research Centre, Canberra 2601, Australia
ABSTRACT
Granulocyte-macrophage colony-stimulating factor (GM-CSF) secretion from epithelial cells lining the female reproductive tract is induced during early pregnancy by ovarian steroid hormones and constituents of seminal plasma. In this study we have investigated the influence of GM-CSF on development of preimplantation mouse embryos. Blastocyst-stage embryos were found to specifically bind 125I-GM-CSF and analysis of GM-CSF mRNA receptor expression by reverse transcriptase-polymerase chain reaction indicated expression of the low-affinity
subunit of the GM-CSF receptor, but not the affinity-converting ß subunit (ßc), or GM-CSF ligand. GM-CSF receptor mRNA was present in the fertilized oocyte and all subsequent stages of development, and in blastocysts it was expressed in both inner cell mass and trophectoderm cells. In vitro culture of eight-cell embryos in recombinant GM-CSF accelerated development of blastocysts to hatching and implantation stages, with a maximum response at a concentration of 2 ng/ml (77 pM). Blastocysts recovered from GM-CSF-null mutant (GM-/-) mice on Day 4 of natural pregnancy or after superovulation showed retarded development, with the total cell number reduced by 14% and 18%, respectively, compared with GM+/+ embryos. Blastocysts generated in vitro from two-cell GM-/- and GM+/+ embryos were larger when recombinant GM-CSF was added to the culture medium (20% and 24% increases in total cell numbers in GM+/+ and GM-/- blastocysts, respectively). Incubation of blastocysts with recombinant GM-CSF elicited a 50% increase in the uptake of the nonmetabolizable glucose analogue, 3-O-methyl glucose. In conclusion, these data indicate that GM-CSF signaling through the low-affinity GM-CSF receptor in blastocysts is associated with increased glucose uptake and enhanced proliferation and/or viability of blastomeres. Together, the findings implicate a physiological role for maternal tract-derived GM-CSF in targeting the preimplantation embryo, and suggest that defective blastocyst development contributes to compromised pregnancy outcome in GM-CSF-null mutant mice.
conceptus, cytokines, developmental biology, growth factors, implantation/early development
INTRODUCTION
Ex vivo embryo culture, even of relatively robust murine embryos, can retard growth [1] and compromise subsequent developmental potential [2, 3]. The survival and rate of development of embryos in vitro is density-dependent [4] and is enhanced in coculture with endometrial cells or cells of other lineages [5]. These effects are mediated largely by diffusable peptide growth factors acting through specific receptors expressed in the embryo to promote cell division and viability. Several different factors that can act to stimulate or constrain embryo development in vitro have now been identified, including cytokines described originally for their actions in the lymphohemopoietic system, such as colony-stimulating factor (CSF)-1 and leukemia inhibitory factor (LIF) [6–8]. With members of the insulin family, specifically transforming growth factor (TGF)- and epidermal growth factor (EGF), accelerated cell division is accompanied by increased metabolic activity and protein synthesis [9]. A physiological role for growth factors identified in in vitro experiments is indicated by their expression in luminal and glandular epithelial cells in the oviduct and uterus during early pregnancy [10, 11].
The cytokine granulocyte-macrophage colony stimulating factor (GM-CSF) is secreted by epithelial cells lining the female reproductive tract in mice [12] and several other species. In mice, a surge in uterine GM-CSF synthesis and release into the luminal compartment occurs during the preimplantation period, when specific factors in seminal plasma, including TGF-ß, interact with luminal epithelial cells after mating [13]. Expression declines at implantation under the inhibitory influence of progesterone [14], but bioactive GM-CSF can be detected in placental and decidual tissues for the duration of pregnancy [12]. In myeloid leukocytes, where GM-CSF is a well-characterized regulator of cell survival, differentiation and functional activation [15], GM-CSF exerts its effects at the target cell surface through binding to heterodimeric receptor complexes comprising an subunit and a ß subunit [16]. The subunit, known as GM-CSF-R, is unique to GM-CSF and binds ligand with low affinity. The ß subunit (ßc; originally designated AIC2B) is shared by the interleukin (IL)-5 and IL-3 receptors [17] and converts the low-affinity interaction to a high-affinity, signal transducing event [16, 18].
Nonlymphohemopoietic cells, including certain tumor cells and placental trophoblast cells, have been reported to express GM-CSF-R and exhibit biological responsiveness to this cytokine, even though ßc expression cannot be detected [19, 20]. An eliciting mechanism involving GM-CSF stimulation of cell metabolism is suggested by experiments in Xenopus oocytes and human melanoma cells showing that ligation of the low-affinity GM-CSF-R can stimulate glucose uptake through a kinase-independent pathway [21, 22].
Fertility is compromised in genetically GM-CSF deficient mice. Despite normal numbers of implantation sites in early pregnancy, fetal loss in late gestation and death of pups during the early postnatal period cause a 25% reduction in litter size by the time of weaning [23]. Surviving pups are smaller, and in males, reduced size persists into adulthood. Changes in placental architecture associated with diminished nutrient transfer function appear to contribute to this growth retardation.
In a preliminary study, we found that hemopoietic cell-conditioned media known to be rich in GM-CSF are effective in promoting the development of murine zygotes to blastocyst stages and beyond, and recombinant GM-CSF appeared to replicate the effect [5]. GM-CSF has since been reported to promote blastocyst development in in vitro-produced bovine embryos [24], and to elicit secretory activity in ovine embryos [25]. Human two- to four-cell embryos cultured in recombinant GM-CSF are more than twice as likely to develop to the blastocyst stage or beyond, develop faster, and have an increased compliment of inner cell mass cells [26].
In the current study, we have employed the murine model to further investigate the embryotrophic actions of GM-CSF. The expression of GM-CSF receptors in preimplantation embryos has been assessed by radio-labeled ligand binding and reverse transcription-polymerase chain reaction (RT-PCR). Using in vitro culture experiments and genetically GM-CSF deficient mice, we have demonstrated a role for GM-CSF in increasing cell numbers in blastocyst-stage embryos, and have identified improved glucose transport as a likely underlying mechanism.
MATERIALS AND METHODS
Materials
Female Balb/c x C57Bl/6 F1 mice (Balb/c F1, 4 wk old) and male CBA x C57Bl/6 F1 mice (CBA F1, 3–6 mo old) were obtained from the Central Animal House at the University of Adelaide. Mice homozygous for a disrupted GM-CSF gene (GM-/-) were generated using gene targeting techniques in 129 embryonic stem cells, and were propagated from founder mice by mating with C57Bl/6 mice, as described in [27]. After rederivation by embryo transfer to achieve specific pathogen-free (SPF) status, GM-/- mice were bred from parents proven by PCR to be homozygous for the GM-CSF-disrupted gene. Control (GM+/+) mice of an equivalent genetic background were derived from the F2 offspring, proven by PCR to be homozygous for the wild-type gene, of a GM-/- female crossed with a 129 wild-type male obtained from the Walter and Elisa Hall Institute for Cancer Research, Melbourne, Australia. All mice were provided with food and water ad libitum, and were housed in 12L:12D in an SPF-barrier facility at the University of Adelaide. Approval for these experiments was granted by the Animal Ethics Committee of the University of Adelaide.
Escherichia coli-derived recombinant murine GM-CSF (rGM-CSF; specific activity = 1 x 105 U/µg), yeast-derived recombinant murine CSF-1 (rCSF-1), and iodinated rGM-CSF (125I-GM-CSF; specific activity 50 000 cpm/ng) prepared using a chloramine-T-catalyzed two-phase method [28], were all provided by N. Nicola (Walter and Eliza Hall Institute of Cancer Research, Melbourne, Australia). Recombinant murine interleukin (rIL)-2, rIL-3, and rIL-10 were purchased from R&D Systems (Minneapolis, MN) and recombinant human insulin-like growth factor (rIGF)-1 was from Gropep (Adelaide, Australia). A goat polyclonal neutralizing antibody to murine recombinant GM-CSF (goat GM-CSF) was kindly provided by J. Scheurs (DNAX, Palo Alto, CA).
Embryo Collection and Culture
For superovulation, 3- to 4-wk-old Balb/c F1, GM+/+ or GM-/- mice were injected intraperitoneally at 1200–1300 h with 5 IU of pregnant mare serum gonadotropin (Folligon; Intervet, Artarmon, NSW, Australia) and 48 h later with 5 IU of human chorionic gonadotropin (hCG; Chorulon, Intervet). Following hCG injection females were caged 1:1 with CBA F1 males of proven fertility (or, for GM+/+ and GM-/- females, with males of the same genotype). Females were examined the following morning for the presence of a vaginal plug, indicating Day 1 of pregnancy. In some experiments, adult GM+/+ or GM-/- females (8–12 wk) were housed 2:1 with adult males of the same genotype and allowed to mate naturally. Superovulated females were killed by cervical dislocation at 27 h post-hCG injection (Day 1 of pregnancy) to collect fertilized oocytes, at 48 h post-hCG (Day 2 of pregnancy) to collect two-cell embryos, at 72 h post-hCG (Day 3 of pregnancy) to collect eight-cell embryos, or at 97 h post-hCG (Day 4 of pregnancy) to collect blastocysts. Naturally mated mice were killed at 1930 h on Day 4 of pregnancy to collect hatching blastocysts.
Media used for embryo culture included human tubal fluid (HTF) [29] and low-glucose Dulbecco modified Eagle medium (DMEM) supplemented with 10% fetal calf serum (FCS) and antibiotics (DMEM-FCS) in the presence or absence of rGM-CSF (0.4–10 ng/ml). In some experiments, embryos were cultured in the presence of goat GM-CSF at a dilution (1:25 000) shown previously to neutralize 2 ng/ml GM-CSF [12]. Both media were adjusted to an osmolarity of 275 to 280 mOsm/kg prior to addition of serum and antibiotics. HTF and DMEM-FCS used for embryo collection and handling contained 20 mM HEPES. Embryos were cultured in 1-ml tissue culture wells (Nunc, Roskilde, Denmark) or in 25-µl droplets under liquid paraffin oil (BDH Laboratory Supplies, Poole, United Kingdom) at 37°C in a humidified atmosphere of 5% CO2, 5% O2 and 90% N2.
The developmental stage of embryos in culture was assessed visually with the aid of an inverted or dissecting microscope at 8-h intervals from 96 h post-hCG. Preimplantation embryos were scored as 2PN (two pronuclei present), two-cell, four-cell, eight-cell, morulla, early or expanded blastocyst, hatching blastocyst, or hatched blastocyst according to conventional criteria [30]. Blastocysts were defined as ‘hatched’ when they were completely free of the zona pellucida. Blastocysts were defined as ‘attached’ when hatched or partially hatched blastocysts were not dislodged by gentle swirling of the culture dish. Values for ‘t50 hatch’ and ‘t50 attach’, defined as the time (hours post-hCG) at which 50% of blastocysts were hatched or attached, respectively, were calculated by linear regression using SigmaPlot software (Jandel Scientific, Software, Corte Madera, CA). Categorical blastocyst development data were compared by chi-square analysis, using CHITEST and CHIDIST procedures in Microsoft (Redmond, WA) Excel 5.0.
Determination of Cell Numbers
The numbers of cells in blastocysts developed either in vivo or in vitro were determined by staining with bisbenzimide. In vivo blastocysts were flushed from the uteri of GM+/+ and GM-/- mice at 1930 h on Day 4 of pregnancy after natural mating, or at 97 h post-hCG after superovulation. In vitro blastocysts were generated from two-cell embryos recovered at 48 h post-hCG from the oviducts of GM+/+ and GM-/- mice and cultured in HTF until 118 h post-hCG. After brief incubation in acid Tyrodes solution containing 8 mg/ml PVP (molecular weight 35 000; Sigma, St. Louis, MO) to remove zonae pellucidae, blastocysts were washed in PBS supplemented with 10 mg/ml BSA (PBS-BSA) and incubated in bisbenzimide (Sigma; 1 µg/ml, 5 min at room temperature). Following a final wash in PBS-BSA, embryos were fixed briefly in 1% paraformaldehyde in PBS, then mounted on a microscope slide under a coverslip and viewed with the aid of an Olympus BH-2 fluorescent microscope fitted with a 400 nm excitation filter. Nuclei were counted with the aid of video image analysis software (Video Pro, Faulding, Australia). Cell number data were compared using one-way ANOVA and Bonferroni t-tests (SPSS software, Chicago, IL).
125I-GM-CSF Binding
Blastocysts flushed from the uteri of superovulated Balbc F1 mice at 94 h post-hCG were stripped of their zona pellucidae in 0.5% pronase in PBS, then cultured for 8 h in DMEM-FCS in 1-ml wells (Nunc). Blastocysts were incubated with 50 ng/ml 125I-GM-CSF for 60 min at 20°C in 25-µl drops of DMEM-FCS. After rapid washing in 5 x 1 ml DMEM-FCS, blastocysts were fixed in 2.5% gluteraldehyde in PBS for 10 min, and dried onto gelatin-coated microscope slides. Slides were dipped in NTB-2 (Eastman Kodak Co., Rochester, NY) nuclear emulsion at 42°C to 45°C under a Kodak #2 dark red safelight. Following incubation in light-tight boxes at 4°C for 14 days, slides were developed for 4 min in D19 (Kodak) and fixed in Hypam Rapid-fix (1:4 in H2O, Ilford, Mobberly, United Kingdom), then counterstained in Giemsa (Gurr, BDH), and photographed under brightfield and darkfield illumination.
Controls included blastocysts incubated in the absence of 125I-GM-CSF, and blastocysts incubated with a 10x excess of cold rGM-CSF plus 125I-GM-CSF. As a further control for the specificity of binding, GM-CSF-responsive FD 5/12 myeloid cells or nonresponsive gibbon T-lymphocytes (MLA 144 cells) were incubated with 125I-GM-CSF (50 ng/ml, with 5 x 107 cells/ml in PBS-FCS, 60 min at 20°C) as described in [31].
Isolation of Inner Cell Mass Cells
Inner cell mass (ICM) cells were prepared for cDNA extraction by removal of trophectoderm (TE) cells from blastocysts flushed from the uteri of superovulated Balb/c F1 mice at 97 h post-hCG essentially as previously described [32]. After zona removal and lysis, blastocysts were incubated sequentially in rabbit anti-mouse serum (Sigma) diluted 1:5 in PBS-BSA and guinea pig complement serum (Sigma) diluted 1:5 in PBS-BSA (both for 30 min at room temperature). Lysed TE cells were then sloughed away by pippetting through successively smaller-bore glass pipettes. Complete removal of TE-associated mRNA was achieved by brief digestion of extracellular RNA with RNAse A and RNAse T1 (Ambion Inc., Austin, TX; 0.1 U/ml and 20 Uml, respectively, in PBS-BSA for 10 min at room temperature). Inner cell mass cells were washed in PBS containing 8 mg/ml PVP (PBS-PVP), snap-frozen in liquid N2, and stored at -80°C prior to RNA extraction and cDNA preparation as described below. To confirm the purity of ICM cDNA preparations, RT-PCR was carried out with primers specific for urokinase plasminogen activator (uPA) mRNA [33].
Reverse Transcription-Polymerase Chain Reaction
Total RNA was isolated from oocytes; ICM cells; and one-cell, two-cell, eight-cell, and blastocyst-stage embryos using an adaptation of the phenol-chloroform extraction method described previously [34]. Pools of approximately 100 denuded oocytes or embryos, or ICM cells from approximately 200 blastocysts were collected in PBS-BSA, washed in PBS-PVP, snap-frozen in liquid N2, and stored at -80°C prior to RNA extraction. One hundred microliters each of extraction buffer (0.2 M NaCl, 25 mM Tris-HCl pH 7.4, 1 mM EDTA) containing 10 µg of tRNA (Sigma), phenol (equilibrated in TE: pH 7.4 Tris-HCl, 1 mM EDTA) and Sevag's solution (24:1 chloroform:isoamyl alcohol) were added sequentially to the frozen pellet at 4°C and vortexed three times in 10-sec bursts. After centrifugation (10 min/13 000 x g at room temperature) the aqueous phase was re-extracted with 200 µl of Sevag's solution, then precipitated in 75% ethanol overnight at -20°C. First-strand cDNA synthesis was achieved by RT of RNA primed with random hexamers (Bresatec, Adelaide) using a Superscript RNase H-reverse transcriptase kit (Gibco) essentially according to the manufacturer's instructions, as detailed previously [14].
PCR was performed using reagents supplied in a TaqI DNA polymerase kit (Biotech International, Bentley, Australia) in 25 µl volumes as detailed previously [14]. The sequences of primers (Table 1) specific for ß-actin, GM-CSF, GM-CSF-R, ßc (internal), and uPA cDNAs were designed with the aid of Primer Designer software (Scientific and Educational Software). The external primer pair specific for ßc cDNA do not amplify the closely related IL-3 receptor ß subunit (AIC2A) cDNA [35]. For ß-actin, amplification was for 30 cycles using an annealing temperature of 60°C. For GM-CSF, GM-CSF-R, and uPA, PCR was for 38 cycles using annealing temperatures of 60°C, 60°C, and 62°C, respectively. For ßc nested PCR, cDNA was amplified with external primers for 30 cycles at an annealing temperature of 65°C, diluted 1:10 in TE, and reamplified with ßc (internal) primers for 25 cycles at an annealing temperature of 65°C. PCR products were analyzed by agarose gel electrophoresis in parallel with HpaII digested pUC19 DNA as molecular weight markers.
|
Glucose Transport Assay
Glucose transport assays were performed in blastocysts using the nonmetabolizable glucose analogue, 3-O-methyl-D-glucose (3-OMG) essentially as described previously [36]. Initially, the rate of uptake of 25 mM 3-OMG over time was examined. In each of three separate experiments, 3-OMG uptake was linear over 1- to 5-min incubation periods at 37°C and decelerated thereafter (data not shown), so a standard incubation time of 4 min was used to measure the initial rate of uptake in subsequent experiments. To examine the effects of cytokines on glucose transport, groups of 8–10 blastocysts collected from superovulated mice at 97 h post-hCG injection were incubated (1 h, 37°C) in droplets of HTF alone or HTF containing rGM-CSF (0.08–10 ng/ml), rCSF-1, rIL-2, rIL-3, rIL-10, or rIGF-1 (all 2 ng/ml). Blastocysts were then transferred into HTF containing 25 mM 3-OMG (Sigma) and 0.3 mM 3-O-methyl-D-[1-3H] glucose (3H-3-OMG; 1 mCi/ml, Amersham International, Buckinghamshire, United Kingdom) and incubated for 4 min at 37°C. After transfer in a minimum volume through four washes in ice cold glucose-free HTF, blastocysts were placed individually into vials containing 1 ml scintillation fluid (Opti Phase Hi Safe'3, Wallace, United Kingdom) and counted for 10 min in a Beckman LS 6000 ß-emission counter. Data were compared by one-way ANOVA followed by Bonferroni t-test using SPSS software.
RESULTS
GM-CSF Receptor Protein and mRNA Expression in Embryos
To investigate GM-CSF receptor expression in embryos, blastocysts harvested from superovulated Balb/c F1 mice at 92 h post-hCG were stripped of zona pellucidae in 0.5% pronase, cultured for 8 h, and incubated in 125I-GM-CSF. Labeled blastocysts were fixed onto glass microscope slides and exposed for 14 days to autoradiographic emulsion. Silver grains were found to localize over blastocysts (>50 grains per blastocyst; Fig. 1). GM-CSF binding was evident in the plasma membrane of TE cells, without evidence of preferential distribution of silver grains over the inner cell mass or either embryonic pole. The specificity of 125I-GM-CSF binding was demonstrated by incubating additional blastocysts with 125I-GM-CSF in the presence of a 10-fold higher concentration of cold ligand.
|
To determine whether preimplantation-stage embryos express mRNA for either component of the GM-CSF receptor, mRNA was harvested from fertilized oocytes, two-cell, eight-cell, and blastocyst-stage embryos, reverse transcribed using random hexamers, and analyzed by RT-PCR with primers specific for GM-CSF-R, ßc, and ß-actin cDNAs. One-cell, two-cell, eight-cell, and blastocyst-stage embryos all expressed GM-CSF-R mRNA (Fig. 2A). In contrast, mRNA encoding ßc was not present in embryos at any developmental stage. Reaction product was not detected even when highly sensitive, nested PCR was employed (Fig. 2A), or when 32P-dATP was included in the reaction mix, and gels were autoradiographed after capillary transfer to nylon filters (data not shown). Identity between the GM-CSF-R and ßc fragments amplified from FD5/12 cell cDNA and their cognate templates was confirmed by sequencing, and restriction enzyme digestion with SmaI and Sau3A, respectively gave fragments of the predicted sizes (data not shown).
|
To determine whether preimplantation-stage embryos express mRNA for GM-CSF ligand, cDNA preparations from one-cell, two-cell, eight-cell, and blastocyst-stage embryos were analyzed by RT-PCR with primers specific for GM-CSF cDNA. GM-CSF mRNA expression was not evident in embryos at any developmental stage (Fig. 2B). Sequencing analysis confirmed the identity of the GM-CSF cDNA fragment amplified from mouse uterus. Identical results for GM-CSF receptor and ligand mRNA expression were obtained from two sets of preparations from cleavage-stage embryos and from four blastocyst preparations (data not shown).
To determine whether GM-CSF mRNA is expressed in ICM cells, cDNA was prepared from ICM cells after antibody and complement-mediated lysis of TE cells. ICM preparations were judged sufficiently free of contaminating TE-derived mRNA because transcripts for the TE-specific gene product, uPA [34], were not detectable (Fig. 3). Messenger RNA encoding GM-CSF-R, but not ßc, was clearly expressed in each of two different ICM preparations.
|
These data together indicate that murine embryos express the low-affinity GM-CSF-R throughout preimplantation development. In blastocysts, expression is evident in both trophectoderm and inner cell mass cells. In contrast, the high-affinity conferring ß subunit, ßc, appears not to be expressed in preimplantation embryos at any developmental stage.
Effect of Exogenous rGM-CSF on In Vitro Development of Eight-Cell Embryos to Blastocyst and Postblastocyst Stages
To examine the effects of GM-CSF on murine embryo development in vitro, eight-cell embryos were recovered from superovulated Balb/c F1 mice at 72 h post-hCG and cultured in DMEM-FCS in the presence or absence of 2 ng/ml rGM-CSF. Embryo development was scored at 8-h intervals until 148 h post-hCG. The proportion of eight-cell embryos reaching the blastocyst stage of development was not influenced by inclusion of GM-CSF in the culture media. However, blastocysts grown in the presence of GM-CSF were more likely to hatch from the zona pellucida and attach to the culture dish by the end of the culture period (Table 2). A dose-response effect was evident, such that the proportion of eight-cell embryos or blastocysts developing to hatched or attached blastocyst stage was greatest when rGM-CSF was used at a concentration of 2 ng/ml (87.3% vs. 77.2% of blastocysts grown in GM-CSF or control media, respectively; P = 0.001). When embryos were cultured in 2 ng/ml rGM-CSF in the presence of goat GM-CSF, the proportion of eight-cell embryos developing to attached blastocyst stage was not different to that seen in media alone (Table 2).
|
GM-CSF also elicited small but consistent increases in the rate of development of blastocysts to hatching and attached stages. In each of five experimental replicates, development was always faster in the presence of 2 ng/ml GM-CSF. The median (range) reduction in t50 attach and t50 hatch values exerted by GM-CSF was 2.3 (1.3–3.6) h and 1.2 (0.1–2.2) h, respectively (n = 429 and 347 embryos in control and GM-CSF-treated groups, respectively). The effect of dose of GM-CSF on t50 attach values was examined in three of the five experiments. Maximal effects were evident when GM-CSF was added to culture medium at a concentration of 2 ng/ml (Table 3).
|
Effect of Genetic GM-CSF Deficiency on Blastocyst Development In Vivo
To investigate the effect of maternal tract GM-CSF deficiency on embryonic development to blastocyst stage in vivo, superovulated or naturally ovulating GM-/- and GM+/+ mice were mated with males of the same genotype, and killed at 1300 h and 1730 h on Day 4 of pregnancy, respectively. GM-/- females responded to the superovulation treatment, and ovulation and fertilization rates were comparable in both naturally ovulating and superovulated mice irrespective of GM-CSF status. The proportions of embryos developed to the blastocyst and hatched blastocyst stages at the time of recovery were not significantly altered by GM-CSF deficiency (Table 4).
|
The number of blastomeres was determined in blastocysts from GM-/- and GM+/+ mice by bisbenzimide staining. The total cell number was significantly altered in blastocysts from both superovulated and naturally mated GM-/- mice, with the number of blastomeres reduced by 18% (P < 0.02) and 14% (P < 0.01), respectively (Fig. 4).
To investigate whether retarded blastocyst development in GM-/- mice might be due to delayed ovulation, fertilization, or first cleavage, fertilized oocytes were recovered from naturally mated GM-/- and GM+/+ mice at 1500 h on Day 1 of pregnancy and cultured in HTF. The kinetics of development to pronuclear and two-cell stages was investigated by morphological assessment of oocytes during the subsequent 18-h culture period. The genotype of mice had no significant effect on the number of oocytes recovered, or the proportion or rate of their development to first cleavage (Table 5).
|
Effect of Exogenous rGM-CSF on Blastocyst Development and Cell Number In Vitro
To further examine the effect of GM-CSF on blastocyst development and cell numbers in vitro, two-cell embryos were harvested at 48 h post-hCG from superovulated GM+/+ and GM-/- mice, and cultured in HTF with or without 2 ng/ml rGM-CSF. The number of cells in blastocysts were determined by bisbenzimide staining at 118 h post-hCG.
Culture in the presence of rGM-CSF did not significantly alter the rate or extent of in vitro blastocyst development, with 64/89 (71.9%) and 86/116 (74.1%) of GM+/+ two-cell embryos, and 112/131 (82.5%) and 85/105 (81.0%) of GM-/- two-cell embryos reaching blastocyst stage by 118 h post-hCG after culture in the absence and presence of GM-CSF, respectively. Blastocysts from both GM+/+ and GM-/- mice were larger (21% and 26% increases in total cell numbers, respectively, P < 0.02) when cultured with rGM-CSF (Fig. 5). Furthermore, there was a significant effect of genotype on blastomere number at the completion of the culture period, with two-cell embryos from GM+/+ mice generating larger blastocysts than those from GM-/- mice (15% fewer cells in GM-/- blastocysts, P < 0.001; Fig. 5).
|
Effect of Exogenous rGM-CSF on Glucose Transport in Blastocysts In Vitro
In other cell lineages, including Xenopus oocytes and melanoma cells, ligation of the low-affinity GM-CSF-R has been reported to stimulate glucose uptake through a kinase-independent pathway [21]. The possibility that GM-CSF exerts its growth-promoting influence in murine blastocysts through facilitating glucose metabolism was investigated in experiments using the nonmetabolizable glucose analogue, 3-OMG. Uptake was significantly increased in blastocysts following incubation in rGM-CSF (Fig. 6A). The response to GM-CSF was dose-dependent with a maximum effect (50% increase) at 2 ng/ml. To examine the specificity of the effect of GM-CSF, blastocysts were cultured with the same concentration of various other cytokines, including IGF-1, CSF-1, IL-2, IL-10, and IL-3. Of these factors, only IGF-1 was found to exert any effect (Fig. 6B), with a 25% increase in 3H-3-OMG uptake being comparable to the previously reported effect of this cytokine [37].
|
DISCUSSION
These experiments show that murine preimplantation embryos express GM-CSF receptors and that GM-CSF acts to promote the proliferation or viability of cells in blastocyst-stage embryos. Increased cell number is associated with and likely to be the consequence of enhanced metabolic activity because glucose uptake is stimulated in embryos exposed to GM-CSF. Experiments in genetically GM-CSF-deficient mice show that blastomere cell numbers are diminished when GM-CSF is absent from the local milieu in vivo, indicating that the growth-promoting activity of GM-CSF is physiologically important. The absence of GM-CSF expression in the embryo itself indicates that maternal GM-CSF emanating from the uterine and oviductal epithelium must act in a paracrine manner to regulate the developing embryo as it traverses the tract during early pregnancy.
Maximal effects of GM-CSF were attained at a concentration of 2 ng/ml (77 pM), which is in the same order of magnitude as the concentrations eliciting biological effects in other cell lineages in vitro. This concentration is comparable with the GM-CSF content of uterine luminal fluids recovered during the preimplantation period, for which median values range from 15–20 U/uterus (equivalent to 60–80 pM, assuming a uterine luminal fluid volume of 100 µl) [12]. The refractoriness of embryos to higher concentrations of cytokines is reminiscent of that observed in other target cells and may result from receptor down-modulation or other negative feedback mechanisms [38].
Blastocysts cultured from two-cell embryos in the presence of GM-CSF were found to contain approximately 15 cells more than control blastocysts by 118 h post-hCG (21% and 26% increases in cell number in GM+/+ and GM-/- blastocysts, respectively), indicating that GM-CSF can partially ameliorate the delayed development characteristic of embryos grown in vitro [1, 39]. The extent of this improvement in cell number is comparable to that reported for blastocysts exposed in vitro to members of the insulin and IGF family [40, 41], CSF-1 [7], or autocrine growth factors, including TGF- [42]. It is interesting that the current experiments show that comparatively retarded development in the absence of GM-CSF was associated with a delay in hatching and implantation in vitro, but was not reflected in altered hatching kinetics in vivo in GM-CSF-deficient mice. This is consistent with the view that hormonally regulated maternal tract factors are the predominant force in the timing of hatching and implantation in the intact animal [43]. However in vitro, GM-CSF may have a direct or indirect role in initiating expression of the trophectoderm-derived proteases and attachment molecules involved in the processes of hatching and implantation.
Responsiveness to GM-CSF in other cell lineages depends on expression of functional GM-CSF receptors at the plasma membrane. Studies in ßc-null mutant mice show clearly that the ßc subunit is required for high-affinity ligand binding and signal transduction in hemopoietic cells [44, 45]. Whereas blastocysts clearly bind 125I-GM-CSF in a specific manner and RT-PCR analysis confirms that GM-CSF-R mRNA is expressed throughout the first 4 days of embryo development, ßc mRNA was undetectable in embryos, even using a highly sensitive, nested RT-PCR technique. It is therefore concluded that murine embryos express only low-affinity GM-CSF-R from fertilization to blastocyst stage. Moreover, the signaling capabilities of various unusual splice variants of the GM-CSF-R molecule identified in human cells have not been fully explored and the possibility of expression of variant GM-CSF-R forms or alternative, as-yet-unidentified receptor components in the preimplantation embryo, remains to be investigated.
As a target for GM-CSF, the murine embryo therefore resembles other nonhemopoietic cell lineages, including melanoma cells [19] and placental trophoblast cells [20], which exhibit biological responsiveness to GM-CSF despite expressing only low-affinity receptors. In contrast to ßc, GM-CSF-R has only a short cytoplasmic domain and cannot on its own transduce proliferative signal in hemopoietic cells [18]. There is a lack of consensus on the functional capabilities of the GM-CSF-R molecule in isolation. The only signaling effect attributed to GM-CSF-R is the stimulation of glucose metabolism. Initially, this was shown in Xenopus laevis oocytes injected with mRNA encoding human GM-CSF-R [21]. The oocytes expressed low-affinity binding sites for GM-CSF and demonstrated increased uptake of the hexose analogue, 2-deoxyglucose, through a phosphorylation-independent effect on endogenous glucose transporters. Similar effects were seen in HL-60 cells and normal human neutrophils [21] and in melanoma cells, which endogenously express GM-CSF-R in the absence of ßc [22]. However, these observations are difficult to reconcile with experiments in ßc-null mutant mice showing that the effect of GM-CSF on glucose transport in bone marrow cells is dependent on the presence of ßc [46].
The most evident effect of GM-CSF in blastocysts was an increase in the total number of blastomeres. Ongoing differential staining experiments indicate that the inner cell mass compartment is principally affected by GM-CSF deficiency (unpublished results). The mechanisms underlying this presumably involve either accelerated cell division or diminished cell death. A role for GM-CSF in regulating blastomere metabolic activity, specifically in promoting glucose uptake, was identified using the nonmetabolizable glucose analogue, 3-OMG. Glucose is the major energy source in embryos from the time of compaction of morullae, and there is a clear relationships between exposure of the embryo to other growth factors that are known to promote glucose metabolism, particularly insulin and IGF-1, and cell number in blastocysts [41, 47–49]. GLUT1 and GLUT3 are the key facilitative glucose transporters acting in blastocysts, with GLUT1 being present in the basolateral plasma membrane of TE cells and in the ICM, and GLUT3 located exclusively in the apical surface of TE cells [37]. Insulin and IGF-1 are believed to act through promoting translocation of GLUT1 or the recently described GLUT8 to the plasma membrane [37, 50]. Further experiments are required to characterize the effect of GM-CSF on specific glucose transporter molecules in blastocysts, and while an effect on GLUT1 is possible, it is relevant to note that GM-CSF-regulated glucose uptake in murine macrophages appears to be elicited through GLUT3 [51]. The current experiments show that exposure to GM-CSF from the eight-cell stage is sufficient to influence blastocyst outcomes, but the precise time of maximum responsiveness to GM-CSF in embryos was not defined. An effect of this cytokine on glucose transporters prior to compaction may also be important because the metabolic switch to glucose is preceded by GLUT1-mediated signaling of environmental glucose availability [37].
Growth factors are also key determinants of the extent of blastomere cell death through apoptosis. In the mouse, a wave of apoptosis occurs in the 60- to 110-cell stage of blastocyst formation and is most predominant in the inner cell mass [52]. Death in inner cell mass cells is a major contributing factor in the retardation of blastocyst development seen during culture in vitro, particularly when the culture environments provides less than optimal concentrations of metabolic substrates [42]. Insulin, IGF-1, and TGF- appear to preferentially maintain the inner cell mass and at least in the case of TGF-, this effect is associated with diminished apoptosis [40–42]. Thus, the effect of GM-CSF on blastomere number may also be associated with a reduced incidence of apoptosis, either as a consequence of altered metabolic substrate availability or resulting from other, as-yet-undefined actions of the cytokine.
Although the current experimental results are consistent with a primary function for GM-CSF in the embryo from the eight-cell stage onward, it could also be argued that disruptions in ovarian physiology contribute to altered blastocyst development in GM-/- mice. GM-CSF is expressed in the ovary and experiments in GM-/- mice implicate this factor in folliculogenesis [53] and in the development and steroidogenic function of the corpus luteum during early pregnancy [54]. Our finding of reduced cell number in blastocysts grown in vitro from GM-/- two-cell embryos, irrespective of GM-CSF supplementation of the media, is consistent with the view that oocytes derived from a GM-CSF-deficient follicular environment have diminished developmental competence. Embryo transfer experiments are required to fully determine the physiological significance of any ovarian influence.
Embryotrophic effects of GM-CSF have been described in other species, including human [26] and bovine [24]. In both species, GM-CSF elicited approximately a twofold increase in the proportion of embryos reaching the blastocyst stage. In human embryos, this was accompanied by a faster rate of development, an increase in the number of blastomeres, particularly in the inner cell mass, and more frequent progression to hatching and implantation. In ovine blastocysts, GM-CSF did not alter the rate of development but did elicit increased synthesis of interferon-
[25]. In vitro embryo development is characteristically more difficult to achieve in livestock and human species than in the mouse, so it is not surprising that the effects of this cytokine are more dramatic when constitutive development is lower. However, it will be of interest to ascertain whether the metabolism-regulating effects of GM-CSF are conserved across species.
Like other growth factors implicated in the regulation of murine embryo development, GM-CSF clearly has a facilitating rather than an essential role. However, the physiological importance of growth factors in ensuring optimal development of the preimplantation embryo should not be underestimated because even minor disruptions to development at this early time can critically influence growth parameters in the resulting fetus, as well as its long-term health and viability [55]. Blastocyst cell number at implantation, particularly the size of the inner cell mass, has been identified as a pivotal determinant of subsequent placental development and fetal growth trajectory. Fetal growth retardation is clearly associated with in vitro embryo culture [2], and this effect can be alleviated by embryo culture with growth factors that promote blastomere viability [9], or amplified when blastomere viability is constrained [56]. Endogenous perturbations in growth factor environment can elicit similar outcomes [48, 49]. In view of these findings, it is reasonable to propose that retarded blastocyst development may contribute to the altered placental structure, diminished fetal growth, and increased fetal death seen late in gestation in GM-CSF-deficient mice [23]. However, unraveling the precise mechanistic basis of this compromised fertility is complicated by the additional actions of GM-CSF in the maternal immune compartment [31] and in regulating placental trophoblast cell differentiation and function [20].
|
ACKNOWLEDGMENTS
The authors thank Drs. Marie Pantaleon and Peter Kaye for assistance with glucose transport assays and Ms. Vicki Mau for technical assistance.
FOOTNOTES
First decision: 3 August 2000.
1 Supported by NH&MRC (Australia) project grants to S.A.R. and R.F.S. and the NH&MRC Fellowship Scheme to S.A.R. 
2 Correspondence. FAX: 618 8303 4099; sarah.robertson{at}adelaide.edu.au
Accepted: November 28, 2000.
Received: June 28, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  A. N. Sferruzzi-Perri, A. M. Macpherson, C. T. Roberts, and S. A. Robertson Csf2 Null Mutation Alters Placental Gene Expression and Trophoblast Glycogen Cell and Giant Cell Abundance in Mice Biol Reprod, July 1, 2009; 81(1): 207 - 221. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. V. Ingman and R. L. Jones Cytokine knockouts in reproduction: the use of gene ablation to dissect roles of cytokines in reproductive biology Hum. Reprod. Update, March 1, 2008; 14(2): 179 - 192. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Robertson Seminal fluid signaling in the female reproductive tract: Lessons from rodents and pigs J Anim Sci, March 1, 2007; 85(13_suppl): E36 - E44. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Nakasato, Y. Shirakura, M. Ooga, M. Iwatsuki, M. Ito, S.-i. Kageyama, S. Sakai, M. Nagata, and F. Aoki Involvement of the STAT5 Signaling Pathway in the Regulation of Mouse Preimplantation Development Biol Reprod, October 1, 2006; 75(4): 508 - 517. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kawamura, J. Fukuda, J. Kumagai, Y. Shimizu, H. Kodama, A. Nakamura, and T. Tanaka Gonadotropin-Releasing Hormone I Analog Acts as an Antiapoptotic Factor in Mouse Blastocysts Endocrinology, September 1, 2005; 146(9): 4105 - 4116. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Sjoblom, C. T. Roberts, M. Wikland, and S. A. Robertson Granulocyte-Macrophage Colony-Stimulating Factor Alleviates Adverse Consequences of Embryo Culture on Fetal Growth Trajectory and Placental Morphogenesis Endocrinology, May 1, 2005; 146(5): 2142 - 2153. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S O'Leary, M J Jasper, G M Warnes, D T Armstrong, and S A Robertson Seminal plasma regulates endometrial cytokine expression, leukocyte recruitment and embryo development in the pig Reproduction, August 1, 2004; 128(2): 237 - 247. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Emond, L. A. MacLaren, S. Kimmins, J. A. Arosh, M. A. Fortier, and R. D. Lambert Expression of Cyclooxygenase-2 and Granulocyte-Macrophage Colony-Stimulating Factor in the Endometrial Epithelium of the Cow Is Up-Regulated During Early Pregnancy and in Response to Intrauterine Infusions of Interferon-{tau} Biol Reprod, January 1, 2004; 70(1): 54 - 64. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Gutsche, M. von Wolff, T. Strowitzki, and C.J. Thaler Seminal plasma induces mRNA expression of IL-1{beta}, IL-6 and LIF in endometrial epithelial cells in vitro Mol. Hum. Reprod., December 1, 2003; 9(12): 785 - 791. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. T. Jubinsky, M. K. Short, G. Mutema, and D. P. Witte Developmental Expression of Magmas in Murine Tissues and Its Co-expression with the GM-CSF Receptor J. Histochem. Cytochem., May 1, 2003; 51(5): 585 - 596. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Dhar-Mascareno, J. Chen, R. H. Zhang, J. M. Carcamo, and D. W. Golde Granulocyte-Macrophage Colony-stimulating Factor Signals for Increased Glucose Transport via Phosphatidylinositol 3-Kinase- and Hydrogen Peroxide-dependent Mechanisms J. Biol. Chem., March 21, 2003; 278(13): 11107 - 11114. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Sjoblom, M. Wikland, and S. A. Robertson Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) Acts Independently of the Beta Common Subunit of the GM-CSF Receptor to Prevent Inner Cell Mass Apoptosis in Human Embryos Biol Reprod, December 1, 2002; 67(6): 1817 - 1823. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Carcamo, O. Borquez-Ojeda, and D. W. Golde Vitamin C inhibits granulocyte macrophage-colony-stimulating factor-induced signaling pathways Blood, May 1, 2002; 99(9): 3205 - 3212. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. G. Chung, M. R.W. Mann, M. S. Bartolomei, and K. E. Latham Nuclear-Cytoplasmic "Tug of War" During Cloning: Effects of Somatic Cell Nuclei on Culture Medium Preferences of Preimplantation Cloned Mouse Embryos Biol Reprod, April 1, 2002; 66(4): 1178 - 1184. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
