|Home | About | Journals | Submit | Contact Us | Français|
Growth hormone (GH) in rhesus macaque in vitro oocyte maturation (IVM) has been shown to increase cumulus expansion and development of embryos to the 9–16 cell stage in response to 100 ng/ml recombinant human GH (r-hGH) supplementation during IVM. Although developmental endpoints for metaphase II (MII) oocytes and embryos are limited in the macaque, gene expression analysis can provide a mechanism to explore GH action on IVM. In addition, gene expression analysis may allow molecular events associated with improved cytoplasmic maturation to be detected. In this study, gene expression of specific mRNAs in MII oocytes and cumulus cells that have or have not been exposed to r-hGH during IVM was compared. In addition, mRNA expression was compared between in vitro and in vivo-matured MII oocytes and germinal vesicle (GV)- stage oocytes. Only two of 17 genes, insulin-like growth factor 2 (IGF2) and steroidogenic acute regulator (STAR), showed increased mRNA expression in MII oocytes from the 100 ng/ml r-hGH treatment group compared with other IVM treatment groups, implicating insulin-like growth factor (IGF) and steroidogenesis pathways in the oocyte response to GH. The importance of IGF2 is notable, as expression of IGF1 was not detected in macaque GV-stage or MII oocytes or cumulus cells.
In order for an oocyte to be developmentally competent it must complete nuclear and cytoplasmic maturation. This includes possessing the molecular components needed to complete meiosis upon fertilization, enter the mitotic cycle and initiate transcription of the embryonic genome during development in addition to supporting homeostatic and metabolic processes (Zheng et al. 2005a). High levels of mRNA, proteins, substrates and nutrients are accumulated in the oocyte during maturation and are associated with oocyte developmental competence (Song and Wessel 2005; Watson 2007). In vitro maturation can alter the abundance of certain MII oocyte mRNAs compared with MII oocytes matured in vivo (Lee et al. 2008; Zheng et al. 2005b). While in vitro-matured MII oocytes are capable of successful fertilization and oocyte activation, their capacity to support early development is decreased, an effect usually attributed to insufficient cytoplasmic maturation (Krisher 2004; Sirard et al. 2006; Watson 2007). The molecular events associated with successful cytoplasmic maturation and high developmental competence are largely unknown and warrant investigation.
Current culture conditions for non-human primate IVM are inadequate, and further research is needed in order to understand the environment surrounding the oocyte and to reproduce and optimize those conditions in culture, as an aid to improving human IVM. There is evidence for a role for GH in oocyte maturation in a variety of mammalian species. Growth hormone has stimulatory effects on oocyte maturation in cattle, rats, horses, pigs and rabbits (Apa et al. 1994; Izadyar et al. 1997, Marchal et al. 2003; Yoshimura et al. 1993). The GH receptor (GHR) is present in rhesus macaque GV-stage and MII oocytes and on the surface of cumulus cells from GV-stage cumulus oocyte complexes (COCs) (Nyholt de Prada and VandeVoort, 2008). In addition, GH increases cumulus expansion during macaque IVM, as well as the percentage of macaque embryos developing to the 9–16 cell stage (Nyholt de Prada and VandeVoort, 2008). The presence of GHR and increased cumulus expansion in rhesus macaque COCs support the hypothesis that GH is involved in oocyte maturation. These studies relied on the developmental capability of embryos after in vitro fertilization (IVF) to assess oocyte competence, though this endpoint may not be sensitive enough to detect subtle changes or improvements that occur in relation to cytoplasmic maturation. Examining gene expression patterns provides a more complete evaluation of the effects of altered IVM conditions, such as incorporation of GH.
The expression of specific mRNAs in MII oocytes and cumulus cells that have or have not been exposed to r-hGH during IVM was analyzed in this study. Immature, FSH-primed COCs cultured with 0, 10 or 100 ng/ml r-hGH were analyzed after in vitro maturation (IVM). In addition, mRNA expression profiles were compared with MII oocytes from FSH- and human chorionic gonadotropin- (hCG) primed females (in vivo matured) and GV-stage oocytes from FSH-primed females (FSH-primed). We report here effects of GH during IVM on the expression of IGF2, STAR, and other genes related to cell survival and transcriptional regulation.
We hypothesized that r-hGH supplementation during IVM would improve cytoplasmic maturation by altering gene expression in selected genes. The genes selected for this study have key roles in cell cycle progression, DNA repair, cell survival, the IGF pathway, transcriptional regulation, steroidogenesis, intercellular communication, metabolic pathways or are up-regulated following GH exposure in other cell types (Table 1).
Insulin-like growth factor 2 (IGF2) mRNA levels were increased up to 3-fold in MII oocytes from the 100 ng/ml r-hGH IVM treatment group compared with the control and 10 ng/ml r-hGH IVM treatment groups (Figure 1). Expression of insulin-like growth factor 1 (IGF1) was not detected in any of the oocyte or cumulus cell samples. The insulin-like growth factor 1 receptor (IGF1R) mRNA and insulin-like growth factor 2 receptor (IGF2R) mRNA expression increased by up to 5-fold in all IVM MII oocyte groups compared with GV-stage oocytes (Figure 1).
Gap junction protein, alpha 1 (GJA1) mRNA expression appeared to increase with maturation, but was only significantly higher in IVM MII oocytes from the 100 ng/ml r-hGH treatment group compared with GV-stage oocytes (Figure 2). Steroidogenic acute regulator (STAR) mRNA expression was greater than 4-fold elevated in IVM MII oocytes from the 100 ng/ml r-hGH treatment group compared with all other MII oocyte groups (Figure 2). The GJA1 and STAR mRNAs were variably elevated in cumulus cells with 10 ng/ml r-hGH treatment.
Expression of stem-loop binding protein (SLBP) mRNA was not statistically different between any oocyte or cumulus cell groups (Figure 3). Endonuclease G (ENDOG), heat shock 27kDa protein 1 (HSPB1) and YY1 mRNA expression in IVM MII oocytes was not significantly different amongst the r-hGH treatment groups and the control group (Figure 3). ENDOG and YY1 mRNA was variably elevated in cumulus cells with 10 ng/ml r-hGH treatment, though this did not reach statistical significance. Heat shock 70kDa protein 4 (HSPA4) mRNA expression was > 1.5-fold higher in MII oocytes from in vivo-matured compared with IVM MII oocytes in the control and 10 ng/ml r-hGH treatment group (Figure 3). The expression level of HSPA4 mRNA in the 100 ng/ml r-hGH treatment group was between the IVM control and in vivo-matured MII oocytes levels and was not statistically significant from either. The GV-stage oocytes had HSPA4 expression levels of >12–fold higher than all MII oocytes, regardless of treatment group.
Metaphase II oocytes in the 100 ng/ml r-hGH IVM treatment group had more than 1.5-fold lower t-complex 1 (TCP1) mRNA expression compared with the 10 ng/ml r-hGH group (Figure 4). Isocitrate dehydrogenase 1 (IDH1) (Figure 4), ataxia telangiectasia mutated (ATM), breast cancer 1 (BRCA1) and ubiquitin B (UBB) mRNAs were not significantly different amongst MII oocytes or cumulus cells from different r-hGH treatment groups (Figure 4). In general, the pattern of mRNA expression levels of MII oocytes was not paralleled by cumulus cell gene expression.
This study is the first to investigate differences in gene expression between macaque oocytes and cumulus cells following IVM with or without varying levels of GH supplementation. We reported that 100 ng/ml r-hGH added to IVM medium did not affect nuclear maturation, however r-hGH did improve subsequent embryo development in the macaque (Nyholt de Prada and VandeVoort, 2008). Here we show changes in gene expression correlating with this effect and therefore potentially indicating improved cytoplasmic maturation (IGF2, STAR, GJA1, HSPA4, TCP1). Notably, IGF2 expression is implicated in mediating the GH response in macaque IVM. Although correlations have been shown between GH supplementation during IVM and IGF gene expression in other species, this is the first report of this association in the macaque (Apa et al. 1994, Barreca 1993).
Out of 17 genes assayed in this study, only two genes, IGF2 and STAR, showed increased expression in MII oocytes from the 100 ng/ml r-hGH treatment group compared with the other IVM treatment groups. Since IGF2 and STAR mRNA levels were only increased after the 100 ng/ml r-hGH IVM treatment, we concluded the lower dose was insufficient to stimulate changes in MII oocyte gene expression. In addition the increased expression of these genes may be associated with improved developmental competence (Nyholt de Prada and VandeVoort, 2008).
According to the somatomedin hypothesis, the effects of GH are mediated by IGF (Le Roith et al. 2001). The expression of IGF1R and IGF2R mRNA is increased in mature compared with immature rhesus macaque oocytes providing support for the hypothesis that IGF1 and/or IGF2 play a role in macaque IVM. The presence of insulin-like growth factor 1 receptor mRNA in human oocytes and granulosa cells is similar to the macaque as demonstrated here (Lighten et al. 1997; Wang et al. 2006; Zhou and Bondy 1993). Further support for the somatomedin hypothesis is increased expression of IGF2 in the high GH treatment group, suggesting this dosage is effective in activating the IGF signaling pathway that may be responsible for mediating the effects of GH on oocyte maturation and developmental competence. The lack of IGF1 expression suggests that the any involvement of the IGF pathway on macaque oocyte maturation after GH stimulation is dependent on IGF2 mediation. Similarly, the importance of IGF2 mRNA in humans has been demonstrated. In the human, IGF2 mRNA has been reported as more abundant than IGF1 and potentially more important in follicular development (el-Roeiy et al. 1993; Zhou and Bondy 1993). Higher levels of IGF2 mRNA in human follicular fluid were correlated with increased follicular diameter and improved embryo development (Thierry van Dessel et al. 1996; Wang et al. 2006) and the effect of GH on human granulosa cell culture has been reported to be mediated in part by IGF2 (Barreca et al. 1993). In an earlier study, macaque IGF1 mRNA was not detected in GV-stage or MII oocytes (Zheng et al. 2007) and the absence of IGF1 in rhesus cumulus cells and GV-stage or MII oocytes in the present study supports this finding and strengthens the role of IGF2 in macaque oocyte maturation. While mediation of GH by IGF2 is suggested in the macaque and human, IGF1 mediates the GH response during IVM in the rat and GH acts via a direct mechanism in the bovine (Apa et al. 1994; Izadyar et al. 1997). The expression of IGF2 mRNA provides a possible avenue for IGF mediated GH effects. Improvement in embryo development in the 100 ng/ml r-hGH treatment group may thus be mediated in part by the increased IGF2 expression observed in the current study however the mechanism of GH action on oocyte maturation warrants further investigation.
In addition to increasing IGF2 levels, the addition of r-hGH to IVM culture increased STAR expression in the 100 ng/ml r-hGH treatment group. Increased STAR expression indicates GH supplementation may aid in steroidogenesis as well. Steroidogenic acute regulator facilitates transport of cholesterol to the inner mitochondrial membrane and is the rate-limiting step in steroidogenesis (Christenson and Strauss 2000). Elevated expression of STAR occurs after the LH surge in human granulosa cells, and higher levels of expression are associated with nuclear status in human cumulus cells (Feuerstein et al. 2007; Rimon et al. 2004; Sasson et al. 2004). Incomplete follicular maturation is observed in STAR-deficient mice (Hasegawa et al. 2000). Growth hormone has been associated with increased STAR mRNA levels in sheep luteal tissues and rat leydig cells indicating the importance of GH in regulating steroidogenesis in the reproductive tissues of both sexes (Juengel et al. 1995; Kanzaki and Morris 1999). Steroidogenic acute regulator mRNA expression was not detected in macaque GV-stage oocytes, suggesting STAR expression is elevated during oocyte maturation in the rhesus macaque to quantifiable levels. Expression of STAR mRNA was higher in MII oocytes in the 100 ng/ml r-hGH treatment group compared with all other groups of MII oocytes. The higher expression of STAR mRNA in the 100 ng/ml r-hGH treatment group indicates that GH may influence oocyte maturation and developmental competence by aiding steroidogenesis. The expression of STAR MII oocytes matured in vivo did not match the 100 ng/ml r-hGH treatment group, indicating the higher levels of STAR mRNA induced by GH does not parallel in vivo conditions. These results are supported by a separate study where STAR mRNA was expressed at higher levels in IVM versus in vivo-matured macaque MII oocytes (Lee et al. 2008). Higher levels of STAR in IVM versus in vivo-matured oocytes may reflect poor oocyte quality but may also represent the need of IVM oocytes to utilize steroidogenesis to compensate for sub-optimal culture conditions. The ability of GH to increase STAR expression may aid in optimizing steroidogenesis in vitro which may be unnecessary or redundant in vivo. The high expression of STAR in MII oocytes can be interpreted differently and further research is necessary to elucidate the downstream pathway and overall effect on MII oocyte quality.
Together the increased expression of IGF2 and STAR mRNA in the 100 ng/ml r-hGH IVM group provide some insight into the role of GH on oocyte maturation and development. While increased expression of IGF2 and STAR in the macaque oocyte was apparent in the presence of GH, other genes exhibited differences between oocyte groups that are not easily interpreted. Expression levels of GJA1, HSPA4 and TCP1 showed at least one difference between a GH treatment group and another oocyte group. While some of these differences may be attributed to the presence of GH during IVM, those distinctions are not adequately supported with the current data.
Germinal vesicle-stage macaque oocytes expressed lower levels of GJA1 mRNA compared with IVM MII oocytes from the 100 ng/ml r-hGH treatment group. It seems likely that the increased expression in the 100 ng/ml r-hGH treatment group was a result of both oocyte maturation and GH supplementation as the other IVM treatment groups were not statistically different. Gap junction protein, alpha 1 provides routes for the diffusion of molecules and functions in intercellular communication (Grazul-Bilska et al. 1997). The resumption of meiosis is generally associated with a reduction in GJA1 protein as seen in porcine cumulus cells, bovine COCs, equine COCs and mouse ovarian follicles (Marchal et al. 2003; Norris et al. 2008; Shimada et al. 2001; Sutovsky et al. 1993). The reduction in gap junctions between somatic cells is sufficient to initiate resumption of meiosis although another meiosis-stimulatory mechanism may act in parallel (Norris et al. 2008). While expression of GJA1 mRNA and protein have been studied in cumulus cells during oocyte maturation in a variety of species we found no quantitative difference between GJA1 mRNA in macaque cumulus cells in the various IVM treatment groups. The expression pattern of GJA1 in oocytes before and after maturation is lacking making comparisons with our data difficult. The addition of 100 ng/ml r-hGH to IVM culture increased GJA1 mRNA expression in macaque MII oocytes compared with GV-stage oocytes potentially indicating the presence of more gap junctions and altered intercellular communication within the COC. The presence of more gap junctions and altered intercellular communication may be a contributing factor to improved cytoplasmic maturation and developmental competence but is speculative at this stage.
Expression of HSPA4 mRNA was higher in macaque in vivo-matured MII oocytes compared with IVM (control and 10 ng/ml r-hGH) MII oocytes, however, it should be noted that all MII oocytes had low levels of HSPA4 mRNA. Heat shock proteins function in regulating protein translocation and folding and protecting the cell from stress-induced damage (Beere and Green 2001; Jurisicova and Acton 2004). Synthesis of HSPA4 during oocyte maturation is important as mature porcine oocytes were unable to induce HSPA4 synthesis (Lanska et al. 2006). As IVM MII oocytes are generally of lower quality than in vivo-matured MII oocytes, the lower HSPA4 expression in IVM MII oocytes suggests an inadequate response to stress or alternatively inadequate storage of HSPA4 mRNA for early embryogenesis and development (Krisher 2004; Rizos et al. 2002; Sirard et al. 2006; van de Leemput et al. 1999). The expression of HSPA4 in the 100 ng/ml r-hGH treatment group was somewhere between the IVM control and in vivo-matured MII oocyte groups, demonstrating GH might be regulating HSPA4 expression to the levels observed in an in vivo-matured MII oocyte.
Expression of TCP1 was higher in the 100 ng/ml r-hGH treatment group compared with the 10 ng/ml treatment group. T-complex 1 is involved in actin and tubulin folding, implicating a cytoskeletal function (Sternlicht et al. 1993; Valpuesta et al. 2002; Yaffe et al. 1992; Yokota et al. 2000). The importance of TCP1 in mitotic spindle assembly has also been noted in yeast, demonstrating the importance of this gene in microtubule-mediated processes (Ursic and Culbertson 1991). Higher expression of TCP1 from the 100 ng/ml r-hGH treatment group suggests that differences in GH dose may result in differences at the cytoskeletal level and ultimately the ability of the maturing oocyte to undergo microtubule mediated processes such as meiosis.
In general, the pattern of mRNA expression levels of MII oocytes was not paralleled by cumulus cell gene expression. Previous results indicated that the addition of 10 ng/ml r-hGH to IVM medium increased cumulus expansion in the macaque COC while 100 ng/ml r-hGH did not (Nyholt de Prada and VandeVoort, 2008). Cumulus expansion cannot be correlated with gene expression in the current study and additional genes should be investigated to elucidate the pathway involved in cumulus expansion.
The mRNA expression analysis presented here explored the potential benefits of supplementing IVM media with r-hGH at the molecular level. The addition of 100 ng/ml r-hGH to IVM medium increased mRNA expression levels of IGF2 and STAR in macaque MII oocytes, supporting the notion that GH mediation by the IGF pathway and involvement of steroidogenesis improves oocyte developmental competence. Furthermore, the importance of IGF2 is implicated, as expression of IGF1 was not detected in macaque oocytes or cumulus cells. Expression levels of GJA1, HSPA4 and TCP1 showed at least one difference between a GH treatment group and another oocyte group. While some of these differences may be attributed to the presence of GH during IVM, additional research is needed to support the role of GH in altering gene expression of GJA1, HSPA4 and TCP1.
Adult female rhesus macaques (Macaca mulatta) were housed and cared for at the California National Primate Research Center as described previously (Nyholt de Prada and VandeVoort, 2008). Females were administered FSH for the collection of immature oocytes. The gonadotropin releasing hormone (GnRH) antagonist, Antide (Ares-Serono, Randolph, MA), was administered subcutaneously (0.5 mg/kg body wt.) once daily in the morning on days when recombinant macaque FSH (r-mFSH) was given to prevent endogenous gonadotropin secretion. Details regarding hormonal stimulation, oocyte aspiration and oocyte collection have been described (Nyholt de Prada and VandeVoort, 2008).
Retrieved immature COCs from six females (105 COCs total) were put randomly into 70 µl drops of M1A medium (Boatman 1987; Nyholt de Prada et al. 2008) with either 0, 10 or 100 ng/ml of r-hGH (R and D Systems, Minneapolis, MN) and incubated in a humidified atmosphere of 5% CO2 in air for 28–30 hours at 37°C (2–14 COCs per group per female). After 28–30 hours of incubation in IVM medium, COCs were placed in 10 mg/ml hyaluronidase (MP Biomedicals, Solon, OH) in Tyrodes lactate (TL)-HEPES medium containing 0.1 mg/ml polyvinyl alcohol (PVA) that was pre-equilibrated to 37°C. Cumulus oocyte complexes were stripped of cumulus cells using a Stripper® (MidAtlantic Diagnositcs Inc., Mount Laurel, NJ). Denuded oocytes were visually assessed, and only oocytes with polar bodies were used for this study. Cumulus cells separated from COCs in each treatment were also used for this study (50–100 cells).
For collection of mature COCs, females were given recombinant hCG in addition to the hormonal treatment outlined for immature COC collection. Hormonal stimulation for the collection of mature oocytes along with oocyte aspiration and oocyte collection has been described (Nyholt de Prada and VandeVoort, 2008; VandeVoort and Tarantal 2001). The COCs were put in pre-equilibrated 70 µl drops of chemically-defined, protein-free hamster embryo culture medium 9 (HECM-9) under oil and incubated at 37°C in a humidified atmosphere of 5% CO2 in air (McKiernan and Bavister 2000). After 6–8 hours of incubation, COCs were placed in TL-PVA pre-equilibrated to 37°C, and a Stripper® was used to ensure all cumulus cells were removed. Denuded oocytes were examined visually, and only oocytes with polar bodies were used.
Quantitative gene expression studies of small numbers of oocytes and embryos have been made possible with a novel reverse transcription-polymerase chain reaction (RT-PCR) approach (Rambhatla et al. 1995). This approach enables comparison of gene expression patterns and relative copy number of many different mRNAs from a common set of samples, which is crucial for non-human primate studies where resources are limited. This methodology allows for quantitative comparison of gene expression in MII oocytes that have been matured in vitro versus in vivo, providing insight into molecular differences among oocyte quality. In addition, gene expression analysis of MII oocytes matured in the presence of specific growth factors allows us to investigate at the molecular level the potential benefits of supplementing IVM media. The Primate Embryo Gene Expression Resource (PREGER; http://www.preger.org) has developed a novel RT-PCR approach to support gene expression analysis of non-human primate oocytes (Zheng et al. 2004a; Zheng et al. 2004b; Zheng et al. 2005b).
Denuded MII oocytes were briefly put in an acid Tyrodes solution (TL-PVA, pH 2.0) to remove the zona pellucida and then rinsed in TL-PVA. Zona-free oocytes were added in a minimal volume to the RT buffer (Rambhatla et al. 1995). Between 50 and 100 cumulus cells were put into a 1.5 ml tube and centrifuged at 6,000 × g for 5 minutes. Supernatant was discarded, and pelleted cells were washed with 100 ul of Dulbecco’s phosphate buffered saline (DPBS) and centrifuged again for 5 minutes. Supernatant was discarded, and the RT buffer was added directly to pelleted cells. A quantitative RT-PCR was performed as described (Rambhatla et al. 1995; Zheng et al. 2004b). This method (Brady et al. 1990; Brady and Iscove 1993) uniformly amplifies the entire mRNA population, preserving relative abundances of cDNA sequences. Quantitative amplification and dot blotting (QADB) was used as described (Rambhatla et al. 1995; Zheng et al. 2004b). The sensitivity and reliability of the QADB method have been discussed extensively elsewhere (Latham 2006; Rambhatla et al. 1995; Zheng et al. 2004b).
The samples used in these experiments were added to the PREGER resource. The COCs were collected from six FSH-primed females and five FSH- and hCG-primed females. The COCs from FSH-primed females were used for IVM. One to two oocytes were collected per sample and one to three samples of oocytes were collected for each treatment per female (0 ng/ml r-hGH, 14 samples; 10 ng/ml r-hGH, 12 samples; 100 ng/ml r-hGH, eight samples). A cumulus cell sample was collected for each treatment group per female. Cumulus oocyte complexes from FSH-and hCG-primed females were collected (10 samples). Germinal vesicle- stage oocytes from two FSH-stimulated females were added to the sample set from the PREGER resource (six samples, one oocyte each).
Plasmids containing cDNA inserts were obtained from Open Biosystems (Integrated Sciences, Australia). Primers and the amplified cDNA regions are listed in Table 2. Probe preparation was performed as described for the majority of genes (Latham 2006; Rambhatla et al. 1995; Zheng et al. 2004b). Complementary DNA probes for IGF1 and IGF2R were obtained by RT-PCR of RNA from rhesus macaque liver and the identities of the amplified cDNAs were confirmed by sequencing. Blot preparation, hybridization and quantitative analyses were performed as described (Latham 2006; Rambhatla et al. 1995; Zheng et al. 2004b). Data are expressed as the mean ± SEM cpm bound value for each treatment group of oocytes or cumulus cells.
Differences in gene expression between treatment groups and cell types were evaluated with two-sample t-tests assuming equal variances. The analysis of oocyte mRNA expression encompassed five different oocyte groups. In the cumulus cell analysis three IVM treatment groups were compared. Assuming the different groups were independent, the expression levels of each gene were tested in each of the four possible comparisons for oocytes or the two possible comparisons for cumulus cells. Non-human primate oocytes are both expensive and difficult to obtain and the limited sample size in this study may limit the differences observed, therefore, all statistical differences that pass the uncorrected marginal significance value of P < 0.05 are reported. In addition, comparisons that achieve statistical significance after applying the Bonferroni correction for the performance of four or two tests are reported; P < 0.0125 or P < 0.025 respectively (Armitage and Berry 1994).
The authors would like to thank Dana Hill, Namdori Mtango and Joe Dutra for technical assistance. In addition the authors would like to thank Dr. Miller, Dr. Buckpitt and Dr. Lyons for general use of their lab equipment.
This research was supported by NIH grants RR13439 (CAV), RR00169 (CNPRC), and RR15253 (KEL)