| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gamete Biology |
AgResearch Crown Research Institute, Ruakura Campus, Hamilton 2001, New Zealand
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
5-mm) and small (
2-mm) follicles to isolate messenger RNA associated with the attainment of developmental competency. RNA was amplified in a linear fashion and then subjected to the SSH procedure to produce a library enriched for genes associated with competency. One thousand clones of this library were subjected to a differential screening approach to identify 31 potentially upregulated isolates. Sequencing revealed these to represent 21 genes. To rigorously identify the degree of upregulation and reproducibility thereof, we examined the expression of these genes in three separate pools of developmentally competent and incompetent oocytes by quantitative real-time PCR. Results indicated that upregulation varied from zero to threefold, showing that accurate quantification is essential for the interpretation of such differential screening experiments. Furthermore, it appears that the molecular causes for poor developmental capacity may be highly complex and be reliant on many small changes. We further characterized a selection of these novel and known maternally expressed genes for their absolute expression levels during maturation in the presence or absence of an inhibitor of transcription and during preattachment development. Last, the effect of nuclear transfer on the levels of these genes was assayed. Nuclear transfer was found to differentially affect transcript levels of genes expressed after embryonic genome activation but did not prevent the degradation of maternal transcripts or result in activation of maternal genes that are silent at blastocyst stages.
bovine, development, early development, embryo, oocyte, oocyte development, subtractive hybridization, transcription
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Transcription drops markedly once the ovarian follicle has grown to 3 mm [9, 14] and the oocyte remains transcriptionally nearly quiescent thereafter [15]. The shutdown of the bulk of transcription at a follicular size corresponding to the observed difference in competency suggests that oocytes from antral follicles smaller than 3 mm have not yet attained the full RNA complement required for developing past the timepoint when embryonic transcription sets in. One study has examined the feasibility of applying PCR-based subtraction techniques to isolate mRNA associated with oocyte maturation and developmental competency based on follicular size [16]. We describe here a novel approach involving linear amplification followed by suppressive subtraction hybridization (SSH) to identify genes with higher levels of transcripts in competent oocytes. We determined the levels of these transcripts in the competent and noncompetent oocyte using real-time reverse transcription-polymerase chain reaction (RT-PCR). Furthermore, the expression of these genes was characterized during meiotic maturation, preattachment development, and after nuclear transfer.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cow ovaries from slaughterhouses were collected and kept at 30°C in 0.9% saline for 2–4 h before aspiration. For low-competency oocytes, cumulus-oocyte complexes (COC) were collected by aspiration from follicles with a diameter 2 mm and within 2–3 h of slaughter, whereas for high-competency oocytes, follicles 5 mm and a 3.5- to 4.5-h interval between slaughter and collection were used, as described by Blondin et al. [17]. The size classification procedure was verified to be accurate to within 0.5 mm by prior trials comparing estimated follicle size and dissected measured follicle size. Only COCs with a homogeneous or slightly granulated ooplasm and more than four layers of cumulus cells showing no expansion were kept [7]. Cumulus cells were rigorously removed by mechanical aspiration and several washing steps in HEPES-buffered TCM199. Oocytes were flash frozen in liquid nitrogen and stored at –80°C until RNA extraction.
In vitro maturation, fertilization, and culture were performed using the synthetic oviduct fluid system [18]. For 
-amanitin-treated matured oocytes, -amanitin was added at a concentration of 20 µg/ml to all media, including the aspiration medium. This treatment resulted in greater than 70% of oocytes not extruding a polar body [19].
Cumulus cells were stripped from matured oocytes destined for freezing as described above. In vitro-produced embryos for collection were examined microscopically for normal morphology and snap frozen in batches according to their developmental stage. Morula refers to the compacted morula stage, blastocyst to prehatching Day 7 expanded blastocysts.
Generation of Nuclear Transfer Blastocysts
Nuclear transfer was performed according to standard procedures using skin fibroblasts derived from two different animals (nos. 1 and 2). In vitro maturation of oocytes derived from 3- to 8-mm follicles was performed as described above. Nuclear transfer was performed as described in detail in [20], with an interval of 3–5 h between activation and fusion. Blastocysts were collected before hatching, at Day 7.
RNA Extraction and Linear Amplification
RNA was extracted from 500 oocytes using Trizol reagent (Invitrogen, Auckland, New Zealand) according to the manufacturer's protocol and resuspended in 10 µl H2O. RNA was treated with 1 U RF-DNAaseI at 37°C for 30 min in the presence of 5 mM DTT and 1 µl RNasin (Promega, Madison, WI) in a 20-µl reaction. RNA was phenol/CHCl3 extracted, ethanol precipitated, and redissolved in 10 µl H2O. Linear amplification was performed as described [21], with several modifications: First-strand cDNA was synthesized by mixing in a thin-walled PCR tube 9 µl total RNA with 1 µl 0.1 µg/µl PAGE-purified SP6T19 primer [5'-CGGCCAGTGAATTGTACATTTAGGTGACACTATAGAAGTAC(T)19VN-3'; italicized is SP6 RNA polymerase promoter] and incubation in a PCR machine with heated lid: 3 min 70°C; ramp over 10 min to 4°C; 5 min 42°C to allow primer annealing. For reverse transcription with template switching, 4 µl first-strand buffer (Invitrogen), 2 µl 0.1 M DTT, 2 µl 10 mM dNTP, 1 µl RNacin, 1 µl TempSw1 primer (0.1 µg; 5'-AAGCAGTGGTAACAACGCAGAGTACGCGGG), 2 µl Superscript II (Invitrogen) were added and the reaction incubated at 42°C for 90 min. For second-strand synthesis, 106 µl water, 15 µl Advantage PCR buffer, 3 µl 10 mM dNTP, 1 µl RnaseH (Promega), and 3 µl Advantage II polymerase (Clontech, Palo Alto, CA) were added and incubated for 20 min 37°C, 3 min 94°C, 3 min 65°C, and 30 min 75°C. The reaction was stopped with 7.5 µl freshly diluted 1 M NaOH, 2 mM EDTA for 10 min at 65°C, phenol/CHCl3, then CHCl3 extracted and washed three times with 500 µl of H2O on a Microcon YM-100 column (Millipore, North Ryde, Australia), leaving an elution volume of 10–16 µl.
The cDNA was amplified-transcribed using the SP6-Megascript kit (no. 1330; Ambion, Austin, TX) in a double (40-µl) reaction and an incubation time of 5 h at 37°C. The RNA was desalted on a Microcon YM-100 column as described above, eluting in a volume no greater than 12 µl. One microliter was electrophoresed on a 1.2% agarose gel with an RNA molecular ladder to estimate quantity and quality.
The antisense SP6-transcribed amplified RNA (aRNA) was converted to DNA as follows: 12 µl of aRNA and 1 µl (2 µg) of random hexamers (Roche Diagnostics, Auckland, New Zealand) were incubated for 10 min at 70°C, rapidly chilled on ice, and equilibrated at room temperature (21°C) for 10 min before addition of 5 µl first-strand buffer (5x; Invitrogen), 2 µl 0.1 M DTT, 2 µl 10 mM dNTP, 1 µl RNasin, and 2 µl SuperscriptII, left for 5 min at room temperature, and incubated for 90 min at 42°C. For second-strand synthesis, 103 µl H2O, 1 µl 0.5 µg/µl SP6T19 primer, 15 µl Advantage PCR buffer, 3 µl 10 mM dNTP, 1 µl RnaseH, and 3 µl Advantage II polymerase were added and incubated at 37°C 20 min, 94°C 3 min, 55°C 3 min, and 75°C 30 min. The reaction was phenol/CHCl3, then CHCl3 extracted, washed on a Microcon YM-100 column as before, and analyzed on an agarose gel.
Suppressive Subtractive Hybridization (SSH) and Differential Screening of Cloned Genes
SSH was performed using the amplified DNA and the PCR Select kit starting at the RsaI step (protocol #PT1117-1, p16; Clontech). Briefly, tester DNA corresponding to DNA amplified from high-competency oocytes and driver DNA of low-competency oocytes was digested with RsaI to obtain shorter, blunt-ended molecules. The tester was then divided into two portions and each was ligated with a different adaptor having a common stretch of sequence corresponding to the T7RNA polymerase promoter site at the 5' end. This common stretch is required for the amplification suppression effect of unwanted hybrids during the later PCR steps. The critical ligation step was optimized as suggested in the protocol but using primers specific for bovine GAPDH: sense, 5'-TATCATCCCTGCTTCTACTG-3'; antisense, 5'-CTGTTGAAGTCGCAGGAGAC; using an annealing temperature of 65°C, giving a product size of 247 basepairs (bp) on cDNA template or 324 bp when using the sense primer in conjunction with the Clontech PCRprimer1 on adaptor-annealed cDNA.
In the first hybridization reaction, following a heat-denaturation step, an excess of driver was annealed to each tester, generating tester-driver hybrids for species common to both populations of DNA. Subsequently, the two primary hybridization products were mixed together without denaturation but with more denatured driver, allowing remaining single-stranded adaptor-ligated testers to form double-stranded hybrids bearing a different adaptor at each end. Upon subsequent PCR amplifications using primers specific for each adaptor, these hybrids are the only ones that could be exponentially amplified under the conditions used. Thirty cycles were run in the first PCR reaction, 12 in the second, nested reaction. Positive controls included in the Clontech kit gave us confidence that the subtraction had worked optimally.
However, as SSH can only result in an enrichment of differentially expressed DNA, we assessed all clones by differential screening. The PCR products were cloned into the pGEM-T Easy vector (Promega) using 3 µl SSH PCR product and 3 U T4DNA ligase in a 10-µl reaction overnight at 4°C. High-competency TOP10 cells (Invitrogen) were transformed as per manufacturer's instructions and plated on Luria broth (LB)/ampicillin/ X-gal agar plates. White colonies were picked and grown with shaking in 100 µl LB/ampicillin for 4 h at 37°C in 96-well plates. Inserts were PCR amplified using 1 µl of the culture and 19 µl of a PCR mastermix containing primers annealing to the adaptors (NP1 and NP2R; Clontech) with the following conditions: 95°C 3 min, then 23 cycles of 95°C 10 sec, 68°C 3 min followed by 10 min at 72°C. Five microliters of PCR product was denatured with 5 µl fresh 0.6 N NaOH and 1 µl spotted on each of four Hybond-N membranes (Amersham Biosciences, Auckland, New Zealand). Membranes were hybridized according to standard protocols to one of four radioactively labeled probes: 1) high-competency oocyte amplified DNA, 2) low-competency oocyte-amplified DNA, 3) PCR product from the primary SSH reaction representing DNA enriched for high-competency material, and 4) PCR representing DNA enriched for low-competency material from a SSH reaction in which the tester and driver were exchanged. Adaptors were removed from the probes by RsaI digestion followed by purification on YM30 columns. Only clones showing significantly higher signals with probes 1 and/or 3 were deemed enriched and used for further analysis. Clones were sequenced and identified by blast searches against public and AgResearch databases.
Real-Time Reverse Transcription Polymerase Quantitation
RNA isolation was performed as described above but after addition of 400 µl Trizol reagent and 0.1 pg rabbit -globin mRNA (R-1253, Sigma, Auckland, New Zealand) per oocyte or embryo was added. RNA was subjected to DNAaseI, purified (described above), and redissolved in 11 µl of water. One microliter was kept as reverse transcription control; the other 10 µl were mixed with 1 µl (500 ng) oligo(dT)12–18 primer (Invitrogen) and 1 µl 10 mM dNTP and incubated at 65°C for 5 min, then rapidly chilled on ice. Four microliters of first-strand buffer, 2 µl 0.1 M DTT, and 1 µl RNacin were added, and after 2 min at 42°C, 1 µl SuperscriptII (Invitrogen) was added and the reaction incubated for a further hour. The reaction was stopped at 70°C for 15 min.
Real-time PCR was performed on an ABI 7700 using the Applied Biosystems SYBR-Green PCR master mix (AP 4309155; Applied Biosystems, Auckland, New Zealand) in 15-µl reactions. Where possible, primers were designed spanning putative introns, as determined by sequence comparisons with genomic sequences of the human or mouse homologous gene, using the GCG program Prime. The final concentration of primers was 100 nM and cDNA from about half an embryo equivalent was added per reaction. Samples were run in duplicate or triplicate. Samples not exposed to reverse transcriptase (RT–) were included as negative controls to monitor DNA contamination. The PCR conditions were 95°C 5 min, then 40 cycles of 95°C 15 sec, 56°C 30 sec, 72°C 30 sec, 78°C 10 sec. Readings were taken at 72°C and 78°C. Subsequent to the PCR, a dissociation curve analysis program was run and representative PCR samples were electrophoresed on agarose-TAE gels to ensure the absence of nonspecific PCR products. Standard curves were established as follows. Gel-purified PCR-derived DNA fragments of all genes analyzed were quantitated fluorometrically with Picogreen (Molecular Probes, Eugene, OR) and copy number per x grams of fragment calculated as
 |
|
Copy numbers of samples were determined from the standard curves, which ranged from 0.5 to 500 fg, each point in duplicate, correlation coefficient range from 0.983 to 0.998. The copy numbers were normalized for RNA recovery and reverse transcription using the Ct values derived for the -globin spike: correction factor r for sample x of i samples containing n(x) embryo equivalents = relative -globin copy number per average -globin copy number measured for all samples on a plate:
where a, the negative slope in a graph plotting Ct against the log of the relative copy number of -globin was consistently measured to be 1.8 (Table 2). The Ct (minimal threshold value) of -globin was always measured for each sample on the same plate as the genes of interest. To minimize artifactual variation, samples to be compared and standard curves were run wherever possible on the same plate using a PCR master mix containing all reaction components apart from the sample.
Statistical Analysis
Analysis of variance of the log ratio of the expression levels (5 mm versus 2 mm) was performed and the significance of each mean analyzed by t-test using pooled standard deviations.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As a bovine oocyte contains only about 2.4 ng total RNA, of which only 1% represents translatable polyA+ mRNA [22, 23], we pooled RNA from 500 oocytes corresponding to 10 ng of polyA+ RNA, which was needed to perform an amplification step, as SSH requires at least 500 ng of starting material. We chose to use linear amplification, as this has been shown to introduce minimal bias [21]. However, we designed a SP6 promoter-containing primer for initial reverse transcription and used SP6 RNA polymerase for transcription (see Methods) instead of the published T7 RNA polymerase-based protocols, as we intended to use this material for subsequent SSH. The SSH method [24] employs a primer containing a T7 RNA polymerase promoter site, which would have interfered with T7 polymerase-amplified DNA. We obtained 1.8 and 2.2 µg of amplified DNA from 500 oocytes of 2- and 5-mm follicles, respectively, representing an amplification factor of around 200-fold.
One thousand clones from the forward subtraction (noncompetent 2 mm subtracted from competent 5-mm population) were picked and PCR amplified (Fig. 1). More than 95% contained an observable insert. The insert size ranged from 100 to 1000 bp, with a mean of 450 ± 25 bp. SSH, similar to other subtraction procedures, will also yield false positives. We therefore subjected our isolated clones to a dot blot differential hybridization screening assay to obtain a rapid semiquantitative index of the relative abundance of these clones. The probes used for screening were derived from the two amplified cDNA populations (5 and 2) and from the forward (5-enrich) as well as a reverse (2-enrich) subtraction. We compared relative signal intensities in the duplicate blots by eye, taking into consideration the overall intensities of neighboring signals. An example of such blots is shown in Figure 2. Two types of clones were selected— those showing greater signals with the 5 and 5-enrich probes, representing abundant genes associated with competent oocytes, and those yielding a signal primarily with the 5-enrich probe, representing clones that are not abundant enough in the original 5 probe to allow detection. Our subjective, but stringent, judgment led to the selection of only 31 of the 1000 clones examined, or 3%. These were sequenced and compared against public and private databases to determine their identity and are listed in Table 1. Five genes were found 2–4 times, suggesting that the complexity of the library was approached [24].
|
|
|
Analysis of Genes Enriched in Competent Oocytes
Considering that the differential screen is based on the premise that no bias has occurred during linear amplification and/or the subtractive hybridization, we subjected the identified clones to a detailed quantitative analysis using fluorescent (SYBR-green)-based real-time RT-PCR. As the abundance of housekeeping genes often used as internal standards can be misleading, samples were also spiked with a fixed amount of rabbit -globin RNA per oocyte (or for subsequent studies, per embryo), allowing correction for RNA recoveries and reverse transcription efficiencies. The slopes of nearly all the standard curves were comparable and within 10% of the ideal amplification efficiency of 2.0 (Table 2). Three batches of oocytes from 2-mm and 5-mm follicles were split in two, with one half of the oocytes collected for real-time PCR analysis, whereas the other half was scored for development to the blastocyst stage (Fig. 3). The developmental competency varied slightly from batch to batch (Fig. 3A); however, the difference in competency between small and large follicles was always at least threefold. Apart from clone 410, all genes isolated showed greater expression in oocytes from 5-mm follicles, indicating that the method used had been successful (Fig. 3B). Interestingly, few of the genes showed more than a twofold (100%) increase in expression in oocytes that were at least threefold enriched for developmental competency (Fig. 3, A and B).
|
We determined the absolute number of transcripts in the oocytes of the genes we had isolated to examine whether we had only selected abundantly expressed genes. These estimates of the absolute number of transcripts represent a minimum value, as we have corrected for intersample differences but have not measured absolute RNA recoveries and reverse transcription efficiencies. The abundance as determined crudely by the differential screen correlated well with the levels seen by real-time PCR. The ranges in abundance of messages differed widely (Fig. 3, C and D), ranging from several hundred copies (clone 1166) to around 73 000 copies per embryo (SLBP), thus demonstrating that rare as well as abundant messages had been enriched for. Notably, Cyclin A and NDFIP1, which are expressed at only around 1500 copies per oocyte, were isolated multiple times in the subtractive screen and indeed showed 100% and 75% increased expression in developmentally enriched oocytes.
We also examined the differential expression and abundance of three genes not isolated in our differential screen. GDF9 and BMP15 (also known as GDF9B) are two highly related oocyte-specific members of the TGFß superfamily of growth factors that have been implicated in folliculogenesis. Whereas GDF9 mRNA was found to be twice as abundant in oocytes of larger follicles (Fig. 3B), BMP15 transcripts exhibited slightly higher levels in smaller, less competent oocytes. Though GDF9 message was abundant at 55 000 copies per oocyte, BMP15 transcripts were more than 20-fold as abundant. We further isolated bovine c-mos, using degenerate primers, as a candidate gene for differential expression in competent versus noncompetent cells. C-mos was expressed very consistently at a 50% increased level in competent oocytes.
Regulation of Maternally Stored Genes During Early Embryogenesis
We further analyzed the expression of four of the known genes showing significant enrichment in competent oocytes and two randomly selected unknown genes, as well as the growth factors GDF9 and BMP15, and ß-actin during in vitro maturation and early embryogenesis (Fig. 4). All maternal transcripts examined, apart from BMP15, are 2- to 11-fold reduced after in vitro maturation, indicating that transcripts are degraded and/or deadenylated, thereby escaping our oligo-dT primed reverse transcription (Figs. 3, B and C, and 4). The differential behavior of GDF9 (11-fold decrease) and BMP15 (no change) is noteworthy.
|
We next examined the influence of de novo RNA transcription during this period by inhibiting RNA polymerase II-mediated transcription with -amanitin, added already upon aspiration of cumulus oocyte complexes. No change in transcript levels was seen in most of the genes. However, for BMP15, severalfold less RNA was detected when transcription was inhibited. In marked contrast, the precipitous drop in GDF9 transcripts during maturation is significantly reduced when transcription is inhibited.
Examination of RNA levels during preimplantation stages indicated that, for all genes examined, maternal messages disappeared, reaching background levels by the 12- to 16-cell stage (Fig. 4). After the 12- to 16-cell stage, as embryonic transcription gains momentum, those genes required for further development are transcribed and accumulate always to higher levels in Day 7 blastocysts than in morula stages. The transcription of clones 410 and 725 as well as oocyte growth factors GDF9 and BMP15 is not reactivated, suggesting that the high levels seen in the oocyte reflect a specific function at the oocyte or early cleavage stages. In contrast, Oct4 and NDFIP1 RNA levels are more abundant in the blastocyst than in the MII-stage oocyte (Fig. 4) in spite of the known restriction of Oct4 to inner cell mass tissue. The ß-actin profile is also of interest. This gene shows a dramatic rise in transcription between the morula and blastula stages, a result very similar to that seen by others [15].
Effect of Nuclear Transfer on Gene Expression at the Blastocyst Stage
We found that those maternal genes that are not reactivated upon embryonic genome activation also remain transcriptionally silent after nuclear transfer (Fig. 4, clones 410 and 725, GDF9, and BMP15). However, the nuclear transfer-derived blastocysts exhibited markedly higher variation in expression. Whereas ß-actin and NDFIP1 levels are very similar in the three populations of blastocysts; Oct4, Znf198, and SLBP expression is upregulated two- to fourfold in one or both of the nuclear transfer-derived batches of embryos.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Many studies, including the present one, have highlighted the observation that oocytes derived from bovine antral follicles smaller than 2–3 mm are less likely to develop to the blastocyst stage than those from follicles larger than 3 mm. From the 3-mm follicle stage until embryonic genome activation at the eight-cell stage, oocytes/embryos have been shown by 3H-uridine incorporation to be transcriptionally practically quiescent [9, 14, 15]. This suggests that, if maternal RNA stores contribute to developmental competency, significant qualitative and/or quantitative differences must exist between the RNA of oocytes from follicles larger than 3 mm and that of the small-follicle oocyte. Methods based on subtractive procedures (SSH, differential display) allow the enrichment for such differentially expressed genes and have been applied to populations of bovine oocytes [16] or early cleavage-stage embryos [25] differing in developmental competency. However, the lists of genes derived by such procedures have to be treated with caution, as a more quantitative analysis of three likely candidates of the isolated genes revealed that only one showed an increase (100%) in abundance [25]. Bearing this in mind, we have rigorously tested the expression levels of all the genes isolated in our screen by quantitative real-time PCR. Furthermore, we have developed a new method combining linear amplification (but using SP6 instead of T7 RNA polymerase) with SSH, thereby avoiding the bias introduced when amplifying small amounts of starting material using PCR-based methods [26]. The success of this method was verified by the observation that 90% of the isolated clones were at least 30% more abundant in three batches of developmentally competent oocytes with one unknown gene showing a greater than 200% enrichment (threefold more abundant). Furthermore, by measuring the absolute copy number per embryo for these genes, we could show that this procedure does not select only abundant messages, as we isolated genes expressed at levels ranging from hundreds to tens of thousands of copies per oocyte.
Maternal Genes Associated with Developmental Competency
Among the eight known genes for which we found the transcripts to be significantly enriched in competent oocytes are three coding for transcription factors, namely Oct4, Msx1, and Znf198. The homeobox-containing genes Oct4 and Msx1 are not very abundant at less than 2000 copies per oocyte, unlike the zinc-finger containing Znf198 transcripts, which are present at 10 times higher copy number. Loss of function of Oct4 in the mouse leads to preimplantation lethality [27] and mouse Msx1 homozygous mutants die at birth [28], thus precluding studies of the maternal role of these genes. Znf198, containing a zinc-finger DNA binding motif, has not been studied much but was recently found to bind Smad2/3 [29]. These Smads are the nuclear effectors of TFGß and activin signaling. Both TGFß and activin have important roles during folliculogenesis and on nuclear maturation [3], thus suggesting that the accumulation of Znf198 may be of functional significance.
Apg3p, NDFIP1, and CMAS are involved in the posttranslational modification of proteins. Though all three were seen to be more abundant in competent oocytes, only NDFIP1 was significantly enriched (P < 0.05). NDFIP1, also known as N4WBP5, binds to the ubiquitin-protein ligase Nedd4, acting as an adaptor to control ubiquitin-dependent protein sorting and trafficking [31, 32]. The enrichment of factors involved in protein modifications is interesting considering that early embryos have to rely on nontranscriptional regulatory mechanisms to control development before embryonic genome activation.
Cyclin A2 mRNA, though not very abundant at around 1000 copies/oocyte, was twice as abundant in competent oocytes. In the mouse, maternal cyclin A2 was shown to be required for the regulation of S-phase during the first mitotic cell division [33]. The regulation of the cyclin A2 protein was controlled by selective polyadenylation of the maternal mRNA, in the absence of which DNA replication was arrested at the one-cell stage. If a similar control takes place in cattle, an accumulation of Cyclin A2 RNA would be expected to correlate with an increase in developmental competency.
The stem-loop binding protein (SLBP) is a RNA binding protein that plays a central role in the stabilization and translation of mRNAs encoding the replication-dependent histones [34]. SLBP protein has been shown in the mouse to accumulate during oocyte meiotic maturation and via its effect on histone mRNA control the availability of histone proteins necessary for repackaging the sperm chromatin and subsequently the replicated DNA [35]. This increase in protein is dependent on translation of SLBP RNA stored in the oocyte [35]. In this context, the identification of the bovine homologue of histone stem loop binding protein SLBP as being transcriptionally enriched by 75% in developmentally competent oocytes is intriguing. Furthermore, the levels of SLBP RNA in such oocytes are very high, in excess of 70 000 copies. Hence, the accumulation of stored SLBP mRNA may be paramount to achieving developmental competency by allowing the subsequent regulation of histone protein, essential to chromatin remodeling occurring from fertilization stages onward.
The present collection of 10 maternally expressed genes (clone 1166, CycA, Dja4, Msx1, NDFIP1, Oct4, SLBP, Trappc, Znf198, and GDF9), which are significantly more abundant in oocytes of large follicles, represents the largest assembly of maternal bovine genes verified to be associated with developmental competency. Yet apart from clone 1166, the changes in expression level tend not to exceed 100% or twofold.
How do these changes compare with those of genes known to be required for folliculogenesis or meiotic maturation? C-mos translation and/or activation is required for the suppression of interphase between meiosis I and II and later for the metaphase arrest at meiosis II [36]. We isolated bovine c-mos and showed that, while its RNA is 1.5-fold as abundant in oocytes of larger follicles, this increase was not statistically significant. This correlates with observations that c-mos activity in the mouse is regulated mainly at the level of protein and RNA translation as opposed to transcription [36].
The closely related TGFß growth factors GDF9 and BMP15 are essential for folliculogenesis in the mouse and sheep, respectively [37, 38]. Whereas GDF9 was indeed upregulated in competent oocytes, BMP15 was present at slightly reduced levels, suggesting that some genes do not accumulate in the 2- to 5-mm antral follicle window of development. In line with our results, no differences in BMP15 transcript levels were found in developmentally superior and inferior two-cell embryos [25]. We noticed a remarkable abundance of BMP15 mRNA. BMP15 transcripts were more than 20-fold as abundant as those of GDF9, itself expressed at substantial levels. This is interesting in view of the observation that sheep homozygous for the null mutations in the BMP15 gene do not phenocopy BMP15 knockout mice but instead resemble GDF9 knock-out mice [39]. We suggest that this difference in phenotype may simply be due to a species-specific difference in relative expression levels of these two closely related genes.
In conclusion, though this is a small representation of all genes expressed in the oocyte, our results suggest that developmental competency may be a quantitative trait, being dependent on small changes in the RNA levels of many genes. In the future, microarray approaches may yield further insights into this question; however, presently, the dearth of bovine maternal and preattachment cDNAs make such an approach unfeasible. Furthermore, the use of available heterologous cDNA arrays from humans or mice is of limited use, as linearly amplified RNA is biased to 3' sequences, which are not well conserved among different species.
Developmental Profile of Maternal Genes Associated with Competency
We analyzed the fate of some of the isolated maternal messages during meiotic maturation and development up to the blastocyst stages. In the mouse, global measurements have indicated that more than half the mRNA is deadenylated (thereby making RNA refractile to oligo-dT-mediated reverse transcription) or degraded during meiotic maturation [40, 41]. Both RNA degradation and deadenylation prevent translation, thus precluding protein synthesis. Similar to the mouse, all the genes we examined, apart from BMP15, exhibited reduced transcript levels/polyadenylation between germinal vesicle-stage (>5-mm follicle-derived) and MII-stage (in vitro-matured) oocytes. The maintenance of BMP15 RNA levels is likely to be due to continued transcription during maturation, as BMP15 RNA levels dropped dramatically (sixfold) when transcription was inhibited during maturation. Seven other genes analyzed were unaffected by the presence of the transcriptional inhibitor -amanitin in line with the documented transcriptional near quiescence of such oocytes [9, 14]. In contrast, transcriptional inhibition during maturation resulted in threefold higher levels of GDF9 RNA, reflecting either a reduction in RNA degradation and/or an increase in the proportion of GDF9 RNA that is polyadenylated. This effect appears to be specific to GDF9 and may depend on the activity of a factor that is normally actively transcribed during meiotic maturation.
We see a decrease of the polyadenylated mRNA of all nine genes analyzed from MII-stage oocytes to the 12- to 16-cell embryo stage. Transcripts are barely detectable by the 12- to 16-cell stage. Oct4, Znf198, NDFIP1, and SLBP are reexpressed at levels that increase from morula to blastula stages. In contrast, GDF9 and BMP15 and the unknown genes, clones 410 and 725, are not reexpressed significantly at morula or blastocyst stages, indicative of oocyte-specific genes. Interestingly, all nine genes examined show the same pattern of degradation during embryogenesis whether their messages are enriched in developmentally competent oocytes (Oct4, Znf198, NDFIP1, SLBP, GDF9) or not (clones 410, 725, and BMP15).
We examined the effect of nuclear transfer (NT) on the expression of the characterized oocyte-specific genes as well as on those maternal genes normally reexpressed by the blastocyst-stage embryo. NT embryos derived from the transfer of two different sources of fibroblast donor nuclei did not contain RNA at the blastula stage of the oocyte-specific genes GDF9, BMP15, clone 410, and 725. This indicates, first, that the degradation of maternal RNA stores has occurred normally and, second, that these genes were not reexpressed erroneously after the process of nuclear transfer. Considering that the maternal nucleus has been replaced by a differentiated nucleus, these results suggest that the degradation/poly- and deadenylation events driving the stability of maternal RNAs is controlled by cytoplasmic as opposed to nuclear components.
Examining those genes that are normally reexpressed in NT blastulae, we noted excessive Oct4 and SLBP expression in the NT2 and NT1 pools. Upregulation and ectopic trophoblast expression of Oct4 has been noted before in mouse NT embryos [42]. Misregulation of Oct4 is likely to lead to embryonic lethality, as both the levels and location of Oct4 expression are critical for correct embryonic development [27]. In this small set of genes examined, one of five genes shows grossly altered levels (over threefold), yet this is not the same gene in the two NT experiments. This differential behavior in gene expression from nuclear transfer to nuclear transfer experiment probably reflects the incomplete and stochastic nature of reprogramming of the fibroblast donor nuclei upon nuclear transfer [42].
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Peter L. Pfeffer, AgResearch Crown Research Institute, Ruakura Campus, East Street, Hamilton 2001, New Zealand. FAX: +64 7 8385628; peter.pfeffer{at}agresearch.co.nz
Received: 19 May 2004.
First decision: 9 June 2004.
Accepted: 22 July 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Adjaye, R. Herwig, T. C. Brink, D. Herrmann, B. Greber, S. Sudheer, D. Groth, J. W. Carnwath, H. Lehrach, and H. Niemann Conserved molecular portraits of bovine and human blastocysts as a consequence of the transition from maternal to embryonic control of gene expression Physiol Genomics, October 19, 2007; 31(2): 315 - 327. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-W. Dessie, F. Rings, M. Holker, M. Gilles, D. Jennen, E. Tholen, V. Havlicek, U. Besenfelder, V. L Sukhorukov, U. Zimmermann, et al. Dielectrophoretic behavior of in vitro-derived bovine metaphase II oocytes and zygotes and its relation to in vitro embryonic developmental competence and mRNA expression pattern Reproduction, May 1, 2007; 133(5): 931 - 946. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Wood, D. A. Dumesic, D. H. Abbott, and J. F. Strauss III Molecular Abnormalities in Oocytes from Women with Polycystic Ovary Syndrome Revealed by Microarray Analysis J. Clin. Endocrinol. Metab., February 1, 2007; 92(2): 705 - 713. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Smith, D. Berg, S. Beaumont, N. T Standley, D. N Wells, and P. L Pfeffer Simultaneous gene quantitation of multiple genes in individual bovine nuclear transfer blastocysts Reproduction, January 1, 2007; 133(1): 231 - 242. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Somers, C. Smith, M. Donnison, D. N Wells, H. Henderson, L. McLeay, and P L Pfeffer Gene expression profiling of individual bovine nuclear transfer blastocysts. Reproduction, June 1, 2006; 131(6): 1073 - 1084. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP |
