Abstract

BACKGROUND

Immature human oocytes matured in vitro, particularly those from gonadotrophin stimulated ovaries, are developmentally incompetent when compared with oocytes matured in vivo. This developmental incompetence has been explained as poor oocyte cytoplasmic maturation without any determination of the likely molecular basis of this observation.

METHODS

Replicate whole human genome arrays were generated for immature and mature oocytes (matured in vivo and in vitro, prior to exposure to sperm) recovered from women undertaking gonadotrophin treatment for assisted reproduction.

RESULTS

More than 2000 genes were identified as expressed at more than 2-fold higher levels in oocytes matured in vitro than those matured in vivo (P < 0.05, range 4.98 × 10−2 –2.22 × 10−4) and 162 of these are expressed at 10-fold or greater levels (P < 0.05, range 4.98 × 10−2–1.38 × 10−3). Many of these genes are involved in transcription, the cell cycle and its regulation, transport and cellular protein metabolism.

CONCLUSIONS

Global gene expression profiling using microarrays and bioinformatics analysis has provided a molecular basis for differences in the developmental competence of oocytes matured in vitro compared with in vivo. The over-abundance of transcripts identified in immature germinal vesicle stage oocytes recovered from gonadotrophin stimulated cycles and matured in vitro is probably due to dysregulation in either gene transcription or post-transcriptional modification of genes. Either mechanism would result in an incorrect temporal utilization of genes which may culminate in developmental incompetence of any embryos derived from these oocytes.

Introduction

Mammalian ovarian folliculogenesis is characterized by growth of follicles and the associated oocytes. The growing oocyte remains arrested in Prophase I of the first meiotic division and is highly transcriptionally and translationally active (Bachvarova, 1985). During the growth phase, the oocyte accumulates both RNA and proteins required for completion of the meiotic cell cycle, events such as chromatin remodeling and cell cycle activation associated with fertilization, the first mitotic cell cycles, the establishment of an embryonic genome and normal metabolic and homeostatic processes (Zheng et al., 2005). Oocytes that fail to complete the growth phase or fail to accumulate and regulate these molecules resulting in an incorrect temporal utilization would be expected to exhibit delays or failure in progression through preimplantation development following fertilization. The ability to complete preimplantation development to the blastocyst stage and successfully implant in a receptive uterus is generally referred to as developmental competence and this may be completely independent of meiotic competence (Trounson et al., 2001). Human oocytes acquire the ability to complete nuclear maturation when the follicle reaches ∼5 mm in diameter (Wynn et al., 1998) and the oocytes are ∼90 µm in diameter (Durinzi et al., 1995) but complete developmental competence is probably not achieved until the follicle reaches a diameter of 10–12 mm (Trounson et al., 2001). Transcripts accumulated throughout the human oocyte growth phase control development until a transcriptionally active embryonic genome is established at the 4- to 8-cell stage of embryo development (Braude et al., 1988).

Much of our knowledge of transcription and translation in oocytes has been acquired from the study of gametes from laboratory animal species. Little is known about the corresponding molecular events in the human oocyte due to the limited number of oocytes available for study. Understanding of the molecular events that result in a developmentally competent human oocyte and the extent to which gene expression profiles can be altered by pathologies that underlie infertility or the cellular manipulations that are used for assisted reproduction procedures, are critical for understanding and addressing improvements in fertility. The advent of commercial microarrays and high-fidelity RNA amplification techniques now makes it possible to profile the polyadenylated [poly(A)] RNA transcripts in the rarely available human oocyte.

Several groups have used microarrays to profile the transcriptome of the human oocyte (Bermudez et al., 2004; Dobson et al., 2004; Assou et al., 2006; Kocabas et al., 2006; Li et al., 2006a; Zhang et al., 2006; Gasca et al., 2007). Although much of the information is useful and concurs with previous findings in the oocytes of laboratory animal species, interpretation of gene expression patterns is nonetheless restricted because of the dependence on a number of critical factors that themselves influence expression, e.g. oocytes aged in vitro were used to represent the mature metaphase II (MII) oocyte in some or all instances (Bermudez et al., 2004; Dobson et al., 2004; Assou et al., 2006; Gasca et al., 2007), a limited coverage of the human transcriptome on the microarray (Bermudez et al., 2004; Dobson et al., 2004; Li et al., 2006a), a lack of sufficient biological replication to produce statistically meaningful results for all samples studied (Bermudez et al., 2004; Dobson et al., 2004; Assou et al., 2006; Li et al., 2006a; Zhang et al., 2006; Gasca et al., 2007) and the use of single oocytes for some arrays (Bermudez et al., 2004; Dobson et al., 2004; Li et al., 2006a). It has been demonstrated previously that the amplification and microarray methodologies employed in the present study fails to produce a fully representative gene expression profile from a single oocyte because of the bias introduced when amplifying from low template RNA samples (Jones et al., 2007). There is no published data to suggest that other amplification and microarray methodologies are not similarly limited. Furthermore, the majority of the studies of the oocyte transcriptome in the medical literature report on an isolated developmental stage and make comparisons with unrelated, or loosely related, material such as preimplantation embryos (Dobson et al., 2004; Li et al., 2006a), cumulus cells (Assou et al., 2006), somatic tissues and embryonic stem cells (Kocabas et al., 2006; Zhang et al., 2006) and as such, it is difficult to derive meaningful data about the gene expression profile of a developmentally competent human oocyte.

The present study was undertaken to comprehensively identify genes associated with maturing human oocytes and biological processes or regulation of the transcriptome that are associated with the gain or loss of developmental competence. The majority of oocytes recovered from infertile women following superovulation are at the mature meiotic metaphase II (MII) stage and 70–80% of these are capable of fertilization with ∼50% of zygotes able to complete development to the blastocyst stage (Jones, 2000). Approximately 35% of transferred embryos/blastocysts derived from these oocytes develop to term (Blake et al., 2005). In contrast, when oocytes with an immature nucleus are recovered following superovulation and are matured in vitro, only 12% of the resulting zygotes develop to the blastocyst stage (Chen et al., 2000) and only 14% of these embryos when transferred develop to term (Veeck et al., 1983; Nagy et al., 1996; Edirisinghe et al., 1997; Tucker et al., 1998; De Vos et al., 1999; Chen et al., 2000; Vanhoutte et al., 2005). The majority of the reported pregnancies have been isolated case reports resulting from in vitro maturation of partly mature metaphase I (MI) oocytes.

The present study compared the transcriptome of oocytes with relatively high developmental competence (in vivo matured oocytes following superovulation) with the transcriptome of oocytes with very low developmental competence (in vitro matured oocytes following superovulation for IVF) in an attempt to identify a molecular basis for the difference in developmental competence. Identification of the molecular basis for oocyte developmental competence may result in improvements to maturation protocols for immature oocytes recovered not only following superovulation but also for immature oocytes recovered in the absence of gonadotrophin stimulation. The former would be important for the few patients for whom the majority of oocytes recovered following superovulation is immature and therefore not suitable for microinjection and also for human somatic cell nuclear transfer, as these oocytes presently represent the largest pool of donor oocytes available for research and autologous stem cell therapy.

Materials and Methods

Human oocyte collection

Human oocytes in excess to the requirements of treatment for assisted reproduction were donated to research following informed consent from patients undertaking a long down-regulation superovulation protocol with GnRH agonists and gonadotrophin for treatment of infertility at Monash IVF Pty Ltd, Melbourne, Australia, and SISMER, Bologna, Italy. The oocytes were obtained under ethics approval from Monash Private Surgical Hospital Human Research Ethics Committee (#02044), SISMER Institutional Review Board and Monash University Standing Committee on Ethics in Research involving Humans (#2004/469MC and #2005/272ED). Oocyte–cumulus complexes were evaluated for maturity in the embryology laboratories within 1–4 h of collection following removal of cumulus cells using hyaluronidase (25 IU/ml Hyalase; CP Pharmaceuticals, Wrexham, UK, or 40 IU/ml; Medicult, Jyllinge, Denmark). Immature, germinal vesicle (GV) and GV breakdown/MI oocytes and mature MII oocytes in excess to requirements were donated to research within 4–6 h of collection. Oocytes donated by Monash IVF patients were transported to Monash Immunology and Stem Cell Laboratories (MISCL) in HEPES buffered culture medium (Sydney IVF oocyte wash buffer; William A. Cook Australia, Eight Mile Plains, Qld or Quinns Advantage® medium; Sage, Cooper Surgical Co., Trumbull, CT, USA) at 37°C and were processed to preserve RNA immediately. Oocytes donated by SISMER patients were processed to preserve RNA on site, stored in liquid nitrogen and transferred to MISCL by air freight on dry ice.

Human oocyte development in vitro

A proportion of oocytes at the GV stage showing clear cytoplasm and no obvious inclusions were not processed immediately for RNA but were placed in pre-equilibrated in vitro maturation medium (Medicult) supplemented with 0.075 IU/ml human recombinant FSH (Gonal-F; Serono Australia, Frenchs Forest, NSW, Australia), 0.1 IU/ml hCG (Profasi; Serono Australia) and 5 mg/ml human serum albumin (HSA; Sage, Cooper Surgical Co.) and incubated at 5% CO2 for 24 h at 37°C. At the end of the incubation period, oocytes were assessed for maturity and oocytes with a visible polar body in the perivitelline space were processed for RNA as in vitro matured MII oocytes (IVM).

Oocyte processing for RNA preservation

Oocytes were re-evaluated to ensure the maturational status and morphological quality prior to processing to preserve RNA. The zona pellucida was removed chemically by exposure to 0.2% Pronase (Sigma Chemical Company, St Louis, MO, USA) in HEPES buffered culture medium or acidified Tyrode's solution and the naked oocyte immediately washed 3× in HEPES buffered culture medium supplemented with 5 mg/ml HSA (Sage, Cooper Surgical Co.). Each naked oocyte was then washed 4× in Dulbecco's phosphate buffered saline (Gibco BRL, Invitrogen Corporation, Grand Island, NY, USA) without protein and transferred in a minimal volume of medium to an RNA–DNA-free PCR tube containing 5 µl of Picopure extraction buffer (Arcturus Bioscience Inc., Mountain View, CA, USA). Tubes were snap frozen in liquid nitrogen and stored at a minimum temperature of −80°C until required.

RNA extraction, labeling and hybridization on Codelink whole human genome microarrays

Prior to RNA extraction, oocyte lysates were pooled in groups of five according to the oocyte maturational stages: GV, MI, MII and IVM, as it has been demonstrated that pooling in these numbers generates a representative gene expression profile with high fidelity and sensitivity (Jones et al., 2007). In order to remove any bias associated with the donor's age or etiology of infertility and generate a gene expression profile typical of the maturation conditions and/or developmental stage, oocytes from different donors were pooled wherever possible.

Total RNA from the pooled lysates in Picopure extraction buffer was isolated using the Picopure RNA Isolation Kit (Arcturus Bioscience Inc.) according to the manufacturer's protocol. Conversion of the T7 tagged complementary DNA (cDNA) was achieved in two rounds of synthesis using the RiboAmp HS RNA Amplification Kit (Arcturus Bioscience Inc.). The T7 tagged cDNAs were purified through Micro-Spin S-400HR columns (GE Healthcare Biosciences, Little Chalfont, England) and biotinylated cRNA generated by in vitro transcription using the Codelink Expression Assay Reagent Kit (GE Healthcare Biosciences, Piscataway, NJ, USA). Biotin-labeled cRNA was purified using an RNeasy Mini Kit (QIAGEN GmbH, Hilden, Germany) and the quantity and purity evaluated by UV spectrophotometry at 260 and 280 nm. The total yield of biotinylated cRNA was 30–68 µg for GV, 35–46 µg for MI, 10–72 µg for MII and 26–32 µg for IVM oocytes. The final cRNA concentrations were adjusted to 1 µg/µl.

Prior to hybridization, 10 µg biotinylated cRNA was fragmented in 25 µl of 1× fragmentation buffer (Codelink Expression Assay Reagent Kit) at 95°C for 20 min. Hybridization reaction buffer (260 µl) was prepared by mixing 25 µl of fragmented cRNA, 78 µl of hybridization buffer component A, 130 µl of hybridization component B and 27 µl of nuclease-free water (Codelink Expression Assay Reagent Kit). The mixture was incubated at 90°C for 5 min, chilled on ice and injected into the inlet port of the hybridization chamber of Codelink Whole Human Genome Bioarrays printed with 54 840 discovery probes representing 18 055 human genes and an additional 29 378 human expressed sequence tags (EST) (GE Healthcare Biosciences). The chamber ports were sealed with 1 cm sealing strips and hybridized for 18 h at 37°C on an orbital shaker (Innova 4080; New Brunswick Scientific Co., Edison, NJ, USA) at 300 rpm. After hybridization, the hybridization chamber was removed from each slide and arrays were briefly rinsed with 0.75× TNT buffer (0.1 M Tris–HCl, pH 7.6, 0.15 M NaCl, 0.05% Tween-20; Sigma Aldrich Corporation, St. Louis, MO, USA) at room temperature, followed by an hour at 46°C in 0.75× TNT buffer. The signal was developed by incubating the arrays in a 1:5000 Cy5-streptavidin working solution (GE Healthcare Biosciences) at room temperature for 30 min followed by four successive 5 min washes in 1× TNT buffer. Arrays were dried by centrifugation and scanned on an Axon GenePix Array Scanner (Molecular Devices Co., Sunnyvale, CA, USA) with the laser set at 635 nm, the photomultiplier tube at 600 V and the scan resolution at 5 µm and images captured as TIFF files. Codelink Expression Analysis version 4.2 software (GE Healthcare Biosciences) was used to analyze images for each slide. Spots with intensities below that of the negative control (absence of an oligonucleotide probe) were excluded, as were those with irregular shapes or near-background intensity or oligonucleotides masked as part of the quality control process during manufacture. Spot quality and signal intensities were exported to Genespring compatible report format.

Microarray data analysis

The Codelink Expression Analysis output was loaded into Genespring GX 7.3.1 (Agilent Technologies, Santa Clara, CA, USA) and values below 0.01 were set to 0.01 with per chip normalization to the 50th percentile and per gene normalization to the median.

A principal components analysis based on all genes was applied to all microarrays in the experiment prior to any detailed statistical analysis to identify any samples that showed significant variability and to identify whether there was discrimination between the different developmental stages. Prior to statistical analysis of MII oocytes matured in vivo versus in vitro filtering was conducted to eliminate data of poor quality. Data were progressively filtered: less precise measurements based on control strength were removed, measurements for each condition with <80% confidence were removed, probes recorded as absent in all samples were removed and the measurements for probes representing the positive and fiducial controls rather than probes representing the 54 840 Codelink discovery probes were removed. A Welch t-test with a false discovery rate of 0.05 followed by a Benjamini and Hochberg multiple testing correction was applied to the quality data to identify probes that were expressed significantly differently at the 5% level between oocytes matured in vivo and in vitro. Genespring GX 7.3.1 Bioscript Library 2.2 Biological Pathways analysis was used to identify the probes associated with particular gene ontology (GO) biological processes represented on the microarrays that were significantly over-represented at the 5% level within the list of probes identified as having significantly different expression levels between in vivo and in vitro matured oocytes.

Results

Global characteristics of human oocyte gene expression according to maturational stage and maturation conditions

A total number of 5 GV, 3 MI, 11 MII and 3 IVM microarrays, representing five pooled oocytes for each array, were completed. The total number of probes detected varied from 22 532–31 837 probes for GV oocytes (10 017–12 734 genes), 27 789–28 843 probes for MI oocytes (12 011–12 697 genes), 12 238–29 593 probes for MII oocytes (4801–12 031 genes) and 20 522–22 624 probes for IVM oocytes (9237–9933 genes). Principal components analysis of gene expression based on all genes on the arrays identified variability in gene expression among the in vivo matured MII oocytes (Fig. 1). Nevertheless the developmental stages and maturation conditions for the most part clustered together with immature, GV and MI oocytes showing similar gene expression profiles and in vivo and in vitro matured oocytes showing distinctly different gene expression profiles with very little overlap between the two maturation conditions. To minimize false positives, only those probes detected in the majority of replicates for each developmental stage/maturation condition were taken as expressed. In the majority of GV oocytes 24 562 probes (10 962 genes) were detected, and 27 980 probes (12 329 genes) were detected in the majority of MI oocytes. Fewer probes were detected in the majority of MII oocytes regardless of maturation condition (17 650 and 21 336 probes representing 7546 and 9479 genes for in vivo and in vitro matured oocytes, respectively; Table I).

Figure 1:

Principal components analysis of the 22 human stage-specific oocyte microarrays based on all genes: GV stage oocytes (closed squares), GV breakdown or metaphase I oocytes (open squares), in vitro matured metaphase II oocytes (open circles) and in vivo matured metaphase II oocytes (closed circles).

Figure 1:

Principal components analysis of the 22 human stage-specific oocyte microarrays based on all genes: GV stage oocytes (closed squares), GV breakdown or metaphase I oocytes (open squares), in vitro matured metaphase II oocytes (open circles) and in vivo matured metaphase II oocytes (closed circles).

Table I.

The number of probes and genes detected in the majority of microarrays for immature (GV and MI) oocytes and oocytes matured in vivo (MII) and in vitro (IVM).

Developmental stage Number of independent samples Number of probes detected Number of genes 
GV 24 562 10 962 
MI 27 980 12 329 
MII 11 17 650 7546 
IVM 21 336 9479 
Developmental stage Number of independent samples Number of probes detected Number of genes 
GV 24 562 10 962 
MI 27 980 12 329 
MII 11 17 650 7546 
IVM 21 336 9479 

GV, germinal vesicle; MI, metaphase I; MII, metaphase II; IVM, in vitro matured MII oocytes.

Comparative gene expression for human oocytes matured in vivo or in vitro

When the data for in vivo and in vitro matured oocytes were filtered for quality data and the data further limited to probes represented in the majority of MII (6/11) or the majority of IVM (2/3) replicates, there was a significant difference in gene expression (P < 0.05) between in vivo and in vitro matured oocytes for 4226 probes (2766 genes). When this data were further limited to those probes that were statistically significantly different at the 2-fold or higher level, 3250 probes (2390 genes) were identified (Table II). The majority of probes (3166 probes representing 2348 genes; Table II) were expressed at greater than 2-fold higher levels in oocytes matured in vitro. This list was subjected to further analysis to determine which of the functional categories assigned by GO terms (Gene Ontology Consortium, 2001) were over-represented at the 95% confidence level with a minimum overlap set to two genes. Over-representation does not evaluate gene expression levels but instead groups genes according to processes that occur more often in the gene list of interest than could be predicted by the distribution among all genes for a particular process represented on the array. One hundred and ninety-eight GO biological processes were identified to be significant at the 95% level which were re-categorized into 15 main groups (Table III) including in order of significance probes involved in transcription, the cell cycle, transport and cellular protein metabolism.

Table II.

Fold change differences in gene expression for the majority of in vivo (MII) and in vitro matured (IVM) oocyte microarrays (*P < 0.05).

Trend Gene expression Significantly* different >2-fold >5-fold >10-fold >25-fold >50-fold 
IVM>MII Genes 2627 2348 1025 162 
 Probes 3776 3166 1280 197 12 
IVM< MII Genes 139 42 15 
 Probes 450 84 23 
Trend Gene expression Significantly* different >2-fold >5-fold >10-fold >25-fold >50-fold 
IVM>MII Genes 2627 2348 1025 162 
 Probes 3776 3166 1280 197 12 
IVM< MII Genes 139 42 15 
 Probes 450 84 23 
Table III.

Functional characterization of transcripts more than 2-fold higher in in vitro matured oocytes and in vivo matured oocytes.

Rank Category Number of overlapping probes in gene list 
In vitro matured oocytes 
Nucleobase, nucleoside, nucleotide and nucleic acid metabolism 469 
 A. Transcription 324 
 B. DNA metabolism 71 
 C. RNA metabolism 109 
Cell cycle 135 
Transport 254 
Cell division 49 
Cellular protein metabolism 382 
Response to stress 74 
 A. Response to DNA damage stimulus 57 
 B. Response to oxidative stress 
Cell death 78 
Signal transduction 259 
Cell proliferation 61 
10 Generation of precursor metabolites and energy 22 
11 Cell organization and biogenesis 60 
 A. Cytoskeleton organization and biogenesis 22 
 B. Chromosome organization and biogenesis 10 
12 Biological process unknown 67 
13 Reproduction 28 
14 Cellular lipid metabolism 35 
15 Development 72 
In vivo matured oocytes 
Transport 
Cell growth 
Signal transduction 
Rank Category Number of overlapping probes in gene list 
In vitro matured oocytes 
Nucleobase, nucleoside, nucleotide and nucleic acid metabolism 469 
 A. Transcription 324 
 B. DNA metabolism 71 
 C. RNA metabolism 109 
Cell cycle 135 
Transport 254 
Cell division 49 
Cellular protein metabolism 382 
Response to stress 74 
 A. Response to DNA damage stimulus 57 
 B. Response to oxidative stress 
Cell death 78 
Signal transduction 259 
Cell proliferation 61 
10 Generation of precursor metabolites and energy 22 
11 Cell organization and biogenesis 60 
 A. Cytoskeleton organization and biogenesis 22 
 B. Chromosome organization and biogenesis 10 
12 Biological process unknown 67 
13 Reproduction 28 
14 Cellular lipid metabolism 35 
15 Development 72 
In vivo matured oocytes 
Transport 
Cell growth 
Signal transduction 

The most frequent categories of GO biological processes were identified using GeneSpring GX 7.3.1 Bioscript Library 2.2 Biological Pathways Analysis. Only GO biological processes where two or more probes overlapped with a likelihood of a random overlap <5% are shown. These biological processes were further re-categorized so that the most frequent GO biological processes were grouped according to statistical significance.

Only a small number of probes were identified at greater than 2-fold higher levels in oocytes matured in vivo (84 probes representing 42 genes; Table II). This list was also subjected to analysis to determine the GO biological processes most over-represented at the 95% confidence level with a minimum overlap set to two genes. Four GO biological processes were identified to be significant at the 95% level which were re-categorized into three main groups (Table III) including, in ranked order, transport, cell growth and signal transduction.

Differentially expressed genes 10-fold or more different in oocytes matured in vitro or in vivo

In order to clarify a gene list that might better reflect differences between oocytes matured in vivo and in vitro a cut-off of 10-fold different was applied to the list of 4226 probes (2766 genes) that were demonstrated to be statistically different. The subset of these probes expressed at 10-fold or greater levels in oocytes matured in vivo or in vitro together with the National Center for Biotechnology Information accession number for all known genes, EST and corenucleotide sequences are provided in Supplementary Table 1. Note that some probes on the Codelink microarrays represent more than one nucleotide sequence and occasionally represent more than one gene. Supplementary Table 1 also includes the official gene name and gene symbol for each of the genes represented, the chromosomal location, the known associated biological processes, the mean normalized expression value from the microarrays for oocytes matured in vitro and in vivo, the P-value and the relative ratio of expression differences.

For oocytes matured in vitro, 197 probes (162 genes) were expressed at 10-fold greater levels than in oocytes matured in vivo. Genes from all 23 chromosomes were represented in this list. Additionally, the biological processes over-represented in the list of genes that were expressed at 2-fold higher levels in oocytes matured in vitro are similarly represented in the list of genes that are expressed at 10-fold higher levels. For oocytes matured in vivo, only six probes (three genes) were expressed at 10-fold greater levels than oocytes matured in vitro.

Discussion

The present study has defined the global gene expression profile of oocytes recovered at all maturational stages of development following gonadotrophin stimulation for assisted reproduction and prior to exposure to sperm. In addition, the global gene expression profile of immature GV stage oocytes recovered from gonadotrophin stimulated cycles and subsequently matured in vitro in culture medium supplemented with gonadotrophins has also been defined and this profile compared with the profile of more developmentally competent in vivo matured oocytes. The major difference between oocytes matured in vivo (MII) and in vitro (IVM) was that a large number of genes were more highly expressed in oocytes matured in vitro and many of these genes are involved in the biological processes of transcription, the cell cycle and its regulation, transport and cellular protein metabolism.

Immature GV stage oocytes recovered from stimulated cycles may be competent to resume meiosis and reach the MII stage of the meiotic cell cycle but have very limited developmental competence once fertilized (Veeck et al., 1983; Nagy et al., 1996; Edirisinghe et al., 1997; Tucker et al., 1998; De Vos et al., 1999; Chen et al., 2000; Vanhoutte et al., 2005). Human oocytes matured in vitro have an increased incidence of spindle abnormalities and chromosomal misalignments compared with human oocytes matured in vivo regardless of whether the immature oocytes are recovered from unstimulated (Racowsky and Kaufman, 1992; Li et al., 2006b) or hormonally stimulated ovaries (Cekleniak et al., 2001; Wang and Keefe, 2002). Contributing to the developmental incompetence, oocytes matured in vitro have a very high incidence of aneuploidy (Magli et al., 2006) and resulting embryos have a higher incidence of nuclear fragmentation, anuclear blastomeres (DeScisciolo et al., 2000), multinuclear blastomeres and aneuploidy compared with embryos resulting from oocytes matured in vivo (Nogueira et al., 2000). These physiological observations indicate a critical failure in cell cycle regulation, particularly spindle assembly and the cell cycle checkpoint and many of the genes involved in these processes were identified to be expressed at higher levels in oocytes matured in vitro.

Mammalian oocytes accumulate large amounts of RNA during the growth phase but transcription effectively ceases once the oocyte resumes the meiotic cell cycle (Bachvarova, 1985; Tomek et al., 2002). Many of the transcripts accumulated during the growth phase exist in a stable but translationally inactive form with short poly(A) tails (Bachvarova, 1992). Most stored messenger RNA (mRNA) is activated following the addition of several hundred adenine residues to the 3′ poly(A) tail at various times during maturation and early embryo development (Bachvarova, 1992). During meiotic maturation, some maternal mRNAs are adenylated whereas others are deadenylated or degraded (Paynton and Bachvarova, 1994). In fact, almost 50% of the accumulated maternal mRNA is degraded during meiotic maturation (De Leon et al., 1983; Lequarre et al., 2004) which correlates with the findings in the present study where significantly fewer transcripts were detected in mature MII oocytes regardless of the maturation conditions.

It could be argued that the dysregulation of gene expression observed in the oocyte matured in vitro originates in the GV oocyte that has failed to mature in vivo despite exposure to the ovulatory dose of hCG. GV oocytes recovered from the monkey following FSH/hCG stimulation (hCG; high developmental competence), FSH alone (FSH; moderate developmental competence) and no hormonal stimulation (NS; low developmental competence) showed remarkably little difference in the expression levels of 23 maternal mRNA's selected on the basis of high expression levels in oocytes and preimplantation embryos (Zheng et al., 2005). However, these same genes were expressed at significantly higher levels in oocytes matured in vitro than in oocytes matured in vivo, which is similar to the findings in the present study and suggests that the dysregulation occurs during maturation and does not arise in the GV oocyte. FSH oocytes and embryos showed a less severe alteration in gene expression than NS oocytes and embryos when compared with hCG oocytes and embryos (Zheng et al., 2005). For most of the genes examined in oocytes with low developmental competence, there is an over-abundance initially in the MII oocyte but this pattern is reversed with a substantial reduction in relative gene expression by the 2-cell stage (Zheng et al., 2005). For genes known to be required later in preimplantation development, there is parity in expression at the oocyte stage with a significant increase in relative expression levels by the pronucleate stage of development indicating that these genes are precociously up-regulated (Zheng et al., 2005). It is not ethically permissible to generate human embryos for the purpose of research so these observations cannot be directly confirmed in the human. Zheng et al. (2005) suggested that the reduced developmental competence of non-human primate oocytes matured in vitro was a result of a failure of these oocytes to undergo the normal pattern of transcript silencing. As we also identified over-abundance of a large number of genes in human oocytes matured in vitro compared with in vivo, we propose a failure in the normal post-transcriptional regulatory processes as an explanation of the relatively poor developmental competence of human oocytes matured in vitro.

What is the origin of the increased gene expression in oocytes matured in vitro compared with in vivo? An increase in gene expression detected by mRNA microarray can be due to either new transcription or polyadenylation of existing dormant transcripts. In the oocyte, differential mRNA stability is an important way to regulate the availability of transcripts for translation and is dependent on the presence in the 3′ untranslated region of the nuclear polyadenylation signal (AAUAAA) and the cytoplasmic polyadenylation signal (CPE) at a variable distance 5′ to the AAUAAA sequence (Bachvarova, 1992). Stored mRNAs are bound by protein complexes which include the cytoplasmic polyadenylation element binding protein (CPEB) which inhibit translation until CPEB is phosphorylated by Aurora A (Mendez et al., 2000). Phosphorylation of CPEB interacts with the cleavage and polyadenylation specific factor bound to the nuclear polyadenylation signal recruiting poly(A) polymerase to form the active polyadenylation complex thus initiating polyadenylation and translation (Mendez et al., 2000). Deadenylation on the other hand appears to happen as a default pathway on re-initiation of meiosis and occurs to transcripts lacking a CPE (Varnum and Wormington, 1990; Paynton and Bachvarova, 1994). Transcript abundance that was observed in oocytes matured in vitro could therefore be explained by a failure of the default deadenylation pathway that occurs in oocytes matured in vivo. Now that these genes have been identified it will be possible to follow the poly(A) tail length during the course of maturation and determine whether their transcripts arise by new transcription, precocious polyadenylation of existing transcripts that are normally dormant until required later in development or a failure of the normal default deadenylation pathway. Experiments are underway to determine the mechanism for the increased gene expression observed for oocytes matured in vitro. We are also proposing to examine the large number of genes expressed more than 2-fold higher in oocytes matured in vitro to determine if there is any sequence similarity in the promoter and non-translated regions that may help to explain why these transcripts are not silenced as in oocytes matured in vivo.

In conclusion, global gene expression profiling using whole human genome arrays and subsequent data mining has provided a molecular basis for the relative developmental incompetence of oocytes matured in vitro compared with in vivo. Immature GV stage oocytes recovered from gonadotrophin stimulated patients and matured in vitro have a large number of genes expressed at more than 2-fold higher levels than oocytes matured in vivo. This is probably due to dysregulation in gene transcription or post-transcriptional modification of genes resulting in an incorrect temporal utilization of transcripts which could manifest as developmental incompetence of any embryos resulting from these oocytes. Similar application of global gene expression profiling using microarrays may reveal the molecular basis for differences in developmental competences of oocytes recovered from women with different etiologies of infertility.

Supplementary Data

Supplementary data are available at http://humrep.oxfordjournals.org/.

Funding

This research was supported by a grant from the National Institutes of Health (U01 HD044778-01) as part of the National Institute of Child Health and Human Development Cooperative Program on Female Health and Egg Quality.

Acknowledgements

The authors wish to thank the embryology staff at Monash IVF, Victoria, Australia, and Dr Eleanora Borghi and Dr Natasha Griaco, SISMER, Bologna and Pieve di Cadore, Italy, for their kind assistance in the collection of human oocytes for the present study. We would also like to thank Ms Joanna McLeod for her assistance with gathering the gene specific literature pertinent to the present study. We would also like to thank MediCult and Mr David Thompson of JDC-bio for the donation of in vitro maturation medium used throughout the present study.

References

Assou
S
Anahory
T
Pantesco
V
Le Carrour
T
Pellestor
F
Klein
B
Reyftmann
L
Dechaud
H
De Vos
J
Hamamah
S
The human cumulus-oocyte complex gene-expression profile
Hum Reprod
 , 
2006
, vol. 
21
 (pg. 
1705
-
1719
)
Bachvarova
R
Browder
LW
Gene expression during oogenesis and oocyte development in mammals
Developmental Biology—A Comprehensive Synthesis
 , 
1985
, vol. 
1
 
New York
Plenum
(pg. 
453
-
524
)
Bachvarova
RF
A maternal tail of poly(A): the long and the short of it
Cell
 , 
1992
, vol. 
69
 (pg. 
895
-
897
)
Bermudez
MG
Wells
D
Malter
H
Munne
S
Cohen
J
Steuerwald
NM
Expression profiles of individual human oocytes using microarray technology
Reprod BioMed Online
 , 
2004
, vol. 
8
 (pg. 
325
-
337
)
Blake
D
Procton
M
Johnson
N
Olive
D
Cleavage stage versus blastocyst stage embryo transfer in assisted conception
Cochrane Database Syst Rev
 , 
2005
, vol. 
4
  
CD002118
Braude
P
Bolton
V
Moore
S
Human gene expression first occurs between the four- and eight-cell stages of preimplantation development
Nature
 , 
1988
, vol. 
332
 (pg. 
459
-
461
)
Cekleniak
NA
Combelles
CMH
Ganz
DA
Fung
J
Albertini
DF
Racowsky
C
A novel system for in vitro maturation of human oocytes
Fertil Steril
 , 
2001
, vol. 
75
 (pg. 
1185
-
1193
)
Chen
S-U
Chen
H-F
Lien
Y-R
Ho
H-N
Chang
H-C
Yang
Y-S
Schedule to inject in vitro matured oocytes may increase pregnancy after intracytoplasmic sperm injection
Arch Androl
 , 
2000
, vol. 
44
 (pg. 
197
-
205
)
De Leon
V
Johnson
A
Bachvarova
R
Half-lives and relative amounts of stored and polysomal ribosomes and poly(A)+ RNA in mouse oocytes
Dev Biol
 , 
1983
, vol. 
98
 (pg. 
400
-
408
)
De Vos
A
Van de Velde
H
Joris
H
Van Steirteghem
A
In-vitro matured metaphase-I oocytes have a lower fertilization rate but similar embryo quality as mature metaphase-II oocytes after intracytoplasmic sperm injection
Hum Reprod
 , 
1999
, vol. 
14
 (pg. 
1859
-
1863
)
DeScisciolo
C
Wright
DL
Mayer
JF
Gibbons
W
Muasher
SJ
Lanzendorf
SE
Human embryos derived from in vitro and in vivo matured oocytes: analysis for chromosomal abnormalities and nuclear morphology
J Assist Reprod Genet
 , 
2000
, vol. 
17
 (pg. 
284
-
292
)
Dobson
AT
Raja
R
Abeyta
MJ
Taylor
T
Shen
S
Haqq
C
Reijo Pera
RA
The unique transcriptome through day 3 of human preimplantation development
Hum Mol Genet
 , 
2004
, vol. 
13
 (pg. 
1461
-
1470
)
Durinzi
KL
Saniga
EM
Lanzendorf
SE
The relationship between size and maturation in vitro in the unstimulated human oocyte
Fertil Steril
 , 
1995
, vol. 
63
 (pg. 
404
-
406
)
Edirisinghe
WR
Junk
SM
Matson
PL
Yovich
JL
Birth from cryopreserved embryos following in-vitro maturation of oocytes and intracytoplasmic sperm injection
Hum Reprod
 , 
1997
, vol. 
12
 (pg. 
1056
-
1058
)
Gasca
S
Pellestor
F
Assou
S
Loup
V
Anahory
T
Dechaud
H
De Vos
J
Hamamah
S
Identifying new human oocyte marker genes: a microarray approach
Reprod BioMed Online
 , 
2007
, vol. 
14
 (pg. 
175
-
183
)
Gene Ontology Consortium
Creating the gene ontology resource: design and implementation
Genome Res
 , 
2001
, vol. 
11
 (pg. 
1425
-
1433
)
Jones
GM
Growth and viability of human blastocysts in vitro
Reprod Med Rev
 , 
2000
, vol. 
8
 (pg. 
241
-
287
)
Jones
GM
Song
B
Cram
DS
Trounson
AO
Optimization of a microarray based approach for deriving representative gene expression profiles from human oocytes
Mol Reprod Dev
 , 
2007
, vol. 
74
 (pg. 
8
-
17
)
Kocabas
AM
Crosby
J
Ross
PJ
Otu
HH
Beyhan
Z
Can
H
Tam
W-L
Rosa
GJM
Halgren
RG
Lim
B
, et al.  . 
The transcriptome of human oocytes
PNAS
 , 
2006
, vol. 
103
 (pg. 
14027
-
14032
)
Lequarre
AS
Traverso
JM
Marchandise
J
Donnay
I
Poly(A) RNA is reduced by half during bovine oocyte maturation but increases when meiotic arrest is maintained with CDK inhibitors
Biol Reprod
 , 
2004
, vol. 
71
 (pg. 
425
-
431
)
Li
SS-L
Liu
Y-H
Tseng
C-N
Singh
S
Analysis of gene expression in single human oocytes and preimplantation embryos
Biochem Biophys Res Comm
 , 
2006
, vol. 
a 340
 (pg. 
48
-
53
)
Li
Y
Feng
H-L
Cao
Y-J
Zheng
G-J
Yang
Y
Mulllen
S
Critser
JK
Chen
Z-J
Confocal microscopic analysis of the spindle and chromosome configurations of human oocytes matured in vitro
Fertil Steril
 , 
2006
, vol. 
b 85
 (pg. 
827
-
832
)
Magli
MC
Ferraretti
AP
Crippa
A
Lappi
M
Feliciani
E
Gianaroli
L
First meiosis errors in immature oocytes generated by stimulated cycles
Fertil Steril
 , 
2006
, vol. 
86
 (pg. 
629
-
635
)
Mendez
R
Kanenganti
GKM
Ryan
K
Manley
JL
Richter
JD
Phosphorylation of CPEB by Eg2 mediates the recruitment of CPSF into an active Cytoplasmic polyadenylation complex
Mol Cell
 , 
2000
, vol. 
6
 (pg. 
1253
-
1259
)
Nagy
ZP
Cecile
J
Liu
J
Loccufier
A
Devroey
P
Van Steirteghem
A
Pregnancy and birth after intracytoplasmic sperm injection of in vitro matured germinal-vesicle stage oocytes: case report
Fertil Steril
 , 
1996
, vol. 
65
 (pg. 
1047
-
1050
)
Nogueira
D
Staessen
C
Van de Velde
H
Van Steirteghem
A
Nuclear status and cytogenetics of embryos derived from in vitro-matured oocytes
Fertil Steril
 , 
2000
, vol. 
74
 (pg. 
295
-
298
)
Paynton
BV
Bachvarova
R
Polyadenylation and deadenylation of maternal mRNAs during oocyte growth and maturation in the mouse
Mol Reprod Dev
 , 
1994
, vol. 
37
 (pg. 
172
-
180
)
Racowsky
C
Kaufman
ML
Nuclear degeneration and meiotic aberrations observed in human oocytes matured in vitro: analysis by light microscopy
Fertil Steril
 , 
1992
, vol. 
58
 (pg. 
750
-
755
)
Tomek
W
Torner
H
Kanitz
W
Comparative analysis of protein synthesis, transcription and cytoplasmic polyadenylation of mRNA during maturation of bovine oocytes in vitro
Reprod Dom Anim
 , 
2002
, vol. 
37
 (pg. 
86
-
91
)
Trounson
AO
Anderiesz
C
Jones
GM
Maturation of human oocytes in vitro and their developmental competence
Reproduction
 , 
2001
, vol. 
121
 (pg. 
51
-
75
)
Tucker
MJ
Wright
G
Morton
PC
Massey
JB
Birth after cryopreservation of immature oocytes with subsequent in vitro maturation
Fertil Steril
 , 
1998
, vol. 
70
 (pg. 
578
-
579
)
Vanhoutte
L
De Sutter
P
Van der Elst
J
Dhont
M
Clinical benefit of metaphase I oocytes
Reprod Biol Endocrinol
 , 
2005
, vol. 
3
 (pg. 
71
-
76
)
Varnum
SM
Wormington
WM
Deadenylation of maternal mRNAs during Xenopus oocyte maturation does not require specific cis-sequences: a default mechanism for translational control
Genes Dev
 , 
1990
, vol. 
4
 (pg. 
2278
-
2286
)
Veeck
LL
Wortham
JWE
Jr
Witmyer
J
Sandow
BA
Acosta
AA
Garcia
JE
Jones
GS
Jones
HW
Jr
Maturation and fertilization of morphologically immature human oocytes in a program of in vitro fertilization
Fertil Steril
 , 
1983
, vol. 
39
 (pg. 
594
-
602
)
Wang
W-H
Keefe
DL
Prediction of chromosome misalignment among in vitro matured human oocytes by spindle imaging with the PolScope
Fertil Steril
 , 
2002
, vol. 
78
 (pg. 
1077
-
1081
)
Wynn
P
Picton
HM
Krapez
JA
Rutherford
AJ
Balen
AH
Gosden
RG
Pretreatment with follicle stimulating hormone promotes the numbers of human oocytes reaching metaphase II by in vitro maturation
Hum Reprod
 , 
1998
, vol. 
13
 (pg. 
3132
-
3138
)
Zhang
P
Kerkela
E
Skottman
H
Levkov
L
Kivinen
K
Lahesmaa
R
Hovatta
O
Kere
J
Distinct sets of developmentally regulated genes that are expressed by human oocytes and human embryonic stem cells
Fertil Steril
 , 
2006
 
doi:10.1210/jc.2006–2123
Zheng
P
Patel
B
McMenamin
M
Moran
E
Paprocki
AM
Kihara
M
Schramm
RD
Latham
KE
Effects of follicle size and oocyte maturation conditions on maternal messenger RNA regulation and gene expression in rhesus monkey oocytes and embryos
Biol Reprod
 , 
2005
, vol. 
72
 (pg. 
890
-
897
)