The impact of maternal age on gene expression during the GV to MII transition in euploid human oocytes.

STUDY QUESTION: Are there age-related differences in gene expression during the germinal vesicle (GV) to metaphase II (MII) stage transition in euploid human oocytes? SUMMARY ANSWER: A decrease in mitochondrial-related transcripts from GV to MII oocytes was observed, with a much greater reduction in MII oocytes with advanced age. WHAT IS KNOWN ALREADY: Early embryonic development is dependent on maternal transcripts accumulated and stored within the oocyte during oogenesis. Transcriptional activity of the oocyte, which dictates its ultimate developmental potential, may be influenced by age and explain the reduced competence of advanced maternal age (AMA) oocytes compared with the young maternal age (YMA). Gene expression has been studied in human and animal oocytes; however, RNA sequencing could provide further insights into the transcriptome profiling of GV and in vivo matured MII euploid oocytes of YMA and AMA patients. STUDY DESIGN, SIZE, DURATION: Fifteen women treated for infertility in a single IVF unit agreed to participate in this study. Five GV and 5 MII oocytes from 6, 21-26 years old women (YMA cohort) and 5 GV and 6 MII oocytes from 6, 41-44 years old women (AMA cohort) undergoing IVF treatment were donated. The samples were collected within a time frame of 4 months. RNA was isolated and deep sequenced at the single-cell level. All donors provided either GV or MII oocytes. PARTICIPANTS/MATERIALS, SETTING, METHODS: Cumulus dissection from donated oocytes was performed 38 h after hCG injection, denuded oocytes were inserted into lysis buffer supplemented with RNase inhibitor. The samples were stored at -80°C until further use. Isolated RNA from GV and MII oocytes underwent library preparation using an oligo deoxy-thymidine (dT) priming approach (SMART-Seq v4 Ultra Low Input RNA assay; Takara Bio, Japan) and Nextera XT DNA library preparation assay (Illumina, USA) followed by deep sequencing. Data processing, quality assessment and bioinformatics analysis were performed using source-software, mainly including FastQC, HISAT2, StringTie and edgeR, along with functional annotation analysis, while scploid R package was employed to determine the ploidy status. MAIN RESULTS AND THE ROLE OF CHANCE: Following deep sequencing of single GV and MII oocytes in both YMA and AMA cohorts, several hundred transcripts were found to be expressed at significantly different levels. When YMA and AMA MII oocyte transcriptomes were compared, the most significant of these were related to mitochondrial structure and function, including biological processes, mitochondrial respiratory chain complex I assembly and mitochondrial translational termination (false discovery rate (FDR) 6.0E-10 to 1.2E-7). These results indicate a higher energy potential of the YMA MII cohort that is reduced with ageing. Other biological processes that were significantly higher in the YMA MII cohort included transcripts involved in the translation process (FDR 1.9E-2). Lack of these transcripts could lead to inappropriate protein synthesis prior to or upon fertilisation of the AMA MII oocytes. LARGE SCALE DATA: The RNA sequencing data were deposited in the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo), under the accession number: GSE164371. LIMITATIONS, REASONS FOR CAUTION: The relatively small sample size could be a reason for caution. However, the RNA sequencing results showed homogeneous clustering with low intra-group variation and five to six biological replicates derived from at least three different women per group minimised the potential impact of the sample size. WIDER IMPLICATIONS OF THE FINDINGS: Understanding the effects of ageing on the oocyte transcriptome could highlight the mechanisms involved in GV to MII transition and identify biomarkers that characterise good MII oocyte quality. This knowledge has the potential to guide IVF regimes for AMA patients. STUDY FUNDING/COMPETING INTEREST(S): This work was supported by the Medical Research Council (MRC Grant number MR/K020501/1).

cDNA cloning and expression of the human NOBOX gene in oocytes and ovarian follicles.

Nobox is a homeobox gene that is preferentially expressed in the oocytes and is essential for folliculogenesis and the regulation of oocyte-specific gene expression in the mouse. The likely human homologue has been identified in silico but has not as yet been confirmed experimentally. Here, we present the first cDNA cloning and transcript expression analysis of the human NOBOX gene. Using RT-PCR, we reveal that expression within adult human tissues is limited to the ovary, testis and pancreas. Expression within the ovary is oocyte specific, with expression observed from the primordial stage ovarian follicle through to the metaphase II (MII) oocyte. In complementary studies, we reveal dynamic expression profiles of 14 additional homeobox genes throughout human oogenesis and early development. The expression of HOXA10 is restricted to primordial and early primary follicles. HOXB7 is expressed from primordial and early primary stage follicles through to germinal vesicle (GV) oocytes. Gastrulation brain homeobox 1 (GBX1) and HOXA7 genes are homeobox markers preferentially expressed by GV oocytes. HOXA1 and HEX are homeobox markers preferentially expressed by MII oocytes. In summary, the homeobox gene transcripts that are detected in ovarian follicles and oocytes are distinct from those expressed in human blastocysts (HOXB4, CDX2 and HOXC9) and granulosa cells (HOXC9, HOXC8, HOXC6, HOXA7, HOXA5 and HOXA4).

Expression of Polycomb-group genes in human ovarian follicles, oocytes and preimplantation embryos.

Mammalian oocytes possess unique properties with respect to their ability to regulate and reprogram chromatin structure and epigenetic information. Proteins containing the conserved chromodomain motif that is common to the Polycomb-group (Pc-G) proteins and the heterochromatin-associated protein HP1, play essential roles in these processes and more specifically, in X-chromosome inactivation in female zygotes and extra-embryonic tissues and in the regulation of genomic imprinting. To characterize the potential role of these proteins in the regulation of epigenetic events during early human development, we utilized a degenerate PCR priming assay to assess the expression of mRNAs of chromodomain proteins in cDNA samples derived from the human female germline and preimplantation embryos. Expression of mRNAs of HP1 genes was observed in ovarian follicles, (HP1 (HSalpha), HP1 (HSbeta), HP1 (HSgamma)), mature oocytes (HP1 (HSalpha), HP1 (HSbeta)), cleavage stage preimplantation embryos (HP1 (HSalpha), HP1 (HSbeta), HP1 (HSgamma)) and blastocysts (HP1 (HSalpha), HP1 (HSgamma)). Transcripts for three Pc-G genes, which are essential for early mammalian development (Yin Yang 1 (YY1), Enhancer of Zeste-2 (EZH2) and Embryonic Ectoderm Development (EED)) and that are essential for the regulation of X-inactivation and certain imprinted genes (EED) were revealed by gene-specific-PCR expression analysis of human ovarian follicles, oocytes and preimplantation embryos. YY1 and EZH2 transcripts were additionally detected in metaphase II oocytes.

Expression of mRNAs for DNA methyltransferases and methyl-CpG-binding proteins in the human female germ line, preimplantation embryos, and embryonic stem cells.

Recent evidence indicates that mammalian gametogenesis and preimplantation development may be adversely affected by both assisted reproductive and stem cell technologies. Thus, a better understanding of the developmental regulation of the underlying epigenetic processes that include DNA methylation is required. We have, therefore, monitored the expression, by PCR, of the mRNAs of DNA methyltransferases (DNMTs), methyl-CpG-binding domain proteins (MBDs), and CpG binding protein (CGBP) in a developmental series of amplified cDNA samples derived from staged human ovarian follicles, oocytes, preimplantation embryos, human embryonic stem (hES) cells and in similar murine cDNA samples. Transcripts of these genes were detected in human ovarian follicles (DNMT3A, DNMT3b1, DNMT3b4, DNMT1, MDBs1-4, MeCP2, CGBP), germinal vesicle (GV) oocytes (DNMT3A, DNMT3b1, DNMT1, MDBs1-4, MeCP2, CGBP), mature oocytes (DNMT3A, DNMT3b1, DNMT1, CGBP), and preimplantation embryos (DNMT3A, DNMT3b1, DNMT1, DNMT3L, MBD2, MDB4, CGBP). Differential expression of DNMT3B gene transcripts in undifferentiated (DNMT3b1) and in vitro differentiated human ES cells (DNMT3b3) further demonstrated an association of the DNMT3b1 transcript variant with totipotent and pluripotent human cells. Significantly, whilst the murine Dnmt3L gene is both expressed and essential for imprint establishment during murine oogenesis, transcripts of the human DNMT3L gene were only detected after fertilisation. Therefore, the mechanisms and/or the timing of imprint establishment may differ in humans.

Growth and maturation of oocytes in vitro.

The development of technologies to grow and mature oocytes from the most abundant primordial follicles holds many attractions for clinical practice, animal production technology and research. However, despite much research attention, it has proved difficult to grow follicles from early stages to maturity in vitro, as relatively little is known about the biology of oogenesis. It is clear that throughout oocyte development in vivo, follicle cell support is fundamental to provide the germ cell with nutrients and growth regulators to ensure progression through the protracted growth phase. Conversely, the oocyte actively promotes growth and differentiation of the follicular cells. Both of these characteristics must be mimicked in vitro. Replication of the normal follicular growth span from the primordial to Graafian follicle stages and the changes in the trophic requirements of the cells, cellular interactions, morphogenesis and the sheer increase in bulk as the antrum forms present major challenges for follicle culture technology. These observations could explain why methods that have proved successful for the culture of isolated rodent follicles are unable to support the growth of larger human and ruminant follicles in vitro and are incompatible with the requirements for primordial follicle growth activation. At present, the best option available for the complete growth and maturation of oocytes in vitro is to develop an extended multistage culture strategy which will provide a complex support system that closely resembles the ovary in vivo. In an attempt to achieve this goal primordial follicle growth is first initiated and maintained to the preantral stages through the culture of thin slices of ovarian cortex. The isolation and continued culture of these preantral follicles will support antral cavity formation and the induction of differentiated function in the somatic cell compartment. Finally, after exposure to an appropriate steroid milieu in vitro it should be possible to induce nuclear and cytoplasmic maturation in the fully grown oocytes. The prospects of succeeding at each stage, and of finally producing a fertile gamete, are likely to be increased by preserving cellular interactions and the phenotype of follicle cells as these provide the physiological environment in which oocytes develop. Although the technology for the in vitro maturation (IVM) of fully grown oocytes has been exploited successfully in ruminants, in human assisted reproduction IVM is still experimental as the efficiency of IVM is low and only a small number of pregnancies and live births have been reported. Thus, although complete in vitro growth and maturation may be achieved eventually, immediate goals must include the optimization of methods for isolating and culturing oocytes at both ends of the size spectrum and the full evaluation of the normality of the oocytes grown for extended periods in vitro.

Lack of involvement of known DNA methyltransferases in familial hydatidiform mole implies the involvement of other factors in establishment of imprinting in the human female germline.

BACKGROUND: Differential methylation of the two alleles is a hallmark of imprinted genes. Correspondingly, loss of DNA methyltransferase function results in aberrant imprinting and abnormal post-fertilization development. In the mouse, mutations of the oocyte-specific isoform of the DNA methyltransferase Dnmt1 (Dnmt1o) and of the methyltransferase-like Dnmt3L gene result in specific failures of imprint establishment or maintenance, at multiple loci. We have previously shown in humans that an analogous inherited failure to establish imprinting at multiple loci in the female germline underlies a rare phenotype of recurrent hydatidiform mole. RESULTS: We have identified a human homologue of the murine Dnmt1o and assessed its pattern of expression. Human DNMT1o mRNA is detectable in mature oocytes and early fertilized embryos but not in any somatic tissues analysed. The somatic isoform of DNMT1 mRNA, in contrast, is not detectable in human oocytes. In the previously-described family with multi-locus imprinting failure, mutation of DNMT1o and of the other known members of this gene family has been excluded. CONCLUSIONS: Mutation of the known DNMT genes does not underlie familial hydatidiform mole, at least in the family under study. This suggests that trans-acting factors other than the known methyltransferases are required for imprint establishment in humans, a concept that has indirect support from recent biochemical studies of DNMT3L.

Isolation, characterization and expression of the human Factor In the Germline alpha (FIGLA) gene in ovarian follicles and oocytes.

The Factor In the Germline alpha (FIGalpha) transcription factor regulates expression of the zona pellucida proteins ZP1, ZP2 and ZP3 and is essential for folliculogenesis in the mouse. Using the published mouse Figla sequence, BLAST searches identified a human chromosome 2 BAC clone with high sequence identity. Using PCR primers derived from this clone, amplicons derived from ovarian follicles and mature oocytes revealed 100% identity with the appropriate human BAC clone, the expected homology with the mouse Figla gene sequence, and homology on translation with the FIGalpha protein identified in the Japanese rice fish, medaka (Oryzias latipes). PCR expression profiling of this transcript revealed FIGLA mRNA expression in cDNA derived from ovarian follicles (5/5 samples from the primordial through to the secondary stage) mature oocytes (6/9 samples), and less frequently in preimplantation embryos (2/7 samples). Subsequent BLAST searches revealed the predicted full length coding sequence of the human FIGalpha protein which demonstrates 68 and 25% similarity overall to mouse and medaka proteins respectively, with 96 and 57% identity respectively within the basic helix-loop-helix region. This confirms our identification of the human homologue for this gene which maps to chromosome 2p12. Further work is required to understand its role in normal human oocyte development and the potential involvement in human infertility.

The use of amplified cDNA to investigate the expression of seven imprinted genes in human oocytes and preimplantation embryos.

Imprinted genes are characterized by expression of only one of the two alleles according to its inheritance from the mother or the father. This mono-allelic expression must arise from primary differential epigenetic modification of the parental alleles of the imprinted gene in the spermatozoon and the oocyte. Most of the information on the onset of imprinted gene expression, and on the molecular mechanisms regulating mono-allelic expression, have been derived from studies in the mouse. In this paper, we investigate the expression of seven imprinted genes in human preimplantation development. Due to limitations imposed by the rarity of human embryos available for research, our approach has been to screen amplified cDNA preparations prepared from human unfertilized oocytes and individual embryos at each of the 4-cell, 8-cell and blastocyst stages. Gene-specific primers were used to investigate expression of the imprinted genes by polymerase chain reaction (PCR) analysis of these amplified cDNA. We found that expression is inherently variable in the amplified cDNA from embryo to embryo but the use of several samples at each stage showed that the SNRPN, UBE3A and PEG1 genes are expressed throughout human preimplantation development. This was confirmed by direct analysis by gene-specific reverse transcription-PCR on a limited number of lysed embryos (one gene analysed per embryo). Thus, the amplified cDNA may be used to rapidly identify those imprinted genes expressed in preimplantation development and, hence, those genes amenable to investigation of the epigenetic mechanisms regulating mono-allelic expression. Confirmation of preimplantation expression also identifies those imprinted diseases amenable to preimplantation diagnosis, and the imprinted genes which may be used in assessment of possible perturbations of imprinting following new procedures in assisted reproduction. Our series of single embryo amplified cDNA are established as a valuable resource for comparative studies of gene expression within one embryo and between embryos throughout early human development. The amplified cDNA thus circumvent the need for a continuous supply of human embryos for studies on embryonic gene expression.

Imprinted expression of SNRPN in human preimplantation embryos.

Prader-Willi syndrome (PWS) and Angelman syndrome (AS) are two clinically distinct neurogenetic disorders arising from a loss of expression of imprinted genes within the human chromosome region 15q11-q13. Recent evidence suggests that the SNRPN gene, which is defective in PWS, plays a central role in the imprinting-center regulation of the PWS/AS region. To increase our understanding of the regulation of expression of this imprinted gene, we have developed single-cell-sensitive procedures for the analysis of expression of the SNRPN gene during early human development. Transcripts of SNRPN were detected in human oocytes and at all stages of preimplantation development analyzed. Using embryos heterozygous for a polymorphism within the SNRPN gene, we showed that monoallelic expression from the paternal allele occurs by the 4-cell stage. Thus, the imprinting epigenetic information inherited in the gametes is recognized already in the preimplantation embryo. The demonstration of monoallelic expression in embryos means that efficient preimplantation diagnosis of PWS may be made by analysis for the presence or absence of SNRPN mRNA.

A methylation-dependent DNA-binding activity recognising the methylated promoter region of the mouse Xist gene.

Differential methylation of CpG sites in the promoter region of the mouse Xist gene is correlated with Xist expression and X-chromosome inactivation in the female. Using oligonucleotides encompassing the differentially methylated sites as probes in band-shift assays, we have identified a nuclear protein which binds to a specific region of the promoter (between base pairs -45 and -30 upstream from the transcription start site) only when CpG sites within the CG rich region (GCGCCGCGG, -44 to -36) are methylated. Competition experiments with methylated or unmethylated heterologous oligonucleotides demonstrate that the activity is sequence-specific as well as methylation-dependent. Analysis by Southwestern blot identifies a protein of approximately 100 kDa molecular weight and confirms strong binding to the methylated Xist promoter oligonucleotide. Using a 233bp Xist-promoter luciferase construct in which the cytosines in the three CpG sites in the -44 to -36 region are mutated to thymine, we have established that this region is required for transcription from the mouse Xist promoter. Therefore, we suggest that the binding of the 100kDa protein to the methylated sequence leads to repression of transcription from the methylated Xist allele, thus suggesting a role in the regulation of both imprinted and random Xist transcription and X-chromosome inactivation.

Mice lacking Snrpn expression show normal regulation of neuronal alternative splicing events.

The SmN protein is closely related to the constitutively expressed RNA splicing protein SmB but is expressed only in brain and heart tissue. Mice which lack expression of SmN die shortly after birth suggesting a critical role for this protein possibly in the regulation of neuronal-specific alternative splicing events. We show here however that the neuronal-specific alternative splicing of the RNAs encoding several different classes of protein proceeds normally in mice lacking SmN expression. The potential role of SmN and the reasons for the lethal effect observed in non-expressing mice are discussed.

Antagonistic effects of retinoic acid and thyroid hormone on the expression of the tissue-specific splicing protein SmN in a clonal cell line derived from rat heart.

The SmN protein is a tissue-specific splicing factor which is expressed only in adult heart and brain but not in other tissues. Although a role for SmN in modulating tissue specific alternative splicing decisions in the heart and brain has been suggested, its precise function and the processes regulating its expression remain unclear. We show here that SmN is expressed in the H9c2 clonal cell line derived from rat heart and that its expression in these cells is affected by a variety of members of the steroid/thyroid hormone family. In particular thyroid hormone and retinoic acid exhibit antagonistic effects on SmN expression with thyroid hormone treatment producing large increases in expression whilst retinoic acid treatment virtually abolishes expression. These findings are discussed in terms of the regulation of SmN expression and the potential role of this factor in the response of cardiac cells to members of the steroid/thyroid hormone family.

Enhanced transcription of the gene encoding the SmN autoantigen in patients with systemic lupus erythematosus does not result in enhanced levels of the SmN protein.

The SmN, protein is closely related to the constitutively expressed SmB and SmB' autoantigens and can also act as a target for human autoimmune sera. In contrast to the single gene encoding SmB and SmB' which is expressed in all tissues, the distinct gene encoding SmN is expressed at high levels only in brain and heart tissue. We show that the SmN gene is transcribed at significantly elevated levels in peripheral blood mononuclear cells (PBMCs) from SLE patients compared to normal controls. In contrast no significant elevation in transcription of the genes encoding SmB/B' or the U1-associated 70kD RNP autoantigen is observed in these patients. The elevation in SmN gene transcription in patient PBMCs does not result however, in enhanced levels of the SmN protein in the PBMCs of these patients. The significance of transcriptional and post-transcriptional processes in regulating the expression in SLE patients of SmN and other autoantigens is discussed.

The snRNP core protein SmB and tissue-specific SmN protein are differentially distributed between snRNP particles.

The SmN protein is a tissue specific component of the small nuclear ribonucleoprotein particle which is closely related to the ubiquitously expressed SmB protein but is expressed only in the brain and heart. To investigate the function of SmN, its localisation within different snRNP particles was investigated using a range of anti-snRNP monoclonal antibodies. SmN and SmB were found to exhibit different patterns of association with snRNP particles in two cell lines, ND7 and F9 which express SmN. In both cases, SmN was found to be present in the U-2 snRNP but was excluded from the U-1 snRNPs whereas SmB was present in both U-1 and U-2 snRNPs. Data from transfected 3T3 mouse fibroblasts cell lines artificially expressing a low level of SmN also confirm this observation. In contrast, SmN was found to be an integral component of both the U-1 and U-2 snRNPs in both 3T3 cells artificially expressing high levels of SmN and in adult rat brain which has a naturally high level of SmN expression. Taken together, the results suggest that the pre-U1 snRNP particle has a lower affinity for SmN than for SmB. Thus, SmN expressed at low levels incorporates into U2, but SmN expressed at high levels incorporates into both U1 and U2 snRNPs and replaces SmB. The significance of these effects is discussed in terms of the potential role played by SmN in constitutive and alternative splicing pathways in neuronal cells.

Lupus autoantibodies discriminate between the highly homologous Sm polypeptides B/B' and SmN by binding an epitope restricted to B/B'.

The ubiquitous Sm polypeptides B/B' (28 and 29 kD) and the highly homologous tissue-specific Sm N polypeptide (29 kD) share several autoepitopes recognized by systemic lupus erythematosus (SLE) sera. Previous studies on the antigenicity of nuclear antigens recognized by human autoantibodies have not discriminated between ubiquitous and tissue-specific forms. We set out to examine whether a tissue-specific nuclear antigen, Sm N, is autoantigenic in SLE by comparing the immunoreactivity of the most unique sequences in this polypeptide. Synthetic peptides from the two regions of least sequence homology that occur between Sm N and Sm B/B', a dodecamer (amino acid residues 179-190 containing five substitutions) and an undecamer (residues 203-213 containing four substitutions) were coupled to a carrier protein. These conjugates were used to quantify IgG anti-peptide antibodies in sera from patients with SLE. Of 43 sera with anti-Sm specificity, six bound to the B/B' 179-190 peptide but not to the N version. None of 17 anti-Sm-negative SLE sera bound these peptides. The second region of least sequence homology between N and B/B' (203-213) was not antigenic. Our data suggest that a subset of SLE patients with anti-Sm reactivity have IgG autoantibodies capable of discriminating between Sm N and SmB/B' polypeptides by binding a previously unreported SmB/B'-specific autoepitope. The data also indicate that brain and heart-specific anti-Sm antibodies do not exist in SLE sera, suggesting that these tissues do not participate in the induction or maintenance of the autoimmune anti-Sm response.