If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corona cells surround the oocyte and maintain a close relationship through transzonal processes and gap junctions, and may be used to assess oocyte competence. In this study, the corona cell transcriptome of individual cumulus oocyte complexes (COCs) was investigated. Isolated corona cells were collected from COCs that developed into euploid blastocysts and were transferred in a subsequent frozen embryo transfer. Ten corona cell samples underwent RNA-sequencing to generate unique gene expression profiles. Live birth was compared with negative implantation after the transfer of a euploid blastocyst using bioinformatics and statistical analysis. Individual corona cell samples produced a mean of 21.2 million sequence reads, and 307 differentially expressed transcrpits (P < 0.05; fold change ≥2). Enriched pathway analysis showed Wnt signalling, mitogen-activated protein kinases signalling, focal adhesion and tricarboxylic acid cycle to be affected by implantation outcome. The Wnt/beta-catenin signalling pathway, including genes APC, AXIN and GSK3B, were independently validated by real-time quantitative reverse transcription. Individual, corona cell transcriptome was successfully generated using RNA-sequencing. Key genes and signalling pathways were identified in association with implantation outcome after the transfer of a euploid blastocyst in a frozen embryo transfer. These data could provide novel biomarkers for the non-invasive assessment of embryo viability.
). For many infertile couples, treatment will involve IVF and a subsequent embryo transfer. The oocyte is a major contributor to the success of an IVF cycle, with oocyte competence tightly linked to embryo viability. Numerous factors determine oocyte competence, including mitochondrial health, metabolomic processes, functional spindle assembly and successful meiotic division (
). Consequently, additional non-invasive methods to measure implantation outcome could represent a valuable asset to improve embryo selection and IVF outcomes.
Follicular fluid is a potential non-invasive source of information indicating oocyte competence. The fluid is aspirated at the time of oocyte retrieval and discarded after oocyte isolation. Follicular fluid is contained within the follicular antrum surrounding the oocyte, and the composition of this fluid changes throughout folliculogenesis. Protein analysis of follicular fluid revealed diverse protein compositions containing mostly plasma proteins and proteins associated with lipid transport, complement pathways and blood coagulation (
). A recent study identified a subset of 75 follicular fluid proteins that were correlated with IVF outcome. Thirteen of these proteins are involved in acute response signalling, coagulation, prothrombin activation, complement system and growth hormone pathways, uniquely associated with IVF outcome (
). Metabolic analysis of follicular fluid has revealed metabolites that are correlated with oocyte maturity based on the depletion and appearance of specific amino acids: arginine, glutamate, glutamine, isoleucine and valine (
). In addition to amino acids, carbohydrates and fatty acid ratios have been shown to correlate with developmental milestones, including embryogenesis, and form an overall predictive value of oocyte quality and developmental potential (
). Additional studies have linked oocyte quality with cytokines and growth factors in follicular fluid, including increased BMP2, interleukins (6, 8, 12 and 18), GDF-9, granulocyte-colony stimulating factor and amphiregulin (
). Although follicular fluid molecular profiling seems to be promising, to date, no prospective studies using a panel of follicular fluid biomarkers to determine potential clinical value in identifying oocyte competence have been published.
Another promising area of research into oocyte competence has been the molecular analysis of follicular cells. Corona cells surrounding the oocyte maintain a close relationship through transzonal processes and gap junctions, providing key nutrients and other factors essential for oocyte maturation and future developmental competence (
). Follow-up investigations identified 45 corona cell genes of interest to be considered biomarkers of successful embryo development and pregnancy outcomes that were tested in a prospective study. The test group underwent a day-3 embryo transfer based on the 45 corona cell gene expression profiles, whereas the control group underwent a day 3 embryo transfer based on traditional embryo grading selection. Both implantation and ongoing pregnancy rates were significantly higher in the test group after embryos were chosen based upon their corona cell gene expression profile. Of the 267 corona cells samples studied, 27% had a corona cell gene expression profile that predicted positive pregnancy outcome, 42% that predicted negative pregnancy outcome and 13% that predicted early arrested development. Remarkably, no relationship between traditional morphological grading and corona cells gene expression was observed (
). Other retrospective studies have shown a significant correlation between the cumulus transcriptome and subsequent pregnancy outcomes where NRP1, UBQLN1, PSMD6, DPP8, HIST1H4C, CALM1, PTGS2, EFNB2 and CAMK1D gene expression were correlated with pregnancy and TOM1 with negative implantation (
). Recently, the combinatorial expression of 12 corona cell genes involved in glucose metabolism, transcription, gonadotrophin regulation and apoptosis were used to predict pregnancy outcome for 55 patients with 78% accuracy (
). Together, these studies indicate cumulus cells are a potential valuable source of information on oocyte competence. Inconsistency to date in the corona cells transcriptome data, however, may be due to differences in experimental design and methodology, including factors such as maternal age, ovarian stimulation protocols and various in-vitro laboratory protocols.
In the present study, the individual corona cells transcriptome of chromosomally normal (euploid) embryos was investigated using RNA sequencing to identify novel gene transcription associated with implantation outcome. Results demonstrated that the canonical Wnt-signalling pathway and AXIN transcription are strong indicators of oocyte developmental competence and subsequent chromosomally normal live birth.
Materials and methods
Corona cell collection
After routine oocyte retrieval, cumulus cells were mechanically separated from each individual cumulus oocyte complex without disturbing the inner corona cells. Oocyte maturity was visually confirmed with the presence of a polar body and individual oocyte corona complexes were allowed to recover for 4 h in an incubator at 7% CO2, 5% O2 and 37°C. Corona cells were removed before intracytoplasmic sperm injection using a media solution containing 0.33 mg/ml hyaluronidase and rinsed through sequential 3 × 20 µL drops of phosphate buffered saline and bovine serum albumin solution before transfer into 10 µL of extraction buffer (PicoPure® RNA-isolation kit, Life Technologies, Carlsbad CA). All corona cells samples were individually snap-frozen in liquid nitrogen and stored at −80°C until later analysis.
Corona cells were donated from individual cumulus oocyte complexes (n = 10), with patient consent. The study was approved by HCA – HealthONE Institutional Review Board on 5 November 2012 (reference number 231845-7). Female participants (n = 10) presented with normal ovarian reserve (anti-Müllerian hormone >1.0 ng/ml, FSH <10 mIU/ml, antral follicle count >6) and advanced maternal age (36–43 years) but no other infertility diagnosis, including male factor, to minimize potential inter-patient variables. Patients underwent routine IVF with morphologically mature oocytes (second metaphase stage) fertilized by intracytoplasmic sperm injection, and embryos were individually cultured to the blastocyst stage for a trophectoderm biopsy before vitrification. Culture conditions were identical, using established routine laboratory protocols. Only blastocysts graded 3BB or better were chosen for trophectoderm biopsy, based on the Gardner Schoolcraft grading system (
). Single-nucleotide polymorphism microarray-based comprehensive chromosome screening was carried out on biopsied trophectoderm cells. After the identification of a euploid blastocyst, patients were scheduled for a routine frozen embryo transfer (
). Corona samples were subsequently grouped on the basis of implantation outcome (live birth n = 5 and negative implantation n = 5), with negative implantation defined as a complete lack of biochemical pregnancy. Both implantation groups had an equal number of day 5 and day 6 biopsied euploid blastocysts transferred.
Individual corona cell samples were isolated for RNA-sequencing using the PicoPure® RNA Isolation Kit (Life Technologies, Carlsbad CA) with modifications. Briefly, samples were lysed and bound to a silica-based filter, treated with RNase-free DNase I (Qiagen, Valencia CA) and washed several times before being recovered in 20 µL elution solution. Purified cDNA libraries were constructed from 50 ng total RNA, which was rRNA-depleted using the Ion Total RNA-Seq Kit v2, Whole Transcriptome Library Prep protocol (Life Technologies, Carlsbad CA). rRNA-depleted total RNA was fragmented using RNase III and purified by Agencourt Ampure XP micro-beads (Beckman Coulter Inc., Brea CA). The fragmented RNA was then hybridized and ligated before reverse transcription and addition of multiplexing barcode adaptor. The subsequent cDNA was then amplified by polymerase chain reaction and purified again. Barcoded libraries were equalized and pooled into an evenly represented final library consisting of 100 pM cDNA, which was then coupled onto templated capture beads and enriched using the Ion OneTouch 2 and OneTouch ES systems (Life Technologies, Carlsbad CA). Final libraries were prepped for loading on the ION PI v2 chip and then sequenced on the Ion Proton with a P1 200 v2 Sequencing kit (Life Technologies, Carlsbad CA).
RNA-sequencing data and bioinformatic analysis
Raw reads in FASTQ format were trimmed and filtered with FastQC such that only reads with Phred Q-scores over 20 and read lengths over 35 bp were retained. Duplicate reads were removed, thereby retaining unique reads with higher average base quality. Strand NGS v.2.1 (Strand Life Sciences Inc., Aurora CO) was used for subsequent RNA-Seq data analysis. The trimmed and filtered reads were aligned to NCBI RefSeq human reference genome (GRCh37/hg19) and human transcriptome. Transcript isoform assembly, abundance estimation and quantification were carried out with Strand NGS. Differential gene expression was carried out with DESeq v.3.0 normalization (
) and fold change >2.0 (bioconductor.org). Strand NGS identified significantly differentially expressed up- and down-regulated genes using the Benjamini-Hochberg multiple test correction at P < 0.05. Gene annotations were provided by NCBI Entrez Gene database. Biological pathway analysis was carried out with Strand NGS based on WikiPathways Analysis, Reactome, GenMAPP and other pathway databases and the BioCyc pathway database. Significantly enriched pathways were identified with a threshold fold change >2.0.
Additional, independent corona cells samples (n = 10) were collected for quantitative real time polymerase chain reaction (qRT-PCR) to validate the RNA-sequencing data. The PicoPure® RNA Isolation Kit (Life Technologies, Carlsbad CA) was used for RNA isolation as previously described. Reverse transcription was carried out using the High Capacity Reverse Transcription cDNA Kit (Life Technologies, Carlsbad CA) to generate cDNA template for qRT-PCR.
RNA-sequencing was successfully carried out for individual corona cell samples, with a mean number of sequence reads of 21.2 million per individual corona cell sample (range = 10.2 million to 29.8 million reads) to generate unique gene expression profiles in association with implantation outcome. After corona cell gene expression quantification with DESeq normalization, an unpaired t-test was carried out comparing live birth (n = 5) with negative implantation (n = 5), which identified 306 significantly differentially expressed genes (P < 0.05; fold change ≥1.5) between the two groups. Of these 307 differentially expressed corona cell genes, 67 were observed to have increased expression and 240 corona cell genes decreased expression in association with a live birth (Supplementary Figure S1a and S1b). A volcano plot was generated to visualize differences in transcript levels between live birth and negative implantation corona cells samples. Each dot on the volcano plot represents a transcript, red dots indicate transcripts that showed statistically significant differential expression between the two groups of corona cells samples, with P-value on the y-axis. The x-axis represents increased and decreased fold change reflecting differential gene expression observed between live birth and negative implantation (Figure 1). Upon further examination of these differentially expressed genes, 87% were observed to be protein-coding, with the remaining 13% being either pseudo or small regulatory RNAs. Several genes were identified in the RNA-sequencing data that displayed stable expression within each group, and differential expression between groups, including DUSP16, TXNIP and TNFRSF10A. These genes were targeted for validation using qRT-PCR using the additional independent corona cell samples. DUSP16 and TNFRSF10A displayed significantly increased transcription in association with negative implantation (P < 0.05) (Figure 2), and a trend towards increased expression was observed for TXNIP (Figure 2).
Gene ontology analysis revealed significantly enriched biological processes among the differentially expressed transcripts, including response to hormone stimulus, regulation of cellular protein metabolic processes, as well as transcription and transmembrane transport, in association with a live birth (P < 0.05; ≥ two-fold). Pathway analysis of the differentially expressed transcripts identified enriched downstream biochemical signalling pathways. Table 1 shows the enriched signalling pathways with increased differentially expressed transcripts in association with live birth that included Wnt signalling, mitogen-activated protein kinases signalling, focal adhesion and tricarboxylic acid cycle (TCA). In contrast, Table 2 shows the enriched signalling pathways identified with decreased differentially expressed transcripts in association with live birth, including degradation of beta-catenin, apoptosis, DNA damage response and the detoxification of reactive oxygen species.
Table 1Enriched signalling pathways with increased differentially expressed transcripts in association with live birth.
Enriched signalling pathways with increased expression
Examination of the two enriched pathways revealed Wnt signalling to be significantly associated with a live birth outcome, and the degradation of beta-catenin by the destruction complex pathway to be significantly associated with negative implantation. These pathways are linked as components of the larger intricate Wnt-canonical pathway. Further validation was carried out on additional independent corona cells samples that were collected and analysed individually using qRT-PCR for APC, AXIN1 and GSK3B gene transcription relative to RPL19. These three genes are part of the beta-catenin destruction pathway and showed consistent expression within each group and significantly increased expression in association with negative implantation compared with live birth, validating the RNA-sequencing data (P < 0.05) (Figure 3).
In this study, RNA-sequencing technologies were used to generate an individual profile of corona cell gene expression in association with implantation outcome and successful live birth after the transfer of a euploid blastocyst. This is the first study to investigate corona cells gene expression in correlation with the success or failure of a frozen euploid blastocyst upon transfer, thereby controlling for embryonic chromosomes and an unstimulated uterine environment. As expected, embryo morphology is not always a good predictor of live birth and, although many of the embryos in this study had similar blastocyst grades, implantation success was not guaranteed, revealing the need for additional embryo selection methods (Table 3). The ongoing research surrounding corona cell gene expression biomarkers for non-invasive assessment of embryo competence could represent an additional tool for selection.
Table 3Oocyte and embryo morphology with embryo transfer outcome.
In the present study, RNA-sequencing was used to characterize the corona cells transcriptome, and has allowed us to investigate more genes than traditional microarray methods. To validate our approach, qRT-PCR validation was carried out on three genes identified by the RNA-sequencing data: DUSP16, TXNIP and TNFRSF10A. DUSP16 is responsible for encoding a mitogen-activated kinase phosphatase, a subfamily of the dual specificity phosphatase protein. DUSP16 gene expression has been shown to potentially regulate epidermal growth factor signalling, and is associated with oocyte quality (
). TXNIP is a protein coding gene. It is believed to act as an oxidative stress mediator, by limiting the bioavailability of thioredoxin. TXNIP is involved in the crosstalk between the oocyte and its somatic compartment in the preconception period, and has been implicated as a non-invasive, quality biomarker in the corona cells of bovine oocytes (
). TNFRSF10A is a member of the TNF-receptor superfamily. The TNFRSF10A receptor is activated by the tumour necrosis factor-related apoptosis inducing ligand (TNFSF10/TRAIL), which is responsible for inducing apoptosis and cell death. In this study, decreased TNFRSF10A expression levels were observed in relation to live birth. Overexpression of TNFRSF10A is indicative of disrupted cell regulation, which leads to increased apoptosis, and decreased developmental competence. Apoptosis is an essential pathway during folliculogenesis; however, a balance between cell death and division is crucial, with increased levels of apoptosis shown to be associated with impaired oocyte maturation, as well as fertilization and pregnancy outcome (
Differential expression of transcripts in the Wnt-signalling pathway was associated with live birth in our study. This pathway is known to be functionally critical during pre-implantation and implantation (
) (Figure 4). Wnts are extracellular molecules that are secreted locally to exert control over a multitude of developmental processes, including cell-to-cell interactions, differentiation, cell-fate specification and regulation. Wnt signalling is vital for development, and is modulated by cytoplasmic, extracellular and nuclear regulations. Within the Wnt-signalling pathway, WNT4 regulates a number of genes involved in late follicular development, and is required for normal antral follicular development (
In the absence of Wnt stimulus, cytoplasmic beta-catenin is targeted for proteolysis by a large multi-protein assembly called the beta-catenin destruction complex. Additionally, the relative levels of the destruction proteins AXIN, APC and GSK3 are critical for both homeostasis as well as responsiveness to Wnt (
AXIN, APC and GSK3 within the beta-catenin destruction complex pathway were observed to have stable, increased expression in corona cell samples associated with negative implantation outcome. AXIN, which is expressed mostly within the cytoplasm, can function as a negative regulator of the Wnt-signalling pathway and is capable of inducing apoptosis. AXIN acts by recruiting beta-catenin into plasma membranes and promotes the formation of adherens junctions (
). The increase of AXIN observed in corona cell samples with negative implantation may be indicative of a degradation of beta-catenin, causing a reduction in Wnt-signalling, cellular proliferation, cellular migration and an increase in apoptosis.
In this study, individual corona cell samples resulting in negative implantation displayed a significant increase in GSK3B activity, which may be indicative of beta-catenin instability and a reduction of nt signalling. GSK3B, a beta isoform of the GSK3 gene, is an evolutionarily conserved serine-threonine kinase that has been implicated in body pattern formation and energy metabolism and has been shown to interact with AXIN1 (
). The binding of Wnt by a seven trans-membrane domain receptor results in disheveled activation, down regulating GSK3 activity. This down-regulation of GSK3 contributes to the stabilization of beta-catenin, along with several other signalling actions. In the absence of a Wnt signal, GSK3 interacts with beta-catenin, AXIN and APC by phosphorylation of the associated proteins (
APC is a tumour suppressor gene that acts as a negative regulator. APC controls the concentrations of nuclear beta-catenin. It has been suggested that APC may actively protect beta-catenin from dephosphorylation by PP2A (
Additional enriched corona cell signalling pathways in association with a live birth included the mitogen-activated protein kinase (MAPK) signalling pathway, which moderates multiple cellular processes including differentiation, growth and proliferation (
). In oocytes, the MAPK signalling pathway is associated with maintaining a metaphase II arrest, although recently it has been discovered that when it is inhibited during the transition from meiosis I to meiosis II, meiosis I is accelerated, leading to an increase in aneuploidy in the metaphase II stage of development (
). The focal adhesion pathway similarly showed increased expression in corona cells, resulting in live birth. This pathway is responsible for the transmission of mechanical force and regulatory signalling between extracellular matrices and cells (
). The focal adhesion pathway has been linked to oocyte maturation, thereby impacting oocyte developmental competence, and is partially involved in the communication between corona cells and the developing oocyte (
). In corona cells, disruption of the MAPK/focal adhesion signalling pathways may affect critical processes involved in corona-oocyte communication, which may result in a less competent oocyte and embryo.
Finally, the TCA cycle and respiratory electron transport pathway were observed to have increased corona cell gene expression in association with a live birth. These pathways are responsible for generating energy in eukaryotic cells in the form of adenosine triphosphate. Energy production can be glucose derived, or extracellular pyruvate may be metabolized by the mitochondria, producing most of the adenosine triphosphate found in the oocyte (
). In the mouse, the oocyte controls intercellular metabolomic cooperation for energy production through the TCA cycle between the corona cells and the developing oocyte. Sufficient energy production during development is crucial for oocyte competence (
In conclusion, this study has demonstrated that RNA-sequencing technologies can successfully identify a unique, individual, corona cell transcriptome profile. The strength of this study is highlighted by the transfer of known euploid blastocysts, cultured and transferred under identical conditions, into an unstimulated uterine environment, specifically in relation to true negative implantation. Ongoing research to elucidate the biological and signalling pathways of corona cells associated with oocyte competence could contribute to advancements in embryo selection. A corona cell transcription assay developed for embryo viability, used in conjunction with advanced morphological algorithms and comprehensive chromosome screening, could result in overall improved IVF outcomes, and the routine use of single embryo transfers.
Appendix. Supplementary material
The following is the supplementary data to this article:
Jason Parks leads the Embryology Team at the National Foundation for Fertility Research. He is currently pursuing his PhD in genetics through the University of Kent in the UK. Jason's research is focused on the window of implantation, the critical time when an embryo implants into the uterus and non-invasive assays for IVF.
Published online: February 24, 2016
Received in revised form:
Declaration: The authors report no financial or commercial conflicts of interest.