|Home | About | Journals | Submit | Contact Us | Français|
Rhesus macaque and human preimplantation embryos display similar rates of chromosomal abnormalities. The aim of this study was to determine whether embryos developing from MI oocytes that mature post-retrieval display more chromosomal anomalies than those embryos that are generated from oocytes that are at MII at the time of retrieval.
Rhesus macaque oocytes were obtained after hormonal ovarian stimulation. Immediately after retrieval, the oocytes were classified according to their maturational status. Following in vitro fertilization, Day 3 embryos with good morphology and development derived from oocytes maturing post-retrieval and those from oocytes that were mature at the time of retrieval were cytogenetically assessed using a five-color fluorescent in situ fluorescent hybridization assay developed for rhesus macaque chromosomes homologous to human chromosomes 13, 16, 18, X and Y.
Blastomeres from 53 embryos were analyzed. Of the 27 embryos that developed from oocytes that were mature at collection, 18 embryos were chromosomally normal (66.7%), while from the 26 embryos that developed from oocytes that matured post-retrieval, only 9 embryos were chromosomally normal (34.6%).
These results indicate that embryos developing from oocytes maturing post-retrieval display high rates of chromosomal abnormalities and have therefore a reduced developmental competence. As a result, the clinical relevance of using immature oocytes that are retrieved after stimulated cycles in human IVF warrants further investigation.
Typically, during human assisted reproductive technology (ART), 15–27% of oocytes fail to resume maturation or to reach the mature (MII) stage within 36 h after the administration of human chorionic gonadotrophin (hCG) (Junca et al., 1995; De Vos et al., 1999). Immature oocytes may be retrieved at the germinal vesicle (GV) stage or may have undergone nuclear membrane breakdown without extrusion of the first polar body.
After a period of in vitro culture, some immature oocytes can extrude a polar body and subsequently may be used for in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI). Most studies reveal that immature human MI and GV oocytes maturing overnight or even within a few hours have lower fertilization rates in comparison to oocytes that were mature at retrieval (De Vos et al., 1999; Huang et al., 1999; DeScisciolo et al., 2000; Nogueira et al., 2000; Balakier et al., 2004; Strassburger et al., 2004; Shu et al., 2007). Embryos originating from human immature GV and MI oocytes that were incubated overnight to attain maturity are of inferior quality in comparison to those resulting from oocytes that were mature at the time of retrieval (DeScisciolo et al., 2000; Nogueira et al., 2000). These poor quality embryos exhibit a significant amount of multinucleation and chromosomal abnormalities (DeScisciolo et al., 2000; Nogueira et al., 2000; Emery et al., 2005). The preceding morphological and cytogenetic observations, however, cannot be extrapolated to embryos that have developed from MI oocytes, which matured within a few hours after retrieval. In many cases, these immature MI oocytes from stimulated human IVF cycles can be fertilized, develop into ‘morphologically normal’ embryos (Junca et al., 1995; De Vos et al., 1999; Huang et al., 1999; Strassburger et al., 2004) but nevertheless generally lead to reduced pregnancy and live birth rates (De Vos et al., 1999; Vanhoutte et al., 2005). It is important to point out that, since these embryos develop within the same time frame as embryos resulting from oocytes that were mature at the time of collection and may be morphologically indistinguishable from one another, they may be selected for transfer. It may be questioned whether embryos with ‘good morphology’, originating from MI oocytes that matured shortly after retrieval, have low implantation rates (De Vos et al., 1999; Vanhoutte et al., 2005) because of significant incidences of chromosomal abnormalities.
Because this question is difficult to address in humans, a suitable animal model should be employed. Non-human primates more closely resemble humans in anatomy and physiology than do commonly used laboratory animals, and are therefore better clinical models for reproductive research. Humans and rhesus macaques have similar menstrual cycles and utilize comparable ART procedures to generate embryos. Furthermore, because in vitro-produced human and rhesus macaque preimplantation embryos display similar chromosomal error rates (Dupont et al., 2008), the rhesus macaque has been established as an optimal model to assess cytogenetic issues in human preimplantation embryos. The aim of this study was to determine whether rhesus macaque embryos with good morphology originating from oocytes maturing post-retrieval display more chromosomal abnormalities than morphologically equivalent embryos derived from sibling MII oocytes that were mature at the time of retrieval. These cytogenetic data may provide insights into the etiology of the lower implantation and pregnancy rates for human embryos originating from oocytes maturing within a few hours after retrieval (De Vos et al., 1999; Vanhoutte et al., 2005).
Unless otherwise acknowledged, all chemicals and reagents were obtained from Sigma-Aldrich (St Louis, MO, USA).
All procedures were performed according to the institutional animal care and use committee (IACUC) protocols approved by the Oregon National Primate Research Center (ONPRC). The protocol for ovarian stimulation has been previously described (Zelinski-Wooten et al., 1995). In brief, 10 female rhesus macaques received sequential injections of recombinant human gonadotrophins. Young female macaques between 4 and 13 years of age were treated with recombinant human follicle-stimulating hormone (rhFSH, Organon, Oss, The Netherlands; 30 IU per injection, twice daily, i.m.) for 7 days, followed by 2 days of rhFSH and recombinant human luteinizing hormone (rhLH, EMD Serono, Rockland, MA, USA; 30 IU rhFSH and 30 IU rhLH per injection, twice daily, i.m.). To prevent a spontaneous LH surge, the gonadotrophin-releasing hormone (GnRH) antagonist acycline (0.075 mg/kg animal body weight) was added to the gonadotrophin injections on the final 3 days of the rhFSH administration. Oocyte maturation was induced by a single injection of recombinant human hCG (rhCG; 750–1000 IU, i.m.) on the last day of rhFSH injections. Follicular fluids were aspirated 36 h after the rhCG administration.
Procedures for the collection of male and female gametes, insemination and embryo culture have been described previously (Bavister et al., 1983,1984; Wolf et al., 1989). Aspirated follicular contents were collected in TALP-Hepes medium containing 0.3% bovine serum albumin (BSA). Oocytes were retrieved from the follicular fluid using a cell filter (Becton-Dickinson, Franklin Lakes, NJ, USA; Falcon, 70 µm pore size). Immediately after collection, cumulus cells were removed using 0.03% hyaluronidase and the oocytes were rinsed and classified according to their maturational status (MII, MI and GV). After classification, oocytes were placed in TALP medium culture drops supplemented with BSA (Bavister and Yanagimachi, 1977) and incubated in 5% CO2 in air at 37°C until insemination. Semen was obtained by electroejaculation (Mastroianni and Manson, 1963). Spermatozoa, separated from seminal plasma using the standard protocol, were activated with cyclic AMP and caffeine (1 mM each) and used for insemination at a final concentration of 1.5–2.0 105 sperm/ml (Bavister et al., 1983; Boatman and Bavister, 1984). The maturity of the oocytes was re-evaluated at insemination. Eight to ten hours after retrieval all oocytes (MII and MI at retrieval), except for oocytes at the GV stage, were inseminated. Oocytes at the GV stage were collected for other experiments. The presence of pronuclei and polar bodies was assessed 18–20 h post-insemination. Presumptive zygotes were cultured in amino acid-supplemented HECM-9 medium drops (McKiernan and Bavister, 2000) and incubated under a 5% CO2 in air atmosphere at 37°C.
Only ‘morphologically good’ embryos derived from oocytes that were successfully fertilized 18–20 h after insemination were included in this study. Within the cohort of oocytes retrieved from each female, similar numbers of embryos resulting from oocytes that were mature at retrieval or maturing post-retrieval were analyzed. Embryos with an equivalent cleavage rate were selected between 54 and 78 h after insemination. After removal of the zona pellucida using 1 mg/ml pronase in TL-Hepes containing 1 mg/ml polyvinyl alcohol, zona-free embryos were dissociated in calcium/magnesium-free phosphate-buffered saline (PBS, Invitrogen, San Diego, CA, USA). The individual blastomere nuclei were fixed on a slide using 0.01 M HCl/0.1% Tween 20 (Coonen et al., 1994). Following fixation, slides were dehydrated in an ethanol series and stored at −20°C until used for analysis.
Fluorescent in situ hybridization (FISH) probes identifying macaque chromosomes X, Y, 17, 18 and 20 (homologous to human chromosomes X, Y, 13, 18 and 16, respectively) were created using previously described pooled human bacterial artificial chromosomes (BAC) (Froenicke et al., 2007; Dupont et al., 2008). BAC DNA from E. coli cultures was extracted using the PhasePrep BAC DNA kit. Following the DNA extraction, BAC DNA was pooled and directly labeled by nick translation. The nick translation mixture consisted of a 1× enzyme mix from a BioNick translation kit (Invitrogen), nucleotides (0.02 mM each of dATP, dCTP and dGTP and 0.01 mM dTTP, Promega, Madison, WI, USA), 50 mM reaction buffer (Tris–HCl pH 7.8, 5 mM MgCl2, 10 mM β-mercaptoethanol, 10 µg/ml nuclease-free BSA) and 0.01 mM fluorophore labeled dUTPs (Alexa488-dUTP, Invitrogen; Cy3-dUTP, Amersham-GE Healthcare, Piscataway, NJ, USA). The nick-translation was carried out for 2 h at 16°C. Immediately after the reaction, nick-translation enzymes were inactivated by incubating the reaction mixtures for 10 min at 65°C. Differentially labeled probes were combined, co-precipitated with Cot-1 DNA and ssDNA (Froenicke et al., 2002) and subsequently resuspended in hybridization solution (50% deionized formamide, 10% dextran sulphate, in 2× SSC).
The complete FISH assay was performed in two successive hybridization experiments. After thawing, slides were placed in ice-cold Carnoy’s fixative for 1 min, rinsed in PBS and water, dehydrated in an ethanol series and air-dried. Probe mixtures were applied on designated areas of the slide. After sealing of the targeted areas with a cover slip and rubber cement, slides were denatured for 5 min at 75°C and incubated overnight at 37°C. Subsequent to the hybridization, cover slips were removed and slides were rinsed twice for 5 min in 0.05× sodium chloride-sodium citrate buffer (SSC) at 39°C and then once for 2 min in 2× SSC at room temperature. Finally, the slides were air-dried, counterstained with DAPI, visualized and recorded. In order to remove the DAPI and prepare the slides for the second hybridization; slides were washed in 2× SSC for 1 min at room temperature, ethanol dehydrated and air-dried. After application of the probe mixture on the designated area followed by a 3 min co-denaturation, all procedures for the first and the second hybridization were identical.
Slides were analyzed using an Olympus BX41 microscope equipped with single bandpass filters (Semrock, Rochester, NY, USA) for DAPI, FITC and Cy3 and documented using an Olympus DP71 Digital Camera and its associated software. Probes for human homologous chromosomes X, Y and 13, labeled respectively with Alexa488, Alexa488+Cy3 and Cy3, were visualized after the first round of FISH. After the second hybridization, probes for human homologous chromosomes 16 and 18, labeled respectively with Cy3 and Alexa488, were assessed. Recorded pictures were superimposed and recolored to allow visualization of the probes in false colors. The scoring criteria were dependent on the diameter of the fixed nuclei as previously described (Dupont et al., 2008). Blastomeres were scored as normal diploid (Fig. 1a and b), haploid (Fig. 2a and b), triploid or aneuploid (Fig. 2c and d). Furthermore, multinucleated blastomeres from which the total complement of all nuclei was normal diploid were scored as normal (Fig. 1c and d). After analysis of each blastomere, the embryos were classified as normal, aneuploid, mosaic, chaotic, polyploid or haploid (Table I) (Dupont et al., 2008).
The cytogenetic results from ‘morphologically good’ embryos resulting from oocytes that were mature at retrieval and from oocytes maturing post-retrieval were statistically compared using a one tailed χ2-test. Only embryos of which at least 50% of their blastomeres were successfully analyzed were included in this study. Differences with P-values ≤0.05 were considered to be statistically significant.
Ten cycles from 10 different rhesus macaque females were included in this study. The average female age was 8 years old, and at the time of stimulation no female had undergone more than five previous hormonal stimulations. On average, 46 oocytes were collected per IVF cycle, of which 38.6% were mature MII’s, 31.8% were MI’s, 24% were GVs and 5.7% were considered atretic.
Twenty-seven embryos derived from oocytes mature at collection (Group 1) and 26 embryos resulting from oocytes maturing post-retrieval (before or after insemination; Group 2) were analyzed. For Groups 1 and 2, respectively, whole embryo dissociation yielded 247 and 249 individual blastomeres, of which 235 and 236 were successfully fixed and 224 and 217 revealed results for human homologous chromosomes X (Alexa488), Y (Alexa488-Cy3) and 13 (Cy3) after the first hybridization. A second FISH hybridization was successfully achieved on 196 and 213 blastomeres for human homologous chromosomes 16 (Cy3) and 18 (Alexa488), for Groups 1 and 2, respectively.
All embryos were classified according to the results from the five chromosomes examined. Within Group 1, 16 embryos were cytogenetically normal (59.3%), 2 embryos were normal with multinucleation (7.4%), 6 embryos were mosaic (22.2%) and 3 embryos were mosaic with multinucleation (11.1%). Of the 26 embryos resulting from Group 2 oocytes, 7 embryos were normal (26.9%), 2 embryos were normal with multinucleation (7.7%), 7 embryos were mosaic (26.9%), 3 embryos were mosaic with multinucleation (11.4%), 2 embryos were chaotic (7.7%), one was chaotic with multinucleation (3.9%), 2 embryos were aneuploid (7.7%), one was aneuploid with multinucleation (3.9%) and one was triploid (3.9%). These data are shown in Table II. Significantly fewer embryos generated from Group 2 oocytes were chromosomally normal (with and without multinucleation) compared with the embryos resulting from Group 1 oocytes (P = 0.019).
Of the embryos resulting from oocytes maturing post-retrieval (Group 2), 16 embryos resulted from oocytes that had reached the MII stage at insemination, while 10 embryos were derived from oocytes that matured after insemination. From the 16 embryos reaching the MII stage at insemination, 4 were normal without multinucleation, 2 were normal with multinucleation, 5 embryos were mosaic, 3 embryos were mosaic with multinucleation, one embryo was chaotic and one embryo was aneuploid. From the 10 embryos developing from oocytes still at the MI stage at insemination, 3 embryos were normal without multinucleation, 2 embryos were mosaic, one embryo was chaotic, one embryo was chaotic with multinucleation, one embryo was polyploid, one embryo was aneuploid and one embryo was aneuploid with multinucleation.
The percentage of chromosomally normal blastomeres in mosaic embryos in Group 1 as well as in Group 2 amounted to 46%. Most mosaic embryos possessed a combination of meiotic and post-zygotic division errors (anaphase lagging, mitotic non-disjunction, nucleus fragmentation).
The results of this study indicate that ‘morphologically normal’ rhesus macaque embryos resulting from oocytes maturing post-retrieval (before or after insemination) display higher rates of chromosomal abnormalities than embryos derived from oocytes that were mature at retrieval. In addition, chromosomal errors were more severe in the embryos derived from oocytes that matured after retrieval. Indeed, while all chromosomally abnormal embryos resulting from Group 1 were mosaic with normal diploid blastomeres (33.3%), abnormal embryos from Group 2 could be either mosaic (38.5%) or completely abnormal (26.9%). While Group 2 comprised embryos resulting from oocytes immature at retrieval that reached maturity before and after insemination, more embryos with a completely abnormal chromosomal complement were observed when the embryos were derived from oocytes maturing after insemination.
In humans, oocytes that mature in vivo (metaphase II at retrieval) result in the highest pregnancy rates, whereas MI (prophase I) oocytes that subsequently mature in vitro rarely result in a viable pregnancy (De Vos et al., 1999; Vanhoutte et al., 2005). The present study indicates that the poor viability of human embryos resulting from oocytes that matured shortly after retrieval is probably due to their predisposition for cytogenetic instability. The competence of oocytes to accurately resume meiosis and support early embryonic development is acquired gradually and encompasses both nuclear and cytoplasmic maturation programs. Nuclear maturation allows oocytes to resume and complete meiosis, whereas cytoplasmic maturation enables oocytes to be fertilized successfully and support embryo development through the blastocyst stage and beyond. Cytoplasmic immaturity, generally associated with poor fertilization rates and reduced developmental potential, may potentially be reflected into an elevated susceptibility to post-meiotic chromosomal errors. Since in this study, meiotic as well as post-meiotic chromosomal errors were elevated in the group of embryos originating from oocytes with delayed polar body extrusion, it seems that both the nuclear and the cytoplasmic maturation programs were suboptimal in oocytes maturing after retrieval. These maturation programs may have been compromised on various levels.
Both nuclear and cytoplasmic maturation are normally acquired during follicular growth. When a number of follicles have reached an optimal size after hormonal ovarian stimulation, the resumption of meiosis in human and rhesus macaque oocytes is induced by the administration of hCG. In both humans and rhesus macaques, meiosis resumes within 18 h and attains metaphase II between 28 and 38 h after hCG exposure (Seibel et al., 1982; Wolf et al., 1996). Given that all follicles do not develop synchronously during hormonal ovarian stimulation, follicles of various sizes are present at the time of retrieval. Human oocytes retrieved from small follicles are more often immature in comparison to oocytes collected from larger follicles (Tsuji et al., 1985; Ectors et al., 1997). Since cytoplasmic and nuclear maturity is normally acquired synchronously during follicular growth, oocytes that are meiotically immature at retrieval are likely to be cytoplasmically immature as well. This may explain why Group 2 embryos with a ‘good morphology’ display in addition to elevated meiotic segregation errors more post-meiotic chromosomal errors as well.
Another factor that may influence oocyte competence is whether immature oocytes are denuded or are co-cultured with their cumulus cell layer intact prior to IVF or ICSI. Intercellular communication between oocytes and cumulus cells is essential for full nuclear and cytoplasmic oocyte maturation (Ebner et al., 2006) and consequently determines the developmental potential of oocytes (Gilchrist et al., 2008). No studies have yet investigated whether immature oocytes denuded after retrieval are more susceptible to aneuploidy than oocytes surrounded by cumulus cells. Because in this present study all of the cumulus cells were removed at the time of retrieval, it is impossible to assess whether these higher rates of chromosomal anomalies are only associated with insufficient maturation during follicular growth or if they are confounded by the absence of attached cumulus cells needed for superior cytoplasmic and nuclear maturation of oocytes (Goud et al., 1998).
An additional aspect of maturation is the fact that cytoplasmic maturation is also partially acquired post-meiotically during MII arrest. Mature oocytes have higher fertilization rates and better development when they are inseminated or undergo ICSI 3–6 h after the first polar body (PB1) extrusion. Oocytes inseminated earlier than this are more likely to have cytoplasmic deficiencies manifested by lack of activation and failure of early preimplantation development (Trounson et al., 1982; Rienzi et al., 1998; Balakier et al., 2004). It seems that the final maturation of human MII oocytes is crucial for assembly and normal function of the meiotic spindle. Indeed, previous studies have shown that not all MII oocytes present a meiotic spindle shortly after extruding a polar body (Montag et al., 2006) and that oocytes without a spindle at insemination have a reduced developmental potential in comparison to oocytes that do possess a spindle (Wang et al., 2001; Eichenlaub-Ritter et al., 2002). In the present study, pre-incubation times before insemination were not taken into account, so it may be argued that the spindle function was not optimal for the oocytes retrieved at the MI stage. As a result, it is not surprising that embryos derived from oocytes in the same cohort that were MI or MII at retrieval had different rates of chromosomal abnormalities. It may have been worthwhile to delay insemination until the meiotic spindle was visualized using a polarizing microscope (Shen et al., 2008).
This study suggests that ‘morphologically normal’ embryos (Fig. 3) derived from MI oocytes maturing after retrieval may have a reduced implantation potential because they frequently display chromosomal abnormalities. This predisposition may originate through defects in maturation programs caused partially during suboptimal follicular growth but also partially after retrieval. These maturation defects may cause deficiencies in genes and proteins necessary for the coordinated removal of cohesion proteins from chromosome arms during meiosis and mitosis and may therefore predispose oocytes that mature after retrieval to aneuploidy (Kalitsis et al., 2000; Homer et al., 2005; Yin et al., 2006; Niault et al., 2007). Since previous studies in humans and rhesus macaques have shown that oocytes maturing after oocyte retrieval have a different gene transcription profile in comparison to oocytes that matured in vivo (Jones et al., 2008; Lee et al., 2008), it may be worthwhile to examine whether oocytes with delayed polar body extrusion exhibit reduced levels of genes and proteins involved in the prophase pathway and/or the spindle assembly checkpoint. Alternatively, since the integrity and activity of proteins involved in the cell cycle and chromosome segregation are regulated through coordinated phosphorylation and dephosphorylation by kinases or phosphatases, respectively (Swain et al., 2003,2008; Swain and Smith, 2007; Uzbekova et al., 2008); proteins important for cytogenetic stability may be malfunctioning because of inappropriate protein modifications. Although the present findings are derived from rhesus macaque embryos, it is very likely that human embryos resulting from oocytes with delayed PB1 extrusion are similarly more susceptible to chromosomal errors (De Vos et al., 1999; Vanhoutte et al., 2005). This study reinforces the value of non-human primates as an essential clinical translational model to study human embryology and the potential adverse outcomes of ART.
This study was supported by the National Institutes of Health (HD045966 and RR015395 to B.D.B.; HD046553 and RR021881 to C.A.B.) and by the Intramural Research Program of the Eunice Kennedy Shriver, National Institute of Child Health and Human Development, National Institutes of Health, DHHS.
We would like to acknowledge Dr Lutz Froenicke for his valuable advice and Dr Mary Zelinski for her donation of a cohort of oocytes from a rhesus monkey cycle for this study. Additionally, we would like to thank Organon, The Netherlands, for their generous support by providing human recombinant FSH hormones.