Abstract
Are oocyte size, chromatin remodelling, transcriptional activity and mitochondrial distribution in human immature oocytes from early antral follicles retrieved for in vitro maturation (IVM) associated with the acquisition of meiotic competence?
Oocyte size, chromatin compaction, cessation of RNA synthesis and mitochondria rearrangement around the nucleus are associated with the oocyte's potential to resume meiosis in vitro.
Information on oocyte features that confer meiotic competence in human mainly derives from germinal vesicle (GV) oocytes that failed to resume meiosis following an hCG trigger after ovulation induction cycles. Characterization of cumulus-enclosed GV oocytes from small antral follicles prior to IVM provides knowledge on the initial oocyte status and suggests culture requirements in order to promote oocyte competence in vitro.
Prospective collection of 107 oocytes immediately after retrieval (before IVM) and of 293 GV oocytes that had failed to resume meiosis (after IVM).
Human oocytes were collected from women with polycystic ovary syndrome (PCOS), receiving in total 450 IU of highly purified-hMG for IVM treatment (patients) or who donated oocytes for IVM research (donors). Oocytes at GV-stage were retrieved from follicles <10 mm (range 2–10 mm) diameter, before IVM (oocytes at retrieval) or those that failed to mature after IVM (meiotically incompetent). Oocytes were allocated for either mitochondrial staining, by incubating in mitotracker red and then fixed; or for nascent RNA staining, which was assessed by fluorescent labelling (Click-iT® RNA Assay). In every case, oocyte diameter was recorded and chromatin was stained after oocyte fixation. GV-stage oocytes were analysed by confocal laser-scanning microscopy and their characteristics were compared and related to their meiotic competence.
Analysis of oocytes at the immature GV-stage revealed that oocytes at retrieval were significantly larger than those that failed to resume meiosis after IVM (112.7 versus 109.6 µm, P < 0.0001). Oocytes assessed at retrieval showed that 50.6% had a condensed chromatin configuration (perinucleolar chromatin rim) and were consistently transcriptionally silent. This rate matched maturation rates in our current in vitro culture system (49%). However, oocytes that had not reinitiated meiosis after 30 h IVM demonstrated, apart of being smaller in diameter, significantly higher rates of dispersed or intermediate chromatin (P = 0.005). Analysis of mitochondrial distribution revealed that many oocytes at retrieval displayed mitochondrial internalization towards the nucleus (12/30) or a perinuclear mitochondrial distribution (6/30). These mitochondrial patterns were observed more rarely in GV incompetent oocytes following 30 h IVM (16/98 and 1/98, respectively).
Most of the analyses involved the use of invasive techniques. Hence, despite the fact that these data deliver essential information on the intrinsic oocyte maturational and developmental status, a direct match with embryological outcomes could not be established.
The evidence described here can aid in tailoring IVM systems in order to promote completion of nuclear and cytoplasmic maturation of unexpanded cumulus-oocyte complexes.
This study was supported by research grants by the Institute for the Promotion of Innovation by Science and Technology in Flanders, project numbers IWT 130327 and 110680; the Fund for Research Flanders, project number FWO G.0343.13, the Belgian Foundation Against Cancer (HOPE project) and COOK Medical. None of the authors has any competing interest to declare.
Introduction
In vitro maturation (IVM) of oocytes is a method that involves the retrieval and culture of immature oocytes to promote their maturation under defined culture conditions. In human assisted reproductive technology (ART), IVM treatment requires minimal or no exogenous gonadotrophin administration (Jurema and Nogueira, 2006; Nogueira et al., 2012). IVM is the only strategy which can completely circumvent complications associated with conventional IVF and can contribute to fertility preservation programmes, where time constrains and/or health risks limit the use of controlled ovarian hyperstimulation (COH) for IVF/ICSI (Uzelac et al., 2012; Chian et al., 2013; De Vos et al., 2014).
IVM typically involves the aspiration of antral follicles before the leading follicle reaches a diameter of 10–12 mm in order to prevent negative interfollicular effects by a dominant follicle (Son and Tan, 2010). In the case of FSH (or hMG)-primed cycles, oocytes retrieved from antral follicles without prior hCG injection are uniformly enclosed in a compact mass of cumulus cells and are generally found at the immature germinal vesicle (GV)-stage (De Vos et al., 2011; Guzman et al., 2012; Ortega-Hrepich et al., 2013). Whilst this non-hCG triggered approach is designed to avoid the collection of a heterogeneous cohort of cumulus-oocyte complexes (COCs) at different stages of nuclear maturation, the oocyte maturation rate in vitro is low (i.e. ∼50%) when compared with COH followed by hCG trigger, where ∼80% of collected oocytes become mature in vivo. With currently applied maturation systems, only rarely will more than 50–60% of IVM oocytes have extruded the first polar body after 30–40 h of in vitro culture (Mikkelsen and Lindenberg, 2001; Fadini et al., 2009; De Vos et al., 2011; Guzman et al., 2012; Ortega-Hrepich et al., 2013). Furthermore, pregnancy and implantation rates in IVM cycles are significantly lower than in COH cycles (Gremeau et al., 2012; Das et al., 2014). Low maturation rates in currently available IVM systems suggest that a considerable proportion of retrieved oocytes from 2 to 10 mm follicles are still meiotically incompetent; hence, there is a need to characterize nuclear and cytoplasmic competence in human oocytes using methods validated in other species.
Throughout follicle development, oocytes are arrested at prophase I of meiosis, a stage at which its nucleus is known as GV. During this stage, oocytes grow (in size) and their nucleus and cytoplasm acquire maturity in preparation for ovulation. Intrinsic oocyte quality determines the oocyte's ability to overcome meiotic arrest and progress to the metaphase II (MII)-stage (meiotic competence), undergo fertilization and support early embryonic stages (developmental competence). Although knowledge is yet incomplete, fundamental studies in different mammalian species have demonstrated that meiotic and developmental competence are acquired gradually during oogenesis and that this involves key events occurring within the growing oocyte. Oocyte-cumulus interactions appear to play an important role in the acquisition of nuclear and cytoplasmic competence (Zuccotti et al., 1995; De La Fuente and Eppig, 2001; Eppig, 2001; Luciano et al., 2011; Lodde et al., 2013).
Volumetric growth of an oocyte is accompanied by large-scale modifications in chromatin organization (Tan et al., 2009; Luciano et al., 2012). The latter has important biological implications in the control of global gene expression within the oocyte: oocytes with chromatin dispersed throughout the nucleus are transcriptionally active, whereas those with condensed chromatin located around the nucleolus are transcriptionally inactive (Mattson and Albertini, 1990; Bouniol-Baly et al., 1999; Christians et al., 1999; Miyara et al., 2003; Lodde et al., 2008). Fully-grown transcriptionally silent oocytes rely on stored mRNAs, which are post-transcriptionally regulated and translated, to undergo successful maturation, fertilization and to support early embryo development (Bachvarova et al., 1985). Therefore, as reported for different species, both chromatin condensation and transcriptional quiescence are essential to confer meiotic and developmental competence to an oocyte (Zuccotti et al., 2002; Lodde et al., 2007; Tan et al., 2009; Luciano et al., 2012). In human, postovulatory GV-stage oocytes with large diameters (≥115 µm) were found to correspond with stages of advanced chromatin compaction (Combelles et al., 2002; Miyara et al., 2003) and transcriptional repression (Miyara et al., 2003). Although the relationship between meiotic competence and human oocyte size has been documented in a few studies (Durinzi et al., 1995; Combelles et al., 2002; Cavilla et al., 2008), there is a need to further explore the relationship between oocyte size, chromatin conformation, transcriptional activity and the ability to resume meiosis. Increased knowledge of these correlations should result in the identification of distinct subgroups of immature oocytes obtained during aspiration from small follicles.
Apart from nuclear parameters, it has emerged that mitochondrial distribution is also associated with oocyte competence (Nagai et al., 2006; Van Blerkom, 2011). In murine and porcine oocytes, mitochondrial aggregation around the nucleus occurs before and during oocyte maturation (Sun et al., 2001; Van Blerkom et al., 2002, Dumollard et al., 2006; Yu et al., 2010). In human, information on the spatial distribution of mitochondria in immature oocytes and its association with nuclear competence is scarce (Wilding et al., 2001; Liu et al., 2010b). Spatiotemporal subcellular and morphological studies in human oocytes have been hampered by the limited availability of donated research material and have mainly been performed in discarded immature oocytes that failed to resume meiosis after ovarian stimulation.
In the frame of our studies towards improved culture systems for in vitro maturation of GV oocytes from 2 to 10 mm follicles in human, we set out to investigate how oocyte size, chromatin condensation, transcriptional activity and distribution of mitochondria may relate to oocyte capacity to resume meiosis in vitro. Given that these features can indicate the level of oocyte differentiation, which is crucial for acquisition of competence in immature oocytes, we aimed to analyse and compare oocytes at an equivalent meiotic (GV) stage: the starting pool of oocytes at retrieval, before IVM, and those that had failed to resume meiosis after IVM. Improved understanding of these correlations may ultimately guide our choice towards oocyte-tailored culture conditions.
Materials and Methods
Ethical approval
This study was approved by the ethical committee of our university-based hospital (UZBrussel, project 2008/068). Signed informed consent was obtained from all participants.
Collection of human immature GV-stage oocytes
Human oocytes at the immature stage were collected from women with polycystic ovaries (PCO or PCOS), who underwent mild stimulation with gonadotrophins (highly purified-hMG) for IVM treatment (‘patients’) or who donated oocytes for IVM research (‘donors’) (project IWT 110680, P.I. J. Smitz). In total, 400 GV oocytes, collected from 61 patients/donors, were evaluated. Two sources of immature GV-stage oocytes can be distinguished: (i) donated oocytes obtained immediately after egg retrieval (freshly isolated group, N = 107, from 12 donors), (ii) GV oocytes that failed to mature in vitro after 30 h (meiotically incompetent group, N = 293, from 49 patients). The latter group represents a fraction of the total pool of 1127 oocytes that were placed in culture for IVM (from 49 patients); however due to their immature (GV) stage these were destined to be discarded. Oocytes were retrieved from follicles <10 mm (range 2–10 mm) diameter.
The following parameters were analysed in GV-stage oocytes, before (freshly isolated) and after (meiotically incompetent) IVM, to investigate their association with meiotic competence: (i) oocyte diameter, (ii) chromatin configuration, (iii) transcriptional activity and (iv) mitochondrial distribution. Oocyte diameter and chromatin configuration were assessed in a total of 400 and 311 GV oocytes, respectively. For practical reasons the transcriptional activity and mitochondrial distribution pattern were assessed in a smaller subset of these oocytes (n = 135 and 128, respectively).
In all cases, media used for oocyte collection (HEPES-buffered human tubal fluid—HTF, Lonza, Verviers, Belgium) and culture (α-MEM, Invitrogen, Merelbeke Belgium) were supplemented with 0.5% human serum albumin (HSA, Vitrolife; Göteborg, Sweden). To avoid spontaneous germinal-vesicle breakdown (GVBD) during handling and assessments, 200 µM IBMX (Sigma-Aldrich; Diegem, Belgium) was added to the media.
Assessment of oocyte diameter
Oocyte size was calculated as an average after measuring the maximum and minimum oocyte diameter, without including the zona pellucida.
Taking into account the possibility of heterogeneity of our study population with regard to oocyte size, and based on the reported relationship between diameter and meiotic competence (Durinzi et al., 1995; Combelles et al., 2002; Cavilla et al., 2008), three size categories were considered in our evaluations: small oocytes, measuring <105 µm; medium oocytes, measuring ≥105 and <110 µm; and large oocytes, measuring ≥110 µm.
Evaluation of chromatin configuration
In each size category, chromatin was revealed after staining either with medium containing Hoechst 33258 (Sigma) at a concentration of 5 µg/ml, during 5 min; or PicoGreen® (Life Technologies; Gent, Belgium) during 10 min. Nuclear chromatin conformation was assessed under a fluorescence microscope (IX70; Olympus).
Classification of chromatin configuration was performed based on previous reports (Combelles et al., 2002; Miyara et al., 2003, 2008). Briefly, chromatin configuration was classified into three groups according to its condensation status and its distribution around the nucleolus (or nucleolar-like body, NLB): dispersed, intermediate or condensed chromatin. The latter configuration exhibited a chromatin ring around the NLB (perinucleolar rim) (Fig. 1).
Chromatin configuration in human germinal vesicle-stage oocytes from small antral follicles. Chromatin was classified in three categories according to its condensation status and in relation to the nucleolus-like bodies (NLBs): dispersed, intermediate or condensed chromatin (displaying a single chromatin ring around a NLB = perinucleolar rim).
Detection of transcriptional activity
Transcriptional activity was determined by fluorescent labelling of nascent RNA by means of the Click-iT® RNA Alexa Fluor® 594 HCS Assay (Life Technologies; Gent, Belgium). Briefly, denuded GV oocytes were incubated for 90 min in supplemented α-MEM medium (see above) containing 4 mM 5-Ethynyl Uridine, at 37°C, 5% CO2, 100% humidity. After incubation, oocytes were once washed in phosphate-buffered saline (PBS, Sigma-Aldrich; Diegem, Belgium) followed by fixation in 4% paraformaldehyde (PFA) for 30 min at 37°C. After fixation, oocytes were once washed in PBS, placed in PBS containing 0.5% Triton X-100 and 1% BSA, for 15 min at room temperature, followed by a washing step in PBS, 1% BSA. Click-iT reaction was performed according to manufacturer's instructions followed by DNA counterstaining with PicoGreen®. A 1:100 dilution of Picogreen® was prepared with PBS.
Stained oocytes were analysed under a Fluoview confocal microscope system (Olympus IX 70) equipped with an Argon–Krypton laser (488 and 568 nm) and band pass 510–540 nm and long pass 610 nm filters, by sequentially scanning each channel to visualize Picogreen® and Alexa Fluor® 594, respectively.
Transcriptional activity was classified as: strong (+), moderate (+/−) or absent (−) (Fig. 2).
Transcriptional activity in human germinal vesicle (GV)-stage oocytes. Nascent RNA was assessed by fluorescent labelling by means of the Click-iT® RNA Alexa Fluor® 594 HCS Assay followed by a DNA counterstaining with PicoGreen®. Representative examples of human GV oocyte showing (A) dispersed chromatin and strong (+) transcriptional activity; (B) intermediate chromatin condensation and moderate (+/−) transcriptional activity, and (C) condensed chromatin and no RNA synthesis (−). Scale bars represents 50 µm.
Evaluation of mitochondrial distribution
Immature denuded oocytes, classified according to their diameter range, were first washed in supplemented α-MEM medium (see above), containing 0.5 µM MitoTracker Red CM-H2XRos (Molecular Probes, Invitrogen, Merelbeke Belgium) and cultured for 30 min at 37°C and under 5% CO2 at 100% humidity. This reduced MitoTracker probe diffuses across the plasma membrane of living cells, where they are oxidized to the correspondent fluorescent probe, sequestered and accumulated in active mitochondria. After incubation, oocytes were washed thoroughly three times (×10 min) in α-MEM medium, fixed in a 2% PFA solution in PBS for 30 min at 37°C and rinsed two times (× 5 min) in PBS containing 1% bovine serum albumin (BSA). Finally, oocytes were stained with PicoGreen® during 15 min and washed two times (× 5 min) in PBS/BSA. At all times during staining and fixation, oocytes were carefully handled to avoid exposure to light and reduce photo bleaching.
Stained oocytes were imaged with an Olympus IX 70 Fluoview confocal laser scanning microscope system equipped with an Argon–Krypton laser (488–568 nm); and band pass filter 510–540 and long pass filter 585 to visualize Picogreen and Mitotracker respectively, using 40× UplanApo objective.
Mitochondrial distribution pattern was assessed in all oocytes showing a normal morphology, intact zona pellucida and uniformly distributed cytoplasm. An image was taken in an optical cross section through the oocyte nucleus.
Statistical analysis
Statistical analysis was done and frequency histograms were constructed using GraphPad Prism version 4.00 software (GraphPad, San Diego, CA, USA). Differences in diameter between oocytes at retrieval and meiotically incompetent oocytes were assessed by Mann–Whitney U-test. Data on chromatin configuration were analysed by a generalized linear model for binary data, with a probit link with subsequent comparisons between groups. Data on transcriptional activity were analysed by a generalized linear mixed model for binary data with a probit link and patient as random factor with subsequent comparisons between groups. In both cases, P-values were corrected for simultaneous hypothesis testing according to Sidak. P < 0.05 was considered to be significant. The latter two analyses were performed by S-Plus 8.0.4 for Linux.
Results
Meiotic competence of oocytes obtained without HCG injection from small antral follicles
Of 1127 GV oocytes (retrieved from 49 patients) placed in culture for IVM, 552 (49%) matured to the MII stage after 30 h culture. Of the latter, 389 (70%) were fertilized normally after ICSI. From the fertilized oocytes, 215 (55%) were able to develop into embryos of good morphological quality on Day 3 (grade EQ1 or EQ2 in Papanikolaou et al., 2005) based on the number of blastomeres (at least 6), rate of fragmentation (<10–25%), no evidence of multi-nucleation of the blastomeres, and early compaction. Among the 575 (51%) non-matured oocytes, 182 (32%) were at MI stage and 319 (55%) remained arrested at GV stage. The latter subgroup of oocytes was the source of meiotically incompetent oocytes used in the current study. The 74 remaining oocytes (13%) were degenerated after IVM culture.
Diameter of GV-stage oocytes before and after IVM
A total of 107 immature GV-stage oocytes freshly isolated right at retrieval (i.e. before IVM) and 293 meiotically incompetent GV-stage oocytes after 30 h IVM (coming from 12 donors and 49 patients, respectively) had their diameters measured.
A large heterogeneity in oocyte diameter was observed in each group of oocytes analysed (Fig. 3). In general, GV oocytes at retrieval were significantly larger (P < 0.0001) than GV oocytes that failed to mature in vitro, with mean diameters (± SD) of 112.6 ± 5.14 and 109.6 ± 4.36 µm, respectively.
Frequency histograms of oocyte diameter in human small antral follicles: (A) at the time of oocyte retrieval (N = 107) and (B) following 30 h in vitro maturation (meiotically incompetent, N = 293). Mean diameter of freshly isolated oocytes was significantly larger than meiotically incompetent oocytes (112.6 ± 5.14 µm versus 109.6 ± 4.36 µm, P < 0.0001). (C and D) Number and proportions of oocytes analysed at retrieval and meiotically incompetent oocytes in the three different size categories. Oocyte size was calculated as the average after measuring maximum and minimum diameters, without including the zona pellucida.
Before IVM, most oocytes (78.5%) showed diameters ≥110 µm. In this size category, the average diameter was 114.5 µm. Following IVM, 55.6% oocytes that had remained at the GV-stage had diameters ≥110 µm, but the average diameter in this category was smaller, 112.7 µm. Oocytes in the medium-size range (≥105 and <110 µm) represented 17.8 and 33.4% of the freshly isolated and meiotically incompetent groups, respectively, and in both groups had an average diameter of 107.0 µm. The fraction of oocytes <105 µm represented 3.7% in oocytes collected at retrieval (group average 98.4 µm) and 11% in meiotically incompetent GV oocytes (group average 102 µm).
Chromatin configuration of GV-stage oocytes before and after IVM
Seventy-nine oocytes retrieved before IVM and 232 incompetent GV oocytes after IVM (deriving from 10 donors and 35 patients, respectively) were evaluated for chromatin conformation.
Table I summarizes the number and proportion of oocytes with dispersed, intermediate and compact chromatin conformation in both groups of oocytes collected before IVM (at retrieval) and meiotically incompetent ones.
Chromatin configuration in germinal vesicle-stage oocytes before (at retrieval) and after (meiotically incompetent) 30 h in vitro maturation.
| Chromatin configuration | ||||
|---|---|---|---|---|
| Oocyte diameter | Dispersed (%) | Intermediate (%) | Condensed (%) | |
| At retrieval | ||||
| Small | <105 µm | 3 (75) | 0 | 1 (25) |
| Medium | ≥105 <110 µm | 3 (27.3) | 3 (27.3) | 5 (45.5) |
| Large | ≥110 µm | 20 (31.3) | 10 (15.6) | 34 (53.1) |
| Total (N = 79) | 26 (32.9) | 13 (16.5) | 40 (50.6) | |
| Meiotically incompetent | ||||
| Small | <105 µm | 17 (65.4) | 6 (23.1) | 3 (11.5) |
| Medium | ≥105 <110 µm | 46 (59.7) | 21 (27.3) | 10 (13.0) |
| Large | ≥110 µm | 49 (38.0) | 51 (39.5) | 29 (22.5) |
| Total (N = 232) | 112 (48.3) | 78 (33.6) | 42 (18.1) | |
| Chromatin configuration | ||||
|---|---|---|---|---|
| Oocyte diameter | Dispersed (%) | Intermediate (%) | Condensed (%) | |
| At retrieval | ||||
| Small | <105 µm | 3 (75) | 0 | 1 (25) |
| Medium | ≥105 <110 µm | 3 (27.3) | 3 (27.3) | 5 (45.5) |
| Large | ≥110 µm | 20 (31.3) | 10 (15.6) | 34 (53.1) |
| Total (N = 79) | 26 (32.9) | 13 (16.5) | 40 (50.6) | |
| Meiotically incompetent | ||||
| Small | <105 µm | 17 (65.4) | 6 (23.1) | 3 (11.5) |
| Medium | ≥105 <110 µm | 46 (59.7) | 21 (27.3) | 10 (13.0) |
| Large | ≥110 µm | 49 (38.0) | 51 (39.5) | 29 (22.5) |
| Total (N = 232) | 112 (48.3) | 78 (33.6) | 42 (18.1) | |
Chromatin configuration in germinal vesicle-stage oocytes before (at retrieval) and after (meiotically incompetent) 30 h in vitro maturation.
| Chromatin configuration | ||||
|---|---|---|---|---|
| Oocyte diameter | Dispersed (%) | Intermediate (%) | Condensed (%) | |
| At retrieval | ||||
| Small | <105 µm | 3 (75) | 0 | 1 (25) |
| Medium | ≥105 <110 µm | 3 (27.3) | 3 (27.3) | 5 (45.5) |
| Large | ≥110 µm | 20 (31.3) | 10 (15.6) | 34 (53.1) |
| Total (N = 79) | 26 (32.9) | 13 (16.5) | 40 (50.6) | |
| Meiotically incompetent | ||||
| Small | <105 µm | 17 (65.4) | 6 (23.1) | 3 (11.5) |
| Medium | ≥105 <110 µm | 46 (59.7) | 21 (27.3) | 10 (13.0) |
| Large | ≥110 µm | 49 (38.0) | 51 (39.5) | 29 (22.5) |
| Total (N = 232) | 112 (48.3) | 78 (33.6) | 42 (18.1) | |
| Chromatin configuration | ||||
|---|---|---|---|---|
| Oocyte diameter | Dispersed (%) | Intermediate (%) | Condensed (%) | |
| At retrieval | ||||
| Small | <105 µm | 3 (75) | 0 | 1 (25) |
| Medium | ≥105 <110 µm | 3 (27.3) | 3 (27.3) | 5 (45.5) |
| Large | ≥110 µm | 20 (31.3) | 10 (15.6) | 34 (53.1) |
| Total (N = 79) | 26 (32.9) | 13 (16.5) | 40 (50.6) | |
| Meiotically incompetent | ||||
| Small | <105 µm | 17 (65.4) | 6 (23.1) | 3 (11.5) |
| Medium | ≥105 <110 µm | 46 (59.7) | 21 (27.3) | 10 (13.0) |
| Large | ≥110 µm | 49 (38.0) | 51 (39.5) | 29 (22.5) |
| Total (N = 232) | 112 (48.3) | 78 (33.6) | 42 (18.1) | |
A higher proportion of oocytes with condensed chromatin configuration was observed with increasing oocyte diameters. In general, small oocytes (both in freshly isolated and meiotically incompetent groups) and medium-size oocytes (in the meiotically incompetent group) revealed significantly higher rates of dispersed chromatin compared with large oocytes (P < 0.05).
Furthermore, regardless of oocyte diameter, the proportions of the three-chromatin patterns differed significantly between oocytes collected at retrieval and those meiotically incompetent after IVM. Oocytes at retrieval were mainly in the more advanced chromatin stage (50.6% showing a perinucleolar chromatin rim); however, oocytes that had remained at the GV-stage after 30 h IVM had significantly higher rates of dispersed or intermediate chromatin (P = 0.005). This difference was mainly observed in large oocytes (P = 0.0002).
Transcriptional activity of GV-stage oocytes before and after IVM
Thirty-six freshly isolated oocytes and 99 incompetent oocytes (derived from 4 donors and 14 patients, respectively) were analysed for transcriptional activity by means of staining of nascent RNA.
Table II summarizes the number of oocytes with strong (+), moderate (+/−) or no (−) transcriptional activity at each range of oocyte diameter in both groups.
Transcriptional activity in germinal vesicle-stage oocytes before (at retrieval) and after (meiotically incompetent) 30 h in vitro maturation.
| Transcriptional activity | ||||
|---|---|---|---|---|
| Oocyte diameter | Strong + (%) | Moderate +/− (%) | Absent − (%) | |
| At retrieval | ||||
| Small | <105 µm | 0 (0) | 1 (100) | 0 (0) |
| Large | ≥110 µm | 6 (17.1) | 9 (25.7) | 20 (57.1) |
| Total (N = 36) | 6 (16.7) | 10 (27.8) | 20 (55.6) | |
| Meiotically incompetent | ||||
| Small | <105 µm | 8 (72.7) | 0 | 3 (27.3) |
| Medium | ≥105 <110 µm | 9 (33.3) | 11 (40.7) | 7 (25.9) |
| Large | ≥110 µm | 12 (19.7) | 17 (27.9) | 32 (52.5) |
| Total (N = 99) | 29 (29.3) | 28 (28.3) | 42 (42.4) | |
| Transcriptional activity | ||||
|---|---|---|---|---|
| Oocyte diameter | Strong + (%) | Moderate +/− (%) | Absent − (%) | |
| At retrieval | ||||
| Small | <105 µm | 0 (0) | 1 (100) | 0 (0) |
| Large | ≥110 µm | 6 (17.1) | 9 (25.7) | 20 (57.1) |
| Total (N = 36) | 6 (16.7) | 10 (27.8) | 20 (55.6) | |
| Meiotically incompetent | ||||
| Small | <105 µm | 8 (72.7) | 0 | 3 (27.3) |
| Medium | ≥105 <110 µm | 9 (33.3) | 11 (40.7) | 7 (25.9) |
| Large | ≥110 µm | 12 (19.7) | 17 (27.9) | 32 (52.5) |
| Total (N = 99) | 29 (29.3) | 28 (28.3) | 42 (42.4) | |
Transcriptional activity in germinal vesicle-stage oocytes before (at retrieval) and after (meiotically incompetent) 30 h in vitro maturation.
| Transcriptional activity | ||||
|---|---|---|---|---|
| Oocyte diameter | Strong + (%) | Moderate +/− (%) | Absent − (%) | |
| At retrieval | ||||
| Small | <105 µm | 0 (0) | 1 (100) | 0 (0) |
| Large | ≥110 µm | 6 (17.1) | 9 (25.7) | 20 (57.1) |
| Total (N = 36) | 6 (16.7) | 10 (27.8) | 20 (55.6) | |
| Meiotically incompetent | ||||
| Small | <105 µm | 8 (72.7) | 0 | 3 (27.3) |
| Medium | ≥105 <110 µm | 9 (33.3) | 11 (40.7) | 7 (25.9) |
| Large | ≥110 µm | 12 (19.7) | 17 (27.9) | 32 (52.5) |
| Total (N = 99) | 29 (29.3) | 28 (28.3) | 42 (42.4) | |
| Transcriptional activity | ||||
|---|---|---|---|---|
| Oocyte diameter | Strong + (%) | Moderate +/− (%) | Absent − (%) | |
| At retrieval | ||||
| Small | <105 µm | 0 (0) | 1 (100) | 0 (0) |
| Large | ≥110 µm | 6 (17.1) | 9 (25.7) | 20 (57.1) |
| Total (N = 36) | 6 (16.7) | 10 (27.8) | 20 (55.6) | |
| Meiotically incompetent | ||||
| Small | <105 µm | 8 (72.7) | 0 | 3 (27.3) |
| Medium | ≥105 <110 µm | 9 (33.3) | 11 (40.7) | 7 (25.9) |
| Large | ≥110 µm | 12 (19.7) | 17 (27.9) | 32 (52.5) |
| Total (N = 99) | 29 (29.3) | 28 (28.3) | 42 (42.4) | |
GV-stage oocytes analysed for transcriptional activity before IVM were mainly (35/36) ≥110 µm diameter. The only oocyte <105 µm revealed moderate activity. In the large range, few oocytes (6/35) exhibited strong transcriptional activity and 9/35 oocytes revealed moderate activity, whereas most of them (20/35) showed no detection of RNA synthesis.
After IVM, meiotically incompetent oocytes from the small- and medium-size groups revealed significantly higher rates of oocytes with strong or moderate transcriptional activity compared with large oocytes (P < 0.05). A gradual cessation of RNA synthesis was observed with increasing oocyte diameters (Table II).
In general, irrespective of the oocyte diameter, transcription was suppressed in 55.6% of oocytes collected before IVM and in 42.4% of the oocytes that failed to resume meiosis following 30 h IVM. Moreover, the proportion of oocytes with active transcription was higher in the latter group, although this was not significantly different.
Chromatin configuration in relation to transcriptional activity
Figure 4 shows the pattern of RNA synthesis in relation to the chromatin status in immature GV oocytes collected at retrieval and after IVM (meiotically incompetent). In general, we found that oocytes with dispersed chromatin mainly displayed strong or moderate transcriptional activity. Oocytes with an intermediate stage of chromatin condensation mainly showed no transcriptional activity; and oocytes with fully condensed chromatin (perinucleolar rim) were consistently transcriptionally silent.
Pattern of transcriptional activity: strong (+), moderate (+/−) or absent (−) in relation to the chromatin status in human germinal vesicle-stage oocytes: (A) at retrieval and (B) following 30 h in vitro maturation (meiotically incompetent). Oocytes with a dispersed chromatin pattern mainly displayed strong (black) or moderate (light grey) transcriptional activity whereas oocytes with an intermediate stage of chromatin condensation showed either moderate or no transcriptional activity (dark grey). Oocytes with fully condensed chromatin (perinucleolar chromatin rim) were consistently found to be transcriptionally silent.
Distribution of active mitochondria in GV-stage oocytes before and after IVM
The distribution of active mitochondria was studied in 30 oocytes before IVM and 98 meiotically incompetent oocytes following 30 h IVM (derived from 4 donors and 16 patients, respectively). Classification of mitochondrial distribution in GV-stage oocytes was based on previous reports (Sun et al., 2001; Wilding et al., 2001; Van Blerkom, 2004; Liu et al., 2010b) and adjusted according to our findings.
Mitochondrial aggregation patterns
In general, mitochondrial aggregation patterns were classified as either homogenous or heterogeneous. Oocytes with a homogenous pattern were characterized by mitochondria evenly distributed all over the cytoplasm (smooth appearance, Fig. 5A). Oocytes with a heterogeneous pattern included those showing small or large granules scattered across the cytoplasm. Because it was not uncommon to observe the combination of large granules with a smooth pattern, these were also allocated to the heterogeneous pattern group (Fig. 5B). In most oocytes, granules were found in the subplasma membrane area and, when present, a smooth pattern was observed in the proximity of the nucleus. The heterogeneous pattern was more predominant than the homogenous one, both in oocytes analysed at retrieval (29/30) and in meiotically incompetent oocytes after IVM (70/98).
Mitochondrial aggregation patters in human germinal vesicle-stage oocytes. Examples of (A) homogenous mitochondria distribution: characterized by a smooth uniform mitochondria pattern and (B) heterogeneous distribution: characterized by the presence of small or large granules, predominantly observed in the subplasma membrane (arrows) or scattered in the cytoplasm. In most cases a mixture of both the smooth and granular patterns was observed. An image was taken in an optical cross section through the centre of the oocyte. Scale bar represents 50 μm.
Translocation of active mitochondria to the nucleus area
Of the 30 oocytes analysed before IVM, 12 showed dispersion of mitochondria towards the inner cytoplasm and the GV (Fig. 6B–D). However, this pattern was only observed in 16/98 of meiotically incompetent oocytes following 30 h IVM. Notably, a perinuclear distribution of mitochondria (Fig. 7) was observed in 6/30 freshly isolated oocytes (which had diameters ≥110 µm), while only in 1/98 meiotically incompetent oocytes.
Internalization of active mitochondria towards the nucleus in human immature germinal vesicle (GV) oocytes from small antral follicles at time of retrieval. (A) An example of mitochondrial distributed throughout the cytoplasm. (B–D) Examples of progressive translocation of mitochondria to the inner cytoplasm and the GV area. Arrows indicate the accumulation of active mitochondria in the vicinity of the GVs. Scale bar represents 50 μm.
Mitochondrial distribution in human immature germinal vesicle oocytes. (A) Example of mitochondria distribution spread throughout the cytoplasm, characteristic in meiotically incompetent oocytes (following 30 h in vitro maturation). (B) Example of perinuclear mitochondria distribution in an oocyte analysed at retrieval. In both cases, a chromatin ring was observed (in B showing two nucleoli). Scale bar represents 50 μm.
Regarding mitochondrial distribution pattern and its association with chromatin configuration, no correlation was observed between the formation of a perinucleolar chromatin rim and mitochondrial aggregation pattern or mitochondrial localization towards the nucleus; regardless of the oocyte source analysed (oocytes at retrieval or meiotically incompetent).
Discussion
In this study, we aimed to characterize human immature oocytes from early antral follicles at retrieval (starting pool) and those that failed to resume meiosis after IVM. Analysis of oocytes at this equivalent meiotic stage (GV) was fundamental to delineate specific features that may be present or missing in immature oocytes and are crucial to confer meiotic competence. To our knowledge, this is the largest study analysing subcellular, molecular and morphological characteristics of human immature oocytes aspirated from follicles between 2 and 10 mm diameter, collected without prior hCG trigger. Earlier reports have focused on GV oocytes obtained in standard IVF/ICSI ovulation induction cycles that failed to resume meiosis following hCG trigger (Wilding et al., 2001; Combelles et al., 2002; Miyara et al., 2003; Liu et al., 2010b). In contrast, the current study provides novel information on oocytes from small antral follicles which is essential to guide us to an improved in vitro culture methodology for human oocytes and expands our knowledge on the essential properties conferring competence in human oocytes.
Size of GV-stage oocytes retrieved from small antral follicles prior and following IVM
The association between the diameter of immature oocytes and meiotic competence has been confirmed in many species (Eppig and Schroeder, 1989; Eppig et al., 1992; Fair et al., 1995; Otoi et al., 1997). Likewise, in humans, GV-stage oocytes have been shown to have a size-dependent ability to resume and complete meiosis (Durinzi et al., 1995; Cavilla et al., 2008). In unstimulated patients, undergoing laparoscopy for gynaecological disorders, Durinzi et al. (1995) reported differences in maturation capacity between oocytes measuring 86–105 µm at retrieval and those that measured 106–125 µm. Surprisingly, the smallest diameter at which an oocyte is capable of maturing appears to differ between normo-ovulatory and PCO patients due to the different follicular kinetics related to their underlying endocrine conditions. While Durinzi et al. (1995) found that a minimum diameter for oocytes to mature was 115 µm, Cavilla's study (2008), in unstimulated PCO patients, demonstrated that a diameter of 103 µm was the threshold to reach the MII stage and most oocytes that had matured in vitro were <115 µm.
The study presented here, in PCO(S) patients undergoing IVM without prior hCG administration, shows that the majority (78.5%) of oocytes from follicles measuring 2–10 mm on the day of retrieval had a diameter of ≥110 µm, with an average among these of 114.5 µm. Following 30 h of in vitro culture, nearly half of meiotically incompetent oocytes (representing ∼15% of the initial pool incubated in IVM media) also had diameters ≥110 µm, although these were on average smaller (112.7 µm). This finding, together with the fact that the total pool of immature GV oocytes analysed immediately at retrieval was significantly larger compared with those that remained arrested at the GV-stage following IVM (112.6 µm versus 109.6 µm) supports the notion that larger GV oocytes have a higher potential for meiotic resumption.
Interestingly, Cavilla et al. (2008) reported significant differences in diameter between oocytes collected from stimulated ICSI patients (111 µm) versus unstimulated PCO patients (106 µm), which they attributed to the differences in endocrine environment and follicle size (>10 mm versus <10 mm diameter, respectively). In the current study, although oocytes were also retrieved from PCO(S) patients and exclusively collected from follicles <10 mm, these were larger (∼112.6 µm) than those reported by Cavilla (∼106 µm), possibly due to the short 3 or 4 day HP-hMG priming systematically applied in the current study. Hence, our observations underscore the impact of follicle origin (size), hormonal stimulation and patient characteristics on oocyte size and their relevance for IVM procedures.
Chromatin configuration and transcriptional activity in immature GV-stage oocytes before and after IVM and its association with meiotic competence acquisition
Mammalian oocyte growth (i.e. size) and differentiation is characterized by large-scale modifications in chromatin structure and function (Tan et al., 2009; Luciano et al., 2012; Luciano and Lodde, 2013). Progress to pre-ovulatory stages is accompanied by gradual chromatin condensation and global transcriptional repression within the oocyte (Parfenov et al., 1989; Bouniol-Baly et al., 1999; Christians et al., 1999; De La Fuente and Eppig, 2001; Miyara et al., 2003); these changes confer both meiotic and developmental competence (Debey et al., 1993; Zuccotti et al., 1998, 2002; Liu and Aoki, 2002; Lodde et al., 2007). In the current study, patterns of chromatin condensation and transcriptional activity of human immature GV-stage oocytes from small antral follicles before and after 30 h IVM were evaluated in relation to the different oocyte size categories.
Previous reports on human GV oocytes from large antral follicles that failed to mature after conventional gonadotrophin stimulation following hCG trigger have shown an association between oocyte size, chromatin conformation and transcriptional activity (Combelles et al., 2002; Miyara et al., 2003, 2008). Oocytes with diameters >115 µm mainly displayed compact chromatin and were transcriptional quiescent, whereas those from smaller diameters (<110 µm) displayed dispersed chromatin and were transcriptionally active (Combelles et al., 2002; Miyara et al., 2003, 2008). Our results are in line with previous reports, since a gradual cessation of RNA synthesis was observed with increasing oocyte diameters. A similar association with chromatin condensation and oocyte size was also observed, especially in the oocytes analysed at retrieval.
Overall, regardless of the diameter, most oocytes that remained arrested at the GV-stage following 30 h IVM displayed an immature chromatin status (dispersed throughout the nucleus) and were still transcriptionally active. In contrast, approximately half of oocytes analysed at retrieval (before IVM), exhibited a perinucleolar rim and were transcriptionally silent. Interestingly, these proportions coincide with the oocyte maturation rate in the IVM system used and others similar to this one, which is ∼50% (Fadini et al., 2009; De Vos et al., 2011; Guzman et al., 2012; Ortega-Hrepich et al., 2013). In different mammalian species, oocytes with highly condensed chromatin have a greater capacity to mature in vitro compared with those with dispersed chromatin (Wickramasinghe et al., 1991; Debey et al., 1993; Schramm et al., 1993; Liu and Aoki, 2002; Lodde et al., 2007; Inoue et al., 2008; Wang et al., 2009a). Our findings support previous statements and indicate that in human small antral follicles (2–10 mm) not exposed to hCG, only half of the oocytes retrieved for IVM have attained an advanced stage of differentiation (at the level of global gene transcription and chromatin configuration) and are meiotically competent.
Findings in the present study, however, differ from a previous study by Combelles et al. (2002), who reported that oocytes with a ‘C-pattern’, which is equivalent to the intermediate configuration in the current study, exhibited meiotic competence after IVM. Although this discrepancy may be due to differences in classification, it may as well be attributed to the different oocyte origin (patients' different endocrine profile). For instance, in the current study, follicle and oocyte sizes were smaller, patients' hormonal treatment did not include an hCG trigger and, most importantly, oocytes were always enclosed in a compact cumulus compartment.
In many mammalian oocytes, a temporal correlation has been found between progressive chromatin condensation and gradual cessation of global transcription (Bouniol-Baly et al., 1999; Christians et al., 1999; De La Fuente and Eppig, 2001; Lodde et al., 2007, 2008). Although the mechanisms controlling these events are poorly understood, experiments performed in mice have indicated that, to some extent, they may be regulated by independent mechanisms (De La Fuente et al., 2004; Andreu-Vieyra et al., 2010). In support of this hypothesis, we found that transcriptional silencing was observed independent of the formation of a perinuclear chromatin rim (i.e. in oocytes displaying intermediate chromatin configuration and, to a lower extent, even in those with a dispersed chromatin). However, we did notice that all oocytes with a perinucleolar chromatin rim consistently exhibited transcriptional silencing. Hence, as previously proposed, it is likely that once the mechanisms leading to global transcriptional repression are initiated, they are further reinforced by an increase in chromatin condensation (De La Fuente, 2006).
Another important aspect to consider is the essential role of cumulus-oocyte interactions in the regulation of chromatin remodelling, transcriptional activity and oocyte competence acquisition (Carabatsos, 2000; De La Fuente and Eppig, 2001; Liu and Aoki, 2002; Luciano et al., 2011; Lodde et al., 2013). A functional bidirectional communication between cumulus cells and the oocyte is a prerequisite for oocytes to progress towards their final differentiation status. In the bovine model, premature disruption of gap-junctional communications (GJC) during final oocyte development by means of 1-heptanol (a GJ uncoupler) caused abrupt chromatin condensation and transcriptional repression. However this effect could be abolished by addition of cilostamide, a specific inhibitor of PDE3 (the phosphodiesterase responsible for the degradation of cAMP within the oocyte during meiotic resumption) (Luciano et al., 2011). Although it is not clear yet how companion granulosa cells regulate oocyte transcription and chromatin remodelling in the oocyte, the latter study suggests that the modulation of such events through GJC might partly involve cAMP-dependent mechanisms (Luciano et al., 2011).
In the current study most of the oocytes that failed to resume meiosis after 30 h of in vitro culture had dispersed chromatin and were transcriptionally active. However, some of these oocytes displayed features of a more advanced developmental stage, i.e. condensed chromatin and transcriptional quiescence. Curiously, the proportion of oocytes that had not become mature and were transcriptionally silent (42%) was much higher compared with the proportion of those displaying a perinuclear chromatin rim (18%). On the contrary, in freshly isolated oocytes these two properties were perfectly matching. Possible explanations for these differences may be related to the status of the cumulus-oocyte complex at the moment of oocyte collection: whereas oocytes isolated at retrieval were enclosed in a compact mass of cumulus cells immediately before being analysed, incompetent oocytes were collected from cumulus-oocyte complexes that presented various degrees of cumulus cell dispersion (disconnection) after 30 h exposure to a meiotic stimulus in presence of gonadotrophins. Therefore, our findings indicate that, in yet immature meiotically incompetent oocytes, a premature interruption of oocyte-cumulus communications following the meiotic stimulus might have induced an abrupt termination of gene transcription, and, as consequence, may have resulted in asynchrony between meiotic progression and transcriptional silencing.
Current results are in accordance with the data in the bovine model; however, it should be noted that in the present study, the transcriptional activity assay was carried out in denuded oocytes and so, further studies are required to evaluate the influence of cumulus cells (and the potential factors secreted by these cells) on oocyte transcription during the assay period; and, of particular importance, the functional status of GJCs in cumulus-oocyte complexes from both groups: GV-stage oocytes at retrieval and meiotically incompetent ones.
Overall, these observations illustrate that currently available IVM systems fail to support proper acquisition of both nuclear and cytoplasmic maturation in oocytes that have not yet acquired full competence.
Mitochondrial aggregation in GV-stage oocytes before and after IVM
Although chromatin condensation and cessation of transcriptional activity in the oocyte may predict competence to resume meiosis, other spatiotemporal events occurring within the oocyte cytoplasm, such as organelle distribution, have been associated with oocyte cytoplasmic maturation (Nagai et al., 2006; Van Blerkom, 2011). Considering that existing information on this aspect is scarce (Wilding et al., 2001; Liu et al., 2010b), we investigated the patterns of mitochondria distribution in human immature GV-stage oocytes and their possible association with oocyte competence.
Stage-specific mitochondrial aggregation and translocation to the nuclear area occur during oocyte maturation in rodents and porcine (Sun et al., 2001; Van Blerkom et al., 2002; Nishi et al., 2003; Torner et al., 2004; Yu et al., 2010). Changes in mitochondrial redistribution and density within the oocyte are accompanied by a local increase in ATP production, providing energy to sustain crucial maturational processes (Van Blerkom and Runner, 1984; Yu et al., 2010). Inadequate redistribution of mitochondria has been associated with incomplete oocyte cytoplasmic maturity since it affects completion of meiosis and embryo development (Wilding et al., 2001; Dumollard et al., 2006; Nagai et al., 2006; Wang et al., 2009b; Van Blerkom, 2011; Ou et al., 2012).
Previous reports in human oocytes retrieved after hCG triggering in conventional IVF/ICSI ovulation induction cycles have shown distinct patterns of mitochondrial distribution in GV oocytes (Wilding et al., 2001; Dell'Aquila et al., 2009; Liu et al., 2010a,b). In the current study, we observed that mitochondria with a granular aggregation pattern (predominantly restricted to the subplasma membrane area) were predominant in immature GV oocytes analysed both before and after IVM. Our findings concur with some of the earlier studies (Wilding et al., 2001; Dell'Aquila et al., 2009) but do not support observations of Liu et al. (2010b) who found the peripheral distribution to be predominant.
It is perhaps not surprising that the observations on the patterns of mitochondrial aggregation in human immature (GV) or mature (MII) oocytes are divergent among studies, as data were acquired mainly from unphysiological conditions that may have influenced mitochondrial distribution. In the case of immature oocytes, it is important to realize that previous analyses have mostly been done on oocytes that failed to mature in vivo after conventional (high-dose) gonadotrophin ovulation induction and hCG triggering. Furthermore, differences in stimulation protocols to retrieve human mature oocytes have been shown to influence the mitochondrial distribution pattern of in vivo MII-stage oocytes (Dell'Aquila et al., 2009) and may contribute, at least partly, to the variable patterns observed among different studies.
Our observations in immature GV-stage oocytes revealed a progressive translocation of mitochondria to the nuclear area, mainly in freshly isolated oocytes. In murine and porcine species, mitochondrial aggregation around the GV (perinuclear distribution) is a feature of fully-grown GV-stage oocytes (Sun et al., 2001; Van Blerkom et al., 2002; Dumollard et al., 2006). In porcine small antral follicles, only a minority of oocytes demonstrated a perinuclear distribution, whereas internalization of mitochondria towards the nucleus and especially a strong perinuclear distribution was a typical feature of GV oocytes from larger follicles (Sun et al., 2001). Similar observations have been reported in mouse, where oocytes collected 24 h after PMSG-priming had a reduced proportion of perinuclear mitochondrial distribution compared with fully-grown GV oocytes obtained 48 h post PMSG-priming (Ou et al., 2012).
In human, only a small number of studies provide information on the changes in mitochondrial ultrastructure by using transmission electron microscopy (TEM), which also included localization of mitochondria in immature GV oocytes (Motta et al., 2000; Nogueira et al., 2003a; Morimoto, 2009). Nogueira et al. (2003a) reported a mitochondrial distribution throughout the cytoplasm, but mainly localized in the surrounding of the GV, where they formed aggregates and were mixed with vesicles of endoplasmic reticulum. Interestingly, although few oocytes were analysed, they were derived from follicles with a larger diameter than those in our study (<15 and <10 mm, respectively). In consideration of these scarce reports in human GV oocytes, it is possible that perinuclear mitochondrial localization may be considered (as in other species) a feature of oocyte maturity. This hypothesis is supported by data from the current study where only 1% of meiotically incompetent oocytes demonstrated the perinuclear pattern of mitochondrial distribution, whereas this was 20% in oocytes analysed the day of retrieval. Taking into account the association of mitochondrial distribution and developmental competence, it is conceivable that the currently suboptimal rates of embryo development from in vitro matured oocytes from small antral follicles between 2 and 10 mm are the result of their poor cytoplasmic maturity, which is reflected at the mitochondrial level.
Conclusions
Our study documents heterogeneity at subcellular and molecular levels in the pool of oocytes derived from 2 to 10 mm antral follicles retrieved for IVM without prior hCG trigger (Fig. 8). According to our analysis, a considerable amount of oocytes undergoing IVM are still in the process of acquiring final differentiation/maturation (based on features such as oocyte diameter, chromatin configuration and mitochondria distribution), and are accumulating mRNAs for further development, which may also include transcripts involved in meiotic resumption.
Schematic representation of the heterogeneity, at the subcellular and molecular levels, in the pool of oocytes retrieved from follicles 2–10 mm to undergo IVM (according to the observations in the present study). GV: germinal vesicle, Chr: chromatin, Mt: mitochondria.
Previous reports from our group, using a non-hCG priming IVM strategy, have shown similar rates (∼30%) of GV-arrested oocytes even after a longer (40 h) IVM culture period (Guzman et al., 2012). In general, oocytes that remained at GV-stage after 30 h did not resume meiosis by extending the incubation period to 40 h (personal unpublished data).
In different animal species, including human, a prematuration culture (PMC) has been proposed in order to support acquisition of oocyte competence (Nogueira et al., 2006; Vanhoutte et al., 2007, 2009; Shu et al., 2008; Albuz et al., 2010; Dieci et al., 2013; Franciosi et al., 2014; Richani et al., 2014). In murine, culture of COCs for 24 h (under meiotic arrest) promoted chromatin condensation, which resulted in a significantly higher proportion of oocytes with a SN after PMC (82%) compared with those obtained at retrieval (60%) (Nogueira et al., 2003b). Similarly, in bovine, a 24 h-PMC of COCs retrieved from early antral follicles supported a functional connection between oocytes and cumulus cells, which had a positive impact promoting a gradual chromatin condensation. Notably, oocytes exposed to a 24 h-PMC and subsequently undergoing IVM, showed significantly higher maturation, cleavage and blastocyst rates compared with oocytes that were incubated in IVM maturation media immediately after oocyte isolation (Luciano et al., 2011).
Considering the heterogeneity of GV oocytes aspirated from 2 to 10 mm follicles there is a rationale for a prematuration culture in order to permit completion of nuclear maturation and progression of cytoplasmic maturation of unexpanded cumulus-oocyte complexes before inducing meiotic resumption in vitro.
Authors' roles
F.S. contributed in the study design, experimental work, data analysis and interpretation and writing of the manuscript; S.R. contributed in the study design, experimental work, data analysis and interpretation, and revised the final version of the manuscript. M.D.V. performed patient recruitment and management, supervised the clinical activities and revised the final version of the manuscript; G.V. supervised the laboratory activities in the fertility clinic and revised the final version of the manuscript; J.S. supervised the laboratory activities, participated in the study design, critical discussion and revision of the manuscript, and is the principal investigator of the oocyte maturation project.
Funding
In vitro maturation research at the Vrije Universiteit Brussel has been granted by: the Institute for the Promotion of Innovation by Science and Technology in Flanders (Agentschap voor Innovatie door Wetenschap en Technologie – IWT, project 130327 and 110680); the Fund for Research Flanders (Fonds Wetenschappelijk Onderzoek–Vlaanderen – FWO, project G.0343.13), the Belgian Foundation Against Cancer (HOPE project) and COOK Medical.
Conflict of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Acknowledgements
The authors wish to thank the laboratory technicians, embryologists and paramedical staff at the Centre for Reproductive Medicine (CRG, UZ Brussel) for their kind help during the collection of the oocytes. We are also thankful with Wim Coucke from Scientific Institute of Public Health (Brussels) for performing part of the statistical analysis and Sandra De Schaepdryver for the editorial support.
