Author's address (for correspondence): M Saint-Dizier, AgroParisTech, 16 rue Claude Bernard, 75231 Paris Cedex 05, France. E-mail: firstname.lastname@example.org
In the bitch, oocyte meiosis resumption takes place in the oviduct. Using oocytes from anestrous bitches, in vitro maturation (IVM) generally gives very poor results. To investigate the contribution of oocyte competence to the low IVM yield, we compared in vivo maturation in an optimal environment with conventional IVM. A total of 418 grade 1 cumulus-oocyte complexes (COCs) from 10 anestrous bitches were transferred into the oviducts of recipient bitches either on Day −1 (n = 3 recipients), Day 0 (n = 2) or on Day +1 (n = 2) relative to ovulation. For each donor bitch, 20 grade 1 COCs were also cultured in vitro. After 72 h of in vivo or IVM, the nuclear stage of oocytes was determined after DNA and tubulin staining. Of the 154 oocytes recovered and examined after intratubal transfer, only 2% reached the metaphase I or II stage and 38.3% were degenerated. Oocytes cultured in vitro displayed a higher metaphase rate (7.6%, n = 170) and lower degeneration rate (12.9%) compared with transferred oocytes (p < 0.001). These results clearly demonstrate that the oocyte competence is the major limiting factor of IVM efficiency in the dog. Mimicking the tubal environment may thus not be sufficient to increase IVM yield in this species.
In vitro maturation (IVM) of canine oocytes represents the main limiting factor in further development of assisted reproductive technologies in the dog. Despite considerable efforts made over the last 15 years to improve the IVM yield, no more than 10–20% of cultured canine oocytes reach the metaphase I or M II (MI/II) stage in vitro, and approximately 25% of these degenerate (for reviews, see Luvoni et al. 2005; Chastant-Maillard et al. 2011). To date, little is known about the factors that regulate oocyte maturation in vivo in this species. Meiosis resumption is highly specific in canids compared with other mammals: indeed, canine oocytes are ovulated at the germinal vesicle (GV) stage and complete their nuclear maturation after 2–3 days spent in the oviduct (Tsutsui 1989; Reynaud et al. 2005). Moreover, because of the luteinization of pre-ovulatory follicles, oocyte meiosis resumption occurs in the presence of high circulating concentrations of progesterone (P4) (Chastant-Maillard et al. 2011), placing maturing oocytes in a very specific endocrine environment. For practical reasons, IVM in the canine is usually performed using oocytes collected in the course of routine neutering of anestrous or diestrous bitches. Low maturation rates observed in vitro may be related to the poor quality of oocytes (harvested from small follicles) and/or to inadequate culture conditions. To improve the latter, IVM of canine oocytes was performed either in synthetic oviductal fluid (SOF) (Hewitt and England 1999; Rota and Cabianca 2004; Saikhun et al. 2008), in co-culture with monolayers of oviductal epithelial cells (Hewitt and England 1999; Bogliolo et al. 2002; Vannucchi et al. 2006), or ex vivo in isolated oviducts (Luvoni et al. 2003): yet, even in these culture conditions, no more than approximately 30% of oocytes reached metaphase stages. Nevertheless, none of the earlier studies reproduced entirely the in vivo conditions of canine oocyte maturation. To precisely evaluate the relative importance of oocyte and of environmental conditions in the low yield of IVM, we performed intratubal transfer of oocytes from anestrous bitches into recipient bitches around the time of ovulation. Maturation rates of transferred oocytes were compared with those obtained after conventional in vitro culture.
Materials and Methods
Recovery of oocytes from donor bitches
The experimental protocol was approved by the ethical committee of the Alfort National Veterinary School. Canine ovaries were obtained from 10 mongrel and purebred bitches (3.6 ± 0.8 years) in anestrus. Routine neutering was performed at the Alfort veterinary hospital. Ovaries were placed in warm PBS and processed within 2 h of collection as previously described to release cumulus-oocyte complexes (COCs) (Chebrout et al. 2009). Only grade 1 COCs were used in this study, that is, regularly shaped oocytes, >130 μm in diameter (zona pellucida included), with homogeneous dark cytoplasm, surrounded by at least two layers of cumulus cells.
In vitro maturation
From each donor bitch, a group of 20 COCs randomly chosen were cultured in vitro in 500 μl of M199 supplemented with 20% foetal calf serum (Sigma-Aldrich, Saint-Quentin Fallavier, France) under a 5% CO2 atmosphere in air at 38.5°C for 72 h. The remaining COCs from each donor bitch were transferred in one oviduct of a recipient bitch.
Seven pubertal Beagle bitches (2.8 ± 0.5 years) from our experimental kennel were used as recipient bitches. The day of ovulation (Day 0) was determined using vaginal smears, plasma P4 assay and ultrasound examination of ovaries, as previously described (Reynaud et al. 2005). Intratubal transfer was performed either on Day −1 (range of plasma P4 concentrations [P4] = 3.6–3.9 ng/ml, n = 3 recipients), Day 0 ([P4] = 6.7–8.3 ng/ml, n = 2 recipients) or on Day +1 ([P4] = 13.0–28.0 ng/ml, n = 2 recipients) relative to ovulation. After puncture of the oviductal wall, a catheter (20G, 1 × 1.3 mm; Vasocan Braunule, B. Braun Melsungen AG, Germany) was inserted into the upper part of each oviduct at a site approximately 1 cm from the infundibulum: 22–42 COCs (mean of 30 ± 2 COCs per oviduct) were injected within 2 h after oocyte recovery through a glass capillary passed into the catheter. In a preliminary study, this procedure was found to not induce any COC denudation (data not shown). Finally, a total of 51–82 COCs (mean of 60 ± 4 COCs per recipient, n = 7) was transferred to each recipient bitch.
Recovery and examination of oocytes from recipient bitches
Seventy-two hours after intratubal transfer (i.e. 48, 72 or 96 h post-ovulation), ovaries and oviducts were collected from recipient bitches by routine ovariectomy. Corpora lutea (CL) were counted on each ovary by gross examination, and oviducts were dissected from fat and vessels. Both oviducts were flushed as previously described (Reynaud et al. 2005), and the collected liquids examined under a stereomicroscope (SMZ645; Nikon, Champigny-sur-Marne, France) for morphological assessment and classification of harvested structures.
Nuclear stage determination
All oocytes were denuded (when necessary), then fixed for staining of microtubules and chromatin as previously described (Reynaud et al. 2005). The nuclear stage was evaluated by confocal laser scanning microscopy. Oocytes were regarded as degenerated when DNA was absent or abnormal (Chebrout et al. 2009).
Data are given as mean ± SEM. The proportions of oocytes were compared between maturation conditions, recipients and days of transfer with chi-squared analysis using the FREQ procedure of the SAS software (version 9.2; SAS Institute Inc., Cary, NC, USA). Differences were considered significant when p < 0.05.
Classification of collected oocytes
A total of 222 structures were recovered from recipient bitches: 42 were oocytes surrounded by two to three layers of compact cumulus cells, 161 were denuded oocytes or oocytes with only a few cumulus cells still attached to the zona pellucida, and the remaining 19 were empty zona pellucida (Fig. 1). In each recipient bitch, the number of oocytes surrounded by compact cumulus cells was always lower than or equal to the number of CL (Table 1). Furthermore, 19.2% (n = 26) of surrounded oocytes were at the metaphase I or II (MI/II) stage when tubal flushing was performed at Day +2 relative to ovulation of the recipient bitches, and this proportion reached 100% (n = 13) at both Days +3 and +4 (Table 1, Fig. 2a,b). These kinetics of meiosis resumption were identical to those recorded in ovulated oocytes in vivo (Reynaud et al. 2005). On the basis of all three criteria (morphology, number and maturation kinetics), surrounded oocytes were considered as those that had been ovulated by the recipient bitches, whereas the other structures were considered to be the transferred oocytes from donor bitches.
Table 1. Ovulated oocytes from recipient bitches
Day of transfer relative to ovulation (Day 0)
No. of recipient bitches
Time of recovery relative to ovulation
No. of CL in recipient bitches
No. of surrounded oocytes in recipient bitches
Total no. of ovulated oocytes (recovery rate)
No. of oocytes observed under confocal microscopy
Nuclear stage of observed oocytes (maturation rate)
The recovery rate of ovulated oocytes was calculated as the ratio of the number of recovered surrounded oocytes to the number of corpora lutea. Values in square brackets were combined to calculate percentage of oocytes at MI/II. CL, corpora lutea; GV, germinal vesicle; Deg., degenerated oocytes.
[15, 11, 7]
[13, 9, 7]
Maturation of transferred and in vitro cultured oocytes
On the basis of morphological criteria, transferred oocytes were denuded or persisted only as empty zona pellucida (Fig. 1). The results of maturation of transferred oocytes, excluding empty zona pellucida, and those of in vitro cultured oocytes, are summarized in Table 2. The proportion of oocytes resuming meiosis after intratubal transfer was lower than that after in vitro culture (2.0% vs 7.6% MI/II, respectively; p < 0.001; Fig. 2–c–f). Furthermore, the rate of oocyte degeneration was higher after intratubal transfer than after in vitro culture (38.3% vs 12.9%, respectively; p < 0.001). The day of transfer relative to ovulation in the recipient bitch affected the rates of oocyte maturation and degeneration after transfer: a higher proportion of oocytes resumed meiosis and a lower proportion degenerated when transferred on Day +1 than on either Day −1 or Day 0 (7.1%, 1.2% and 0% of MI/II, respectively, p < 0.001; and 10.7%, 29.9% and 76.9% of degeneration, respectively; p < 0.001; Table 2).
Table 2. Oocytes from donor bitches 72 h after intratubal transfer or in vitro culture
Day of transfer relative to ovulation (Day 0)
No. of recipient bitches
No. of transferred COCs (nb donor bitches)
No. of collected oocytes
No. of oocytes observed under confocal microscopy
Nuclear stage of observed oocytes (maturation rate)
No. of oocytes in vitro cultured.
Empty zona pellucida and ovulated oocytes were excluded from these data. Values in square brackets were combined for statistical analysis. Values with different superscripts within the same column are significantly different (p < 0.001). GV, germinal vesicle; Deg., degenerated oocytes; COCs, cumulus-oocyte complexes.
In contrast to other mammals, meiosis resumption of oocytes in the dog takes place in the oviduct 2–3 days after ovulation (Reynaud et al. 2005). Intratubal transfer around the time of ovulation provides therefore an environment naturally supportive of oocyte meiosis resumption. This study clearly demonstrates that, although placed under these optimal conditions, oocytes recovered from anestrous ovaries were unable to survive and resume meiosis. Our findings show for the first time that oocyte meiotic competence is the major factor limiting the yield of IVM in the canine species.
Intratubal transfer performed around the time of ovulation leads to a mix of oocytes ovulated by the recipient bitch with those transferred from the donor bitch. Well-established characteristics of in vivo matured canine oocytes were used to differentiate these two populations: the presence of two to three layers of cumulus cells in close contact to the zona pellucida (Reynaud et al. 2005, 2009; Chastant-Maillard et al. 2011) and the nuclear stage according to the time elapsed since ovulation, and the low incidence or absence of oocyte degeneration (Tsutsui 1989; Reynaud et al. 2005; Chastant-Maillard et al. 2011). Furthermore, the number of oocytes that fitted the above criteria was equal or similar to the number of CL in each recipient bitch. For all these reasons, surrounded oocytes were considered as ovulated oocytes. Moreover, empty zona pellucida and denuded oocytes were not observed in intratubal contents collected from bitches after ovulation (Tsutsui 1989; Renton et al. 1991; Reynaud et al. 2005). Thus, denuded oocytes and empty zona pellucida were considered as transferred oocytes.
In the present study, oocytes from anestrous ovaries reached degeneration (13%) and meiotic resumption (8%) rates in accordance with those reported in the literature after 72 h of IVM (Vannucchi et al. 2006; Otoi et al. 2007; Chebrout et al. 2009). Surprisingly, transferred oocytes displayed significantly higher degeneration and lower metaphase rates (38% and 2%, respectively) compared with in vitro cultured oocytes. The collection of empty zona pellucida, which reflected a severe oocyte degeneration after a 3 days stay in the recipient oviduct, was also unexpected. Indeed, the lifetime of ovulated oocytes in vivo is longer in the bitch than in other mammals: oocytes may remain fertilizable for as long as 8 days (Tsutsui et al. 2009). In previous in vitro studies, the degeneration rate of canine oocytes was not modified by the co-culture with oviductal cells (Bogliolo et al. 2002), or even lowered when oocytes were cultured in isolated oviducts for a short period of 24–30 h (Luvoni et al. 2003). The intratubal environment around ovulation was expected to provide optimal conditions for oocyte survival, in terms of nutrient supply, hormone concentrations, oxygen tension or cell interactions. However, although an increase in oocyte density was reported to be beneficial during IVM (Otoi et al. 2007), the relatively high density of transferred oocytes within recipient oviducts (30 ± 2/oviduct) may have limited the beneficial effects of the in vivo intratubal environment in the present study.
At the time of recovery in this study, transferred oocytes were partially or totally denuded, although they were selected with at least two layers of cumulus cells prior to intratubal transfer. Spontaneous denudation of oocytes from anestrous and diestrous bitches was reported to occur at a rate of only 12–13% during IVM (Otoi et al. 2007; Chebrout et al. 2009). Denudation was not an artefact associated with the flushing technique, since the surrounding cumulus cells were preserved around ovulated oocytes. Furthermore, the same technique is routinely used without damage in our laboratory to collect in vivo surrounded oocytes after ovulation (Reynaud et al. 2009). It is thus likely that the tubal environment was responsible for the denudation of the transferred oocytes. The muscular contractions of the oviductal wall increase around ovulation in most mammals (Hunter 2012), and several proteases have been identified in the mammalian tubal fluid (Aviles et al. 2010). It remains difficult to determine from our results whether the denudation of transferred oocytes was a cause or a consequence of their degeneration. The impact of oocyte denudation on the rate of degeneration gave divergent results in vitro: the proportion of degenerated oocytes was either increased (Chebrout et al. 2009) or decreased (Otoi et al. 2007) by spontaneous denudation, whereas it was not affected by mechanical denudation (Otoi et al. 2007; Chebrout et al. 2009).
In addition to a high degeneration rate, only 2% of transferred oocytes reached metaphase stages after 72-h maturation in recipient oviducts. Oocytes collected from the same ovaries and cultured in vitro reached a significantly higher metaphase rate (8%). Studies that used SOF medium to mature canine oocytes in vitro did not report any beneficial effects on the rate of meiosis resumption (Hewitt and England 1999; Rota and Cabianca 2004; Saikhun et al. 2008). When co-culture with oviductal cells was compared with the culture medium alone, metaphase rates of oocytes from various stages of the oestrous cycle were either identical (Hewitt and England 1999) or dramatically increased (29–30.4 vs 6% without oviductal cells) (Bogliolo et al. 2002). Using oocyte culture in an isolated oviduct, higher MI/II rate were also achieved compared with conventional drops of culture medium (31.9 vs 3.7% after 30 h) (Luvoni et al. 2003). However, this ex vivo tubal environment was not so efficient when the culture time was slightly extended: the metaphase rate decreased to 3.7% after 48 h in culture (Luvoni et al. 2003), probably due to the dedifferentiation and degeneration of the oviduct explants.
Our findings showed that in vivo culture of canine oocytes from Day +1 to Day +4 relative to ovulation supported higher maturation and lower degeneration rates than those from earlier periods of time (Day −1 to Day +2 or Day 0 to Day +3). The tubal fluid is a complex milieu formed by selective filtration of plasma, de novo synthesis from oviductal epithelial cells, and, presumably, follicular fluid and cells from ovulatory follicles (Hunter 2012). It can be hypothesized that some factors that stimulate the meiotic resumption in the oocyte appear or increase in concentration at Day 3–4 post-ovulation in the canine tubal fluid compared with previous days. However, the low metaphase rate observed after transfer at Day +1 (7.1%) limits the conclusions that can be drawn.
Oocytes obtained from anestrous ovaries in the present study were globally non-competent for meiosis resumption, either after in vitro or in vivo culture. The poor competence of these oocytes may be related to the small size of their respective follicles (0.5–1 mm vs 6–8 mm before ovulation) (Chastant-Maillard et al. 2011). As in other mammals, the meiotic competence of canine oocytes was reported to increase with the diameter of the donor follicle (Songsasen and Wildt 2005). Structural studies demonstrated the marked immaturity of canine oocytes until late in the follicular growth and up to the LH peak (Viaris De Lesegno et al. 2008). The present work clearly confirms the very low meiotic competence of oocytes from anestrous ovaries, the lack of follicular maturation being not overcome by a 3-day maturation within the periovulatory tubal environment in vivo.
In conclusion, the present study provides clear evidence that the meiotic competence of oocytes is the primary factor responsible for the low IVM rates obtained in the bitch. Culture media mimicking the post-ovulatory tubal environment will not suffice to improve the results of IVM with oocytes from anestrous bitches. Rather, a two-step culture, with a first ‘follicular-like’ period leading to the acquisition of meiotic competence, followed by a second ‘oviductal-like’ period for meiosis resumption, has to be evaluated for oocytes from anestrous bitches.
Authors are grateful to Marc Chodkiewicz for the critical review of this paper.
Conflicts of interest
None of the authors have any conflicts of interest to declare.