Cumulus Oocyte Communications in the Horse: Role of the Breeding Season and of the Maturation Medium

Transcription

1 Reprod Dom Anim 39, (2004) Ó 2004 Blackwell Verlag, Berlin ISSN Cumulus Oocyte Communications in the Horse: Role of the Breeding Season and of the Maturation Medium S Colleoni 1, AM Luciano 2 and FGandolfi 2 1 Laboratorio di Tecnologie della Riproduzione (LTR), CIZ, Istituto Sperimentale Italiano Lazzaro Spallanzani, Cremona, Italy; 2 Department of Anatomy of Domestic Animals, University of Milan, Milan, Italy Contents Horse is a seasonal breeder and information on oocyte quality outside the breeding season is very limited. Ovaries obtained at the slaughterhouse are a convenient but often limited source of oocytes in this species. As the low quantity of ovaries leads to an intensive use of all available material, it would be useful to know whether ovaries collected during the non-breeding season are suitable for in vitro maturation (IVM). In an attempt to characterize the effect of season on oocyte quality, we investigated the permeability of the gap junctions (GJ) present between cumulus cells and oocytes because of their important role in oocyte growth and maturation. We also compared the effect of supplementing the maturation medium with bovine serum albumin (BSA) or oestrus mare serum (EMS). A total of 645 oocytes isolated from 158 and 154 ovaries collected during the breeding and the non-breeding season, respectively, were used in this study. Oocytes were matured for 30 h in TCM 199 supplemented either with 10% EMS or with 4 mg/ml BSA. The presence of permeable GJs between cumulus cells and oocytes was investigated with the injection of a 3% solution of the fluorescent dye Lucifer yellow into the ooplasm. No differences in efficiency of oocyte retrieval or oocyte meiotic competence were detected between oocytes collected during the breeding and non-breeding season. The vast majority (90%) of the oocytes collected during the breeding season had fully functional communications with their surrounding cumulus cells but such communications were completely interrupted in 55.3% of the oocytes collected during the nonbreeding season. During the non-breeding season, the proportion of oocytes whose communications with cumulus cells were classified as closed or intermediate at the end of maturation was lower in the group matured with BSA than with EMS (71.4 vs 97.7, p < 0.05). The same trend, although not statistically significant, was observed during the breeding season also. The presence of BSA caused an incomplete cumulus expansion during both seasons. Our data indicate that oocytes collected during the non-breeding season do not show any meiotic deficiency but lack active communication with the surrounding cumulus cells at the time of their isolation from the ovary. No data are available at present for determining the consequences on the developmental competence even if data from other species suggest that this is likely. Introduction The in vitro production of horse embryos would be very useful in the case of female and male infertility as well as for developing laboratory tests for the evaluation of frozen semen or for improving embryo freezing and culture methods (Squires et al. 2003). Moreover, a reliable source of mature horse oocytes is required for developing cloning methods in this species (Galli et al. 2003). However, in vitro production of horse embryos has not been very efficient and is one of the reasons for the limited number of oocytes available for research (Squires et al. 2003). Oocytes can be obtained through transvaginal ultrasound-guided follicular aspiration in mares but this is expensive and a poor recovery rate is often reported (Galli et al. 2001; Bogh et al. 2002; Maclellan et al. 2002). Horse ovaries obtained at the slaughterhouse are a more convenient source of oocytes (Alm et al. 1997), but are often available in small numbers and often entails a tedious trip from the slaughterhouse to the laboratory (Love et al. 2003). Therefore, the limited quantity of available ovaries has lead to an intensive use of all available material. Horse is a seasonal breeder and its reproductive efficiency varies throughout the year. In the northern hemisphere mares enter anestrus and stallion semen quality sharply declines during late fall and early winter (Gerlach and Aurich 2000; Nagy et al. 2000). Information on oocyte quality outside the physiological reproductive season is very limited. Only two studies evaluated meiotic competence of oocytes aspirated from ovaries collected at the slaughterhouse throughout the year (Hochi et al. 1993; Hinrichs and Schmidt 2000). Both reports agree that oocytes collected during the non-breeding season can reach the metaphase II at the same rate as those collected during the rest of the year but no other parameters were available for further evaluating their actual quality. However, the ability of the oocytes to reach the metaphase II does not necessarily mean that they have full developmental competence (Brevini-Gandolfi and Gandolfi 2001; Sirard 2001). In an attempt to further characterize the effect of season on oocyte quality, we investigated the permeability of the junctions between cumulus cells and oocytes because this is a relevant parameter for evaluating oocyte quality for several reasons. Oocytes are coupled with the surrounding follicle cells via gap junction (GJ) from the primary through the antral stage (Kidder and Mhawi 2002) and through these intercellular channels, follicle cells transfer amino acids, glucose metabolites and nucleotides to the growing oocyte (Eppig 1991). Signals that regulate meiotic progression are known to flow through GJs connecting cumulus cells and fully grown oocytes (Tanghe et al. 2002). Finally, communications between cumulus oophorus and oocyte are required for cytoplasmic maturation and for acquiring the capacity to support male pronucleus formation and early embryonic development (Moor et al. 1998). U.S. Copyright Clearance Centre Code Statement: /2004/ $15.000/0

2 Cumulus Oocyte Communication in Horse IVM 71 While oocyte quality is certainly the most important component for successful in vitro embryo production, maturation medium has an important role too. Tissue culture medium 199 (TCM 199) is the most widely used medium for in vitro oocyte maturation in many species including horse (Squires 1996) and the addition of serum has generally proved to better support nuclear maturation than alternatives such as bovine serum albumin (BSA) (Willis et al. 1991) or follicular fluid (Aguilar et al. 2001). In this study, we examined whether the quality of horse oocytes collected at the slaughterhouse for IVM was affected by the time of the year (breeding vs nonbreeding season) and, in turn, whether this caused a different response to the macromolecular supplementation [oestrus mare serum (EMS) vs BSA] of the maturation medium. We evaluated the permeability of communications between oocytes and cumulus cells and oocyte meiotic competence as quality parameters. Materials and Methods Media and chemicals All chemicals were purchased from Sigma Chemical Co. (St Louis, MO, USA) unless otherwise stated. Ovaries were transported to the laboratory and rinsed in Dulbecco modified phosphate balance solution at ph 7.5 supplemented with antibiotic/antimycotic solution (Sigma A9909). Oocytes were aspirated from the follicles and selected in TCM 199 with Earle s salts (Sigma M3769), 0.68 mm L-glutamine, 25 mm Hepes buffer, 9 mm sodium pyruvate, 75 lg/ml penicillin-g, 50 lg/ml streptomycin sulphate (TCM 199H). Oocyte maturation was the same as TCM 199-H but buffered with 2.2 g/l sodium bicarbonate without Hepes (TCM 199M), supplemented with 1 IU/ml of humanrecombinant follicle stimulating harmone (FSH) (Gonal-F, Serono, Italy) and 4 mg/ml fatty acid-free BSA (Sigma A7511) or 10% of EMS. In order to avoid possible effects caused by collection from a single individual, the oestrus mare blood was collected from a pool of three mares. Each mare was monitored through oestrus using a teaser stallion, rectal palpation and ultrasound examination. Collection was performed after the second day of oestrus and before ovulation in order to choose a time of oestrus as similar as possible amongst the three mares. Serum was heat inactivated at 56 C for 30 min, sterilized with 0.22 lm pore Millipore filters and stored at )20 C. The same lot was used for all the experiments. Oocyte collection and IVM Ovaries collected from mares slaughtered during the non-breeding (January February) and breeding (April June) season were transported to the laboratory within 3 h at C. Ovaries were washed three times in phosphate-buffered saline (PBS) supplemented with antibiotic/antimycotic solution and aspirated with a 14-gauge needle after removal of the tunica albuginea. The needle was attached to a collection tube connected at a water vacuum pump and the aspiration pressure was mmhg. Cumulus oocyte complexes (COC) were aspirated from follicles ranging between 10 and 30 mm in diameter. Only oocytes surrounded by at least four cell layers of compact cumulus cells and homogeneous cytoplasm were selected for culture. COCs were matured at 38 C for 30 h in TCM 199M supplemented with either 4 mg/ml BSA or 10% (v/v) EMS. Assessment of nuclear morphology and dye-coupling assay At the time of collection (0 h) and at the end of maturation (30 h) a total of 213 oocytes were mechanically separated by the surrounding cumulus cells and fixed in ethanol : acetic acid (3 : 1) for 24 h. Chromatin was stained with lacmoid solution and its morphology was evaluated under a phase-contrast microscope (Nikon Diaphot, Kawasaki, Japan). The remaining 432 cumulus-enclosed oocytes were injected with a 3% solution of 5 mm LiCl of the fluorescent dye Lucifer yellow (LY) to evaluate the permeability of GJ communications between oocytes and cumulus cells. The oocytes were injected at the time of collection (0 h) and at the end of maturation (30 h). As most COCs (as described below) were classified as open at the time of collection only during the breeding season, a detailed analysis of the GJ functionality during maturation was limited to this group. COCs were microinjected at 6, 9 and 18 h of culture. Some COCs were microinjected at the time of collection in each experiment to confirm that GJ were opened. The COCs were placed in 50 ll droplets of TCM 199H under mineral oil on the heated stage (37 C) of an inverted microscope (Nikon Diaphot) equipped with an epi-fluorescent light (excitation 490 nm, emission 510 nm) and a microinjection apparatus (Narishige Co. Ltd, Tokyo, Japan) to guide the holding and injecting micropipettes. Holding pipettes were made from borosilicate glass capillaries without inner filament [outer diameter (OD), 1 mm, inner diameter (ID), 0,58 mm; Clark, Pangbourne, UK); they were hand pulled and then fire polished with a microforge to have an ID of lm. Injection needles were pulled with a micropipette puller from borosilicate glass capillaries with inner filament (OD 1 mm, ID 0,78 mm; Clark); final ID was <1 lm. Both holding and injection pipette were bent to an angle of 30. The diffusion of LY from the oocyte cytoplasm to the cumulus cells was evaluated 10 min after injection. The COCs were classified as open when at least 80% of corona radiata cells were fluorescent, as intermediate when only a limited number of cells showed signs of dye diffusion between ooplasm and corona radiata cells and as closed when the LY was confined to the cytoplasm or only when few cells showed fluorescence (Fig. 1). Statistical analysis All experiments were replicated at least three times. Oocyte distribution within the different categories was analysed by the chi-square test and the criterion for

3 72 S Colleoni, AM Luciano and F Gandolfi Fig. 1. Representative pictures of horse oocytes injected with the fluorescent dye Lucifer yellow to detect the functional status of the gap junctions (GJs) connection with cumulus cells. On the right, the same cumulus oocyte communication (COC) observed under bright field. (A,a) Fully open GJs; (B,b) partially open GJs; (C,c) closed GJs (scale bars represent 100 lm) significance was set at p < Results are expressed as total numbers and/or percentages. Results A total of 154 and 158 ovaries were used for this experiment during the non-breeding and breeding season, respectively. No significant differences were detected between ovaries collected during the breeding and non-breeding season with respect to the number of follicles aspirated for each ovary or the number of oocytes recovered (Table 1). During the non-breeding season, only 21.3% of the COCs had fully functional communications between oocyte and cumulus cells at the time of their isolation from the follicles. This was significantly different from the breeding season when in 90% of the oocytes, the GJ communications with their surrounding cumulus cells were found to be open (Table 2). After 30 h of maturation the macromolecular supplementation of the culture medium (BSA vs EMS) affected the permeability of COCs as indicated by the difference between the proportion of oocytes whose communications with cumulus cells were classified as closed or intermediate at the end of maturation (30 h; p ¼ 0.05, Table 2). Although this effect was statistically different only during the non-breeding season, a similar trend was observed during the breeding season also. The detailed analysis of GJ functionality during IVM of COCs collected during the breeding season showed that a progressive interruption of communications takes place during maturation. This process is mostly completed within 18 h of maturation (Table 3). The difference in GJ permeability at the time of isolation from the follicles between oocytes collected during the non-breeding and breeding season was not accompanied by a difference in their nuclear morphology (0 h; Table 4), with the vast majority of the oocytes being at the germinal vesicle stage. At the end of maturation, a more complete cumulus expansion was observed in the COCs matured with EMS than with BSA but no differences could be observed when the season was considered. Finally, neither the maturation medium supplement nor the season had any effect on the oocyte maturation rate. Discussion The aim of the present work was to determine whether mare oocytes collected during the non-breeding season could be used for the in vitro production of horse embryos. When we considered the efficiency of oocyte retrieval or oocyte meiotic competence no differences were detected between oocytes collected during the breeding and non-breeding season. This is consistent with previous data (Hochi et al. 1993; Hinrichs and Schmidt 2000) and confirms the fact that ovaries obtained at the slaughterhouse are one of the most convenient source of oocytes in horses also. However, nuclear maturation is not a good predictive marker of oocyte developmental competence particularly in horse where IVM is always followed by low fertilization rates and by the almost total lack of development both in vitro and in vivo. This situation severely limits the possibility of adequately answering the initial question of whether out-of-season oocytes are good material for in vitro embryo production or not. The analysis of communications between cumulus cells and oocyte was performed in order to gain a deeper knowledge of the potential oocyte quality. Small molecules like amino acids, nucleotides and glucose are

4 Cumulus Oocyte Communication in Horse IVM 73 Table 1. Effect of season on efficiency of oocyte retrieval Season Number of ovaries Aspirated follicles per ovary Recovered oocytes per ovary % Recovered oocyte per follicle Breeding Non-breeding Table 2. Effect of season and macromolecular supplementation of IVM medium on mare oocytes cumulus cell communications Time Season IVM medium supplement n Closed Intermediate 0 h Non-breeding* (55.3) 11 (23.4) 10 (21.3) Breeding* (10.0) 27 (90.0) 30 h Non-breeding BSA* (71.4) 8 (22.8) 2 (5.8) Breeding BSA (70.9) 7 (22.6) 2 (6.5) Non-breeding EMS* (97.3) 0 1 (2.7) Breeding EMS (90.9) 3 (9.1) 0 Open *p ¼ Table 3. Effect of IVM medium supplement on cumulus oocyte communications (COCs) during IVM of COCs collected during the breeding season Time (h) IVM medium supplement n Closed Intermediate (10.0) 27 (90.0) 6 BSA 22 2 (9.1) 3 (13.6) 17 (77.3) EMS 20 1 (5.0) 3 (15.0) 16 (80.0) 9 BSA 22 5 (22.7) 8 (36.3) 9 (40.9) EMS 23 9 (39.1) 6 (26.1) 8 (34.8) 18 BSA (68.4) 4 (21.1) 2 (10.5) EMS (79.0) 2 (10.5) 2 (10.5) 30 BSA (71.0) 7 (22.6) 2 (6.4) EMS (90.9) 3 (9.1) 0 Open Table 4. Effect of season and IVM medium supplement on nuclear morphology and oocyte meiotic competence Time (h) Season IVM n GV/GVBD MI MII Deg CE 0 h Non-breeding (94.4) (5.56) ) Breeding (93.4) 1 (3.3) 0 1 (3.3) ) 30 h Non-breeding BSA 48 3 (6.3) 5 (10.4) 35 (72.9) 5 (10.4) ± Breeding BSA 32 1 (3.1) 1 (3.1) 27 (84.4) 3 (9.4) ± Non-breeding EMS 37 3 (8.1) 1 (2.7) 23 (62.2) 10 (27.0) + Breeding EMS (70.0) 9 (30.0) + GV/GVBD, germinal vesicle/germinal vesicle break down; MI, metaphase 1; MII, metaphase 2; Deg, degenerated oocyte; CE, degree of cumulus cells expansion (), no expansion; ±, partial expansion; +, full expansion). transferred from the cumulus cells to the oocytes through the GJs throughout the entire folliculogenesis (Eppig 1991). When the oocyte reaches its full development, these communicating channels are used for the passage of the signals that regulate meiotic maturation (Downs 1995). Consistent with their functional role during follicular development and oocyte maturation, only a minority (21.3%) of oocytes isolated from ovaries collected during the non-breeding season had fully functional communications with their surrounding cumulus cells. A reduced communication between oocyte and cumulus cell was also observed in anestrus bitches. In this species, however, the lack of communication between oocytes and cumulus cells was accompanied by the inability of the oocyte to reach the second metaphase (Luvoni et al. 2001). The difference in COCs was not related to a lower capacity of the oocyte to complete meiosis and no differences were observed in the rate of metaphase II at the end of maturation. In a way this is surprising because a severe disruption of intercellular communication caused by vitrification of horse oocytes was accompanied by a significant reduction in meiotic competence (Hochi et al. 1996). However, other alterations of the oocyte ultrastructure were observed in this work and the additive effect of different lesions may have been the ultimate reason for the defective nuclear maturation observed after vitrification. Indeed horse oocyte maturation, as in any other species, is accompanied by the relocation of cytoplasmic organelles including mitochondria, cortical granules, membrane-bound vesicles and lipids droplets (Grondahl et al. 1995). Cortical granules, in particular, have been studied and it was observed that they are distributed throughout the cytoplasm at the germinal vesicle stage and then migrate to form a monolayer adjacent to the plasma membrane at the metaphase II stage (Goudet et al. 1997; Carneiro et al. 2002). However, in horse oocytes, relocation of cortical granules seems to be always correlated in the stage of nuclear maturation

5 74 S Colleoni, AM Luciano and F Gandolfi (Goudet et al. 1997; Carneiro et al. 2002) and therefore this parameter does not provide additional information on the competence level of the oocyte. Our results indicate that the flux between cumulus cells and oocyte is almost completely interrupted after 18 h of maturation. We are not aware of other data available on the timing of COC in this species but this interval is considerably longer than the one observed in cattle oocytes where communications are interrupted within 6 9 h of maturation (Sutovsky et al. 1993). This reflects the longer time required by horse oocytes to complete the maturation in vitro. The role of the macromolecular supplementation of the maturation medium seems to be more relevant. During the non-breeding season, it could be observed that supplementing the maturation medium with BSA had an inhibitory effect on the complete closure of COC at the end of maturation. The same trend was observed also during the breeding season also. The prolonged communication between oocyte and cumulus observed in the BSA-supplemented medium was accompanied by an incomplete cumulus expansion while the addition of EMS induced both a full cumulus expansion and a complete interruption of communications. This seems consistent with previous data showing that the addition of mare serum, EMS or fetal calf serum to the medium determines an oocyte maturation rate higher than the addition of BSA (Willis et al. 1991). The serum effect seems to be rather unspecific as no differences were observed between EMS and oestrus cow serum (Dell Aquila et al. 1996) and, more recently, the use of newborn calf serum or fetal calf serum has also become very common (Hinrichs and Schmidt 2000; Tremoleda et al. 2003). However, while oocyte and cumulus communications during the early phases of IVM are known to be beneficial to oocyte maturation, their role during later stages is largely unknown. In fact, the prolonged communication between oocyte and cumulus observed in the BSAsupplemented medium was also accompanied by maturation rates that tended to be higher and number of degenerated oocytes that tended to be lower than in the EMS-supplemented medium. Data did not reach statistical significance but might constitute a useful indication. In conclusion, our findings indicate that oocytes collected during the non-breeding season differ from those collected during the breeding season only for the lack of active communication with the surrounding cumulus cells at the time of their isolation from the ovary. No data are available at present for determining if this may have consequences on their developmental competence even if data from other species suggest that this is likely. However, the lack of active communication between the germinal and somatic compartments of the mare ovary could simply reflect the temporary quiescent status of the organ without any consequences on the oocyte as these ovaries resume their normal function within a few weeks. Acknowledgements This study was supported by the MIUR-COFIN grant References Aguilar JJ, Woods GL, Miragaya MH, Olsen LM, Vanderwall DK, 2001: Effect of homologous preovulatory follicular fluid on in vitro maturation of equine cumulus oocyte complexes. Theriogenology 56, Alm H, Torner H, Kanitz W, Becker F, Hinrichs K, 1997: Comparison of different methods for the recovery of horse oocytes. Eq Vet J 25 (Suppl.), Bogh IB, Bezard J, Duchamp G, Baltsen M, Gerard N, Daels P, Greve T, 2002: Pure preovulatory follicular fluid promotes in vitro maturation of in vivo aspirated equine oocytes. Theriogenology 57, Brevini-Gandolfi TAL, Gandolfi F, 2001: The maternal legacy to the embryo: cytoplasmic components and their effects on early development. Theriogenology 55, Carneiro GF, Liu IK, Hyde D, Anderson GB, Lorenzo PL, Ball BA, 2002: Quantification and distribution of equine oocyte cortical granules during meiotic maturation and after activation. Mol Reprod Dev 63, Dell Aquila ME, Fusco S, Lacalandra GM, Maritato F, 1996: In vitro maturation and fertilization of equine oocytes recovered during the breeding season. Theriogenology 45, Downs SM, 1995: The influence of glucose, cumulus cells, and metabolic coupling on ATP levels and meiotic control in the isolated mouse oocyte. Dev Biol 167, Eppig JJ, 1991: Intercommunication between mammalian oocytes and companion somatic cells. Bioessays 13, Galli C, Crotti G, Notari C, Turini P, Duchi R, Lazzari G, 2001: Embryo production by ovum pick up from live donors. Theriogenology 55, Galli C, Lagutina I, Crotti G, Colleoni S, Turini P, Ponderato N, Duchi R, Lazzari G, 2003: Pregnancy: a cloned horse born to its dam twin. Nature 424, 635. Gerlach T, Aurich JE, 2000: Regulation of seasonal reproductive activity in the stallion, ram and hamster. Anim Reprod Sci 58, Goudet G, Bezard J, Duchamp G, Gerard N, Palmer E, 1997: Equine oocyte competence for nuclear and cytoplasmic in vitro maturation: effect of follicle size and hormonal environment. Biol Reprod 57, Grondahl C, Hyttel P, Grondahl ML, Eriksen T, Gotfredsen P, Greve T, 1995: Structural and endocrine aspects of equine oocyte maturation in vivo. Mol Reprod Dev 42, Hinrichs K, Schmidt AL, 2000: Meiotic competence in horse oocytes: interactions among chromatin configuration, follicle size, cumulus morphology, and season. Biol Reprod 62, Hochi S, Choi YH, Braun JW, Sato K, Oguri N, 1993: Factors affecting the recovery of follicular oocytes from horses and their in vitro maturation. Jap J Eq Sci 4, (in japanese). Hochi S, Kozawa M, Fujimoto T, Hondo E, Yamada J, Oguri N, 1996: In vitro maturation and transmission electron microscopic observation of horse oocytes after vitrification. Cryobiology 33, Kidder GM, Mhawi AA, 2002: Gap junctions and ovarian folliculogenesis. Reproduction 123, Love LB, Choi YH, Love CC, Varner DD, Hinrichs K, 2003: Effect of ovary storage and oocyte transport method on maturation rate of horse oocytes. Theriogenology 59, Luvoni GC, Luciano AM, Modina S, Gandolfi F, 2001: Influence of different stages of the oestrous cycle on cumulus-oocyte communications in canine oocytes: effects on the efficiency of in vitro maturation. J Reprod Fertil 57 (Suppl),

6 Cumulus Oocyte Communication in Horse IVM 75 Maclellan LJ, Carnevale EM, Coutinho da Silva MA, Scoggin CF, Bruemmer JE, Squires EL, 2002: Pregnancies from vitrified equine oocytes collected from super-stimulated and non-stimulated mares. Theriogenology 58, Moor RM, Dai Y, Lee C, Fulka J Jr, 1998: Oocyte maturation and embryonic failure. Hum Reprod Update 4, Nagy P, Guillaume D, Daels P, 2000: Seasonality in mares. Anim Reprod Sci 60 61, Sirard MA, 2001: Resumption of meiosis: Mechanism involved in meiotic progression and its relation with developmental competence. Theriogenology 55, Squires EL, 1996: Maturation and fertilization of equine oocytes. Vet Clin North Am Equine Pract 12, Squires EL, Carnevale EM, McCue PM, Bruemmer JE, 2003: Embryo technologies in the horse. Theriogenology 59, Sutovsky P, Flechon JE, Flechon B, Motlik J, Peynot N, Chesne P, Heyman Y, 1993: Dynamic changes of gap junctions and cytoskeleton during invitro culture of cattle oocyte cumulus complexes. Biol Reprod 49, Tanghe S, Van Soom A, Nauwynck H, Coryn M, de Kruif A, 2002: Minireview: Functions of the cumulus oophorus during oocyte maturation, ovulation, and fertilization. Mol Reprod Dev 61, Tremoleda JL, Van Haeften T, Stout TA, Colenbrander B, Bevers MM, 2003: Cytoskeleton and chromatin reorganization in horse oocytes following intracytoplasmic sperm injection: patterns associated with normal and defective fertilization. Biol Reprod 69, Willis P, Caudle AB, Fayrer-Hosken RA, 1991: Equine oocyte in vitro maturation: Influences of sera, time, and hormones. Mol Reprod Dev 30, Submitted: Author s address (for correspondence): Prof. Fulvio Gandolfi, Istituto di Anatomia degli Animali Domestici, via Celoria, Milan, Italy. Tel.: (+39) ; fax: (+39) ; fulvio.gandolfi@unimi.it