We have used confocal microscopy to measure calcium waves and examine the distribution of tubulin in oocytes of the ascidian Ciona intestinalis during meiosis. We show that the fertilisation calcium wave in these oocytes originates in the vegetal pole. The sperm penetration site and female meiotic apparatus are found at opposite poles of the oocyte at fertilisation, confirming that C. intestinalis sperm enter in the vegetal pole of the oocyte. Following fertilisation, ascidian oocytes are characterised by repetitive calcium waves. Meiosis I-associated waves originate at the vegetal pole of the oocyte, and travel towards the animal pole. In contrast, the calcium waves during meiosis II initiate at the oocyte equator, and cross the oocyte cytoplasm perpendicular to the point of emission of the polar body. Immunolocalisation of tubulin during meiosis II reveals that the male centrosome is also located between animal and vegetal poles prior to initiation of the meiosis II-associated calcium waves, suggesting that the male centrosome influences the origin of these calcium transients. Ascidians are also characterised by an increase in sensitivity to intracellular calcium release after fertilisation. We show that this is not simply an effect of oocyte activation. The data strongly suggest a role for the male centrosome in controlling the mechanism and localisation of post-fertilisation intracellular calcium waves.

In the last 3 years, several studies have shown that xenogeneic transplantation of rodent spermatogonia is feasible. The treatment of infertile patients with spermatogenic arrest using the injection of immature germ cells has yielded only poor results. We attempted to establish a complete spermatogenetic line in the testes of mutant aspermatogenic (W/Wv) and severe combined immunodeficient mice (SCID) transplanted with germ cells from azoospermic men. Spermatogenic cells were obtained from testicular biopsy specimens of men (average age of 34.3 ± 9 years) undergoing infertility treatment because of obstructive and non-obstructive azoospermia. Testicular tissue was digested with collagenase to promote separation of individual spermatogenic cells. The germ cells were injected into mouse testicular seminiferous tubules using a microneedle (40 μm inner diameter) on a 10 ml syringe. To assess the penetration of the cell suspension into the tubules, trypan blue was used as an indicator. Mice were maintained for 50 to 150 days to allow time for germ cell colonisation and development prior to them being killed. Testes were then fixed for histological examination and approximately 100 cross-sectioned tubules were examined for human spermatogenic cells. A total of 26 testicular cell samples, 16 frozen and 10 fresh, were obtained from 24 men. The origin of the azoospermia was obstructive (OA) in 16 patients and non-obstructive (NOA) in 8 patients. The concentration of spermatogenic cells in the OA group was 6.6 × 106 cells/ml, and 1.3 ? 106 cells/ml in the NOA group (p < 0.01). The different spermatogenic cell types were distributed equally in the OA samples, ranging from spermatogenia to fully developed spermatozoa, but in the NOA group the majority of cells were spermatogonia and spermatocytes. A total of 23 testes from 14 W/Wv mice and 24 testes from 12 SCID mice were injected successfully, as judged by the presence of spermatogenic cells in histological sections of testes removed immediately after the injection. However, sections from the remaining testes examined up to 150 days after injection showed tubules lined with Sertoli cells and xenogeneic germ cells were not found. The reason why the two strains of mouse used as recipients did not allow the implantation of human germ cells is probably due to interspecies specificity involving non-compatible cell adhesion molecules and/or immunological rejection.

Two centuries ago, Lazzaro Spallanzani, a Professor of Natural Sciences at the University of Pavia, revolutionised the field of reproduction by artificially inseminating a female poodle that subsequently gave birth to pups. This experiment was the first step in manipulating gametes outside the body and paved the way to present-day in vitro techniques used in human and animal fertilisation. As part of the bicentennial celebrations of the death of Spallanzani the Laboratory of Developmental Biology of the University of Pavia, led by Professors Carlo Redi, Silvia Garagna and Maurizio Zuccotti, organised a meeting where leaders in the field of reproduction discussed the most important advances made this century. Lectures were given by R. Schmid, C. Redi, E. Capanna, W. Hilscher, M. Handel, R. Yanagimachi, R. Dallai, B. Dale, A. Byskov, E. Topfer-Peterson, P. Wassarman, K. Swann, R. Schultz, K. Campbell and E. Fox Keller.

On this occasion, in recognition of 35 years of highly innovative and productive research, the North American Editor for Zygote, Ryuzo Yanagimachi, was awarded The Laurea Honoris Causa by the University of Pavia. Yana's studies have had an undisputable impact on both basic research and its application to biomedicine, and range from the first in vitro fertilisation with capacitated sperm and the first intracytoplasmic sperm injection to the successful cloning of mice using cumulus cell nuclei. Congratulations from all at Zygote to Yana.

Sperm capacitation and in vitro fertilisation (IVF) have been achieved in most eutherian mammals and American marsupials under relatively simple culture conditions. In contrast sperm capacitation in Australian marsupials has not been achieved in vitro and attempts at IVF have previously been characterised by a complete lack of sperm–zona pellucida (ZP) binding. Recently, co-culture of sperm with oviduct epithelial cell monolayers or with oviductal explant conditioned media has been shown to prolong the viability and motility of brushtail possum spermatozoa, as well as to induce capacitation-associated changes such as transformation of sperm to the T-shape orientation. In this study we report that these in vitro produced T-shaped sperm, and in vivo derived T-shaped sperm flushed from the oviduct of artificially inseminated possums as a control, are able to bind to and penetrate the ZP of approximately 25% of eggs recovered from PMSG/LH-superovulated possums in vitro. Development of ZP receptivity and penetrability towards sperm was also identified as a major factor affecting the outcome of IVF. Neither in vivo nor in vitro derived T-shaped sperm were able to bind to or penetrate the ZP if eggs were obtained from animals that were treated with pLH less than 76 h after PMSG. Thus this study provides preliminary evidence for the necessity of sperm–oviduct epithelial cell interactions for capacitation in Australian species and lends further support to the suggestion that the T-shape head orientation is indicative of sperm capacitation. Despite the occurrence of sperm–ZP binding and penetration, sperm–egg membrane fusion and egg activation were not observed. Although the factor(s) responsible for the lack of sperm–egg membrane fusion in the possum have not been identified it is possible that egg capacity for membrane fusion develops independently of zona receptivity and is defective in these eggs, or alternatively that membrane fusion requires strictly defined ionic conditions which are not provided by the IVF media used in this study.

Previous studies have suggested that both acrosome-intact and acrosome-reacted guinea pig sperm are capable of binding to the zona pellucida of cumulus-free oocytes, but the acrosomal status of guinea pig sperm during penetration of the cumulus has not been reported. We made video recordings of the interaction between capacitated guinea pig sperm and cumulus-invested guinea pig oocytes. The videotapes were analysed to identify sperm with hyperactivated motility and to classify the acrosomal status of sperm during penetration of the cumulus and after binding to the zona pellucida. The resolution of the video recordings was not sufficient to recognise sperm with swollen acrosomes. However, sperm that had completed the acrosome reaction were easily identified. Acrosome-reacted sperm were found adherent to the outer boundary of the cumulus, but were never observed to penetrate the cumulus. The percentage of acrosome-intact, hyperactivated sperm was higher in the cumulus oophorus than in culture medium, suggesting that changes in motility were elicited in response to contact with the cumulus. Fully acrosome-reacted sperm were found adherent to the zona pellucida, and solubilised guinea pig zona pellucida was capable of inducing acrosome reactions in capacitated guinea pig sperm. Acrosome-intact sperm were also observed on the zona, but they were not tightly bound and did not have hyperactivated motility, suggesting that these sperm were not functionally capacitated. Our observations demonstrate that guinea pig sperm penetrate the cumulus matrix in an acrosome-intact state. Although we did not observe sperm undergoing the acrosome reaction, our observations and experimental data suggest that the acrosome reaction of guinea pig sperm is completed on or near the surface of the zona pellucida.

In this study our aim was to characterise the presence and the role of DNA alterations during sperm decondensation in the mouse. To visualise the changes during decondensation we investigated for the presence of DNase I hypersensitive sites in situ and for a putative role for topoisomerase II by examining the effect of teniposide, a topoisomerase II inhibitor, during fertilisation. In situ nick translation without the previous addition of DNase I failed to reveal the presence of endogenous nicks in decondensing sperm and pronuclei whereas preincubation of fixed oocytes with DNase I indicated that decondensing sperm were sensitive to this enzyme. Addition of 100 μM teniposide did not completely inhibit pronuclei formation but its addition to the fertilisation medium did lead to the presence of endogenous DNA nicks in decondensing sperm. These observations suggest that DNase I hypersensitivity during sperm decondensation is related to the dramatic conformational changes that the chromatin undergoes during the decondensation process, in which topoisomerase II may be implicated.

It is well known that sea urchin eggs, which exhibit quite a low rate of respiration before fertilisation, undergo a sudden increase in the rate of respiration followed by its gradual decrease in about a 15 min period after fertilisation (Ohnishi & Sugiyama, 1963; Epel, 1969), in which the respiration is mediated mainly by Ca2+-activated non-mitochondrial respiratory systems (Foerder et al., 1978; Perry & Epel, 1985a,b). During this short period the rate of mitochondrial respiration gradually increases (Yasumasu et al., 1988) and stabilises at a higher rate than before fertilisation (Warburg, 1908, 1910; Whitaker, 1933; Yasumasu & Nakano, 1963), when the respiration due to non-mitochondrial respiratory systems is turned off. The rate of mitochondrial respiration, once enhanced upon fertilisation, increases further in the period between hatching and the gastrula stage, without any changes in the number of mitochondria or the capacity of electron transport in the mitochondrial respiratory chain (Fujiwara & Yasumasu, 1997; Fujiwara et al., 2000). It is likely that the respiratory rate is reduced by regulation of electron transport in the mitochondrial respiratory chain and increases due to the release of electron transport from the regulation upon fertilisation and after hatching.

A marked increase in the respiratory rate after hatching is accompanied by an evident decrease in the ATP level without any change in the levels of ADP and AMP (Mita & Yasumasu, 1984). In isolated mitochondria, the rate of respiration, estimated in the presence of ADP at the same concentration as in embryos, is reduced by a high concentration of ATP as found in embryos before hatching but is not affected at a concentration as low as in gastrulae (Fujiwara & Yasumasu, 1997; Fujiwara et al., 2000) ATP at a high concentration probably blocks ATP release from mitochondria and consequently inhibits ADP uptake coupled to ATP release in the ATP/ADP translocation reaction in the mitochondrial membrane, causing a shortage of intra-mitochondrial ADP.

The effects of the new cyclin-dependent kinase inhibitors, roscovitine and olomoucine, on oocytes and eggs of Xenopus laevis were investigated and compared with those of 6-dimethylamino purine (6-DMAP). The inhibitory properties of 6-DMAP, olomoucine and roscovitine towards p34cdc2-cyclin B isolated from Xenopus eggs revealed K-IC50 values of 300, 40 and 10 μM respectively. The three compounds inhibited progesterone-induced maturation with M-IC50 values of 200, 100 and 20 μM. These values were consistent with the K-IC50 values but the ratio M-IC50/K-IC50 was higher for roscovitine and olomoucine than for 6-DMAP. The disappearance of spindle and condensed chromosomes without pronucleus formation was observed when 1 mM 6-DMAP was applied for 4 h at germinal vesicle breakdown or at metaphase II, whereas no effect was observed using 1 mM olomoucine or 50 μM roscovitine. Changes in the electrophoretic mobility of p34cdc2 and erk2 were observed only in homogenates of matured oocytes or eggs exposed for 4 h to 1 mM 6-DMAP. When the drugs were microinjected into matured oocytes, olomoucine (100 μM) and roscovitine (50 μM) induced pronucleus formation more efficiently than did 6-DMAP (100 μM). Taken together, these results demonstrate that Xenopus oocytes possess a lower permeability to olomoucine and roscovitine and that these new compounds are suitable for in vivo studies after germinal vesicle breakdown provided they are microinjected.

Primary spermatocytes originating from prepubertal mouse testes were electrofused to metaphase II (MII)-stage oocytes, enucleated either by the conventional micromanipulation method or by chemical treatment with etoposide and cycloheximide. These experiments were followed by assessment of morphological changes in transferred nuclei using light microscopy, by chromosomal analyses and by screening of hybrids for the presence or absence of DNA synthesis using anti-bromodeoxyuridine antibody and immunofluorescence staining of the hybrids. The results show differences between the two types of ooplasts in susceptibility to activation stimuli. However, when activated, both types of ooplasts gave rise to hybrids of similar morphology. From 35.3% to 63% of activated hybrids originating from chemically or microsurgically enucleated oocytes, respectively, contained one large pronucleus in cytoplasm, 62% or 31.6% hybrids from those two groups, respectively, possessed two smaller pronuclei and a few contained three or four pronuclei. No DNA synthesis was detected in any hybrid containing one or more pronuclei. The chromosome spreads of hybrids with premature chromosome condensation (PCC) morphology (before activation) show that most of the hybrids had a diploid (2n) number of chromosomes. The nature and regularity of the cell division cycle in the hybrids are discussed.

The physiology of the early embryo may be indicative of embryo vitality and therefore methods for non-invasively monitoring physiological parameters from embryos could improve preimplantation diagnoses. The self-referencing electrophysiological technique is capable of non-invasive measurement of the physiology of individual cells by monitoring the movement of ions and molecules between the cell and the surrounding media. Here we use this technique to monitor gradients of calcium, potassium, oxygen and hydrogen peroxide around individual mouse preimplantation embryos. The calcium-sensitive electrode in self-referencing mode identified a region of elevated calcium concentration (∼0.25 pmol) surrounding each embryo. The calcium gradient surrounding embryos was relatively steep, such that the region of elevated calcium extended into the medium only 4 μm from the embryo. By contrast, using an oxygen-sensitive electrode an extensive gradient of reduced dissolved oxygen concentration was measured surrounding the embryo and extended tens of micrometres into the medium. A gradient of neither potassium nor hydrogen peroxide was observed around unperturbed embryos. We also demonstrate that monitoring the physiology of embryos using the self-referencing technique does not compromise their subsequent development. Blastocysts studied with the self-referencing technique implanted and developed to term at the same frequency as did unexamined, control embryos. Therefore, the self-referencing electrode provides a valuable non-invasive technique for studying the physiology and pathophysiology of individual embryos without hindering their subsequent development.

The lecture given by Dr Yasumasu should be considered from two points of view, namely its significance in the field of developmental biology and his own personal history as a developmental biologist.

In 1908 Otto Warburg published a paper entitled ‘Beobachtungen üeber die Oxydationsprozesse im Seeigelei’ (Warburg, 1908). This is one of his earliest works, where he measured respiration of eggs with Winkler's method not with his manometer. This was the first paper describing the fact that the respiration in unfertilised eggs was considerably lower than that in fertilised ones. Many researchers confirmed his experiments and extended them. Borei (1948) measured oxygen consumption of oocytes and unfertilised and fertilised eggs and compared his results with those of other researchers. He observed that the respiration of eggs declined after they were removed from the ovary and placed into seawater and that it increased at fertilisation. He observed an exponential increase in respiration of fertilised eggs or embryos from the cleavage stage to the hatching blastula. He also observed an initial burst of respiration but failed to record it exactly. Ohnishi & Sugiyama (1963) measured the initial burst of respiration quantitatively with the oxygen electrode method. Thus, respiration of sea urchin eggs and early embryos was divided into three phases: respiration of unfertilised eggs, the initial burst of respiration at fertilisation and the respiration of fertilised eggs. The respiration of fertilised eggs increased exponentially with progression of development until hatching.

Mitogen-activated protein kinase (MAPK) has been reported to be involved in oocyte maturation in all animals so far examined. In the present study we investigate the expression and localisation of active phosphorylated MAPKs (p44ERK1/p42ERK2) during maturation of pig oocytes. In immunoblot analysis using anti-p44ERK1 antibody which recognised both active and inactive forms of p44ERK1 and p42ERK2, we confirmed that MAPKs were phosphorylatred around the time of germinal vesicle breakdown (GVBD) and the active phosphorylated MAPKs (pMAKs) were maintained until metaphase II, as has been reported. On immunofluorescent confocal microscopy using anti-pMAPK antibody which recognised only phosphorylated forms of MAPKs, pMAPK was localised at the spindle poles in pig mitotic cells. On the other hand, in pig oocytes, no signal was detected during GV stage. After GVBD, the area around condensed chromosomes was preferentially stained at metaphase I although whole cytoplasm was faintly stained. At early anaphase I, the polar regions of the meiotic spindle were prominently stained. However, during the progression of anaphase I and telophase I pMAPK was detected at the mid-zone of the elongated spindle, gradually becoming concentrated at the centre. Finally, at the time of emission of the first polar body, pMAPK was detected as a ring-like structure between the condensed chromosomes and the first polar body, and the staining was maintained even after the metaphase II spindle was formed. The inhibition of MAPK activity with the MAPK kinase inhibitor U0126 during the meiosis I/meiosis II transition suppressed chromosome separation, first polar body emission and formation of the metaphase II spindle. From these results, we propose that the spindle-associated pMAPKs play an important role in the events occurring during the meiosis I/meiosis II transition, such as chromosome separation, spindle elongation and cleavage furrow formation in pig oocytes.

We reported previously that in mouse testis calmodulin-dependent protein phosphatase (calcineurin) is localised in the nuclei of round and elongating spermatids (Cell Tissue Res. 1995; 281: 273-81). In this study, we studied the immunohistochemical localisation of calcium/calmodulin-dependent protein kinase (CaM kinase II) using antibodies against CaM kinase IIγ from chicken gizzard and specific antibodies raised against the amino acid sequence Ileu480–Ala493 of this enzyme, and compared it with the distribution of calmodulin. Indirect immunofluorescence was most concentrated in early spermatocytes and localised in the outermost layer of seminiferous tubules where the calmodulin level was relatively low. Measurements of immuno-gold particle densities on electron micrographs revealed that CaM kinase II is transiently increased in the nucleus of zygotene spermatocytes. These observations suggest the involvement of CaM kinase II in the meiotic chromosomal pairing process. An extremely high concentration of calmodulin in spermatogenic cells undergoing meiosis may not be directly related to activation of calmodulin-dependent kinases and phosphatases.

Freshly ovulated mouse oocytes exposed to 5 mM calcium ionophore A23187 for 5 min and controls (not exposed) were cultured in TYH medium with 10 μg/ml puromycin (the puromycin group) or 2 mM 6-dimethylaminopurine (DMAP; the DMAP group) for 4 h. Among the controls, few oocytes were activated even if they were treated with DMAP or puromycin. In the oocytes exposed to A23187, in contrast, the activation rate, i.e. the rate of oocytes showing at least one pronucleus (PN) after the treatment, was 46.2% (48/104) in the DMAP group and 90.0% (118/131) in the puromycin group. Activation rate in the puromycin group was significantly higher than in the DMAP and control groups (p < 0.0001, respectively). Furthermore, 82.4% (108/131) of the activated oocytes in the puromycin group showed one PN with extrusion of the second polar body (PB). In the puromycin group, the DNA content of the PN of parthenogenones with 1PN2PB was half that of a set of metaphase II chromosomes. Chromosomal analysis was possible in 14 parthenogenones with 1PN2PB in the puromycin group. The parthenogenones possessed a normal set (n = 20) of haploid chromosomes. The combination of A23187 and puromycin proved to be an effective method of producing haploid parthenogenones.

Soybean trypsin inhibitor (SBTI) inhibits the catalytic activity of serine proteases, and has been shown to bind to acrosin, an acrosomal hydrolase which is not exposed on the surface of macaque sperm until after the acrosome reaction. Following activation with caffeine and dibutyryl cAMP, cynomolgus macaque sperm were induced to acrosome react with calcium ionophore A23187 in the presence of SBTI and were fixed for ultrastructural observation. Transmission electron microscopy (TEM) revealed secondary labelling of anti-SBTI-IgG with colloidal gold in association with the acrosomal matrix and fused membranes of sperm undergoing the acrosome reaction, but gold labelling was not observed on acrosome-intact sperm. When SBTI was conjugated with the fluorochrome Alexa 488, labelled (acrosome-reacted) sperm showed bright fluorescence that ranged from a patchy or punctate appearance to solid labelling over the region of the acrosomal cap. Following treatment with ionophore, the percentages of total acrosome-reacted sperm (motile and non-motile) as assessed with Alexa-SBTI, fluorescein isothiocyanate-conjugated Pisum sativum agglutinin (FITC-PSA), and TEM were 54.6%, 51.6% and 61.5%, respectively. Measures of acrosomal status with FITC-PSA and Alexa-SBTI were highly correlated (r = 0.94; n = 3). Macaque zonae pellucidae were co-incubated with activated sperm for 1 min and then rinsed in medium containing Alexa-SBTI and immediately observed with epifluorescence microscopy. The mean percentage of Alexa-SBTI-labelled (acrosome-reacted) motile sperm bound to the zona was 45.7 ± 14 (range: 22–80.4%; n = 4). Fewer than 1% of the motile sperm in suspension surrounding the zonae were acrosome-reacted. Alexa-SBTI had no effect on sperm motility, survival, or zona binding capability.

The problem of how the sperm activates the egg has captivated the attention of cell and developmental biologists since the turn of the century. An early focus concerned species-specific fertilisation and the pioneering work of Lilly and Tyler (Tyler & Tyler, 1966) used immunological analogies to provide explanations of species-specific fertilisation. Contemporary work has focused on the identity of unique receptors on the sperm and the egg as exemplified in the work of Lennarz (Ohlendieck & Lennarz, 1996), Vacquier (Vacquier, et al., 1995) and Wasserman (1999). Lately, this approach has provided unexpected insights on how speciation might occur. Speciation requires reproductive isolation and creative research from the Vacquier laboratory has demonstrated how reproductive barriers might occur through rapid evolution of sperm/egg recognition systems (Lee et al., 1995).

Studies on the cell biology of activation received a major impetus in the 1930s with Mazia's observation of a calcium increase in eggs of the sea urchin following fertilisation (Mazia, 1937). His discovery, however, was a premature one in that there was no satisfactory model at that time for explaining how a calcium increase could affect cell activity. It took almost 40 years to develop a paradigm, and this came from studies on muscle and nerve which revealed how calcium increases could somehow control cell activity. Work in the 1970s rapidly established a similar role for calcium in activation of the egg at fertilisation. The first break-through was the direct demonstration by Steinhardt & Epel (1974) that calcium was involved in egg activation, through manipulation of calcium levels in sea urchin oocytes by use of calcium ionophores.

Microtubules and microfilaments are major cytoskeletal components and important modulators for chromosomal movement and cellular division in mammalian oocytes. In this study we observed microtubule and microfilament organisation in bovine oocytes by laser scanning confocal microscopy, and determined requirements of their assembly during in vitro maturation. After germinal vesicle breakdown, small microtubular asters were observed near the condensed chromatin. The asters appeared to elongate and encompass condensed chromatin particles. At the metaphase stage, microtubules were observed in the second meiotic spindle at the metaphase stage. The meiotic spindle was a symmetrical, barrel-shaped structure containing anastral broad poles, located peripherally and radially oriented. Treatment with nocodazole did not inhibit germinal vesicle breakdown. However, progression to metaphase failed to occur in oocytes treated with nocodazole. In contrast, microfilaments were observed as a relatively thick uniform area around the cell cortex and overlying chromatin following germinal vesicle breakdown. Treatment with cytochalasin B inhibited microfilament polymerisation but did not prevent either germinal vesicle breakdown or metaphase formation. However, movement of chromatin to the proper position was inhibited in oocytes treated with cytochalasin B. These results suggest that both microtubules and microfilaments are closely associated with reconstruction and proper positioning of chromatin during meiotic maturation in bovine oocytes.

Oocytes, in general, are greatly enriched in mitochondria to support higher rates of macromolecular synthesis and critical physiological processes characteristic of early development. An inability of these organelles to amplify and/or to accumulate ATP has been linked to developmental abnormality or arrest. The number of mitochondrial genomes present in mature mouse and human metaphase II oocytes was estimated by fluorescent rapid cycle DNA amplification, which is a highly sensitive technique ideally suited to quantitative mitochondrial DNA (mtDNA) analysis in individual cells. A considerable degree of variability was observed between individual samples. An overall average of 1.59 × 105 and 3.14 × 105 mtDNA molecules were detected per mouse and human oocyte, respectively. Furthermore, the mtDNA copy number was examined in polar bodies and contrasted with the concentration in their corresponding oocytes. In addition, the density of mtDNA in a cytoplasmic sample was estimated in an attempt to determine the approximate number of mitochondria transferred during clinical cytoplasmic donation procedures as well as to develop a clinical tool for the assessment and selection of oocytes during in vitro fertilisation procedures. However, no correlation was identified between the mtDNA concentration in either polar bodies or cytoplasmic samples and their corresponding oocyte.

Nuclear bodies occuring during the 2-cell stage of bovine embryos (obtained either by in vitro fertilisation of in vitro matured ovarian oocytes, or collection after fertilisation and cleavage in vivo) were studied using ultrastructural cytochemistry and immunocytochemistry to determine whether their occurrence may be linked with the onset of embryonic transcription. In addition, the species-specific ultrastructural features of the interchromatin structures of the 2-cell bovine embryo were displayed. Three different types of nuclear bodies were distinguished: (i) nucleolus precursor bodies (NPBs), (ii) loose bodies (LBs) and (iii) dense bodies (DBs). In order to determine their possible functional significance, we considered parallels between these three nuclear entities and interchromatin compartments reported in other cells. As detected by their preferential ribonucleoprotein staining, all types of nuclear bodies contained ribonucleoproteins. In contrast to the other types of nuclear bodies studied, NPBs contained argyrophilic proteins but in no case they did show morphological features of functional nucleoli. Both compact and vacuolated forms of NPBs were seen in both in vivo and in vitro embryos, sometimes simultaneously in the same nucleus. LBs and DBs reacted with antibodies to Sm antigen, indicating the presence of a group of nucleoplasmic, non-nucleolar small nuclear ribonucleoproteins (snRNPs). The immunoreactivity for Sm antigen was more intense and homogeneous in DBs than in LBs. DBs were seen in both categories of embryo. A possible kinship of DBs with the sphere organelle known from oocytes of different animal species or the prominent spherical inclusions of the early mouse embryo nuclei is suggested. The last type of intranuclear body, the LBs, showed a composite structure. Their granular component, occurring in clusters and displaying immunoreactivity for Sm antigen, was similar to interchromatin granules and was therefore named IG-like granules. Another component forming the LBs showed a much finer structure and a lower immunoreactivity with anti-Sm antibodies. We suggest that this amorphous component may be related to the IG-associated zone. All three types of intranuclear bodies were often seen close together, suggesting their possible mutual functional relationship. From these and other observations we conclude that the intranuclear bodies in 2-cell bovine embryos correspond, with the exception of the NPB, to similar structures/compartments supposed to accumulate inactive spliceosomal components in certain phases of somatic cell nucleus functions. Accordingly, the occurrence of such nuclear bodies does not represent cytological evidence for RNA synthesis. In contrast to this, an important morphological feature revealing the status of the bovine 2-cell embryo is the vacuol-isation of the NPB.