The human zygote inherits its mitotic potential from the male gamete

Transcription

1 Human Reproduction vol.9 no.7 pp , 1994 OPINION The human zygote inherits its mitotic potential from the male gamete Gianpiero Palermo, Santiago Munne* and Jacques Cohen 1 The Gamete and Embryo Research Laboratory, Cornell University Medical College, 1300 York Avenue, PO Box 30, New York, NY 10021, USA 'To whom correspondence should be addressed The control of centrosome and centriole duplication and formation of a single bipolar spindle during syngamy is different from that of somatic cells. If centrosomes from both gametes become functional, mosaicism could be induced following the formation of abnormal spindles. The centrosome used for development in most species is paternally inherited, with the exception of the mouse. Single and double centrioles within centrosomes in human zygotes have been described. Interpretation of data suggests that sperm centrioles are associated with further development Here we provide evidence that the human sperm centrosome controls the first mitotic divisions after fertilization. Mosaicism and ploidy were determined in blastomeres derived from four groups of embryos: (i) dispermic zygotes, (ii) enucleated dispermic zygotes, (iii) monospermic zygotes in which the second polar body was retained (digynk), and (iv) enucleated monospermic digynic zygotes. Results prove that dispermic zygotes do not have bipolar spindles and become mosaic. This is determined by an extranudear factor, since removal of a single pronudeus does not correct the abnormal division. Monospermic digynic embryos become triploid, indicating that the three pronuclei are organized in a single spindle and a single centrosome is active. Removal of one pronudeus from such zygotes causes the embryo to revert to a normal diploid state. These findings have consequences for the study of parthenogenesis, androgenesis and gynogenesis in the human. Key words: centrosome inheritance/human gametes/mitotic control Introduction Animal and lower plant cells contain a specialized organelle, the centrosome, which is associated with the organization of spindle fibres that function in mitosis and meiosis (Rappaport, 1969; Gould and Borisy, 1977; Wheatley, 1992). The centrosome was identified as a centre of organization involved in forming the mitotic spindle more than 100 years ago. It reproduces and doubles during interphase in anticipation of cell division. In most cells it consists of two morphologically distinct objects called centrioles: a pair of cylinders enclosed in a complex and 1220 asymmetrical arrangement and the pericentriolar material from which spindle microtubules are generated. These organelles reproduce by forming daughter copies which segregate from their mother. The cell exerts exact control over the number of centrosomes since too few or too many will inevitably lead to abnormal division (Sluder et al., 1989). Centriole-free karyoplasts (nucleated cell fragments) are not able to initiate nuclear membrane disassembly, chromosome condensation, spindle formation or generate centrioles de novo (Maniotis and Schliwa, 1991). The organization of spindle fibres by the centrioles occurs during the early phases of mitosis as well as meiosis. Composed of arrays of microtubules, these fibres play an important role in the movement of chromosomes as they separate during cell division. The control of centrosome duplication and formation of a single bipolar spindle during fertilization is different from that of somatic single-nucleated diploid cells. In the development of fertilized animal eggs, specific mechanisms must exist at the gamete or zygote level to control centrosome inheritance. If centrosomes from both gametes were retained and remained functional, the zygote would enter first mitosis with two double sets of centrosomes and four centrioles, resulting in multipolar or extra spindles. Mosaicism and aneuploidy would be induced by the ensuing three- or four-way division, which would jeopardize the viability of the embryo (Sluder et al., 1989). Wilson and Matthews (1895) discovered nearly a century ago that the centrosome used for development was paternally inherited in the sea urchin. In recent years, it was demonstrated using anti-centrosome antibodies that a similar mechanism exists in sheep (Crozet, 1990), pig (Szollosi and Hunter, 1973), rabbit (Longo, 1963) and bovine (Long et al., 1993) species. Polyspermic bovine zygotes contain several sperm asters with different degrees of chromatin association and the maternal centrosomes in monospermic as well as polyspermic zygotes are lost or degenerated. Pinto-Correia et al. (1992) demonstrated that the active microtubule-generating site in the rabbit was derived from the mid-piece of the spermatozoon and associated with the proximal centriole. Of currently studied mammals the mouse appears to be the exception to sperm-associated microtubule organizing centres, displaying no distinct centrosomal complex (Schatten et al., 1986). During germinal vesicle breakdown in rodent oocytes, fibrillar aggregates form small asters which perform the role of the spindle pole without the presence of centrioles (Szollosi et al., 1972). Human dictyate oocytes also lack centrioles (Hertig and Adams, 1967). The presence of centrioles in fertilized monospermic and dispermic human embryos was recently demonstrated by Sathananthan et al. (1991). Single and double centrioles within centrosomes were detected by transmission electron microscopy. Sperm centrioles Oxford University Press

2 Paternal control of mitosis in embryos appeared closely associated with the male pronucleus. Tripolar spindles were found in tripronuclear zygotes with two centrioles. Interpretation of this data suggests that human centrioles are inherited through the paternal line. Here we provide evidence that the sperm centrosome controls the first mitotic division after fertilization. Mosaicism and ploidy was determined in blastomeres derived from four groups of zygotes: (i) polyspermic zygotes which were penetrated by two sperm cells (dispermic), (ii) dispermic zygotes from which a single pronucleus was removed, (iii) monospermic zygotes derived from intracytoplasmic sperm injection (ICSI) (Palermo et al., 1992) in which the second polar body was retained (digynic), and (iv) monospermic digynic zygotes from which a single pronucleus was removed. Results prove that dispermic zygotes usually do not have bipolar spindles and become mosaic. This situation is determined by an extranuclear factor, since removal of a single pronucleus does not correct the abnormal division pattern. Monospermic digynic embryos, on the contrary, become triploid, indicating that the three pronuclei are organized in a single spindle at syngamy. Removal of one pronucleus from such zygotes reverts the embryo to a normal diploid state. Only female centrosome is active diploidy monospermic triploidy dlplotdy Theoretical models for centrosome inheritance in human zygotes There are a number of theoretical patterns for centrosome inheritance in mammalian embryos. In the first model both gametes have functional centrosomes within the fertilized oocyte, but centrosomal areas fuse before syngamy. Monospermic embryos would usually lead to a single mitotic spindle and result in diploidy, provided the gametes were haploid. In all other common forms of abnormal fertilization, single bipolar spindles would be formed and the embryo would become triploid or diploid. Mosaicism derived from multiple or tripolar spindles at syngamy would rarely occur. Fusion of male and female centrosomes is unlikely since they are intrinsically different from each other (Sluder etal., 1989). In the second model, the female gamete produces the active centrosome or has the ability to organize cell division and the male centrioles are suppressed (Figure 1). This is analogous to the situation in the mouse (Schatten et al., 1986), though it does not explain the ability of some parthenogenetic diploid mouse embryos to develop into fetuses. According to this model, single spindles would be formed in monospermic as well as polyspermic zygotes, whereas multiple active centrosomes would exist when formation of the second polar body would be suppressed. In the third and final model, only the male centrosome is active. The male gamete may have the only centrosome (Figure 2) or the female centrosome may be suppressed at syngamy (Figure 3). In this model, chromosomes in monospermic and digynic zygotes would be organized in a single spindle, whereas dispermic embryos become mosaic. Micromanipulation and fluorescence in-situ hybridization Patient consents were provided through protocols and granted by The Committee of Human Rights in Research of The New York Hospital, Cornell University Medical Center, mosaic monospermic/digynic - 1pn o female centrosome O female pronucleus male centrosome ( ) male pronucleus Fig. 1. A model of maternal centrosome inheritance. Ploidy and mosaicism in embryos resulting from monospermic, dispermic, digynic and enucleated zygotes are dependent on the number of active centrosomes prior to syngamy. Digynic embryos would become mosaic in this model. Removal of a single pronucleus would not have an effect on the mosaicism. USA. Polyspermic embryos from in-vitro fertilization (TVF) or subzonal sperm insertion procedures were assessed for pronuclei (defined by the presence of nucleoli) and polar bodies. Embryos widi more than three pronuclei were excluded from this study and dispermic zygotes were selected. Some were enucleated, while others were left intact and cultured for two more days. ICSI was performed with single spermatozoa and tripronucleate embryos with single polar bodies (monospermic and digynic) were obtained in 4% of fertilized eggs. They were either cultured for 2 days or enucleated prior to culture. The enucleation procedure involved a holding pipette for stabilizing the zona pellucida while the vitellus was penetrated with a small enucleation needle. The egg was first rolled around in order to determine me distances between pronuclei and polar body. Dispermic zygotes with one pronucleus clearly distal from the polar body were selected for enucleation. Microscopes, micromanipulators, suction devices and techniques for enucleation have been described (Malter and Cohen, 1989; Cohen et al., 1992). The second polar body was usually held at the 9 o'clock 1221

3 G.Palermo, S.Munnl and J.Cohen Only male gamete has a centrosome Only male centrosome is active diploidy diploidy monospermic Irlploid triploid dlploid diploid monospermic/digynlc - 1 pn monospermic/digynic - 1pn a lemale centrosome male cenlrosome Q female pronucleus (J) male pronucleus Fig. 2. A model of paternal centrosome inheritance in the human without the presence of maternal centrosomes in the zygote. Ploidy and mosaicism in embryos resulting from monospermic, dispermic, digynic and enucleated zygotes are dependent on the number of active centrosomes prior to syngamy. Digynic embryos would become either triploid or diploid in this model. Dispermic embryos would become mosaic and removal of a single pronucleus would not have an effect on the mosaicism. This model closely resembles the findings presented here. position adjacent to the lumen of the holding pipette. The target pronucleus was located near the periphery of the opposed oocyte area. Recent studies have shown that this is a sperm pronucleus (Tang etal., 1994). Enucleation needles were ~ 10 /tm in diameter and had sharp spears attached to one side of the tip to facilitate penetration. The needle was front-loaded with acidified Tyrode's solution (ph 2.3) and a superficial area of the zona pellucida at the 3 o'clock position was removed without piercing the zona. The holding pipette and the egg were then moved to another position in the droplet to limit exposure to the acid. The tip of the needle was gently positioned inside the target pronucleus, which was aspirated while the needle was being withdrawn. The pronuclear membrane would remain visible for several minutes. Extranuclear material would not be removed. This gentle technique presumably allows centrosomal material to remain within the cytoplasm, although this can only be 1222 Q female pronucleus male centrosome male pronucleus Fig. 3. A model of paternal centrosome inheritance in the human in the presence of an inactive maternal centrosome in the zygote. Ploidy and mosaicism in embryos resulting from monospermic, dispermic, digynic and enucleated zygotes are dependent on the number of active centrosomes prior to syngamy. Dispermic embryos would become mosaic in this model. Removal of a single pronucleus would not have an effect on the mosaicism. definitely confirmed by transmission electron microscopy, when further studies are performed. Tripronucleate zygotes from ICSI were enucleated in a similar fashion, except that the target pronucleus was next to the polar body. Based on recent studies, it is likely that this is the female pronucleus (Tang et al., 1994). Enucleated and intact tripronucleate zygotes were allowed to develop for an extra 3 days, at which time they were biopsied. The number of blastomeres ranged from two to 16 cells, depending on whether the embryos arrested development or not. The presence of X and Y chromosomes and five autosomes (chromosomes 13, 16, 18 and 13/21) was assessed in at least three blastomeres of each > 3-cell embryo with fluorescence in-situ hybridization (FISH) (Munne" etal., 1993). A FISH technique using directly labelled fluorochrome probes for X-, Y-, 16- and 18-chromosomes (Imagenetics, Naperville, IL, USA) was used. The simultaneous analysis of two (X,Y and 18), three (X,Y, 16 and 18), or four chromosome pairs (X,Y, 18 and 13/21) allowed the differentiation of aneuploidy from haploidy or

4 Paternal control of mitosis in embryos Table I. Incidence and onset of mosaicism in monospermic, dispermic and monospermic/digynic embryos Table n. Examples of mosaic embryos for the X, Y, 18 and 16 or 13/21 chromosomes Mostly diploid Developing monospermic Arrested or slow monospermic Enucleated dispermic Enucleated monospermic digynic Mostly polyplokl Arrested or slow monospermic Dispermic Monospermic digynic Embryos (1) Mosaicism Onset mosaicism at first cleavage (*) Embryo Cells analysed/ total cells Karyotypes Mosaicism occurring at first division 1 8/8 XO1818:1616(4)', XXX1818:1616 (4) 2 3/3 XX (2), 8X 6[18] b 6[16] 3 2/2 4X4Y 8[18] 16[13/21], XXXXYY 6[18] 12[13/31] 4 2/2 7X 7[18] 11[13/21], 3(13/21] 5 3/3 XXYYY181818, XXY 6(18], YYY1818 Mosaicism occurring at second division 1 4/4 XX1818 4(13/21] (2), XO1818 4[13/21], XXX1818 4(13/21] 2 7/7 XY1818 (3), YO (2), XX181818, XY /8 XY1818 3[13/21] (2), XXYY 4[18] 8(13/21] (3) 4 5/5 XO18O16O (3), XX (2) 5 3/3 XY , XY 4[18]1616, XY 1616 polyploidy, provided that the frequency of double monosomy and double trisomy for the two chromosome pairs studied was very low. The majority of cells of embryos were analysed. Embryos with less than half of the cells analysed were not included. Chromosomal status of abnormal zygotes Ploidy and mosaicism was assessed in at least three or the majority of blastomeres from 505 embryos using multicolour multiprobe FISH (Table I). The assessment of the extent of mosaicism by analysing two to four chromosome pairs and not necessarily all the cells of some embryos has clear limitations. Accurate detection probably depends on the type of mosaicism. Three classes of mosaicism can be distinguished: the first consists of mosaics produced by mitotic non-disjunction or anaphase lag of one of the chromosome pairs. Detection would depend on the chromosome types analysed. The second group of mosaics consists of two cell lines with different numbers of haploid complements. These embryos can be detected with a few probes if all blastomeres are analysed. The third group of mosaics arises through abnormal DNA replication possibly followed by uneven karyokinesis. Most chromosomes in such embryos would be abnormally distributed, and detection would be feasible, even when a few probes were used. Considering these models and the detection methods employed here, mosaicism is probably underestimated. The onset of mosaicism was determined by calculating the ratio of abnormal blastomeres in each embryo. Mosaicism may originate at syngamy, when all blastomeres are chromosomally abnormal, but the abnormality is not uniform (aneuploidy). Mosaicism may originate at the second cleavage division, when 50 % of the blastomeres are abnormal and at the third division when 25 % of the blastomeres are abnormal. Examples of mosaic embryos with regard to their first occurrence are presented in Table II. Mosaicism was only determined when hyper- and hypoploidy were balanced. Mosaicism occurred in 17% of developing 2-pronucleate monospermic embryos. These embryos were never polyploid. Monospermic but cleavage-arrested or slowly developing embryos could be diploid as well as polyploid. The latter had a high rate of mosaicism of 46%, of which 44% occurred at the first cleavage division. The earliest signs of mosaicism were detected in embryos which divided from the Mosaicism occurring at third or fourth division 1 7/ /10 8/11 15/15 6/6 "Blastomere number. 'Total number of chromosomes 18, 16 or 13/21. XX1818 4(13/21] (5), XX1818 3(13/21], XX1818 5(13/21] XX (8), 4X 4[18] 8(16] (2) XY1818 4(13/21] (7), XXXYYY 6[18] 12(13/21] XY1818 5[13/21] (11), XY1818 3(13/21] (2), XY1818 7(13/21] (2) XY1818 3(113/21] (4), XY 2(13/21], XY (13/21] 2-cell stage to the 4-cell stage. Dispermic embryos had a significantly {P < 0.01) higher frequency of mosaicism (8%), which corroborates the findings of Pieters et al. (1992). Mosaicism originated from an abnormal spindle organization at syngamy in 61 % of these embryos. This mechanism of mosaicism was rarely observed in monospermic embryos. Removal of the extra male pronucleus did not revert the embryo to a non-mosaic state although the number of chromosomes was reduced by onethird. These embryos were uniformly mosaic, a condition which usually originated at syngamy. Only one of five monospermic and digynic embryos resulting from ICSI had signs of mosaicism at the 8-cell stage (Table I). The four non-mosaic embryos were triploid, indicating that the chromosomes of the three pronuclei had organized in a single bipolar spindle at syngamy, and suggesting that only one centrosome is active in such zygotes. The mosaicism in the fifth embryo was balanced and originated from a non-disjunction at the 2-cell stage. Three of the four monospermic and digynic embryos from which the pronucleus next to the polar body was removed became normal diploid, indicating that a single centrosome is active in such embryos. Mosaicism in the fourth embryo was balanced and developed at the 2-cell stage, again suggesting that a single centrosome is active at syngamy in tripronucleate embryos in which the second polar body is retained. Strategies for correction of triploidy and mosaicism in tripronucleate embryos The current findings provide clues for the application of strategies aimed at reversing abnormalities in tripronucleate human embryos. 1223

5 G.Palermo, S.Munne and J.Cohen Enucleation of tripronucleate zygotes may have several advantages. Patients who have only polyspermic or monospermic-digynic zygotes may have a chance of becoming pregnant following microsurgical correction. In addition, experimental enucleation of tripronucleate dispermic and digynic zygotes may provide androgenetic and gynogenetic cell lines, which are important for studies involving cell lineage, stem cell formation and differentiation. Dispermy is the most common fertilization anomaly in the human (Kola and Trounson, 1989). Further cleavage can occur but division patterns are irregular in the majority of such embryos, with the formation of immediate 3-cell stages as the most common form (Kola etal., 1987). The majority of dispermic embryos will arrest prior to differentiation. Such embryos may develop hydatidiform moles in rare cases. Recognition of the parental origin of pronuclei has therefore been considered important, but application has been hindered by the fact that the sperm tail cannot be seen in the zygote with standard light microscopy (Gordon et al., 1989; Malter and Cohen, 1989). Moreover, pronuclear size and morphology seem not to be associated with their general topography in the zygote (Wiker et al., 1990). The distribution of sex chromosomes and autosomes in daughter blastomeres of enucleated and intact dispermic embryos was studied by Tang etal. (1994) using multiprobe multicolour FISH and the polymerase chain reaction. The frequency of Y-bearing embryos diminished when the pronucleus furthest from the second polar body was targeted. From this, it was concluded that the female pronucleus is usually located next to the second polar body. The results presented above, however, indicate that extranuclear components remain unaltered in these enucleated zygotes, resulting in abnormal spindle formation and mosaicism. Potential correction of these embryos may be possible in the future, after removal of the extra centrosome using labelling techniques involving anti-centrosome antibodies. Human monospermic digynic embryos usually become diploid after removal of a single pronucleus, since the spindle is bipolar in these embryos. Considering the current findings, it is safe to replace enucleated tripronucleate digynic embryos. Further studies, however, are needed in order to determine whether a female pronucleus can be selected with high accuracy. Accidental removal of the single male pronucleus from these embryos may result in diploid digynic embryos, which are presumably not compatible with pregnancy. (Hall etal., 1989). This suggests the presence of control mechanisms similar to that for nuclear and mitochondrial DNA. Whether a maternal spindle-organizing centre can be created in human oocytes in the absence of the sperm centriole will need to be further clarified. Human parthenogenetic embryos develop poorly, whereas mouse parthenotes and dispermic embryos may develop and even implant, suggesting maternal inheritance of the centrosome in that species (Schatten etal., 1991). Over 30 parthenogenetic human embryos which were activated by exposure to a mannitol solution never developed to morulae (J.Levron and S.Willadsen, unpublished results). This confirms previous findings by Winston et al. (1991), who activated fresh oocytes parthenogenetically with alcohol or calcium ionophore. Five single pronucleate activated embryos developed into three cells after 4 days of culture. Only two were haploid, whereas the remainder were mosaic. Parthenogenetic single pronucleate embryos following ICSI develop much better. Of 20 such embryos, 11 developed into non-mosaic embryos and some were able to compact. It was found that only 10% of these embryos were fertilized. This confirms a previous report by Winston et al. (1991) who activated fresh oocytes parthenogenetically with alcohol or calcium ionophore. Those embryos never developed beyond the 8-cell stage. These findings also indicate the importance of extranuclear sperm elements for normal embryonic development in the human, even in such adverse genetic conditions as haploidy. In echinoderms, the centrosome is paternally inherited as well and parthenogenetic eggs have low development potential. Parthenogenetic development in the mouse, on the contrary, is compatible with early development, since in this species the maternal centrosome is retained. Mann and Lovell-Badge (1984) proved that the cytosol fraction of a parthenogenetic mouse egg can sustain entirely normal development, once the pronuclei are replaced with those from a fertilized zygote. The current findings suggest that the human maternal gamete has not entirely lost its ability to form a microtubule-organizing area, though female centrosomes have not been clearly identified in these zygotes (Sathananthan etal., 1991). The ability of the microtubuleorganizing centres to double appears to be lost at successive mitoses, indicating that it will be potentially difficult to produce parthenogenetic stem cells. The loss of this doubling capacity is determined by the egg and not by inhibitory factors from the fertilizing spermatozoon. Discussion An entirely new paradigm for cell genetics has arisen over the last two decades, that challenges Mendel's laws. Mitochondrial DNA is maternally inherited, because it is extranuclear in origin and thousands of copies exist in each cell. Most mitochondrial diseases are maternally inherited, some apparently occurring without family history. The potential of cells to undergo mitosis is also inherited in an extranuclear manner. Centrosomes in the early human embryo are paternally inherited and passed on from cell to cell. Whether or not this is associated with adverse conditions in the adult individual will need to be investigated. The molecular basis of centriole duplication in mammals may be similar to that of the single-celled green algae {Chlamydomonas, in which a circular DNA structure is associated with the centriole 1224 Acknowledgements The authors are grateful to Mina Alikani, Adrienne Reing and Dr Glenn Schattman for their assistance. Giles Tomkin is acknowledged for editorial comments and Drs Ledger and Rosenwaks for their support of this study. References Cohen,J., Malter,H.E., Talansky,B.E. and Grifo.J. (1992) Micromanipulation of Human Gametes and Embryos. Raven Press, New York. Croset,N. (1990) Behavior of the sperm centriole during sheep oocyte fertilization. Eur. J. Cell Bioi, 53, GordonJ.W., Grunfeld.L., Garrisi,G.J., Navot.D. and Laufer.N. (1989) Successful microsurgical removal of a pronucleus form tripronuclear human zygotes. Fertil. Steril., 52,

6 Paternal control of mitosis in embryos Gould,R.R. and Borisy.G.G. (1977) The pericentriolar material in Chinese hamster ovary cells nucleates microtubule formation. J. Cell Biol, 73, HalU.L., Ramanis.Z. and Luck,D.J. (1989) Basal body/centriolar DNA: molecular genetics studies in Chlamydomonas. Cell, 59, Hertig.A.T. and Adams,E. (1967) Studies on the human oocyte and its follicle. Ultrastructural and cytochemical observations on the primordial follicle. J. Cell Biol, 34, Kola,I. and Trounson.A. (1989) Dispermic human fertilization: violation of expected cell behavior. In Schatten,H. and Schatten,G. (eds), The Cell Biology of Fertilization. Academic Press, San Diego, CA, pp Kola,I., Trounson,A., Dawson.G. and Rogers,P. (1987) Tripronuclear human oocytes: altered cleavage patterns and subsequent karyotypic analysis of embryos. Biol. Reprod., 37, Long.C.R., Pinto-Correia.C, Duby.R.T., Ponce de Leon.A.F., Boland.M.P., Roche,J.F. and Robl.J.M. (1993) Chromatin and microtubule morphology during the first cell cycle in bovine zygotes. Mol. Reprod. Dev., 36, Longo,F.J. (1963) Sperm aster in rabbit zygotes; its structure and function. J. Cell Biol., 69, Malter.H.E. and Cohen.J. (1989) Embryonic development after microsurgical repair of polyspermic human zygotes. Fertil. Steril., 52, Maniotis.A. and Schliwa.M. (1991) Microsurgical removal of centrosome blocks cell reproduction and centriole generation in BSC-1 cells. Cell, 67, Mann,J.R. and Lovell-Badge.R.H. (1984) Inviability of parthenogenomes is determined by pronuclei, not egg cytoplasm. Nature, 310, Munne\S., Lee,A., Rosenwaks.Z., GrifoJ. and Cohen,J. (1993) Diagnosis of major chromosome aneuploidies in human preimplantation embryos. Hum. Reprod., 8, Palermo.G., Joris.H., Devroey.P. and Van Steirteghem.A.C. (1992) Pregnancies after intracytoplasmic injection of single spermatozoon into an oocyte. Lancet, 340, Pieters.M.H.E.C, Dumoulin,J.C.M., Ignoul-Vanvuchelen,R.C.M., Bras.M., Evers,J.L.H. and Geraedts.J.P.M. (1992) Triploidy after in vitro fertilization: cytogenetic analysis of human zygotes and embryos. J. Assist. Reprod. Genet., 9, Pinto-Correia.C, Collas.P., Ponce de Leon.F.A. and Robl,J.M. (1992) Parental centrosome inheritance in mammals: sperm aster formation in rabbit zygotes. 5th Int. Cong. Cell Biol., p Rappaport.R. (1969) Aster-equatorial surface relations and furrow establishment. J. Exp. ZooL, 171, Sathananthan.A.H., Kola.I., Osborne,J., Trounson,A.O., Ng,S.C, Bongso.A. and Ratnam.S.S. (1991) Centrioles in the beginning of human development. Proc. Natl. Acad. Sci. USA, 88, Schatten,H., Schatten.G., Mazia.D., Balczon,R. and Simerly,C. (1986) Behavior of centrosomes during fertilization and cell division in mouse oocytes and in sea urchin eggs. Proc. Natl. Acad. Sci. USA, 85, Schatten,G., Simerly.C. and Schatten.H. (1991) Maternal inheritance of centrosomes in mammals? Studies on parthenogenesis and polyspermy in mice. Proc. Natl. Acad. Sci. USA, 88, Sluder.G., Miller,F.J., Lewis.K., Davison,E.D. and Rieder.C.L. (1989) Centrosome inheritance in starfish zygotes: selective loss of the maternal centrosome after fertilization. Dev. Biol., 131, Szollosi.D. and Hunter.R.H.F. (1973) Ultrastructural aspects of fertilization in the domestic pig: sperm penetration and pronucleus formation. J. Anat., 116, Szollosi.D., Calarco,P. and Donahue.R.P. (1972) Absence of centrioles in the first and second meiotic spindles of mouse oocytes. J. Cell Sci., 2, Tang,Y.-X., Munne\S., Reing.A., Schattman.G., Grifo,J. andcohen.j. (1994) The parental origin of the distal pronucleus in dispermic human zygotes. Zygote, in press. Wheatley.D.N. (1982) The Centriole: A Central Enigma of Cell Biology. Elsevier, North-Holland Biomedical, New York. Wilson,E.B. and Mathews.A. (1895) Maturation, fertilization and polarity in the echinoderm egg. J. Morphol, 10, Wiker,S., Malter.H., Wright.G. and Cohen.J. (1990) Recognition of paternal pronuclei in human zygotes. J. In Vitro Fertil. Embryo Transfer, 1, Winston,N., Johnson,M., Pickering.S. and Braude.P. (1991) Parthenogenetic activation and development of fresh and aged human oocytes. Fertil. Steril, 56, Received on January 4, 1994; accepted on March 16,