Hunan Reproduction vol.7 suppl.l pp.89-94, 1992
Fertilization abnormalities in human in-vitro fertilization
Premature sperm chromosome condensation
'U 173 Inscrm, Hdpital Neckcr, 149 Rue de Sevres, 75743 Paris CEDEX 15, France 2 INRA Unite Biologie de la Fecondation, 78352 Jouy en Josas France
Premature chromosome condensation (PCC) was first reported in human oocytes by Schmiady et al. (1986) and Plachot et al. (1987). It is characterized by the presence of sperm chromatids in inseminated mature or immature oocytes which failed to form pronuclei and therefore were considered as unfertilized. Johnson and Rao (1970) showed, by inducing PCC by the experimental fusion of mitotic and interphase nuclei, that the morphology of prematurely condensed chromosomes varies according to the stage of the interphase nucleus at the time of fusion. PCC of Gl-phase cells are single chromatids, those of Sphase have a pulverized appearance, and those of the G2-phase, after DNA synthesis, consist of two chromatids. In human oocytes, chromatids and not chromosomes are most often observed, since chromosome condensation occurs before the S-phase which usually takes place in fully grown pronuclei. However, Schmiady and Kentenich (1989) occasionally observed PCC in fertilized eggs displaying 2 or 3 pronuclei. In our IVF programme, we routinely perform a cytogenetic study of unfertilized oocytes according to the method of Tarkowski (1966) to appraise the causes of failure. Sperm PCC was observed in 58 of 321 supposed unfertilized oocytes (18%), 11(19%) having more than one sperm chromosome complement (up to 8), thus indicating a high rate of polyspermy, and suggesting impairment of maturation. Indeed, the incidence of PCC in immature oocytes (metaphase I) 34%, was significantly higher than in mature oocytes (14%) (p < 0.01). Regarding metaphase II oocytes in this short series, we observed no effect of either IVF indication (tubal or unexplained infertility) or maternal age, but a slight increase following ovarian stimulation with FSH (30%) as compared to clomiphene/ human menopausal gonadotrophin (HMG) (11%), HMG alone (11%) or luteinizing hormone - releasing hormone agonist (LHRHa)/HMG (9%). These data show that using the latter stimulation protocol, routinely, 9% of cases are falsely diagnosed as fertilization failures. This phenomenon has been reported by other groups, the overall incidence of PCC ranging from 5% to 22.8% after LHRHa/HMG (Pieters et al., 1989; Ma et al., 1989; Schmiady and Kentenich, 1989; Delhanty and Penketh, 1990; Tarin and Pellicer, 1990; Macas et al., 1990; Janny et al., 1990, Zenzes et al., 1990). In an elegant study, Calafell et al. (1990) confirmed the effect of oocyte immaturity in the mouse, since the incidence of PCC after insemination increased from zero in mature control oocytes to 3.4% in in vitro matured oocytes and to 45.5% in immature oocytes. In humans, more PCC were reported when > 11 oocytes were recovered, suggesting that these oocytes had a relative cytoplasmic immaturity which could also explain
Fertilization abnormalities (premature chromosome condensation of spermatozoa (PCQ, triploidy, haploidy) were analysed in order to determine their origin. PCC occurs in 9% of unfertilized oocytes and seems to be the consequence of a failure of oocyte activation, leading to the continuing presence of cytoplasmic chromosome-condensing factors, causing the sperm nucleus to undergo chromosome condensation prematurely. This anomaly appears to be related to incomplete nuclear and/ or cytoplasmic maturation. Triploid zygotes (6.5% of fertilized oocytes) display an original type of division: half of them divide into 3 and 6 cells, whereas at the same time diploid zygotes divide into 2 and then 4 cells. A cytological study, using both antitubulin antibodies and Hoechst dye, allowed us to demonstrate that they divide into 3 cells by means of a tripolar spindle. Triploidy seems to be correlated with four of 16 clinical or biological parameters examined: semen origin (fresh or frozen), type of stimulation treatment, number of oocytes recovered and embryo morphology. Haploid eggs (1.6% of inseminated oocytes) result from parthenogenetic activation. A correlation was found between a high number of recovered oocytes and triploid zygotes, and the occurrence of oocyte activation. These data show that increasing follicular recruitment decreases the overall oocyte quality and maturity leading to an overall 9% with impaired fertilization.
Key Words: chromosomes, mitotic spindle, triploidy, haploidy, human oocytes, human embryos
Introduction
During the past six years, the cytogenetic analysis of oocytes, zygotes and embryos recovered from in vitro fertilization (IVF) has revealed a high rate of chromosome abnormalities in human, much higher than in other mammals (for review, Plachot et al, 1987). These are a consequence of either meiotic or mitotic non-disjunctions or of errors in fertilization, the latter frequently leading to an arrest of zygote or embryo development. This is why it is of paramount importance to investigate the causes of these anomalies in order to prevent their occurrence. Among the fertilization anomalies, premature sperm chromosome condensation, triploidy and haploidy will be discussed in detail. © Oxford University Press
89
Downloaded from http://humrep.oxfordjournals.org/ at Temple University on November 5, 2014
Michelle Plachotl. and Nicole Crozet2
M.Plachot and N.Croztt
#
4 » ' .'
Triploidy Triploidy in mammals may result from digyny or diandry. The extra chromosome set of digynic triploids has a maternal origin and may derive from fertilization of an egg with a non-reduced (diploid) chromosome number by a normal (haploid) spermatozoon. Conversely, the extra chromosome set in diandric triploids has a paternal origin and may result from the fertilization of normal (haploid) oocytes by diploid (non-reduced) spermatozoa, or from dispermic fertilization. After IVF, the great majority of triploid zygotes (86%) result from diandry, which is slightly higher than the 75% reported after in vivo fertilization (Plachot etal, 1989). Two hundred and nineteen of the 5729 oocytes inseminated in our IVF centre (3.8%) had 3 pronuclei when observed 17 hours after insemination, i.e. 6.5% of the fertilized oocytes. A higher rate was noted when aging oocytes (42 hours after oocyte collection) were reinseminated (50%) demonstrating that aging oocytes have a decreased protection against polyspermy, because of centripetal migration of the cortical granules. At least one tripronucleate zygote was observed in 26% of the cycles, the recurrence rate being 8%. Considering the developmental capacities of triploid zygotes, we observed that only 82% of them proceeded to the first two divisions as compared to 96% for diploid embryos. Moreover, an original type of division has been reported for about half of the triploid zygotes: they divided first into 3 cells and then into 6 cells, whereas diploid zygotes divide into 2 and then 4 cells (Kola et al., 1987; Plachot et al., 1989; Angell, 1989; Kola and Trounson, 1989). An immunocytochemical study was performed in order to analyze the formation of the first mitotic spindle. Tripronucleate eggs resulting from IVF were selected by stereomicroscopy about 17 hours after insemination. They were cultured 90
Fig. la,b,c Different sets of chromosomes observed in 9% of supposed unfertilized oocytes recovered after LHRHa/HMG; (a) polar body, (b) metaphase II chromosomes, (c) sperm chromatids.
further, until the first mitotic division, attested by the disappearance of pronuclei occurring between 27 and 32 hours after insemination. Eggs were then processed for immunofiuorescence microscopy according to Le Guen and Crozet (1989). Briefly, after fixation in acetic acid-ethanol, they were embedded in gelatin and cooled at -130° C. Cryostat sections (10 um) were successively exposed to mouse anti-tubulin antibody, FITC-labelled goat antibody against mouse IgG and Hoescht dye 33342 for visualization of chromosomes. We clearly identified 3 sets of chromosomes at the time of breakdown of the pronuclear envelope (Figure 2a), each of them surrounded by a prominent microtubular network (Figure 2b). Thereafter, instead of assembling on a single metaphase plate, sets of chromosomes remained separated and formed a Y-like arrangement in the centre of the egg (Figure 3a). Emergence of a tripolar spindle was clearly shown by anti-tubulin staining (Figure 3b) in 4 out of 8 eggs analyzed, the other 4 being in the premitotic stage and containing microtubules not yet organized on a spindle. Using electron microscopy, we are currently investigating the mode of formation of these tripolar spindles. Such spindles have already been described by Sathananthan et al. (1991). Moreover, they were observed in 3% of epithelial kidney cells in Microtus agrestis, probably as a consequence of the division of binucleate cells (Schwarzacher and Pera, 1969).
Downloaded from http://humrep.oxfordjournals.org/ at Temple University on November 5, 2014
their low fertilization rate (Tarin and Pellicer, 1990). However, the mechanism of PCC is still unclear. An explanation of the phenomenon in human oocytes was first given by Schmiady et al. (1986) who showed that the situation is dependent upon the permanent arrest of oocytes at metaphase II after sperm penetration, resulting from a failure of oocyte activation. It leads to the continuing presence of cytoplasmic chromosomecondensing factors, causing the sperm nucleus to undergo chromosome condensation prematurely. Angell et al. (1991) demonstrated that the less time the oocyte takes to mature in vivo, the greater is the likelihood of abnormal fertilization in which sperm PCC are formed. Indeed, the incidence of PCC reduced from 38% when oocyte recovery was scheduled because 3>3 follicles measuring > 16 mm in diameter were identified by ultrasound, to 17% when oocyte recovery was delayed by 1 or 2 days, thus indicating again the role of incomplete nuclear and cytoplasmic maturity of oocytes in this phenomenon. The fact that some oocytes classified as unfertilized demonstrated sperm PCC should be taken into account when considering a new IVF cycle. Greater attention should be paid to the choice of ovarian stimulation and to the monitoring of follicular growth and ovulation.
Fertilization anomalies in FVF
Continuing their development, we observed that 53% of these embryos reached the 8-cell stage, 34% the morulae stage and only 13% the blastocyst stage in standard culture conditions (B2+15% maternal serum). In cocultures on various monolayers, about 25% reached the blastocyst stage, but most often with developmental abnormalities (Plachot and Mandelbaum, 1990). The karyotypes of such embryos were unexpected, since only 24% were triploid. Thirty-two per cent were diploid and 5% haploid, probably because of the exclusion of one or two
Downloaded from http://humrep.oxfordjournals.org/ at Temple University on November 5, 2014
Fig. 2a,b. Tripronucleatc egg at the stage of nuclear envelope breakdown, double stained for chroraatin (a) and tubulin (b); (a) three sets of chromosomes; (b) each set of chromosomes is surrounded by microtubules x 1100.
Fig. 3a,b. Tripronucleate egg at first mitotic division, double stained for chromatin (a) and tubulin (b); (a) three sets of chromosomes are in a Y-like arrangement; (b) a tubulin staining delineates the tripolar spindle; two spindle poles are clearly visible on the section, the third is slightly out of focus, x 1100.
pronuclei at the first mitotic division. About 29% of these embryos showed various mosaicisms, probably induced by the anarchic migration of the chromosomes to the poles of the tripolar spindle. Similar results were reported by Michelmann et al. (1986), Angell et al. (1986), Wramsby et al., (1987), Kola and Trounson (1989), Pieters et al. (1990). 91
M.PUchot and N.Croiet
Among 16 clinical or biological parameters of IVF which were analysed (Table 1), four seemed to correlate with the occurrence of triploidy: the semen origin (fresh or frozen), the type of stimulation treatment, the number of oocytes recovered and embryo quality (Table 2). Indeed, when considering patients having tripronucleate zygotes once (in one cycle) or twice (in two cycles), we observed that the use of frozen semen (from husband or donor) was significantly associated with the risk of recurrence (37%) when compared with patients having no tripronucleate zygotes or only once. These data suggest that frozen semen induced a delay in fertilization, and hence oocyte aging. Royere et al. (1988) observed a delay in the appearance of pronuclei when using frozen semen, suggesting either a delay in sperm-egg recognition and interaction, or a delay in pronuclear growth.
Maternal age Type of infertility Sperm origin (fresh or frozen) Stimulation treatment Number of oocytes recovered Number of oocytes with zona pellucida breakage Number of atretic oocytes Number of immature oocytes Incubation time before insemination Fertilization rate Delayed or arrested fertilization Tripronucleate zygotes Parthenogenesis Embryo quality Sperm quality and survival Pregnancy rate
Haploidy
Table II. Parameters associated with the occurrence of triploidy once (one cycle) or twice (2 cycles or more) Cycles with or without triploid zygotes No triploid zygotes
Triploid zygotes in 1 cycle
In > two cycles
607
141
19
Frozen semen
12.1%
10.9%
37%
Fertilization abnormalities in human in-vitro fertilization
Premature sperm chromosome condensation
'U 173 Inscrm, Hdpital Neckcr, 149 Rue de Sevres, 75743 Paris CEDEX 15, France 2 INRA Unite Biologie de la Fecondation, 78352 Jouy en Josas France
Premature chromosome condensation (PCC) was first reported in human oocytes by Schmiady et al. (1986) and Plachot et al. (1987). It is characterized by the presence of sperm chromatids in inseminated mature or immature oocytes which failed to form pronuclei and therefore were considered as unfertilized. Johnson and Rao (1970) showed, by inducing PCC by the experimental fusion of mitotic and interphase nuclei, that the morphology of prematurely condensed chromosomes varies according to the stage of the interphase nucleus at the time of fusion. PCC of Gl-phase cells are single chromatids, those of Sphase have a pulverized appearance, and those of the G2-phase, after DNA synthesis, consist of two chromatids. In human oocytes, chromatids and not chromosomes are most often observed, since chromosome condensation occurs before the S-phase which usually takes place in fully grown pronuclei. However, Schmiady and Kentenich (1989) occasionally observed PCC in fertilized eggs displaying 2 or 3 pronuclei. In our IVF programme, we routinely perform a cytogenetic study of unfertilized oocytes according to the method of Tarkowski (1966) to appraise the causes of failure. Sperm PCC was observed in 58 of 321 supposed unfertilized oocytes (18%), 11(19%) having more than one sperm chromosome complement (up to 8), thus indicating a high rate of polyspermy, and suggesting impairment of maturation. Indeed, the incidence of PCC in immature oocytes (metaphase I) 34%, was significantly higher than in mature oocytes (14%) (p < 0.01). Regarding metaphase II oocytes in this short series, we observed no effect of either IVF indication (tubal or unexplained infertility) or maternal age, but a slight increase following ovarian stimulation with FSH (30%) as compared to clomiphene/ human menopausal gonadotrophin (HMG) (11%), HMG alone (11%) or luteinizing hormone - releasing hormone agonist (LHRHa)/HMG (9%). These data show that using the latter stimulation protocol, routinely, 9% of cases are falsely diagnosed as fertilization failures. This phenomenon has been reported by other groups, the overall incidence of PCC ranging from 5% to 22.8% after LHRHa/HMG (Pieters et al., 1989; Ma et al., 1989; Schmiady and Kentenich, 1989; Delhanty and Penketh, 1990; Tarin and Pellicer, 1990; Macas et al., 1990; Janny et al., 1990, Zenzes et al., 1990). In an elegant study, Calafell et al. (1990) confirmed the effect of oocyte immaturity in the mouse, since the incidence of PCC after insemination increased from zero in mature control oocytes to 3.4% in in vitro matured oocytes and to 45.5% in immature oocytes. In humans, more PCC were reported when > 11 oocytes were recovered, suggesting that these oocytes had a relative cytoplasmic immaturity which could also explain
Fertilization abnormalities (premature chromosome condensation of spermatozoa (PCQ, triploidy, haploidy) were analysed in order to determine their origin. PCC occurs in 9% of unfertilized oocytes and seems to be the consequence of a failure of oocyte activation, leading to the continuing presence of cytoplasmic chromosome-condensing factors, causing the sperm nucleus to undergo chromosome condensation prematurely. This anomaly appears to be related to incomplete nuclear and/ or cytoplasmic maturation. Triploid zygotes (6.5% of fertilized oocytes) display an original type of division: half of them divide into 3 and 6 cells, whereas at the same time diploid zygotes divide into 2 and then 4 cells. A cytological study, using both antitubulin antibodies and Hoechst dye, allowed us to demonstrate that they divide into 3 cells by means of a tripolar spindle. Triploidy seems to be correlated with four of 16 clinical or biological parameters examined: semen origin (fresh or frozen), type of stimulation treatment, number of oocytes recovered and embryo morphology. Haploid eggs (1.6% of inseminated oocytes) result from parthenogenetic activation. A correlation was found between a high number of recovered oocytes and triploid zygotes, and the occurrence of oocyte activation. These data show that increasing follicular recruitment decreases the overall oocyte quality and maturity leading to an overall 9% with impaired fertilization.
Key Words: chromosomes, mitotic spindle, triploidy, haploidy, human oocytes, human embryos
Introduction
During the past six years, the cytogenetic analysis of oocytes, zygotes and embryos recovered from in vitro fertilization (IVF) has revealed a high rate of chromosome abnormalities in human, much higher than in other mammals (for review, Plachot et al, 1987). These are a consequence of either meiotic or mitotic non-disjunctions or of errors in fertilization, the latter frequently leading to an arrest of zygote or embryo development. This is why it is of paramount importance to investigate the causes of these anomalies in order to prevent their occurrence. Among the fertilization anomalies, premature sperm chromosome condensation, triploidy and haploidy will be discussed in detail. © Oxford University Press
89
Downloaded from http://humrep.oxfordjournals.org/ at Temple University on November 5, 2014
Michelle Plachotl. and Nicole Crozet2
M.Plachot and N.Croztt
#
4 » ' .'
Triploidy Triploidy in mammals may result from digyny or diandry. The extra chromosome set of digynic triploids has a maternal origin and may derive from fertilization of an egg with a non-reduced (diploid) chromosome number by a normal (haploid) spermatozoon. Conversely, the extra chromosome set in diandric triploids has a paternal origin and may result from the fertilization of normal (haploid) oocytes by diploid (non-reduced) spermatozoa, or from dispermic fertilization. After IVF, the great majority of triploid zygotes (86%) result from diandry, which is slightly higher than the 75% reported after in vivo fertilization (Plachot etal, 1989). Two hundred and nineteen of the 5729 oocytes inseminated in our IVF centre (3.8%) had 3 pronuclei when observed 17 hours after insemination, i.e. 6.5% of the fertilized oocytes. A higher rate was noted when aging oocytes (42 hours after oocyte collection) were reinseminated (50%) demonstrating that aging oocytes have a decreased protection against polyspermy, because of centripetal migration of the cortical granules. At least one tripronucleate zygote was observed in 26% of the cycles, the recurrence rate being 8%. Considering the developmental capacities of triploid zygotes, we observed that only 82% of them proceeded to the first two divisions as compared to 96% for diploid embryos. Moreover, an original type of division has been reported for about half of the triploid zygotes: they divided first into 3 cells and then into 6 cells, whereas diploid zygotes divide into 2 and then 4 cells (Kola et al., 1987; Plachot et al., 1989; Angell, 1989; Kola and Trounson, 1989). An immunocytochemical study was performed in order to analyze the formation of the first mitotic spindle. Tripronucleate eggs resulting from IVF were selected by stereomicroscopy about 17 hours after insemination. They were cultured 90
Fig. la,b,c Different sets of chromosomes observed in 9% of supposed unfertilized oocytes recovered after LHRHa/HMG; (a) polar body, (b) metaphase II chromosomes, (c) sperm chromatids.
further, until the first mitotic division, attested by the disappearance of pronuclei occurring between 27 and 32 hours after insemination. Eggs were then processed for immunofiuorescence microscopy according to Le Guen and Crozet (1989). Briefly, after fixation in acetic acid-ethanol, they were embedded in gelatin and cooled at -130° C. Cryostat sections (10 um) were successively exposed to mouse anti-tubulin antibody, FITC-labelled goat antibody against mouse IgG and Hoescht dye 33342 for visualization of chromosomes. We clearly identified 3 sets of chromosomes at the time of breakdown of the pronuclear envelope (Figure 2a), each of them surrounded by a prominent microtubular network (Figure 2b). Thereafter, instead of assembling on a single metaphase plate, sets of chromosomes remained separated and formed a Y-like arrangement in the centre of the egg (Figure 3a). Emergence of a tripolar spindle was clearly shown by anti-tubulin staining (Figure 3b) in 4 out of 8 eggs analyzed, the other 4 being in the premitotic stage and containing microtubules not yet organized on a spindle. Using electron microscopy, we are currently investigating the mode of formation of these tripolar spindles. Such spindles have already been described by Sathananthan et al. (1991). Moreover, they were observed in 3% of epithelial kidney cells in Microtus agrestis, probably as a consequence of the division of binucleate cells (Schwarzacher and Pera, 1969).
Downloaded from http://humrep.oxfordjournals.org/ at Temple University on November 5, 2014
their low fertilization rate (Tarin and Pellicer, 1990). However, the mechanism of PCC is still unclear. An explanation of the phenomenon in human oocytes was first given by Schmiady et al. (1986) who showed that the situation is dependent upon the permanent arrest of oocytes at metaphase II after sperm penetration, resulting from a failure of oocyte activation. It leads to the continuing presence of cytoplasmic chromosomecondensing factors, causing the sperm nucleus to undergo chromosome condensation prematurely. Angell et al. (1991) demonstrated that the less time the oocyte takes to mature in vivo, the greater is the likelihood of abnormal fertilization in which sperm PCC are formed. Indeed, the incidence of PCC reduced from 38% when oocyte recovery was scheduled because 3>3 follicles measuring > 16 mm in diameter were identified by ultrasound, to 17% when oocyte recovery was delayed by 1 or 2 days, thus indicating again the role of incomplete nuclear and cytoplasmic maturity of oocytes in this phenomenon. The fact that some oocytes classified as unfertilized demonstrated sperm PCC should be taken into account when considering a new IVF cycle. Greater attention should be paid to the choice of ovarian stimulation and to the monitoring of follicular growth and ovulation.
Fertilization anomalies in FVF
Continuing their development, we observed that 53% of these embryos reached the 8-cell stage, 34% the morulae stage and only 13% the blastocyst stage in standard culture conditions (B2+15% maternal serum). In cocultures on various monolayers, about 25% reached the blastocyst stage, but most often with developmental abnormalities (Plachot and Mandelbaum, 1990). The karyotypes of such embryos were unexpected, since only 24% were triploid. Thirty-two per cent were diploid and 5% haploid, probably because of the exclusion of one or two
Downloaded from http://humrep.oxfordjournals.org/ at Temple University on November 5, 2014
Fig. 2a,b. Tripronucleatc egg at the stage of nuclear envelope breakdown, double stained for chroraatin (a) and tubulin (b); (a) three sets of chromosomes; (b) each set of chromosomes is surrounded by microtubules x 1100.
Fig. 3a,b. Tripronucleate egg at first mitotic division, double stained for chromatin (a) and tubulin (b); (a) three sets of chromosomes are in a Y-like arrangement; (b) a tubulin staining delineates the tripolar spindle; two spindle poles are clearly visible on the section, the third is slightly out of focus, x 1100.
pronuclei at the first mitotic division. About 29% of these embryos showed various mosaicisms, probably induced by the anarchic migration of the chromosomes to the poles of the tripolar spindle. Similar results were reported by Michelmann et al. (1986), Angell et al. (1986), Wramsby et al., (1987), Kola and Trounson (1989), Pieters et al. (1990). 91
M.PUchot and N.Croiet
Among 16 clinical or biological parameters of IVF which were analysed (Table 1), four seemed to correlate with the occurrence of triploidy: the semen origin (fresh or frozen), the type of stimulation treatment, the number of oocytes recovered and embryo quality (Table 2). Indeed, when considering patients having tripronucleate zygotes once (in one cycle) or twice (in two cycles), we observed that the use of frozen semen (from husband or donor) was significantly associated with the risk of recurrence (37%) when compared with patients having no tripronucleate zygotes or only once. These data suggest that frozen semen induced a delay in fertilization, and hence oocyte aging. Royere et al. (1988) observed a delay in the appearance of pronuclei when using frozen semen, suggesting either a delay in sperm-egg recognition and interaction, or a delay in pronuclear growth.
Maternal age Type of infertility Sperm origin (fresh or frozen) Stimulation treatment Number of oocytes recovered Number of oocytes with zona pellucida breakage Number of atretic oocytes Number of immature oocytes Incubation time before insemination Fertilization rate Delayed or arrested fertilization Tripronucleate zygotes Parthenogenesis Embryo quality Sperm quality and survival Pregnancy rate
Haploidy
Table II. Parameters associated with the occurrence of triploidy once (one cycle) or twice (2 cycles or more) Cycles with or without triploid zygotes No triploid zygotes
Triploid zygotes in 1 cycle
In > two cycles
607
141
19
Frozen semen
12.1%
10.9%
37%
