Spontaneous blastomere fusion after freezing and thawing of early human embryos leads to polyploidy and chromosomal mosaicism.

The incidence of blastomere fusion after cryopreservation of early human embryos (day 2 and day 3) was investigated using the standard propanediol technique. The process of fusion was observed in all developmental stages (from 2 to 10 cells) and the frequency of this event was 4.6% in day 2 (41/889) and 1.5% in day 3 (10/646) embryos that survived the thawing (embryos with 50-100% intact cells). Fusion of two, and occasionally of several, blastomeres resulted in the formation of multinucleated hybrid cells, which clearly indicated that the ploidy of these newly created cells had been altered. This event, depending on the number of fused cells per embryo, transformed the embryos into either entirely polyploid embryos (complete fusion at 2- or 3-cell stage) or into mosaics being a mixture of polyploid and normal cells. Chromosomal preparations of embryos affected by blastomere fusion indicated the presence of tetraploid mitotic plates. Also, fluorescence in-situ hybridization (FISH) analysis using DNA probes targeting unique sequences on chromosomes 9, 15, 17 and 22 indicated the existence of tetraploid and diploid fluorescence signals in the interphase nuclei within mosaics. Therefore, observations on live and fixed embryos suggested that tetraploid (4n) or hexaploid (6n) and tetraploid-diploid or more complex aberrations of ploidy might be formed as a consequence of blastomere fusion. Furthermore, this demonstrates that freezing and thawing may induce numerical chromosomal changes in human embryos.

The frequency and developmental capability of human embryos containing multinucleated blastomeres.

The frequency of multinucleated blastomeres (MNB) in 2- and 4-cell stage human embryos was recorded immediately before embryo transfer using a high-power inverted microscope. About 44% of patients (150/338) possessed embryos exhibiting MNB. The appearance of this nuclear abnormality was not correlated with maternal age. Overall, 15% of the otherwise good quality embryos (274/1885) that developed after monospermic fertilization contained several multinuclei (from two to seven) in at least one cell. Quite often MNB were found within all cells of the embryo (50% in 2-cell embryos). Blastomere multinucleation was significantly higher in 2-cell than 4-cell embryos (P <0.0001). This suggests that a considerable number of human embryos become abnormal during the first embryonic division. The embryos containing MNB were usually excluded for uterine transfers, with the exception of 19 cases when only such embryos could be replaced (6%; 19/338 patients). The results demonstrated that embryos with MNB may implant (4/19 cases; 21%) and they can lead to both spontaneous abortions and the successful birth of healthy infants (two cases). The fact that in the successful cases, 2-cell stage embryos with a mononucleated and a binucleated blastomere were transferred also suggests that due to the cell totipotency, development of a healthy baby is possible from one normal blastomere. Since multinucleation in early embryos may reflect gross chromosomal abnormalities or development of mosaic embryos, it is advisable not to replace embryos with MNB. Occasional transfers, however, can be considered because defective embryos may sometimes develop normally.

Color Doppler assessment of folliculogenesis in in vitro fertilization patients.

OBJECTIVES: To characterize hemodynamics of human ovarian follicles in the course of their development and to compare the results in relation to the age of patients, pregnancy, location of the ovary, and the quality of the oocytes. DESIGN: Serial measurements of peak velocity and resistance index by color Doppler imaging. SETTING: The In Vitro Fertilization Clinic, Success Through Assisted Reproductive Technologies, Inc., Toronto, Ontario, Canada. PATIENTS: Fifty-two women undergoing hormonal stimulation for IVF. MAIN OUTCOME MEASURES: The perifollicular peak velocity and resistance index values. RESULTS: The perifollicular peak velocity values were gradually increasing with the increasing size of the follicles. Highly significant elevation of the peak velocity was observed especially after hCG injection. Such rapid rise of blood velocity was greater in the right ovary than in the left in the younger patients (mean age, 34 years) whereas in older patients (mean age, 41 years), the greater values were observed in the left ovary. There was no significant difference in peak velocity or resistance index between pregnant and nonpregnant patients. CONCLUSIONS: There is a strong, positive correlation between the size of ovarian follicles and their peak velocity, which suggests an increase of blood flow around developing follicles in the course of the follicular phase. Human chorionic gonadotropin plays an important role in inducing influx of blood within follicles; however, the significance of this event remains obscure. It appears that the resistance index is not a useful parameter for characterization of the intrafollicular flow. Changes in peak velocity did not correlate with oocyte quality and may have no predictive value for IVF outcome.

Tripronuclear human zygotes: the first cell cycle and subsequent development.

About 6.6% of human oocytes fertilized in vitro were tripronuclear. The first cell cycle and developmental ability of these triploid zygotes were studied by autoradiographic analysis of pronuclear [3H]thymidine incorporation and by morphological observations of living zygotes at various times after insemination. Weak labelling over the pronuclei was first detected at 10 h after insemination and in many zygotes the pronuclei were heavily saturated with radioactive grains by 11-13 h. Thymidine incorporation was not apparent in the majority of tripronuclear zygotes by 21-22 h. At this time, 96% of living triploids (65/68) had visible pronuclei while the others entered the first mitosis and their pronuclei disappeared. By 25 h, 70% of triploid zygotes (37/58) still exhibited visible pronuclei and 26% were already in mitosis. The first divided embryos were seen by 25 h. The results suggest that the length of the complete cell cycle of human 1-cell stage triploids is approximately 25 h. Pronuclear DNA synthesis starts approximately 9-10 h after insemination and seems to last a minimum of 4 h (the precise time of S phase could not be determined). The minimal duration times of G2 phase and mitosis are estimated in the range of 4-6 h and 3.5-4 h, respectively. Tripronuclear zygotes were capable of substantial in-vitro development. About half of examined triploids attained at least the 4-6-cell stage of development (14/32; 44%); however, approximately 19% of these exhibited multinuclear cells. The remaining triploids had reached early morula (8 and 13 cells; 25 and 19%) or blastocyst (6%) stage.

Experimentally induced parthenogenetic activation of human oocytes.

A total of 486 unfertilized, aged human oocytes were exposed to ethanol, calcium ionophore A23187, phorbol ester or puromycin and examined for evidence of activation. Five per cent of control oocytes (3/58) were spontaneously activated. Of the two agents which cause the release of intracellular Ca2+ ions, Ca2+ ionophore induced activation of only 16% of unfertilized oocytes, while ethanol did not have any effect. Phorbol ester, a stimulator of protein kinase C, also resulted in limited activation (14% of oocytes). In contrast, puromycin, an inhibitor of protein synthesis, resulted in activation of 91% of the exposed oocytes. It is speculated that puromycin probably inhibits a specific cytostatic factor or factors which are responsible for maintenance of the metaphase II block. Morphologically activated oocytes usually retained the second polar body and formed subnuclei. The developmental potential of activated oocytes appeared to be reduced, with only some oocytes capable of a single division.

Characterization of abnormal one pronuclear human oocytes by morphology, cytogenetics and in-situ hybridization.

The morphology, chromosomal constitution and developmental capability of abnormal human oocytes (94/3500 oocytes; 2.7%) which after insemination exhibited only one pronucleus were examined. The majority of one pronuclear oocytes exhibited two or more distinct polar bodies. Dividing oocytes showed irregular chromosome distribution from haploid to diploid. Embryos resulting from abnormal oocytes displayed limited developmental potential. Many of them underwent fragmentation or were arrested at the 2- or 8-cell stage of development, and only some reached the morula or blastocyst state (11/35 oocytes). In approximately 45% (15/33) of examined oocytes, decondensed sperm heads or tiny nucleus-like structures were found in addition to a single nucleus. Chromosome Y was also detected in chromosomal preparations in approximately 10% of the oocytes by in-situ hybridization utilizing the human Y chromosome-specific DNA probe (DYZ3). These observations provided strong evidence that many of these oocytes originated from fertilized oocytes. The origin of the other one pronuclear oocytes could not be determined. Parthenogenetic activation of some oocytes cannot be excluded and other explanations concerning the origin of abnormal oocytes are discussed.

Characterization of the first cell cycle in human zygotes: implications for cryopreservation.

OBJECTIVE: To study the temporal pattern of deoxyribonucleic acid (DNA) synthesis leading up to the first mitotic division in human one-cell stage zygotes. SETTING: In vitro fertilization program of a university hospital. PATIENTS: Couples donating spare embryos before the existence of an embryo freezing program. DESIGN: Incorporation of 3H-thymidine was examined between 9 to 27 hours after insemination in 72 untransferred human zygotes containing two pronuclei. Microscopic observations on an additional 978 transferred zygotes extended the 3H-thymidine incorporation data. RESULTS: The first thymidine incorporation was seen 9 to 10 hours after insemination, and the heaviest incorporation occurred between 11 and 13 hours. Subsequently, thymidine labeling declined, and chromosomal condensation and cell division first occurred approximately 19 to 20 hours. At 20 hours after insemination, 89% of the zygotes had two visible pronuclei (PN). In contrast, by 24 hours, 41% had no visible PN, whereas 5% had cleaved to the two-cell stage. By 27 hours, 38% had cleaved to the two-cell stage, and only 25% still had two visible PN. CONCLUSIONS: These results suggest that the DNA-synthetic, S-phase of the human zygote is initiated by approximately 9 to 10 hours after insemination and is completed approximately 3 to 5 hours later. The duration of G2 phase and mitosis is in the range of 4 to 6 and 3 to 3.5 hours, respectively. Because zygotes may be particularly susceptible to damage during the S phase of the cell cycle, these findings suggest that the optimal time for freezing of human zygotes may be approximately 20 to 22 hours after insemination when the majority of zygotes should have entered the G2 phase, before pronuclear dissolution and chromosome condensation.

A morphologic study of unfertilized oocytes and abnormal embryos in human in vitro fertilization.

The morphology of human, unfertilized oocytes and abnormal embryos cultured in vitro for 48-72 hr was examined in an attempt to learn more about oocyte maturation and reproductive failure in in vitro fertilization (IVF). About 21% of the unfertilized oocytes were totally degenerated. The majority (56%) of the remaining oocytes was arrested at the metaphase II stage. They contained coherent chromosomal plates and had extruded the first polar body with nuclear material. About 13% of oocytes underwent spontaneous activation. In most of these cases the second polar body was retained and many subnuclei or one big nucleus was formed. Five percent of metaphase II oocytes penetrated by sperm were not activated, likely as a result of oocyte immaturity. The developmental ability of abnormal embryos was poor. Several one-cell-stage zygotes were arrested at the pronuclear stage or at mitosis of the first mitotic division. Polyspermic embryos, especially those which contained four or more pronuclei, did not divide or formed uneven, multinucleated blastomeres. However, some triploid and tetraploid embryos often appeared normal morphologically despite their lethal chromosomal abnormalities.

The effect of cryopreservation on the development of S- and G2-phase mouse embryos.

The survival rate and development of four-cell-stage mouse embryos frozen and thawed in S phase versus G2 phase was compared. Significantly more G2-phase than S-phase embryos survived freezing and thawing. In both groups, disruption of the zona pellucida, fusion of blastomeres, and dispersion of chromosomes were occasionally observed after thawing. Cryopreservation resulted in a longer delay in cleavage from the four- to the eight-cell stage of S (about 5 hr)- and G2-phase embryos (about 3 hr) compared to unfrozen controls. The number of frozen embryos which developed to the blastocyst stage was reduced compared to controls, and in the case of S-phase embryos, formation of the blastocyst cavity was also delayed. However, the average number of cells in the experimental and control embryos was similar. No increased incidence of chromosome abnormalities was seen. Our results suggest that freezing embryos in G2 is superior to freezing in S phase.

Use of gonadotropin-releasing hormone agonist to trigger follicular maturation for in vitro fertilization.

In spontaneous cycles both LH and FSH are secreted in a surge at midcycle. In in vitro fertilization (IVF) cycles, hCG administration results in elevation of LH-like activity only. The objective of this study was to compare the effectiveness of a single midcycle dose of GnRH agonist with hCG on follicular maturation. Eighteen IVF cycles in 14 women were randomized to receive either 0.5 mg leuprolide acetate or 5000 IU hCG at midcycle. Both groups underwent identical ovarian stimulation and cycle monitoring. On the day of GnRH agonist or hCG administration, estradiol concentrations and the number of follicles 1.5 cm or larger were the same in both groups. Mean serum LH and FSH levels were elevated for 34 h after GnRH agonist administration. In contrast, mean serum hCG levels were elevated for approximately 6 days after the administration of hCG, and serum FSH levels did not change. Mean luteal phase serum estradiol concentrations were lower in the GnRH agonist group than in the hCG group (P less than 0.02). No differences were observed in mean serum progesterone or PRL during the luteal phase or in the length of the luteal phase in the two groups. The mean number of oocytes retrieved and embryo number and quality did not differ between the two groups. Three of nine GnRH agonist cycles and none of nine hCG cycles resulted in clinical pregnancy (P = 0.1). The results of this study indicate that GnRH agonist is able to simulate a midcycle surge of gonadotropins, leading to follicular maturation and pregnancy. Further work is needed to determine whether there is any clinical advantage of GnRH agonist over hCG administration with regard to pregnancy rates.

Infrequent genomic rearrangement and normal expression of the putative RB1 gene in retinoblastoma tumors.

Retinoblastoma (RB) tumors develop when both alleles of a gene (RB1) are mutated and unable to function normally. Recently, Friend et al. [S. H. Friend, R. Bernards, S. Rogelj, R. A. Weinberg, J. M. Rapaport, D. M. Albert, and T. P. Dryja, Nature (London) 32:643-646, 1986] reported the cloning of a gene, 4.7R, with some properties expected for the RB1 gene, namely, a high frequency (30%) of genomic rearrangements in tumors and absence of message in all RB tumors examined. To extend the characterization of this gene, we used 4.7R probes to search for genomic rearrangements of DNA and to study the expression of the 4.7R gene in RB tumors, osteosarcoma (OS) tumors arising in RB patients, and other normal and malignant tissues. In 34 previously unreported RB and OS tumors arising in RB patients, we observed only four (12%) with genomic abnormalities. Transcripts of 4.7R were present in 12 of 17 RB tumors, 2 of 2 OS tumors, and all non-RB tumors and normal tissues tested. We were unable to confirm the high frequency of truncated messages of 4.7R in RB tumors reported by Lee et al. (W. H. Lee, R. Bookstein, F. Hong, L. J. Young, J. Y. Shaw, and E. Y. Lee, Science 235:1394-1399, 1987) and Fung et al. (Y. K. Fung, A. L. Murphree, A. Tang, J. Qian, S. H. Hinrichs, and W. F. Benedict, Science 236:1657-1661, 1987) but did confirm the presence of a truncated transcript in the RB cell line Y79. Of the RB and RB-related OS tumors which appeared normal on Southern blots, 2 of 26 or 12% had abnormal transcripts, giving a combined frequency of 22% abnormalities in the 4.7R gene detectable by Southern and Northern (RNA) blot analyses.

Retinoid-binding proteins in retinoblastoma tumors.

A combination of Western blot, Northern blot, and radiolabeled ligand-binding techniques was used to investigate retinoid-binding proteins in retinoblastoma (RB) cells from fresh tumors and from 19 RB tumor lines cultured in vitro. Using rabbit anti-bovine cellular retinal-binding protein (CRA1BP) antibodies, no CRA1BP could be detected. As determined by [3H]retinol binding, cellular retinol-binding protein was sometimes not detectable but averaged 2.3 +/- 2.7 means +/- SD, n = 7) pmol [3H]retinol bound/mg protein, similar to adult retina cytosol. Using [3H]retinoic acid as ligand, cellular retinoic acid-binding protein was not detectable in some lines and averaged 1.0 +/- 1.2 (means +/- SD, n = 7) pmol [3H]retinoic acid bound/mg protein, well below the adult retina cytosol level of 94.4 +/- 16.3 (means +/- SD, n = 4) pmol [3H]retinoic acid bound/mg. Using rabbit antibovine interstitial retinol-binding protein (IRBP) antibodies, IRBP of the same molecular mass as human IRBP (135,000) was detected in the medium from all cultured RB cells and averaged 75.9 +/- 19.2 pmol/10(8) cells (bovine IRBP immunochemical equivalents). Cytosol levels were less than 1% of the medium. Using a human IRBP complementary DNA probe, levels of IRBP RNA transcripts in 19 RB cell lines were comparable to adult retina. The Y-79 RB cell line was different from the others; the amount of IRBP in the medium was only about 1% of the RB cell lines. Levels of cellular retinol-binding protein were comparable with the other lines, but cellular retinoic acid-binding protein was 9 times more abundant. IRBP RNA transcripts in Y-79 cells were below the limits of detectability but appeared at low levels after induction of differentiation of Y-79 by 10(-6) M retinoic acid. Although this cell line has been in culture longer than the others, it may also have been initiated at an earlier stage of retinal development.

Origin of the inner cell mass in mouse embryos: cell lineage analysis by microinjection.

The mouse inner cell mass is established by cells that are allocated to internal positions after the 8-cell stage. We analyzed the timing of this allocation by microinjecting two cell lineage markers, horseradish peroxidase and rhodamine-conjugated dextran, into mouse blastomeres at the 8- to 32-cell stage. Prospective analysis was performed by coinjection of peroxidase and dextran, followed by 12-22 hr of culture and staining for peroxidase activity; retrospective analysis was performed by injection of peroxidase alone and localization of sister cells without further culture. Both approaches indicated that cells are allocated to internal positions during the fourth and fifth cleavage divisions, but not the sixth cleavage division, of the mouse embryo. Thus, outer cells can have inner descendants until the late morula/early blastocyst (32-cell) stage, but cells remaining outside after the fifth cleavage division are restricted to a trophectoderm fate. This information about cell lineage indicates that the previously observed totipotency of the cleaving mammalian embryo's cells is a regulative attribute that is used in normal development.

Interactions between metaphase and interphase factors in heterokaryons produced by fusion of mouse oocytes and zygotes.

The cytoplasmic factor responsible for chromosome condensation was introduced into mouse zygotes at different times after fertilization by fusion of the zygotes with metaphase I oocytes. In 72% of heterokaryons obtained after fusion of early zygotes (14-18 hr post-human chorionic gonadotrophin (HCG) with oocytes, the male and female pronuclei of the zygote decondensed. At the same time, the oocyte chromosomes became enclosed in a nuclear envelope and decondensed to an interphase state. However, in the rest of the heterokaryons, the chromatin of the pronuclei condensed to metaphase chromosomes, thus resulting in three sets of chromosomes. Fusion of zygotes that had begun DNA synthesis (20-22 hr post-HCG) with oocytes induced chromosome condensation of the pronuclei in 76% of the cases. In some heterokaryons, however, the oocyte chromosome decondensed to an interphase state similar to the zygote pronuclei. Fusion between late zygotes (27-29 hr post-HCG) with oocytes resulted in chromosome condensation of the pronuclei in all heterokaryons. On the basis of these results, the formation of the pronuclei and their progression toward mitosis in the zygote may be explained by changing levels of a metaphase factor in the cell, or by a balance between interphase and metaphase factors.

Chromosome condensation activity in the cytoplasm of anucleate and nucleate fragments of mouse oocytes.

The activity of maturation promoting factor (MPF) which causes chromosome condensation and subsequent oocyte maturation was investigated in mouse oocytes using polyethylene-glycol-mediated cell fusion technique. Fully grown oocytes were bisected at germinal vesicle (GV) stage or shortly after germinal vesicle breakdown (GVBD) into anucleate and nucleate fragments. After 2-3 or 15-17 hr of culture these fragments were fused with interphase blastomeres from two-cell embryos. It was found that almost all the anucleate oocyte fragments cultured for a short term (2-3 hr), regardless of whether they were produced at GV stage or after GVBD, induced premature chromosome condensation in the blastomere nuclei, whereas only about 20% of those cultured for a long term (15-17 hr) could do so. On the other hand, the nucleate fragments always retain the cytoplasmic activity to induce chromosome condensation. Thus we suggested that the MPF initially could appear in mouse oocytes independently of the GV, that the mixing of GV material with the oocyte cytoplasm following GVBD had no effect on the activity of MPF in anucleate fragments, and that oocyte chromosomes or some components associated with them could play a significant role in maintaining the MPF activity.

Endogenous peroxidase in the visceral endoderm of early mouse embryos.

Endogenous peroxidase activity was demonstrated in early mouse embryos by means of the diaminobenzidine staining reaction. This enzyme was observed in visceral endoderm on the seventh to eighth day of gestation in vivo, but was no longer detected on the ninth day of development. In cell layers developing from blastocysts or isolated inner cell masses cultured for 96-144 h (developmental stage equivalent to 6-7.5-day-old embryos), diaminobenzidine product was also observed in visceral endodermal cells. Most of the endogenous peroxidase was localized inside or close to the numerous apical vacuoles in the endoderm. Ectoderm, mesoderm, ectoplacental cone, and trophoblast cells did not contain endogenous peroxidase.

Nucleo-cytoplasmic interactions in cell hybrids between mouse oocytes, blastomeres and somatic cells.

With the help of the technique of Sendai virus-mediated cell fusion, hybrid cells were produced between two maturing oocytes, between maturing oocytes or mature secondary oocytes and interphase blastomeres from 2-cell embryos, and between secondary oocytes and follicle cells (FC). In the first case giant oocytes form and in these the two groups of condensing bivalents join on a common spindle, undergo a first meiotic division, and become arrested in metaphase II; these hybrids are heterozygous diploids. When blastomeres are fused with oocytes undergoing maturation, then the blastomere nucleus undergoes premature chromosome condensation (PCC), and two sets of chromosomes (meiotic and mitotic) come close to each other but the mitotic chromosomes are not fully incorporated into the meiotic spindle. The behaviour of the nuclei from blastomeres or from follicle cells fused with secondary ovulated oocytes, depends on whether or not the oocyte undergoes activation. When the oocyte is not activated, then the introduced nuclei undergo PCC but the chromosomes remain separate from the spindle of metaphase II. When the oocyte is activated, then the introduced nuclei remain in interphase; the FC nuclei may increase their volume eight times and undergo some other structural changes but during the 6 h period after fusion their appearance remains distinct from that of the female pronucleus. Since some oocytes are not activated after fusion has occurred, it follows that the fusion process itself is not sufficient to trigger egg activation.

Interspecific heterokaryons between oocytes and blastomeres of the mouse and the bank vole.

Interspecific hybrid cells were produced using Sendai-virus-mediated fusion of interphase blastomeres from 2-cell embryos with ovulated oocytes of the mouse and bank vole. Under the influence of the cytoplasm of oocytes (arrested in metaphase II) premature chromosome condensation occurs in the nuclei of blastomeres, thus implying that the cytoplasmic factor which induces chromosome condensation is not species specific, at least not between these two species.