Article Duplication of the sperm genome by human androgenetic embryo production: towards testing the paternal genome prior to fertilization

Transcription

1 RBMOnline - Vol 14. No Reproductive BioMedicine Online; on web 20 February 2007 Article Duplication of the sperm genome by human androgenetic embryo production: towards testing the paternal genome prior to fertilization Dr Valeriy Kuznyetsov received his PhD in 1986 from the Russian Institute of Animal Genetics and Breeding, St. Petersburg. He moved to the Reproductive Genetics Institute in His current research interests include assisted fertilization, early embryogenesis and nuclear transfer in mammals. Dr Valeriy Kuznyetsov Valeriy Kuznyetsov, Iryna Kuznyetsova, Mark Chmura, Yury Verlinsky 1 Reproductive Genetic Institute, 2825 North Halsted Street, Chicago, IL 60657, USA 1 Correspondence: Tel: ; Fax: ; rgi@flash.net Abstract There is currently no technique for evaluating the sperm genome before fertilization. However, sperm genome duplication could offer a way forward, whereby one of the sister blastomeres of a 2-cell haploid androgenetic embryo could be analysed. A method was developed for production of human androgenotes by enucleation of oocytes at telophase II (TII) after intracellular sperm injection (ICSI). The results were compared with those obtained via the more usual procedure of oocyte enucleation at metaphase II (MII) prior to ICSI. TII enucleation led to an improvement in the rate of embryo survival, increased the production rate of 1PN-embryos, and also the production of 2- to 8-cell-stage embryos (85.0, 74.9 and 65.8% in TII enucleation, versus 73.8, 48.9 and 33.3% in MII enucleation). Fluorescence in-situ hybridization (FISH) analysis of to 5-cell androgenic embryos for two to seven chromosomes revealed the correct chromosome distribution in 76.7% of haploid human androgenotes. Keywords: FISH, human androgenetic embryos, oocyte enucleation, paternal genome testing, sperm genome duplication Introduction 504 Preimplantation genetic diagnosis (PGD) is currently performed by embryo biopsy at the cleavage stage, or through first (PB1) and second (PB2) polar body sampling, following the fertilization of oocytes. The latter provides information about the maternal genetic contribution to the human embryos, which is applicable for PGD of autosomal recessive, X-linked and maternally-derived dominant conditions (Verlinsky and Kuliev, 2005). PB-based PGD is also particularly useful for testing of chromosomal abnormalities, the majority of which originate from female meiosis (Kuliev et al., 2002). Although the rate of chromosomal abnormalities is much lower in human spermatozoa (Templado et al., 2005), it is increased considerably in males with balanced translocations (Munné, 2002), and in severe oligo-asthenoteratozoospermia (Egozcue et al., 2000). However, there is currently no technique to investigate the genotype of the sperm genome prior to using it for fertilization, as direct testing destroys the spermatozoon. It has recently been suggested that the sperm genome may be evaluated without destruction through human sperm genome duplication, following its transfer into enucleated mouse or cow oocytes (Munné et al., 2003; Willadsen et al., 2003). However, the efficiency and fidelity of sperm genome duplication may be species specific, so this study describes a method for the production of human androgenetic embryos and for checking the resulting chromosomal constitution of the duplicated genome. This method may provide a way of analysing the sperm genotype prior to fertilization Published by Reproductive Healthcare Ltd, Duck End Farm, Dry Drayton, Cambridge CB3 8DB, UK

2 Materials and methods Source of human oocytes Oocytes were obtained after informed consent from patients undergoing assisted reproduction treatment in the authors institute. Germinal vesicle (GV) oocytes were cultured in human tubal fluid (HTF) medium supplemented with 5% Plasmanate (Bayer, USA) for 24 h for in-vitro maturation to metaphase II (MII) stage. Micromanipulation technique and embryo culture Intracytoplasmic sperm injection (ICSI), partial zona dissection with a glass needle, polar body removal and blastomere biopsy were performed as described elsewhere (Verlinsky and Kuliev, 2005). To obtain haploid androgenotes, two different techniques of oocyte enucleation were used (Figure 1). The oocytes were enucleated either at MII (Edwards and Sirlin, 1959; Kono et al., 1993; Lagutina a et al., 2003; Tesarik et al., 2003; Verlinsky and Kuliev, 2005), or at telophase II (TII). For enucleation at MII (Figure 1A), the zona pellucida was dissected using a glass microneedle, followed by removal of PB1 plus the adjacent 20% of ooplasm with a micropipette of outer diameter μm. To confirm the removal of metaphase chromosomes, the oocytes were stained with Hoechst with a few seconds exposure to ultraviolet (UV). Injection of a single spermatozoon was performed 30 min after oocyte enucleation. It is also possible to obtain heterozygous diploid androgenotes by simultaneous injection of two spermatozoa into the enucleated MII oocytes. Oocyte enucleation at TII was previously proposed for the production of cow oocyte ooplasts and subsequent nuclear transfer (Bordignon and Smith, 1998), and has been reproduced by others (Kuznyetsov and Verlinsky, 2005). The technique is based on removing TII chromosomes following oocyte parthenogenetic activation, which is similar to PB2 removal with approximately 10% of the surrounding cytoplasm (Bordignon and Smith, 1998). TII enucleation was performed either after PB1 removal or without PB1 removal (Figure 1B). In both cases, oocytes were monitored every 30 min, 2 h after the single sperm injection. Immediately after PB2 extrusion, PB2 (together with minimal amount of underlying ooplasm) was removed using an enucleating pipette with an outer diameter of μm. No cytochalasin was used to facilitate the enucleation in either procedure, and the resulting oocytes were cultured in Global medium (LifeGlobal, USA) supplemented with 10% Plasmanate for h. On day 1, the resulting androgenetic embryos and small oocyte karyoplasts resulting from TII enucleation were checked for the presence of a male or female pronucleus, respectively. Then after embryo morphological assessment on day 2, blastomere biopsy was performed. Blastomeres were either fixed for FISH analysis in interphase, or further cultured in the same medium with vinblastine sulphate (100 ng/ml; Sigma, USA) for 1 12 h to obtain metaphase plates (see Figure 1, starting from 1-cell embryo). Parthenogenetic activation of oocytes Human oocytes were activated to parthenogenetic development using the XRONOS-1 electrofusion apparatus (RGI, Chicago, USA) via one or three sets (in 30 min) of double direct current (DC) pulses of 1.2 kv/cm in a medium with 0.3 mol/l mannitol, 0.1 mmol/l MgSO 4, 0.05 mmol/l CaCl 2 and 0.05 mg/ml bovine serum albumin. FISH and G-banding analysis Fixed blastomere interphase nuclei were analysed using the FISH technique, as described elsewhere (Verlinsky and Kuliev, 2005), while the metaphase plates were studied either by FISH or G- banding techniques, depending on their quality. FISH analysis was performed using the following probes (Abbott, USA): MultiVysion PB Panel, separately or with addition of CEP X and CEP Y, or CEP 17 and CEP X; MultiVysion PGT with addition of 22q or 8q, 15q, CEP 17; and also CEP 16, CEP 17, CEP X, CEP Y, 13q, 16p probes. Results The data presented in Table 1 show that oocyte survival after enucleation depended on the stage of oocyte maturation during enucleation (MII or TII), as well as on the context of oocyte maturation (in vitro or in vivo). Compared with MII, TII was much more suitable for enucleation of GV-MII oocytes in respect of the rate of survival and the subsequent production of 1PN and 2- to 8-cell haploid androgenotes. Oocyte enucleation at TII also revealed differences between oocytes matured in vitro or in vivo (including frozen thawed MII oocytes) in their capacity for further development. When enucleated in MII, differences between MII and GV-MII oocytes became apparent only in the rate of their survival. Rates of production of 1-cell and 2- to 8-cell embryos were similar for these groups of oocytes and significantly lower compared with enucleation at TII (for P-values, see Table 1). It is interesting to note that following injection with two spermatozoa after enucleation, GV-MII oocytes did not produce two pronuclei; all three 1-cell embryos contained one pronucleus. FISH analysis of two 2-cell androgenotes for chromosomes 16, X and Y revealed a diploid set of these chromosomes. The rates for obtaining 1-cell and cleaved embryos were similar for haploid and diploid androgenotes. PB1 removal before TII enucleation did not affect the efficiency of oocyte enucleation and androgenote developmental capacity. Therefore, these data were not analysed separately. However, it has to be taken into account that PB1 morphology might often change, even during a short period of culture (Verlinsky et al., 2003). It is of note, that after TII enucleation, the androgenetic GV-MII embryos showed the presence of a small male pronucleus as early as 5 h after ICSI, while the female pronucleus in the oocyte karyoplast appeared later, at 6.5 h. A small volume of karyoplast cytoplasm could explain this delay, as the female pronucleus in the intact zygotes generally also becomes visible 5 h after ICSI. During subsequent culture, the increase in size of male pronucleus was observed on day 1, and development of the androgenetic embryos to the 7- to 8-cell stage on day 3 (Figure 2). 505

3 Figure 1. Duplication of sperm genome by human haploid androgenetic embryo production. (A) Oocyte enucleation at metaphase II. (B) Oocyte enucleation at telophase II (see description in text). 506 It has been shown that human PB2 may be extruded not earlier that 1 h after ICSI (Van den Bergh et al., 1995), so PB2 presence was checked for 2 h after ICSI. It was shown that the majority of oocytes extruded PB h after ICSI, depending on the maturation status of oocyte cytoplasm (Table 2). In-vitro matured oocytes had a tendency to extrude PB2 later than in-vivo matured ones. As shown in Table 3, the number of electric pulses had no influence on the activation rate of human oocytes, but affected the type of activation. The rate of haploid parthenogenote (1 PN and 2 PB) production was higher after one set than after three sets of double DC pulses. Increase in the number of DC pulses led to an increase in the frequency of diploid parthenogenotes (1 PN and 1 PB; 2 PN and 1 PB); 38.9 versus 6.7%, obtained after a double DC pulse. Three sets of pulses also promoted an increase in the number of 2 4-cell parthenogenetic embryos, some of which developed to the blastocyst stage. FISH analysis of blastomeres for two to seven chromosomes revealed the correct distribution of chromosomes in 23 (76.7%) of 30 of the 2 5-cell androgenetic embryos (Table 4, Figure 3). Figures 4 and 5 show G-banding of the haploid metaphase chromosome plates from 2-cell human androgenetic embryos obtained with the use of GV-MII oocytes, with normal and abnormal chromosome sets.

4 Table 1. Effect of enucleation technique on human oocyte survival and developmental ability of resulting androgenetic embryos. Stage during Group of No. of oocytes No. (%) of androgenotes at: Ploidy of enucleation oocytes Total Survived 1PN stage* 2- to 8-cell stage** androgenotes enucleation (%) Metaphase II GV-MII a (73.8) 44 e (48.9) 30 i (68.2) Haploid GV-MII 8 6 (75.0) 3 (50.0) 2 (66.7) Diploid MII b (97.0) 17 f (53.1) 11 k (64.7) Haploid Telophase II GV-MII c (85.0) 140 g (74.9) 123 l (87.9) Haploid MII 6 6 (100) 6 (100) 6 (100) Haploid Thawed MII 7 7 (100) 7 (100) 6 (85.7) Haploid Total MII d (100) 13 h (100) 12 m (92.3) Haploid a,b; f,g; f,h; i,l Values are significantly different (P < 0.01), χ 2 test. a,c; a,d; k,l Values are significantly different (P < 0.05), χ 2 test. e,g; e,h; i,m Values are significantly different (P < 0.001), χ 2 test. b,c,d; e,f; i,k; k,m, l,m No significant difference. *Calculated from number of oocytes surviving enucleation. **Calculated from number of 1PN embryos. PN = pronucleate. Note: GV-MII = in-vitro-matured oocytes; MII = in-vivo-matured oocytes. Figure 2. (a) Haploid human androgenetic single-pronucleated embryo. (b) Haploid oocyte karyoplast with female pronucleus (open arrow) and second polar body (solid arrow). (c) Fixed pronucleus of oocyte karyoplast (20 h after ICSI). (d-f) Cleavage stage haploid human androgenetic embryos; (d) 2-cell embryo (29 h after ICSI); (e) 4-cell embryo (45 h after ICSI); (f) 7-cell embryo (67 h after ICSI) ( 200). 507

5 Table 2. Time of second polar body extrusion after intracytoplasmic sperm injection before telophase II enucleation. Group of No. of No. (%) of oocytes with second PB in: oocytes oocytes h h h GV-MII a (25.5) 54 (55.1) 19 (19.4) MII 6 6 b (100) Thawed MII 7 2 (28.6) 5 (71.4) a,b Values are significantly different (P < 0.001), χ 2 test. GV = germinal vesicle; MII = metaphase II; PB = polar body. Table 3. Developmental ability of human oocytes activated to parthenogenesis by electric pulses. Number of No. of No. (%) of activated oocytes No. (%) of parthenogenetic double DC GV-MII Total 1PN 1PN 2PN 3PN embryos in cleavage stages pulses oocytes and 2PB and 1PB and 1PB and 1PB 2- to 4-cell 5- to 8-cell Blastocysts (day 2) (day 3) (day 5) One a (80.0) 10 c (66.7) 1 (6.7) 0 1 (6.7) 9 e (60.0) 4 (26.7) 0 Three* b (83.3) 7 d (38.9) 4 (22.2) 3 (16.7) 1 (5.6) 15 f; (83.3) 6 (33.3) 3 (16.7) a,b; c,d; e,f No significant difference. *30-min interval between each double pulse. GV = germinal vesicle; MII = metaphase II; PB = polar body; PN = pronucleate. Table 4. Fluorescence in-situ hybridization analysis of human haploid androgenetic embryos. Group of Embryo No. of embryos oocytes developmental Total No. with correct No. of chromosomes analysed (%) stage chromosome disjunction in each blastomere (%) GV-MII 2-cell 10 8 (80.0) 6 (75.0) 1 (12.5) 1 (12.5) Thawed MII 2-cell 6 6 (100) 6 (100) MII 2-cell 2 2 (100) 2 (100) Total 2-cell a (88.9) 6 (37.5) 6 (37.5) 1 (6.3) 1 (6.3) 2 (12.5) GV-MII 3-cell 2 0 (0) GV-MII 4-cell 8 7 b (87.5) 3 (42.9) 1 (14.3) 3 (42.9) GV-MII 5-cell 2 0 (0) Total All c (76.7) 6 (26.1) 6 (26.1) 4 (17.4) 2 (8.7) 5 (21.7) a,b,c No significant difference. GV = germinal vesicle; MII = metaphase II. 508

6 Figure 3. Fluorescence in-situ hybridisation (FISH) analysis of blastomeres from human haploid androgenetic embryo (correct chromosome disjunction). Panel (I) Blastomere nucleus of 2-cell embryo ( 600); (a) first hybridization, MultiVysion PGT probe; chromosomes: 13 (red), 18 (aqua), 21 (green), Y (gold), (b) second hybridization, chromosomes: 8q (green), 15q (red), CEP 17 (aqua). Panels (II) and (III) Metaphase chromosome plates of two blastomeres from a 4-cell embryo; (a) Phase contrast image of chromosome plate ( 400), (b) first hybridization, MultiVysion PB panel probe; chromosomes: 13 (red), 16 (aqua), 18 (violet blue), 21 (green), 22 (gold) ( 600), (c) second hybridization; chromosomes: CEP 17 (aqua), CEP X (green) ( 600). 509

7 Figure 4. Karyotyping of haploid metaphase plates of human 2-cell androgenetic embryos by G-banding. (a, b) Normal haploid karyotype of two blastomere chromosome plates from an embryo derived from a Y-chromosome-carrying spermatozoon. Figu ure 5. Karyotyping of haploid metaphase plate of a human 2-cell androgenetic embryo derived from a Y-chromosome-carrying spermatozoon by G-banding. Abnormal haploid karyotype of blastomere chromosome plate with extra chromosomes 18 and

8 Discussion The two basic techniques were previously described for the production of mammalian androgenetic embryos, one of which involved IVF (conventional or by ICSI) of oocytes enucleated at metaphase II (Kono et al., 1993, 1995; Lagutina et al., 2003) and the other, removal of the female pronucleus from the zygote (Barra and Renard, 1988; Hagemann et al., 1998). In this work, TII enucleation was applied after ICSI for androgenetic human embryo production. It is known that the position of the metaphase II spindle cannot be exactly predicted by the location of PB1 in the human oocyte (Hardarson et al., 2000). Therefore, oocyte enucleation at MII stage requires the removal a larger volume of cytoplasm and/or oocyte staining with Hoechst followed by UV treatment, which could result in a detrimental effect on the subsequent development of reconstructed embryos (Zakhartchenko et al., 1997; Bordignon and Smith, 1998; Peura et al., 1998). Telophase II enucleation of activated oocytes allows removal of smaller quantities of cytoplasm and does not require verification of chromatin removal (Bordignon and Smith, 1998). In the present study, in addition to the presence of one pronucleus in the 1-cell androgenetic embryo, an extra criterion for the greater efficiency of TIIenucleation was the presence of a clearly defined female pronucleus in the oocyte karyoplast (Figure 2b). Another known technique for production of androgenotes is the removal of the female pronucleus from the zygote. In mice, female pronuclei can easily be distinguished by their proximal location to PB2 and by their smaller size in comparison with male pronuclei (Kono et al., 1995; He et al., 2003). Dimensional differences between male and female pronuclei were also used for sheep androgenetic embryo production (Hagemann et al., 1998). However, in human as well as in bovine zygotes, female and male pronuclei are not distinguishable and can easily be mixed up (He et al., 2003; Kattera and Chen, 2003; Lagutina et al., 2003). Additionally, this study showed that female and male pronuclei occupy a similar position in relation to the PB2 in some zygotes (data not shown). Because of this, this feature cannot be used for the identification of female and male pronuclei in humans. Therefore, enucleation of oocytes at TII is more appropriate for human androgenetic embryo production. As shown for mice and humans, sperm labelling with adenine-8 14 C (Edwards and Sirlin, 1956) or fluorescent mitochondrial stain (Takeuchi et al., 2005) is effective for male pronucleus identification. However, the application of sperm labelling techniques in human assisted reproduction is questionable because of possible subsequent detrimental effect. As shown in the present study, TII enucleation made it possible to obtain ooplasts, 1PN embryos and cleaving embryos at significantly higher rates than after MII enucleation. This could be explained not only by removal of smaller ooplasm fragments and the absence of Hoechst staining and UV irradiation, but also by higher TII oocyte resistance to the enucleation process. The timing of PB2 extrusion is important not only for the efficient removal of TII chromosomes, but also enables selection of oocytes that respond promptly to activation by extruding PB2 within h of stimulation. Previous reports suggested the use of cytochalasin B and D for oocyte and zygote enucleation, to relax cytoskeleton and increase plasmalemma resistance and cell survivability during micromanipulations (Wakayama and Yanagimachi, 2001; Mitalipov et al., 2002; Tesarik et al., 2003). However, this was avoided for both MII and TII enucleation, because cytochalasin treatment may lead to abnormal embryo development and cause a higher frequency of multiple female pronuclei after human oocyte activation (Wakayama and Yanagimachi, 2001; Tesarik et al., 2003). Isolated mouse pronuclei are shown to be less sensitive to cryopreservation than intact oocytes (He et al., 2003). It is believed that small oocyte karyoplasts obtained after TII enucleation could also be successfully cryopreserved. It is possible to use these karyoplasts for their transfer to the TII-enucleated donor oocytes after previous ICSI. Such reconstructed zygotes will contain nuclear DNA of the patient and her male partner, together with some mitochondrial DNA of the patient and the donor. If 2-cell haploid androgenetic embryos were obtained, one of their blastomeres would be assessed for the presence of chromosome abnormalities or gene mutations, and the other (or its karyoplast) could be cryopreserved. If the first blastomere was genetically normal, the second thawed blastomere could have been fused by electrostimulation with the blastomere of the 2-cell haploid parthenogenetic embryo, in order to obtain a diploid embryo carrying both paternal and maternal genomes. If necessary, activated oocytes and produced reconstructed diploid embryos could be assessed for genetic abnormalities before the embryo transfer. Therefore, electrofusion of the haploid androgenetic and parthenogenetic blastomeres might be comparable with the syngamy of maternal and paternal genomes during fertilization, which may be called NT fertilization (i.e. fertilization by nuclear transfer). Cytoplasm of parthenogenetic mouse eggs was shown to be as effective in accelerating development and improving developmental ability of cloned embryos derived from morulae as the zygote s cytoplasm (Ono and Kono, 2006). Transfer of the male pronucleus obtained after IVF of an enucleated oocyte into the parthenogenetically activated oocyte resulted in mouse live offspring (Kong et al., 2005). It was also shown that isolated blastomeres derived from haploid mouse parthenogenetic and androgenetic embryos at the late 2-cell stage could form a diploid embryo using electrofusion, which developed to term and produced normal offspring (Barra and Renard, 1988). NT fertilization could be also performed by electrofusion of the blastomere (or its karyoplast) isolated from 2-cell haploid androgenetic embryo with a 1-cell haploid parthenogenetic embryo at the 1PN stage. In this case, the cell cycles are synchronized and a normal diploid embryo could be obtained. Surani et al. (1986) showed that it was possible to obtain normal offspring after transfer of haploid paternal nuclei from 2-cell and 4-cell androgenetic mouse embryos back to fertilized eggs from which the male pronucleus had been removed. Experiments on the production of interspecies heterokaryons showed that the electrofusion of human zygotic blastomeres with zona-free mouse zygotes at 2PN stage resulted in synchronization of their cell cycles. This resulted in the simultaneous formation of human 511

9 512 blastomere and mouse zygote chromosome plates, despite the latter entering metaphase earlier. Metaphase plates obtained from human blastomeres were suitable for FISH analysis for the presence of translocations (Cieslak et al., 2004). Oocyte activation is a critical factor for further embryo development in parthenogenesis and NT (Lee et al., 2004; Liu et al., 2004a,b). Among the methods of artificial activation, electrical stimulation is most common for NT embryo production (Lee et al., 2004). Repeated DC pulses (two or three sets at 30 min intervals) are very effective for parthenogenote production in cattle (Collas et al., 1993), pigs (Grupen et al., 1999), monkeys (Marshall et al., 1998; Mitalipov et al., 2001) and rats (Iannaccone et al., 2002). The number of pulses and pulse intervals were chosen for two reasons (Collas et al., 1993): (i) multiple electrical pulses induce multiple Ca 2+ elevations in oocytes of several species and (ii) fertilization results in Ca 2+ transients every min in the majority of oocytes showing multiple Ca 2+ elevations. However, in the present study, three sets of DC pulses (with 30 min intervals) resulted in a higher human diploid parthenogenote rate relative to a single set; therefore, that type of electrostimulation may be inappropriate for the study purpose. The diploid status of parthenogenetic embryos obtained after three sets of DC pulses could also explain their better development. Intact haploid parthenogenotes with one PN and two PB appear to be suitable for transfer of nuclei derived from 2-cell haploid androgenetic embryos as well as parthenogenotes with two PN and one PB after one pronucleus removal. However, it should be taken into account that some embryo reconstruction techniques result in embryos and offspring that may be heteroplasmic for mitochondrial DNA (mtdna) due to mixing mtdna from karyoplasts and cytoplasts (Cohen et al., 1998; Wakayama and Yanagimachi, 2001; Heindryckx et al., 2002; Mitalipov et al., 2002; Tesarik et al., 2003; Kong et al., 2005). The prospect of mtdna heteroplasmy and its associated potential problems require further study (Cummins, 2002; Hall et al., 2006; Spikings et al., 2006). Transfer of karyoplasts with the minimal volume of cytoplasm could significantly decrease the amount of donor mtdna in the recipient cell. Previous FISH analysis has shown an increase in the frequency of mosaicism in human zygotic embryos with the advancement of their development (2- to 4-cell and 5- to 8-cell embryos, morulae and blastocysts), which may be due to post-zygotic errors (Bielanska et al., 2002). The data are in agreement with these results (Table 4), although more experiments are needed to confirm it. Therefore, at the 2-cell stage, sperm-derived haploid embryos may be more suitable for sperm genome analysis using one blastomere, with the data extrapolated to the sister blastomere, a situation analogous to that of non-direct evaluation of the oocyte genome via its PB. As shown in Table 4, FISH analysis for two to seven chromosomes revealed normal chromosome disjunction in 76.7% of the constructed human 2- to 5-cell androgenotes derived from GV-MII oocytes and MII oocytes. Munné and colleagues performed FISH analysis for five chromosomes of 2-cell embryos produced after human sperm injection into MIIenucleated mouse or cow oocytes (Munné et al., 2003). About 97% of mouse human duplicates and 83.9% of cow human duplicates were shown to have daughter cells haploid and identical for the chromosomes tested. Chromosomal abnormalities in blastomeres of human androgenetic embryos may be due to sperm meiotic errors as well as errors during the first mitotic division affected by defects or dysfunction of the sperm centrosome (Silber et al., 2003). In human fertilization, the centrosome originates from spermatozoa, and this centrosomal inheritance pattern is similar to that in other primates, cattle, sheep and pigs, whereas in mice and hamsters, the sperm centrosome is not required for fertilization and embryo development and is supported by the maternal inheritance of the zygotic centrosome (Terada et al., 2000; Nakamura et al., 2002; Liu et al., 2004). In rabbits, the zygotic centrosomes are of both paternal and maternal origin (Terada et al., 2000). The centrosome plays a critical role in the assembly of a microtubule network for pronuclear movement preceding the association of male and female genomes (Nakamura et al., 2002; Liu et al., 2004). As the sperm centrosome is involved in the first mitotic division in humans, centrosome defects or dysfunction may result in abnormal chromosome distribution between sister blastomeres of the embryo (Sathananthan, 1998; Silber et al., 2003). Oocyte maturation conditions (in vivo or in vitro) could also affect the correct chromosome disjunction during the first mitosis in human androgenotes. Confocal microscopic observations revealed that in comparison with in-vivo matured oocytes, human GV oocytes matured in vitro had a higher frequency of abnormal meiotic spindle and chromosomal alignment morphology, which might result in aneuploid embryos (Li et al., 2006). Thus, the maturity of oocyte cytoplasm also involving the reorganization of microtubule-associated proteins and pericentriolar proteins could affect chromosome segregation in the first mitotic division. It is shown that cytoplasm of human oocytes matured in vitro contains a low amount of specific proteins, and the addition of gonadotrophins to the in-vitro maturation medium results in a dramatic increase in cytoplasmic protein content (Anderiesz et al., 2000). In these studies, GV oocytes were generally used after their in-vitro maturation to MII in HTF medium with 5% Plasmanate, which is usually used for short-term culture of MII oocytes before ICSI. Culture medium supplements such as gonadotrophins, steroid hormones and growth factors could have a positive influence on nuclear and oocyte cytoplasmic maturation in vitro (Kuznyetsova et al., 2000). Thus, the possibility of effective production of cleaving human androgenetic embryos has been shown with the use of oocyte enucleation at the TII stage of meiosis. In the first mitotic division, the majority of androgenotes showed the correct chromosome disjunction. It may be postulated that further development of this technique could allow genetic testing of the human sperm genome before NT fertilization. Sperm duplication followed by NT fertilization might be useful for patients with high frequency of sperm genetic abnormalities or/and sperm centrosome defects or dysfunction.

10 Acknowledgements The authors would like to express their gratitude to Professor Anver Kuliev for editing the paper and for critical review of the results of this work. References Anderiesz C, Ferraretti AP, Magli C et al Effect of recombinant human gonadotrophins on human, bovine and murine oocyte meiosis, fertilization and embryonic development in vitro. Human Reproduction 15, Barra J, Renard JP 1988 Diploid mouse embryos constructed at the late 2-cell stage from haploid parthenotes and androgenotes can develop to term. Development 102, Bielanska M, Tan SL, Ao A 2002 Chromosomal mosaicism throughout human preimplantation development in vitro: incidence, type, and relevance to embryo outcome. Human Reproduction 17, Bordignon V, Smith L 1998 Telophase enucleation: an improved method to prepare recipient cytoplasts for use in bovine nuclear transfer. Molecular Reproduction and Development 49, Cieslak JM, Lavin C, Galat V et al Clinical outcome of preimplantation genetic diagnosis for chromosome translocations. Fertility and Sterility 82 (Suppl. 2), O-73. Cohen J, Scott R, Alikani M et al Ooplasmic transfer in mature human oocytes. Molecular Human Reproduction 4, Collas P, Fissore R, Robl J et al Electrically induced calcium elevation, activation and parthenogenetic development of bovine oocytes. Molecular Reproduction and Development 34, Cummins J 2002 The role of maternal mitochondria during oogenesis, fertilization and embryogenesis. Reproductive BioMedicine Online 4, Edwards RG, Sirlin JL 1959 Identification of C 14 -labelled male chromatin at fertilization in colchicines-treated mouse eggs. Journal of Experimental Zoology 140, Edwards RG, Sirlin JL 1956 Labelled pronuclei in mouse eggs fertilized by labelled sperm. Nature 177, 429. Egozcue S, Vendrell M, Garcia F et al Increased incidence of meiotic abnormalities in oligoasthenozoospermic males preselected for intracytoplasmic sperm injection. Journal of Assisted Reproduction and Genetics 17, Grupen CG, Verma PJ, Du ZT et al Activation of in vivo- and in vitro-derived porcine oocytes by using multiple electrical pulses. Reproduction, Fertility and Development 11, Hagemann LJ, Peterson AJ, Weilert LL et al In vitro and early in vivo development of sheep gynogenones and putative androgenones. Molecular Reproduction and Development 50, Hall VJ, Stojkovic P, Stojkovic M 2006 Using therapeutic cloning to fight human disease: a conundrum or reality? Stem Cells 24, Hardarson T, Lundin K, Hamberger L 2000 The position of the metaphase II spindle cannot be predicted by the location of the first polar body in the human oocyte. Human Reproduction 15, He Z, Liu HC, Rosenwaks Z 2003 Cryopreservation of nuclear material as a potential method of fertility preservation. Fertility and Sterility 79, Heindryckx B, Rybouchkin A, Van der Elst J, Dhont M 2002 Serial pronuclear transfer increases the developmental potential of in vitro-matured oocytes in mouse cloning. Biology of Reproduction 67, Iannaccone P, Bader M, Galat V 2002 Cloning of rats. In: Principles of Cloning. Cibelli J, Lanza R, Campbell K, West M (eds) Elsevier, Burlington, MA USA, pp Kattera S, Chen C 2003 Normal birth after microsurgical enucleation of tripronuclear human zygotes: case report. Human Reproduction 18, Kong FY, Zhang G, Zhong ZS et al Transplantation of male pronucleus derived from in vitro fertilization of enucleated oocyte into parthenogenetically activated oocyte results in live offspring in mouse. Zygote 13, Kono T, Carroll J, Swann K, Whittingham DG 1995 Nuclei from fertilized mouse embryos have calcium-releasing activity. Development 121, Kono T, Sotomaru Y, Sato Y, Nakahara T 1993 Development of androgenetic mouse embryos produced by in vitro fertilization of enucleated oocytes. Molecular Reproduction and Development 34, Kuliev A, Cieslak J, Ilkevitch Y, Verlinsky Y 2002 Chromosomal abnormalities in a series of 6733 human oocytes in preimplantation diagnosis for age-related aneuploidies. Reproductive BioMedicine Online 6, Kuznyetsov V, Verlinsky Y 2005 Development of preconception testing for paternally derived mutations by sperm duplication (Abstract). 6th International Symposium on Preimplantation Genetics 2005, London, p. 48. Kuznyetsova IB, Kuznyetsov VE, Kovtun SI 1999 Factors affecting the viability of the bovine embryos produced in vitro (Russian). Biopolymers and Cell 15, Lagutina I, Lazzari G, Duchi R, Galli C 2004 Developmental potential of bovine androgenetic and parthenogenetic embryos: a comparative study. Biology of Reproduction 70, Lee JW, Tian XC, Yang X Optimization of parthenogenetic activation protocol in porcine. Molecular Reproduction and Development 68, Li Y, Feng HL, Cao YJ et al Confocal microscopic analysis of the spindle and chromosome configuration of human oocytes matured in vitro. Fertility and Sterility 85, Liu JL, Kusakabe H, Chang CC et al. 2004a Freeze-dried sperm fertilization leads to full-term development in rabbits. Biology of Reproduction 70, Liu JL, Sung LY, Du F et al. 2004b Differential development of rabbit embryos derived from parthenogenesis and nuclear transfer. Molecular Reproduction and Development 68, Marshall VS, Wilton LJ, Moore HDM 1998 Parthenogenetic activation of marmoset (Callithrix jacchus) oocytes and the development of marmoset parthenogenones in vitro and in vivo. Biology of Reproduction 59, Mitalipov SM, Yeoman RR, Nusser KD, Wolf DP 2002 Rhesus monkey embryos produced by nuclear transfer from embryonic blastomeres or somatic cells. Biology of Reproduction 66, Mitalipov SM, Nusser KD, Wolf DP 2001 Parthenogenetic activation of rhesus monkey oocytes and reconstructed embryos. Biology of Reproduction 65, Munné S 2002 Preimplantation genetic diagnosis of numerical and structural chromosome abnormalities. Reproductive BioMedicine Online 4, Munné S, Willadsen S, Schimmel T, Cohen J 2003 Nuclear sperm duplication as a tool to study mosaicism. First International Symposium on Preimplantation Genetics, Antalya, Turkey, P.31. Nakamura S, Terada Y, Horiuchi T et al Analysis of the human sperm centrosomal function and the oocyte activation ability in the case of globozoospermia, by ICSI into bovine oocytes. Human Reproduction 17, Ono Y, Kono T 2006 Irreversible barrier to the reprogramming of donor cells in cloning with mouse embryos and embryonic stem cells. Biology of Reproduction 75, Peura TT, Lewis IM, Trounson AO 1998 The effect of recipient oocyte volume on nuclear transfer in cattle. Molecular Reproduction and Development 50, Sathananthan AH 1998 Paternal centrosomal dynamics in early human development and infertility. Journal of Assisted Reproduction and Genetics 15, Silber S, Escudero T, Lenahan K et al Chromosomal abnormalities in embryos derived from testicular sperm extraction. Fertility and Sterility 79, Spikings EC, Alderson J, John JCS 2006 Transmission of mitochondrial DNA following assisted reproduction and nuclear 513

11 transfer. Human Reproduction Update 12, Surani MAH, Barton SC, Norris ML 1986 Nuclear transplantation in the mouse: in heritable differences between paternal genomes after activation of the embryonic genome. Cell 45, Takeuchi T, Neri QV, Palermo GD 2005 Construction and fertilization of reconstituted human oocytes. Reproductive BioMedicine Online 11, Templado C, Bosch M, Benet J Frequency and distribution of chromosome abnormalities in human spermatozoa. Cytogenetic and Genome Research 111, Terada Y, Simerly CR, Hewitson L, Schatten G 2000 Sperm aster formation and pronuclear decondensation during rabbit fertilization and development of a functional assay for human sperm. Biology of Reproduction 62, Tesarik J, Martinez F, Rienzi L et al Microfilament disruption is required for enucleation and nuclear transfer in germinal vesicle but not metaphase II human oocytes. Fertility and Sterility 79, Van den Bergh M, Bertrand E, Englert Y 1995 Second polar body extrusion is highly predictive for oocyte fertilization as soon as 3 hr after intracytoplasmic sperm injection (ICSI). Journal of Assisted Reproduction and Genetics 12, Verlinsky Y, Kuliev AM 2005 Atlas of Preimplantation Genetic Diagnosis. Taylor and Francis, London and New York. Verlinsky Y, Lerner S, Illkevitch N et al Is there any predictive value of first polar body morphology for embryo genotype or developmental potential? Reproductive BioMedicine Online 7, Wakayama T, Yanagimachi R 2001 Effect of cytokinesis inhibitors, DMSO and the timing of oocyte activation on mouse cloning using cumulus cell nuclei. Reproduction 122, Willadsen S, Munné S, Schimmel T, Cohen J 2003 Application of nuclear sperm duplication. First International Symposium on Preimplantation Genetics, Antalya, Turkey, O-47. Zakhartchenko V, Stojkovic M, Brem G, Wolf E 1997 Karyoplast cytoplast volume ratio in bovine nuclear transfer embryos: effect on developmental potential. Molecular Reproduction and Development 48, Received 11 December 2006; refereed 3 January 2007; accepted 24 January