Mitochondrial DNA content affects the fertilizability of human oocytes

Transcription

1 Molecular Human Reproduction Vol.7, No.5 pp , 2001 Mitochondrial DNA content affects the fertilizability of human oocytes P.Reynier 1,4, P.May-Panloup 2, M-F.Chrétien 2, C.J.Morgan 1, M.Jean 3, F.Savagner 1, P.Barrière 3 and Y.Malthièry 1 INSERM EMI-U 00-18, 1 Laboratoire de Biochimie et Biologie Moléculaire, CHU d Angers, F Angers, 2 Laboratoire d Histologie-Embryologie-Cytologie et Laboratoire de FIV, CHU d Angers, F Angers and 3 Biologie de la Reproduction, Pavillon de la Mère et de l Enfant, CHU de Nantes, B.P. 1005, F Nantes cedex 1, France 4 To whom correspondence should be addressed. pareynier@chu-angers.fr Mitochondrial DNA content varies considerably in oocytes, even when collected from the same patient. In the present study, real-time quantitative polymerase chain reaction analysis of 113 unfertilized oocytes obtained from 43 patients revealed an average of (range: to ) mitochondrial genomes per cell. We compared several groups of oocytes to investigate the relationship between mitochondrial DNA content and fertilizability. The average mitochondrial DNA copy number was significantly lower in cohorts suffering from fertilization failure compared to cohorts with a normal rate of fertilization. In addition, the mitochondrial copy number of oocytes from patients with fertilization failure due to unknown causes was significantly lower than that of oocytes from patients in which IVF failure was due mainly to a severe sperm defect. The lower mtdna copy number could be due to defective cytoplasmic maturation of oocytes. We conclude that low mitochondrial DNA content, due to inadequate mitochondrial biogenesis or cytoplasmic maturation, may adversely affect oocyte fertilizability. Key words: cytoplasmic maturation/mitochondrial biogenesis/mtdna/oocyte/real-time PCR Introduction Most cases of male infertility can be treated by intracytoplasmic sperm injection (ICSI), but successful IVF is still limited by factors such as oocyte quality. Whereas the nuclear maturation of the oocyte is easily identified by the appearance of the first polar body, the cytoplasmic maturation, which is also essential for successful fertilization and early embryo development, remains difficult to evaluate. Defects in oocyte maturation, suspected of being involved in some cases of IVF failure, are sometimes difficult to overcome by ICSI (Miller et al., 1995; Thomas et al., 1998). A spectacular growth of the oocyte takes place during follicular maturation (Wolf et al., 1995) and many morphological, metabolic and gene expression changes seem to be essential to oocyte maturation (Wassarman and Albertini, 1994; Bell et al., 1997; Gosden et al., 1997; Ji et al., 1997). Large variations in oocyte ATP have been reported and the level of ATP production is believed to be associated with the developmental competence of the embryo (Van Blerkom et al., 1995, 1998). The growth of the oocyte also involves mitochon- drial biogenesis and mitochondrial DNA (mtdna) replication (Piko and Taylor, 1987; Cummins, 1998). Mitochondria are maternally inherited organelles that use oxidative phosphorylation to supply ATP to the cell. Each mitochondrion carries its own multicopy genome, which is a circular, double-stranded DNA molecule with 16.6 kilobases (Anderson et al., 1981). mtdna codes for 13 essential subunits of the respiratory chain complexes that provide the main ATP supply of the cell (Wallace, 1992). Human mtdna is derived exclusively from maternally inherited DNA (Giles et al., 1980). Paternal mitochondria are specifically eliminated after fertilization and before the 4-cell stage (Kaneda et al., 1995), probably by a mechanism involving ubiquitination of sperm mitochondria (Sutovski et al., 1999). It has been suggested that defects in sperm mitochondria may lead to male infertility (Lestienne et al., 1997; Kao et al., 1998; Ruiz-Pesini et al., 1998). However, the significance of the defective mtdna observed in some oocytes remains unclear (Chen et al., 1995; Brenner et al., 1998; Cummins et al., 1998; Reynier et al., 1998). Mammalian somatic cells contain several thousands of mitochondria, each of them containing 1 10 copies of mtdna. It is currently thought that oocytes contain to mitochondrial genomes (Cummins, 1998). The first estimations of mouse oocyte mitochondrial content were made in the 1970s. Electron microscopy morphometric analysis has found an average of mitochondria per mouse pronucleate oocyte (Piko and Matsumoto, 1976). Subsequently, dot blot techniques revealed an average of mtdna copies per European Society of Human Reproduction and Embryology 425

2 P.Reynier et al. cell in mature pooled mouse oocyte (Piko and Taylor, 1987). Preparation of DNA Hybridization techniques on pooled oocytes determined that DNA was extracted from each oocyte by means of the High bovine oocytes contained an average of mtdna copies Pure PCR Template Preparation Kit (Roche Diagnostics, Mannheim, per cell (Michaels et al., 1982). More recently, competitive Germany) according to the manufacturer s recommendations. DNA polymerase chain reaction (PCR) (Chen et al., 1995) and realchaotropic bound specifically to glass fibres following the combined action of a agent (guanidine), a detergent (Triton X-100) and the time PCR (Steuerwald et al., 2000) have quantified the mtdna enzyme proteinase K. After washing, the silica-bound DNA was copy number in unfertilized human oocytes obtained after IVF eluted with 400 µl of pre-warmed (72 C) elution buffer and preserved failure. The former found an average of mtdna at 4 C. copies in nine oocytes, and the latter found an average of mtdna copies in 18 oocytes, showing a surprising Preparation of external standards variation in oocyte mtdna content. The external standard was a 158 bp PCR product. The nucleotide Real-time quantitative PCR, a highly sensitive method, positions of the primers on light strand mtdna were: D41 (3254 is the most suitable technique currently available for the 3277) and D56 ( ). PCR reactions were carried out under quantification of DNA copy numbers in single cells. Unlike standard conditions with 100 ng of total DNA in a 50 µl volume: the studies based on hybridization techniques, which require 1.5 mmol/l MgCl 2, 75 mmol/l Tris HCl(pH9at25 C), 20 mmol/l pooled samples of oocytes, the sensitivity of real-time PCR (NH 4 ) 2 SO 4, 0.01% Tween 20, 50 pmol of each primer, 200 µmol/l allows investigation of inter-oocyte variability of mtdna of each dntp and 2 units of GoldStar DNA polymerase (Eurogentec, Seraing, Belgium). Each of the 30 cycles consisted of: a denaturation content. We have therefore used real-time PCR to quantify the step of 30 s at 94 C, a hybridization step of 30 s at 58 C, and an mtdna content of 113 unfertilized oocytes obtained from 43 extension step of 1 min at 72 C. The PCR products were phenolpatients undergoing IVF. To test the hypothesis of a link chloroform purified from low melting point agarose, precipitated between the mtdna content of oocytes and cytoplasmic in ethanol, diluted in water and quantified by spectrophotometry. maturation, and thus the fertilizability, the oocytes were classi- Purification quality was checked by means of the absorbance 260/ fied into three groups according to the known fertilization 280 ratios, with values between 1.8 and 2.0 being considered characteristics of the patients, i.e. fertilization failure due to a acceptable. It was assumed that 1 ng of a 158 bp product contained severe sperm defect, idiopathic failure or normal fertilization molecules of double-stranded DNA. Serial dilutions were then carefully made in order to assess several concentrations of a Materials and methods known number of templates. This was used as the external standard for real-time PCR. The serial dilutions were all conserved at 20 C Oocyte samples in single-use aliquots. The ethics committee of the University Hospital of Angers approved the plan of the study we intended to perform on oocytes discarded Real-time PCR quantification during IVF procedures. Follicular growth was stimulated by recombin- A Roche LightCycler was used to determine the mtdna copy ant FSH associated with a gonadotrophin-releasing hormone agonist. number. The LightCycler is a combined microliter volume thermal- Ovulation was induced with human chorionic gonadotrophin and cycler and fluorometer suitable for real-time fluorescent PCR. oocytes were collected by a transvaginal probe. For the purpose of Twenty µl PCR reactions were set up with final concentrations: IVF, oocytes were considered to be unfertilized if on the fourth day 1 buffer containing 4 mmol/l MgCl 2, 0.2 mmol/l dntps, following insemination (72 h of culture), there were no pronuclei or 0.5 µmol/l of both primers (D41 and R56), SYBR green I dye, 0.25 U second polar bodies. Oocytes were separated from follicular cells by Taq DNA polymerase, 0.2 µg of TaqStart antibody (Clontech, Palo washing and gentle pipetting in IVF-20 medium (IVF Scandinavian, Alto, CA, USA) and 10 µl of the extracted DNA or 10 µl of Standard Stockholm, Sweden). A total of 113 isolated metaphase II oocytes with a known copy number. The reactions were performed in the (identified by the presence of the first polar body) were individually LightCycler as follows: initial denaturing at 95 C for 2 min and 45 collected from 43 women over a period of 1 year. Each oocyte was cycles at 95 C for 0 s (temperature transition rate of 20 C/s), 58 C placed in 50 µl of IVF-20 medium. The oocytes were preserved at for 5 s (temperature transition rate of 20 C/s), and 72 C for 13 s 20 C until DNA extraction, which was always performed within a (temperature transition rate of 2 C/s). The SYBR green fluorescence month after collection. The average age of the patients was 33 years was read at the end of each extension step (72 C). A melting (range 25 42). The oocyte population collected from a given patient curve was systematically analysed in order to check the absence of defined a cohort. The average number of oocytes per cohort was 2.6 mispriming and the quality of amplifications. This consisted of SYBR (range 1 8). green fluorescence analysis of the temperature transition between 66 The oocytes were classified into three groups. Group 1 consisted and 94 C (temperature transition rate of 0.2 C/s), with continuous of 21 oocytes collected from five patients (average age 32 years, fluorescence readings. The rapid loss of fluorescence at a given range 27-38) where the fertilization failure was evidently due to a temperature indicates the melting temperature of the PCR product. severe sperm defect (insemination with 10 5 motile spermatozoa). For each PCR run, a standard curve was generated using five 10-fold We assumed that this group was relatively homogeneous, consisting serial-dilutions ( copies) of the target mtdna PCR mainly of mature fertilizable oocytes free from defects due to faulty product with the same primers as those used for oocyte mtdna cytoplasmic maturation. Group 2 consisted of 47 oocytes collected amplification. The LightCycler software (version 3.01) generated a from 12 patients (average age 32 years, range 27-42) with idiopathic standard curve, which then allowed the determination of the starting fertilization failure (fertilization rate 20% with normal spermatozoa). copy number of mtdna in each sample. Quantitative PCR was In this group, 10 patients had complete fertilization failure. Group 3 performed on 1/40 of the isolated oocyte total DNA, i.e. on a 10 µl consisted of 45 unfertilized oocytes collected from 26 patients sample taken from 400 µl of eluted DNA. All oocytes were tested (average age 34 years, range years) with normal fertilization twice. The raw data were then increased 40-fold to calculate the total (fertilization rate 20% with normal spermatozoa). mtdna content in each oocyte. 426

3 mtdna content affects oocyte fertilizability Table I. Intra-assay precision method (Figure 1). The reproducibility between runs (interassay) was assessed from five measurements of three different Oocyte oocyte DNA samples. The CV ranged from 9.3 to 12.7% (Table II). These results indicate that the quantitative PCR has good intra-assay precision and reproducibility. Ten experiments Oocyte mtdna quantification mtdna copies The average mtdna copy number of the 113 unfertilized oocytes was found to be ( ) with a high inter oocyte variation ( to ), even within individual cohorts. Since there was a great variation between different oocytes from a given cohort, we were unable to establish any correlation between the average mtdna content and the age Mean of the patients (r 0.10). SD In order to investigate the influence of mtdna content on CV (%) oocyte fertilizability, we compared 21 unfertilized oocytes CV coefficient of variation. from group 1 (fertilization failure mainly due to a serious sperm defect) with 47 unfertilized oocytes from group 2 (fertilization failure in spite of normal spermatozoa). The Extraction efficiency average mtdna copy number of group 1, ( ), Extraction efficiency was evaluated by the extraction of several was significantly different (P ) from that of group 2, samples, containing known copy numbers of the standard, which ( ) (Figure 2). were then diluted in 50 µl of IVF 20 medium under the same We also compared the mtdna content of unfertilized conditions as those used for oocyte extraction. PCR quantification oocytes with regard to the fertilization rate of their cohort of was then performed under the same conditions as those used for DNA from oocytes. origin. The average mtdna copy number of group 2 (cohorts with fertilization failure due to unknown causes) was Statistical analysis ( ) compared with an average of ( ) The three groups of oocytes were compared using the Mann-Whitney in oocytes from group 3 (cohorts with a normal rate of U-test. Differences were considered significant when P fertilization) (Figure 3). The difference between the two groups Statistical analyses were performed with SYSTAT software, version was significant (P 0.02). Interestingly, six out of seven 8.0 (SPSS Inc., Chicago, IL, USA). oocytes containing less than mtdna molecules were found in group 2. However, this result was not significant. There was no correlation between the mtdna content and the Results rate of fertilization in group 2 (r 0.03) or in group 3 (r Extraction efficiency 0.09). The comparison of group 1 with group 3 revealed no significant difference (P 0.07). Since accurate single-cell quantification is highly dependent upon high recoveries of the mtdna templates, we performed DNA extraction with six samples containing 10 4 or 10 5 copies Discussion of the standard in the IVF medium, under the same conditions Since French law does not allow the use of fertilized human as those used for oocyte DNA extraction. We found an average oocytes for the evaluation of mtdna content, we quantified the of copies for the 10 4 extracted molecules and an average number of mtdna molecules in isolated unfertilized average of copies for 10 5 extracted molecules. oocytes. The efficiency of DNA extraction in our study allows Thus, the extraction efficiency ranged from 93 to 108%, the assumption that the average mtdna content in human demonstrating the accuracy of this technique. The DNA binding unfertilized oocytes after IVF is ~ If we exclude the capacity of the glass fibres was much greater than the small presence of the first polar body, the amount of mtdna is amount of single-cell DNA that was extracted. We may considerable and represents ~50% of the total DNA of the therefore suppose that nearly all the DNA molecules present oocyte ( mtdna represent 3.4 pg of DNA, whereas were bound to the chromatography column prior to the washing the nucleus contains 3 pg of DNA per haploid genome). A and elution steps. similar approach (Chen et al., 1995; Steuerwald et al., 2000) found an average mtdna copy number of (n 9 Reproducibility oocytes) and (n 18 oocytes) respectively. The Intra-assay precision was assessed from 10 measurements of difference in the mean mtdna content of oocytes between three different oocyte DNA samples with varying levels of these studies and ours could be explained by the difference in mtdna content. The coefficients of variation (CV) ranged sample size and the large inter-individual variation. It should from 3.9 to 9.1% (Table I). When a DNA sample was amplified be noted that in our study there is a potential bias in the 10 times in the same experiment, the 10 kinetic curves were absolute quantification of mtdna content due to the age of all grouped together, demonstrating good repeatability of the the oocytes analysed but the extended period of culture was 427

4 P.Reynier et al. Figure 1. Intra-assay precision of the quantitative polymerase chain reaction. Samples 1 4: standards ( copies); sample 5: negative control; samples 6 15: mtdna all from the one oocyte, sample 2. Table II. Reproducibility between runs (inter-assay) Oocyte Five experiments mtdna copies Mean SD CV (%) CV coefficient of variation. Figure 3. Comparison of mtdna content of oocytes from group 2 (idiopathic fertilization failure) and group 3 (normal fertilization) (P 0.02). unavoidable to ensure that only unfertilized oocytes were used. However, our results confirm the great variation of mtdna content in oocytes reported by Chen et al. (1995) and Steurwald et al. (2000). Moreover, the mtdna content also varies widely between oocytes belonging to a given cohort. This large variation in mitochondrial biogenesis may reflect 428 Figure 2. Comparison of mtdna content in oocytes from group 1 (fertilization failure due to sperm defect) and group 2 (idiopathic fertilization failure) (P ). Results are presented in box- and-whisker form: the box comprises 50% of the values, the line in the box gives the median value, and the whiskers represent the distribution of values (mean 2SD). *Outliers. the variation in the level of cytoplasmic maturation of the oocytes. A large part of the increase in mitochondrial content occurs during the final stages of the maturation of the oocyte (Jansen and de Boer, 1998; Poulton et al., 1998). Indeed, it has found that immature mouse oocytes have a lower mtdna copy number than mature oocytes (Piko and Taylor, 1987). Moreover, ovarian stimulation in IVF typically leads to cohorts of oocytes at various stages of maturation, with about a quarter of them being immature (Flood et al., 1990). Our results suggest that mitochondrial DNA content could play a role in oocyte fertilizability. A significant reduction in mtdna content was observed in oocytes from group 2 in comparison with oocytes from group 1 (P ). Both groups of oocytes came from cohorts with fertilization failure (fertilization rate 20%). However, group 1 could be assumed to be composed of mature oocytes since fertilization failure was due mainly to severe sperm defects. In contrast, cytoplasmic immaturity could be suspected in oocytes from group 2 since they failed to fertilize despite insemination with normal spermatozoa with apparent nuclear maturity. This suggests that a cytoplasmic maturation defect, in the cohort inseminated with normal sperm, may be related to mitochondrial deficiency. A second analysis also indicates a relationship between

5 mtdna content affects oocyte fertilizability mitochondrial defects and fertilization failure. The cohorts Flood, J.T., Chillic, C.F., VanUem, J.F.H.M. et al. (1990) Ooplasmic transfusion: prophase germinal vesicle oocytes made development suffering from fertilization failure (despite normal spermato- competent by microinjection of metaphase II egg cytoplasm. Fertil. Steril., zoa) had a significantly lower mtdna copy number 63, (P 0.02) compared to cohorts with a normal rate of Giles, R.E., Blanc, H., Cann, H.M. et al. (1980) Maternal inheritance of human mitochondrial DNA. Proc. Natl. Acad. Sci. USA, 77, fertilization (group 3). Moreover, six out of the seven oocytes Gosden, R., Krapez, J. and Briggs, D. (1997) Growth and development of the with a lower mtdna content ( ) were found in cohorts mammalian oocyte. BioEssays, 19, with fertilization failure. This result demonstrates a significant Jansen, R.P.S. and de Boer, K. (1998) The bottleneck: mitochondrial difference in the cytoplasmic maturation level of cohorts, imperatives in oogenesis and ovarian follicular fate. Mol. Cell. Endocrinol., 145, indicating a global mitochondrial immaturity in cohorts with Ji, Y.Z., Bomsel, M., Jouannet, P. et al. (1997) Modifications of the human fertilization failure. oocyte plasma membrane protein pattern during preovulatory maturation. It is worth noting that the average mtdna content of group Mol. Reprod. Dev., 47, (unfertilized oocytes from cohorts with a normal rate of Kaneda, H., Hayashi, J.I., Takahama, S. et al. (1995) Elimination of paternal mitochondrial DNA in intraspecific crosses during early mouse fertilization) appeared to be lower than that of group 1 (mature embryogenesis. Proc. Natl. Acad. Sci. USA, 92, oocytes, unfertilized due to defective spermatozoa), but this Kao, S-H., Chao, H-T. and Wei, Y-H. (1998) Multiple deletions of difference was not statistically significant (P 0.07). The mitochondrial DNA are associated with the decline of motility and fertility of human spermatozoa. Mol. Hum. Reprod., 4, apparently lower mtdna content in group 3 might be due to the Lestienne, P., Reynier, P., Chrétien, M.F. et al. (1997) Oligoasthenospermia absence of mature oocytes following successful fertilization. It associated with multiple mitochondrial DNA rearrangements. Mol. Hum. would therefore be interesting to further explore the difference Reprod., 3, in mtdna content between the two groups with larger numbers Michaels, G.S., Hauswirth, W.W. and Laipis, P.J. (1982) Mitochondrial DNA copy number in bovine oocytes and somatic cells. Dev. Biol., 94, of oocytes. Miller, K.F., Falcone, T., Goldberg, J.M. et al. (1995) Previous fertilization In conclusion, some IVF failures could be due to inadequate failure with conventional in vitro fertilization is associated with poor oocyte maturation related to defective mitochondrial biogen- outcome of intracytoplasmic sperm injection. Fertil. Steril., 69, Piko, L. and Matsumoto, L. (1976) Number of mitochondria and some esis. This hypothesis needs further confirmation in an animal properties of mitochondrial DNA in the mouse egg. Dev. Biol., 49, model where the utilization of fertilized oocytes is possible. Piko, L. and Taylor, K.D. (1987) Amounts of mitochondrial DNA and The measurement of mtdna content in unfertilized oocytes, abundance of some mitochondrial gene transcripts in early mouse embryos. however, could be an interesting tool for the investigation Dev. Biol., 123, Poulton, J., Macaulay, V. and Marchington, D.R. (1998) Mitochondrial genetics of the molecular mechanisms involved in the cytoplasmic 98. Is the bottleneck cracked? Am. J. Hum. Genet., 62, maturation of oocytes. At the clinical level, a low mtdna Reynier, P., Chrétien, M.F., Savagner, F. et al. (1998) Long PCR analysis of copy number could account for some IVF and ICSI failures. human gamete mtdna suggests defective mitochondrial maintenance in spermatozoa and supports the bottleneck theory for oocytes. Biochem. The quantification of mtdna in the corresponding unfertilized Biophys. Res. Commun., 252, oocytes may help in identifying faulty cytoplasmic maturation. Ruiz-Pesini, E., Diez, C., Lapena, A.C. et al. (1998) Correlation of sperm motility with mitochondrial enzymatic activities. Clin. Chem., 44, 1616 Acknowledgements Steuerwald, N., Barritt, J.A., Adler, R. et al. (2000) Quantification of mtdna We are grateful to Prof. D.Chappard for his help with the statistical in single oocytes, polar bodies and subcellular components by real-time analysis and Dr K.Malkani for his critical reading of the manuscript. rapid cycle fluorescence monitored PCR. Zygote, 8, This work was supported by a grant from the Délégation Régionale Sutovski, P., Moreno, R.D., Ramalho-Santos, J. et al. (1999) Ubiquitin tag for sperm mitochondria. Nature, 402, de la Recherche Clinique of the University Hospital of Angers (PHRC Thomas, C., Orava, M., Tuomivaara, L. et al. (1998) Low pregnancy rate is PL 98-01). achieved in patients treated with intracytoplasmic sperm injection due to previous low or failed fertilization in in-vitro fertilization. Hum. Reprod., 13, References Van Blerkom, J., Davis, P. and Lee, J. (1995) ATP content of human oocytes Anderson, S., Bankier, A.T., Barrell, B.G. et al. (1981) Sequence and and developmental potential and outcome after in vitro fertilization and organization of the human mitochondrial genome. Nature, 290, embryo transfer. Hum. Reprod., 10, Bell, J.C., Smith, L.C., Rumpf, R. et al. (1997) Effect of enucleation on Van Blerkom, J., Sinclar, J. and Davis, P. (1998) Mitochondrial transfer protein synthesis during maturation of bovine oocytes in vitro. Reprod. between oocytes: potential applications of mitochondrial donation and the Fertil. Dev., 9, issue of heteroplasmy. Hum. Reprod., 13, Brenner, C.A., Wolny, Y.M., Barrit, J.A. et al. (1998) Mitochondrial DNA Wallace, D.C. (1992) Diseases of the mitochondrial DNA. Annu. Rev. Biochem., deletion in human oocytes and embryos. Mol. Hum. Reprod., 4, , Chen, X., Prosser, R., Simonetti, S. et al. (1995) Rearranged mitochondrial Wassarman, P.M. and Albertini, D.F. (1994) The mammalian ovum. In Knobil, genomes are present in human oocytes. Am. J. Hum. Genet., 57, E. and Neill, J.D. (eds), Physiology of Reproduction, 2nd edn. Raven Press, Cummins, J. (1998) Mitochondrial DNA in mammalian reproduction. New York, 79 pp. J. Reprod. Fertil., 3, Wolf, J.P., Bulwa, S., Rodrigues, D. et al. (1995) Human oocyte cytometry Cummins, J.M., Jequier, A.M., Martin, R. et al. (1998) Semen levels of and fertilization rate after subzonal insemination. Zygote, 3, mitochondrial DNA deletions in men attending an infertility clinic do not correlate with phenotype. Int. J. Androl., 21, Received on November 8, 2000; accepted on March 9,