Long PCR Analysis of Human Gamete mtDNA Suggests Defective Mitochondrial Maintenance in Spermatozoa and Supports the Bottleneck Theory for Oocytes

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.

252, 373–377 (1998)

RC989651

Long PCR Analysis of Human Gamete mtDNA Suggests Defective Mitochondrial Maintenance in Spermatozoa and Supports the Bottleneck Theory for Oocytes Pascal Reynier,* Marie-Franc¸oise Chre´tien,† Fre´de´rique Savagner,* Ge´rald Larcher,* Vincent Rohmer,‡ Paul Barrie`re,§ and Yves Malthie`ry* *Laboratoire de Biochimie et Biologie Mole´culaire A, †Laboratoire d’Histologie-Embryologie et Cytologie, and ‡Service de Me´decine C-Endocrinologie-Nutrition-Me´decine Interne, CHU d’Angers, 49033 Angers Cedex 01, France; and §Biologie de la Reproduction, Pavillon de la me`re et de l’enfant, CHU de Nantes, B.P. 1005, 44093 Nantes Cedex 1, France

Received October 8, 1998

The long PCR and the Southern blot techniques were used to study mitochondrial DNA (mtDNA) in 94 sperm samples, and in 35 oocytes collected from 12 women. The sperm samples were classified in two sets: 37 samples from normal subjects, and 57 samples from patients with oligoasthenospermia. In both sets, most of the spermatozoan mitochondria had multiple mtDNA deletions. The rate of mtDNA mutation, which appears unexpectedly high, considering the short life span of the spermatozoa, may be due to impaired maintenance during differentiation. In contrast, despite the long life span of oocytes and the extended meiotic period, oocyte mitochondria showed few mtDNA rearrangements. However, mitochondria in oocytes from a given donor revealed considerable mutational heterogeneity. This supports the bottleneck theory of rapid segregation of mtDNA genotypes during early oogenesis. The long PCR technique, which allows analysis of the entire mitochondrial genome, provides new information on mtDNA instability in human gametes. Our findings suggest that mtDNA maintenance differs in the two types of gametes. © 1998 Academic Press

MtDNA transmission raises two important questions. The first concerns the mechanism of the specific elimination of paternal mtDNA within a few days of oocyte fertilization (1,2), even though the mid-piece containing mitochondria enters the egg during fertilization (3). This specific elimination may explain the uniquely maternal transmission of mtDNA (4,5). The second question is why individuals, who normally accumulate mtDNA mutations all through their life, transmit only a few, if any, abnormal genomes to their descendants. Indeed, a selection of mitochondrial genotypes occurs over a single generation in humans

(6,7). The bottleneck theory explains this rapid segregation by the selective amplification of a small number of mtDNA molecules during the early stages of oogenesis (8). Oocyte mtDNA thus seems to originate from a very small pool of precursor mtDNA, perhaps as few as 200 copies (7). These molecules are greatly amplified (up to about 100,000 copies) in the mature oocyte. The first ovum divisions consist only in the fragmentation of this oocyte pool of mitochondrial genomes, without any mitochondrial biogenesis. This dramatic reduction in mtDNA numbers during early oogenesis, associated with paternal mtDNA elimination, may be responsible for the rapid shift toward homoplasmy (presence of one genotype of mtDNA) in contrast to heteroplasmy where different genotypes coexist in the same cell (7). Neither the molecular basis nor the precise stage at which this restriction/amplification process occurs is known. Several authors have recently reported the detection of the common 4977 base-pair deletion of mtDNA in spermatic cells (9), in oocytes (10, 11) and in embryos (11). This deletion, which removes about one third of the mitochondrial genome, accumulates in the cells all through life, mainly in post-mitotic tissues or in weakly regenerating tissues. This deletion has been extensively studied by the quantitative PCR technique, which detects the deleted genome alone, in several tissues (12-14). The common deletion becomes detectable at about 30 years of age and reaches an average rate of about 0.1% in skeletal muscle at the age of about 80. Since this low rate is not sufficient to explain the respiratory chain decline characteristic of aging, it has been postulated that many other undetected mutations might be responsible for the mitochondrial impairment that occurs during aging (15). The long PCR (16,17) is a good tool for analyzing multiple mtDNA rearrangements since it is far more exhaustive than the 4,977 base-pair deletion analysis

373

0006-291X/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

Vol. 252, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

alone (18-20). This technique allows amplification of the entire mitochondrial genome (16,569 base pairs) and shows that the common 4,977 base-pair deletion does not account for the multiple mtDNA rearrangements observed (19). The long PCR detects the common deletion as well as all the other main deletions. The qualitative aspect of the long PCR gives a good idea of the multiple rearrangements occurring in the mitochondrial genome. The analysis of multiple mutations should provide a better explanation of mtDNA impairment since such mutations probably act in synergy (15,21). In order to investigate the possible pre-existence of mtDNA deletions in gametes, we performed the long PCR on human spermatozoa and oocytes. This approach provides new insights into gamete mtDNA maintenance, segregation and transmission. MATERIAL AND METHODS Sperm and oocyte samples. The Ethics Committee of the University Hospital of Angers approved the plan of our study. After a series of procedures in Technically Assisted Reproduction, the 94 sperm samples left over were classified, after biological examination, in two groups. Group I contained 37 normal sperm samples and group II contained 57 oligoasthenospermic sperm samples, as defined by the WHO recommendations (22). In each case, 500 ml samples of sperm were conserved at 220° until DNA extraction. The age of patients ranged from 25 to 53 years (average 34 years). During the in vitro fertilization and embryo transfer procedure, we examined the oocytes after insemination with normal sperm, cultured in a humidified atmosphere of 5% CO2. After four days, when there were neither pro-nuclei, nor second polar bodies, nor any spermatozoa on the zona pellucida, they were considered to be unfertilized and were separated from the follicular cells in suspension by washing in a BM1 medium (Ellios Bio-media, France). Thirty-five single-cell oocytes from 12 women were collected, one by one, in 50 ml of PBS (Eurobio, France) by pipeting, and checked for the absence of any spermatozoa fixed to zona pellucida. The oocytes were then conserved at 220° until DNA extraction. The age of the patients ranged from 26 to 37 years (average 28.6 years). DNA preparation. DNA was extracted by means of the High Pure PCR Template Preparation Kit (Boehringer, Mannheim) according to the recommendations of the manufacturer. Spermatozoa (106) were centrifuged (5 min at 4000 rpm) and pellets were collected in 200 ml of PBS for DNA extraction. The 50 ml buffer containing oocytes was directly submitted to extraction. All the DNA samples were conserved in 200 ml of water at 14°, in order to prevent nicking before the long PCR. Long PCR. The long PCR was carried out with 2 ml of spermatozoan or oocyte mtDNA in a 50 ml final volume: 50 mM Tris-HCl (pH 9.2 to 25°C), 14 mM (NH4)2SO4, 2.25 mM MgCl2, 100 pmole of each primer, 200 mM dNTP, 2.5 units of DNA polymerase mixture (Taq DNA and pwo polymerases) according to the manufacturer’s recommendations (Expand Long Template PCR System, Boehringer, Mannheim). The two primers were D24: 59GGC ACC CCT CTG ACA TCC (nucleotide positions on the light strand: 4,830-4,848) and R32: 59TAG GTT TGA GGG GGA ATG CT (4,261-4,242) located respectively in on the ND2 and ND1 genes in a region rarely affected by deletions. These primers allow the amplification of a 15,981 bp PCR product. The reaction was performed as follows: 1.5 minutes at 94°C, 15 cycles of 30 seconds at 94°C, 30 seconds at 56°C and 10 minutes at 72°C followed by the same 15 cycles with an increase in synthesis time of 15 seconds per cycle. Primers were added last according to

the hot-start technique. The thermocycler used was a minicycler manufactured by MJ Research (USA). The reactions were analyzed on 0.8% agarose gels in TBE buffer (pH 8.3). Five ml of the reaction products were subjected to electrophoresis (30 minutes at 100V) and revealed by ethidium bromide. Southern blot. One ml of each of the long PCR products (about 50 ng of DNA) was subjected to electrophoresis in 0.8% agarose gel at 60V for 12 hours. DNA was transferred onto a positively charged nylon membrane (Boehringer, Mannheim) and fixed by a 30-second exposure to short UV (Stratalinker, Stratagene, USA). The mtDNA probes labeled by digoxigenin were obtained by multi-random priming of a mtDNA matrix (1 mg) prepared by the long PCR (Random primed DNA labeling kit, Boehringer, Mannheim). After prehybridization during two hours at 42° in a buffer containing: 5X SSC, 50% formamide, 0.1% sodium-lauroylsarcosine, 0.02% SDS and 2% of blocking agent, the nylon membrane was hybridized overnight at 42° with 100 ng of the probe. The first washing was carried out at room temperature with a 2X SSC and 0.1% SDS solution. The second washing was carried out at 68°, with a SSC 0.1X and SDS 0.1% concentrations. The Southern blot was then revealed with antidigoxigenin antibodies labeled by alkaline phosphatase (Dig DNA labeling and detection kit, Boehringer, Mannheim).

RESULTS Ninety-four spermatozoan samples and 35 singlecell oocytes were analyzed by the long PCR technique followed by the hybridization step. The long PCR allowed the amplification of the entire mitochondrial genome. The PCR products were hybridized in order to increase the specificity as well as the sensitivity of the detection. The 16-kb PCR products were systematically visualized before hybridization, but no deletions were seen at this stage. After the sensitivity of the detection was improved by the hybridization step, the spermatozoan mtDNA showed multiple deletions. These rearrangements occurred in 31/37 (84%) of group I (normal donors) and 49/57 (86%) of group II (patients with oligoasthenospermia). Figure 1 shows examples of the group I PCR products. A few spermatozoan samples showed no deletion (lanes 14 and 18). The specificity of the detection of deletions has been evaluated by means of different technical approaches (20). Some spermatozan samples displayed only one deletion (lanes 7, 9 and 11). All the other spermatozoan samples revealed two to seven deletions. The multiple mtDNA deletions are similar to those observed in the skeletal muscle, myocardium and adipocytes of aged subjects (19,20). With the same technique, cells with a short life span, such as lymphocytes, polynuclear leukocytes or cells in culture, displayed no mtDNA deletions. Oocytes were little affected by these multiple rearrangements. There were no detectable deletions in 21/35 (60%) of the oocytes, and only one to three deletions in 14/35 (40%). Interestingly, when several oocytes were collected from the same donor, each oocyte displayed its own distinct type of deletion. In Figure 2, lanes 1 to 3, show the results for three oocytes collected from one patient. The first oocyte shows mtDNA with

374

Vol. 252, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 1. Southern blot of PCR products from 18 spermatozoan samples. M: Molecular weight marker Lambda/Hind III. The arrow shows the normal PCR product of 16 kb.

three deletions (one deletion of about 4 kb and two deletions of about 8 kb), whereas the other two oocytes show no mtDNA deletions. Lanes 5 to 14 correspond to 10 oocytes from another patient. Only two of the oocytes carry a specific mtDNA deletion of about 9 and 11 kb (lanes 7 and 12). Lanes 15 to 17 correspond to oocytes from yet another patient. Each of these oocytes carries a specific type of mtDNA deletion of about 7, 6.5 and 6 kb. Contamination by spermatozoan mtDNA is unlikely since all the oocytes were washed and checked for the absence of spermtraction. Moreover, even in the event of such contamination, the proportion of mtDNA

in spermatozoa (about 100 copies per cell) and in oocytes (about 100 000 copies per cell) would lead to a 1000-fold dilution of spermatozoan mtDNA, thus making the cross-detection of spermatozoan deletions unlikely. DISCUSSION There are two kinds of multiple mtDNA rearrangement. Firstly, several rare mtDNA diseases are associated with a large number of deleted mitochondrial genomes (15). These deletions are usually visualized

FIG. 2. Southern blot of the PCR products from different single-cell oocytes. M: Molecular weight marker Lambda/Hind III. Several oocytes collected from the same donor: patient 1 (lanes 1 to 3), patient 2 (lanes 5 to 14) and patient 3 (lanes 15 to 17). The deletions are indicated by arrows. Lane 4: negative PCR control. 375

Vol. 252, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

with a single Southern blot. The mendelian transmission of this type of anomaly is linked to the unidentified nuclear genes probably involved in mtDNA maintenance (23-25). Reduced sperm mobility and male infertility are often associated with this type of rare mtDNA disease (26). We previously reported a case of male hypofertility associated with major alteration of the mtDNA in spermatoza and muscle (27,28). In this case, the large numbers of multiple mtDNA rearrangements were detectable by the Southern blot alone. Our present report shows that, as in the case of mtDNA disorders in general, this high proportion of mtDNA mutation is not frequent in male hypofertility. Indeed, none of the 57 patients with oligoasthenospermia had a particularly high rate of mtDNA mutations. Secondly, mitochondria are frequently affected by a lower rate of the same multiple deletions in most tissues of all individuals. These mtDNA deletions occur during aging, and particularly in post-mitotic tissues or in tissues with a low rate of regeneration (29,30). These age-linked deletions are rarer in dividing cells because efficient mitochondria may be selected during cellular divisions by natural selection. Such multiple deletions are generally not identified by the Southern blot but need more sensitive techniques of detection, such as the PCR. Before the use of the long PCR, this age-correlated phenomenon was mainly explored by the analysis of the single common 4,977 base-pair deletion. Gametes display up to about 0.1% of mtDNA common deletion (9,10). Kao et al. have found that the occurrence of the common deletion varies from 2.5%, in the more mobile sperm fractions, to 40% in the less mobile sperm fractions (9). Multiple deletions have been recently associated with the decline of mobility and fertility by the same authors (31). Our results show that a more exhaustive analysis allows the detection of several deletions in 84-86% of sperm samples. We did not detect a greater global prevalence of deletions in spermatozoa from subjects with oligoasthenospermia than in normal spermatozoa. Nevertheless, a more accurate analysis should be performed with the long PCR to confirm the higher rate of deletion especially in the less mobile fractions of spermatozoa. Spermatozoa are not expected to undergo major mtDNA rearrangements since they have a short life span. Moreover, they originate from progenitor cells with a high rate of cellular division, which allows the natural elimination of abnormal mitochondria. However, our study shows that most normal human spermatozoa display large numbers of multiple mtDNA deletions. Another surprising finding is the absence of correlation between the deletions and the age of the patients, in contrast with studies previously performed in other tissues. Our results suggest that numerous deletions are probably not transmitted because of the specific elimination of paternal mtDNA, whereas the

oocyte mtDNA, which is the source of the mtDNA in the embryo, carries only a few deletions, if any. It has been proposed that oocytes might be specifically protected to avoid the accumulation of mtDNA mutations during the long phase of latency preceding fertilization (32). Indeed, the nuclear background of the cell-type dramatically influences mtDNA maintenance and repair mechanisms in cell cultures (33). Spermatozoa, which probably do not have the same protective mechanisms as the oocytes, may thus accumulate large numbers of mutations in a very short time, probably during the differentiation period (approximately five weeks). The high rate of mtDNA mutation in spermatozoa may be correlated with the high rate of chromosomal anomalies observed in spermatozoa and with the high percentage of spermatozoa with obvious structural abnormalities (34). Since the paternal mtDNA is not transmitted to the descendants, some of the mechanisms protecting mtDNA might be absent during the differentiation of spermatozoa, thus leading to rapid accumulation of mtDNA mutations. Another interesting observation is that mtTFA, a factor involved in transcription and replication of mtDNA, is downregulated during spermatogenesis (35). We believe that the decrease of mtTFA leads not only to reduced mtDNA replication in spermatozoa, but also to progressive mtDNA instability. Furthermore, since the main function of spermatozoan mitochondria is to provide ATP for flagellar propulsion, an accurate genetic system may no longer be necessary after differentiation. The capital of oxydative phosphorylation inherited by mature spermatozoa may be sufficient to ensure mobility until fertilization, after which the mtDNA maintenance might well be down-regulated. Spermatozoa could thus serve as a useful model to study the universal process of mtDNA instability, and to investigate the missing mtDNA maintenance factors leading to this defect. Oocytes appear to be less affected by mtDNA deletions than spermatozoa. However, oocytes from the same donor displayed distinctive deletions. This suggests that the deletions might occur before the division of the precursor cells and may be randomly segregated during early oogenesis, according to the bottleneck theory. This random genetic segregation of mtDNA may lead to heteroplasmic oocytes (with deletions) as well as to homoplasmic oocytes (without deletions). It has been shown that oocytes from normal women are frequently heteroplasmic for length polymorphisms in the mtDNA regulatory, non-coding region (36). This implies that the bottleneck arises at some time between the conception and the maturation of oocytes, rather than during early embryogenesis. The occurrence of the bottleneck during early oogenesis has also been suggested by experimental observations on mice (7). Our results support these findings and show that this heteroplasmy might involve not only neutral polymor-

376

Vol. 252, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

phisms but also deleterious mutations such as largescale deletions in humans. Our work confirms the usefulness of the long PCR in the study of mtDNA instability since it provides information not given by the common deletion studies usually performed. It demonstrates that spermatozoa contain mtDNA with numerous deletions in addition to the common 4,977 base-pair deletion. Spermatozoa are the first short-lived cells to display such large numbers of mtDNA deletions. These alterations may be related to a programmed failure of mtDNA maintenance. The accurate detection of multiple mtDNA rearrangements by the long PCR technique reveals mitochondrial heterogeneity among the different cells of a given tissue. Oocyte mtDNA appears to be more stable than in the case of spermatozoa. However, oocytes from the same donor display different sets of deletions. This is in accordance with the bottleneck theory since each oocyte may inherit its mtDNA from a very small pool of maternal molecules. Thus, a pre-existing deleted genome might be selected after the reduction/amplification stage during early oogenesis. This may explain why each oocyte displays either no rearrangement or a distinct individual pattern of mtDNA deletions. ACKNOWLEDGMENTS We are grateful to Mr. M. Benon for his help in the collection of the samples, Mr. K. Malkani for his critical reading of the manuscript, and Mrs. F. Brizard, A. Coutolleau, L. Charoze´, and D. Couturier for their technical assistance. This work was supported by grants from the University Hospital of Angers, the Region of the Pays-de-Loire and the University of Angers.

REFERENCES 1. Kaneda, H., Hayashi, J. I., Takahama, S., Taya, S., FisherLindahl, K., and Yonekawa, H. (1995) Proc. Natl. Acad. Sci. USA 92, 4542– 4546. 2. Shitara, H., Hayashi, J. I., Takahama, S., Kaneda, H., and Yonekawa, H. (1998) Genetics 148, 851– 857. 3. Ankel-Simons, F., and Cummins, J. M. (1996) Proc. Natl. Acad. Sci. USA 93, 13859 –13863. 4. Giles, R. E., Blanc, H., Cann, H. M., and Wallace, D. C. (1980) Proc. Natl. Acad. Sci. USA 77, 6715– 6719. 5. Gyllensten, U., Wharton, D., Josephson, A., and Wilson, A. C. (1991) Nature 352, 255–257. 6. Clayton, D. A. (1996) Nature Genet. 14, 123–125. 7. Jenuth, J. P., Peterson, A. C., Fu, K., and Shoubridge, E. A. (1996) Nature Genet. 14, 146 –151. 8. Hauswirth, W. W., and Laipis, P. J. (1982) Proc. Natl. Acad. Sci. USA 79, 4686 – 4690. 9. Kao, S. H., Chao, H. T., and Wei, Y. H. (1995) Biol. Reprod. 52, 729 –736. 10. Chen, X., Prosser, R., Simonetti, S., Sadlock, J., Jagiello, G., and Schon, E. (1995) Am. J. Hum. Genet. 57, 239 –247.

11. Brenner, C. A., Wolny, Y. M., Barrit, J. A., Matt, D. W., Munne´, S., and Cohen, J. (1998) Mol. Hum. Reprod. 4, 887– 892. 12. Cortopassi, G. A., and Arnheim, N. (1990) Nucleic Acids Res. 18, 6927– 6933. 13. Simonetti, S., Chen, X., DiMauro, S., and Schon, E. A. (1992) Biochim. Biophys. Acta 1180, 113–122. 14. Cortopassi, G., Shibota, D., Soon, N. W., and Arnheim, N. (1992) Proc. Natl. Acad. Sci. USA 89, 7370 –7374. 15. Wallace, D. C. (1992) Ann. Rev. Biochem. 61, 1175–1212. 16. Barnes, W. M. (1994) Proc. Natl. Acad. Sci. USA 91, 2216 –2220. 17. Cheng, S., Fockler, C., Barnes, W. M., and Higuchi, R. (1994) Proc. Natl. Acad. Sci. USA 91, 5695–5699. 18. Melov, S., Shoffner, J. M., Kaufman, A., and Wallace, D. C. (1995) Nucleic Acids Res. 23, 4122– 4126. 19. Reynier, P., and Malthie`ry, Y. (1995) Biochem. Biophys. Res. Commun. 217, 59 – 67. 20. Reynier, P., Pellissier, J. F., Harle, J. R., and Malthie`ry, Y. (1994) Biochem. Biophys. Res. Commun. 205, 375–380. 21. Reynier, P., Figarella-Branger, D., Serratrice, G., Charvet, B., and Malthie`ry, Y. (1994) Biochem. Biophys. Res. Commun. 202, 1606 –1611. 22. World Health Organization. (1992) in WHO Laboratory Manual for the Examination of Human Semen and Sperm–Cervical Mucus Interactions, Cambridge University Press. 23. Suomalainen, A., Kaukonen, J., Amati, P., Timonen, R., Haltia, M., Weissenbach, J., Zeviani, M., Somer, H., and Peltonen, L. (1995) Nature Genet. 9, 146 –151. 24. Kaukonen, J. A., Amati, P., Suomalainen, A., Ro¨tig, A., Piscaglia, M. G., Savi, S., Weissenbach, J., Fratta, G., Comi, G., Peltoene, E., and Zeviani, M. (1996) Am. J. Hum. Genet. 58, 763–769. 25. Barrientos, A., Volpini, V., Casademont, J., Genis, D., Masanares, J. P., Ferrer, I., Cardellach, F., Urbano-Marquez, A., Esrtevill, X., and Nunez, V. (1996) J. Clin. Invest. 97, 1570 – 1576. 26. Folgero, T., Berteussen, K., Tobergsen, T., and Oian, P. (1993) Hum. Reprod. 8, 1863–1868. 27. Lestienne, P., Reynier, P., Chre´tien, M. F., Penisson-Besnier, I., Malthie`ry, Y., and Rohmer, V. (1997) Mol. Hum. Reprod. 3, 811– 814. 28. Reynier, P., Chre´tien, M. F., Penisson-Besnier, I., Malthie`ry, Y., Rohmer, V., and Lestienne, P. (1997) CR. Acad. Sci. III-vie 320, 629 – 636. 29. Corral-Debrinski, M., Horton, T., Lott, M. T., Shoffner, J. M., Beal, M. F., and Wallace, D.C. (1992) Nature Genet. 2, 324 –328. 30. Lee, H. C., Pang, C. Y., Hsu, H. S., and Wei, Y. H. (1994) Biochim. Biophys. Acta 1226, 37– 43. 31. Kao, S. H., Chao, H. T., and Wei, Y. H. (1998) Mol. Hum. Reprod. 4, 657– 666. 32. Kitagawa, T., Suganuma, N., Nawa, A., Kikkawa, F., Tanaka, M., Ozawa, T., and Tomoda, Y. (1993) Biol. Reprod. 49, 730 –736. 33. Holt, I. J., Dunbar, D. R., and Jacobs, H. T. (1997) Hum. Mol. Genet. 6, 1251–1260. 34. David, G., Bisson, J. P., Czyglik, F., Jouannet, P., and Gerningon, C. (1975) J. Gynecol. Obst. Reprod. (Paris) 4, 17–36. 35. Larsson, N. G., Oldfors, A., Garman, J. D., Barsh, G. S., and Clayton, D. A. (1997) Hum. Mol. Genet. 6, 185–191. 36. Marchington, D. R., Hartshorne, G. M., Barlow, D., and Poulton, J. (1997) Am. J. Hum. Genet. 60, 408 – 416.

377