Chromosomal non-disjunction in human oocytes: is there a ...

Human Reproduction, Vol. 15, (Suppl. 2), pp. 160-172, 2000

Chromosomal non-disjunction in human oocytes:is there a mitochondrial connection?

Eric A.Schon1'2'6, Sang Ho Kim1, Jose CFerreira3,Paulo Magalhaes1, Marcy Grace4, Dorothy Warburton2'5

and Susan J.Gross3

department of Neurology, Columbia University, 2Department of Geneticsand Development, Columbia University, 3Department of Obstetrics and

Gynecology, Montefiore Medical Center/Albert Einstein College ofMedicine, Bronx, NY 10461, 4Armed Forces Radiobiology ResearchInstitute, Bethesda, MD 20889-5603, and department of Pediatrics,

Columbia University, 630 West 168th Street, New York, NY 10032, USA

6To whom correspondence should be addressed at: Department of Neurology,Room 4-431 Columbia University, 630 West 168th Street,

New York, NY 10032, USA.E-mail: [email protected]

The frequency of chromosome abnormalit-ies due to non-disjunction of maternal chro-mosomes during meiosis is a function ofage, with a sharp increase in the slope ofthe trisomy-age curve between the ages of30 and 40 years. The basis of this increase,which is a major cause of birth defects,is unknown at present. In recent years,mutations in mitochondrial (mt) DNA havebeen associated with a growing number ofdisorders, including those associated withspontaneous deletions of mtDNA (AmtDNAs). Intriguingly, these pathogenicAmtDNAs, which are present at extremelyhigh levels in certain patients, are alsopresent at extremely low levels (detectableonly by polymerase chain reaction) in nor-mal individuals. The proportion of suchAmtDNAs in normal muscle is a functionof age; the shape of this curve is exponential,with the accelerating part of the curvebeginning at -30-40 years. We postulatethat, as well as muscle and brain, a similar

160

time-dependent accumulation of AmtDNAsalso occurs in normal oocytes. SinceAmtDNAs are functionally inactive, anaccumulation of such aberrant genomescould eventually compromise ATP-depend-ent energy-utilization in these cells. Fur-thermore, these deficiencies would alsoaffect the function of the somatic follicularcells that surround, and secrete importantparacrine factors to, the oocyte. If there isindeed an age-associated relationshipbetween AmtDNAs and oocyte age, perhapserrors in meiosis (which is almost certainlyan energy, and ATP, dependent process) arerelated to mutations in mtDNA (primarilydeletions, but perhaps point mutations aswell) in oocytes and/or the surroundingsomatic cells, which result in deficiencies inboth mitochondrial function in general andoxidative energy metabolism in particular.This hypothesis would explain many ofthe non-Mendelian features associated withmaternal age-related trisomies, e.g. Down'ssyndrome.

© European Society of Human Reproduction & Embryology

Downloaded from https://academic.oup.com/humrep/article-abstract/15/suppl_2/160/619783by gueston 17 February 2018

Oocyte mitochondria and non-disjunction?

Key words: ageing/Down's syndrome/meiosis/mitochondrial disease/mitochondrial DNA

Introduction

The study of mammalian female meiosis pro-vides us with a fascinating challenge: meiosisis not only protracted over several years,but the overall process arrests twice, both inmeiosis I (MI) and meiosis II (Mil). Theprimordial germ cells are first seen in fetallife among the endodermal cells of the yolksac at 4 weeks, and following migration tothe gonadal ridge, undergo rapid mitoses,leading to -7X106 oogonia by 20 weeks.These cells lose their ability to undergo furthermitotic division and subsequently enter mei-osis. The exact mechanism that initiatesmeiosis is not precisely defined. Ovarianstromal (pregranulosa) cells surround theseprimary oocytes, thus forming the primordialfollicles that will remain in the dictyate stageof meiosis until maturation of the hypothal-amic-pituitary axis at puberty. The majorityof these follicles will undergo various stagesof maturation and atresia, so that at birth only2X106 germ cells will remain (Peters et al.,1976). This process of atresia is an ongoingone, proceeding throughout infancy andchildhood.

With the maturation of the human femalereproductive system, a dominant follicle willemerge from within a cohort of early stagefollicles. As a result of precise hormonalcontrol, including autocrine and paracrine sig-nalling, this maturing follicle will continuethrough the full spectrum of follicular develop-ment. These stages include the primary, pre-antral, and antral follicular stages, whereuponthe oocyte is surrounded by the somaticcumulus granulosa cells and the Call-Exnerbodies coalesce into a single antrum containingthe follicular fluid. Following the surge ofluteinizing hormone that presages ovulation,

meiosis resumes, as evidenced by germinalvesicle breakdown and extrusion of the firstpolar body. Meiosis will then proceed untilmetaphase in Mil and will remain arresteduntil fertilization.

Inhibition and resumption of meiosis

For the purposes of understanding meioticcontrol of oocytes, it is helpful to dividefollicular development into two stages, prean-tral and antral. When removed from a latestage antral follicle, the oocyte has the capacityto resume and continue meiosis unhindered toMil (Edwards, 1965), a property that preantraloocytes do not have. Furthermore, co-cultureof the stripped oocytes with granulosa cellsand follicular fluid inhibits meiosis (Tsafririand Pomerantz, 1984; Sirard and Bilodeu,1990). These important observations wouldindicate that the inhibition of meiosis in thepreantral oocyte is a result of incompletedevelopment within the cell itself. Conversely,the mature antral oocyte is responding toinhibitory signals from the surroundingsomatic cells and to factors within the follic-ular fluid. These experiments also demonstratethat the oocyte must undergo inherent changesto become competent to resume meiosis. Inmurine models, nuclear competence appearsto require both the presence of follicularsomatic cells (Byskov et al., 1997) as well asthe acquisition of the constituents that arenecessary for cell cycle transition. Maturationpromoting factor (MPF) is required for theresumption of meiosis; it increases in theoocyte even in the absence of somatic cells(Eppig et al., 1993). However, nuclear matura-tion factors do not act in isolation; rather it isthe interplay of extrinsic cell signalling withintrinsic cell cycle regulators such as MPFthat is critical for the acquisition of nuclearcompetence in the oocyte (Chesnel et al.,1994).

161


E.A.Schon et al.

Age-related changes in maturingoocytes and follicles

One of the recognized changes in the nuclearenvelope of the mature oocyte is the appear-ance of microtubule organizing centres(MTOCs) that will later be used in assemblyof the meiotic spindle. Microtubule motorproteins are associated with the microtubulesand the chromosomes, particularly the kineto-chore (de Pennart et al., 1994) and centro-somes (Battaglia et al., 1996a), and activelyparticipate in the arrangement and preservationof the spindle structure. In ageing humanoocytes, irregularities in spindle formation andchromosome alignment have been identified(Battaglia et al., 1996b). The actual aetiologyfor these changes remains unclear, with anumber of different hypotheses proposed. Thepossibility of inherent time-dependency inthe breakdown of the spindle componentsthemselves has been suggested (Hawley et al.,1994). However, the astute suggestion hasbeen made that it could be the necessarystockpiling of proteins necessary for nucleardevelopment that is impaired and that ulti-mately leads to non-disjunction in ageingoocytes (Hunt and LeMaire-Adkins, 1998). Ofnote, changes not just within the oocyte butin overall follicular functional maturation,such as decreased time to ovulation, have beenobserved in humans (Klein et al., 1996).

Mitochondrial DNA andmitochondrial disease

The human mitochondrial genome (Figure 1)is a 16 569 bp circle of double-stranded DNA(Anderson et al., 1981). It contains genesencoding two ribosomal RNAs and 22 transferRNAs, as well as 13 structural genes thatencode subunits of components of the respirat-ory chain complexes. Of the 13 structuralgenes, seven encode subunits of complex I(NADH-coenzyme Q oxidoreductase), one

Cytb

NDS

ND1 | Sporadic "clonal" deletions (e.g. KSS)Inherited "multiple" deletions (e.g. AD-fc

ISporadic "multiple" deletions (e.g. ageing)

ND2 ND4

COX ICOX III

COX II K A 8 / 6

Figure 1. Map of the human mitochondrial genome (afterAnderson et al., 1981). The genes for the mtDNA-encoded12S and 16S ribosomal RNAs, the 22 transfer RNAs (1-letteramino acid nomenclature), and the NADH-coenzyme Qoxidoreductase (ND), cytochrome c oxidase (COX),cytochrome b (Cyt b) ATP synthase (A) subunits are allshown. The origins of light-strand (OL) and heavy-strand(OH) replication, and of the promoters for initiation oftranscription from the light-strand (LSP) and heavy-strand(HSP), are shown by arrows (see Clayton, 1982, 2000). The'common deletion', a AmtDNA species often found inmtDNA rearrangement disorders and in normal ageing, isshown, as are point mutations associated with the sporadicmyopathies or the maternally inherited MELAS and MERRF.(See text for key to abbreviations.)

encodes the cytochrome b subunit of complexIII (CoQ-cytochrome c oxidoreductase), threeencode subunits of complex IV (cytochromec oxidase, or COX), and two encode subunitsof complex V (ATP synthetase). Each of thesecomplexes also contains subunits encoded bynuclear genes, which are imported from thecytoplasm and assembled, together with themtDNA-encoded subunits, into the respectiveholoenzymes, each of which is located in themitochondrial inner membrane. Complex II(succinate dehydrogenase-CoQ oxidoreduc-tase), of which succinate dehydrogenase(SDH) is a component, is encoded entirelyby nuclear genes; SDH thus serves as anindependent marker for mitochondrial numberand activity.

162



Mitochondria (and mtDNAs) are maternallyinherited (Hutchison et al, 1974; Giles et al,1980). Thus, pathogenic mutations in mtDNAnormally (but, importantly, not always) resultin pedigrees that exhibit maternal inheritance,i.e. the disease passes only through females,and essentially all children (both boys andgirls) inherit the error. In addition, becausethere are hundreds or even thousands of mito-chondria in each cell, with an average of fivemtDNAs per organelle in somatic cells (Satohand Kuroiwa, 1991), mutations in mtDNAtypically result in two populations of mtDNAs(i.e. wild-type and mutated), a conditionknown as heteroplasmy.

An ever-growing number of diseases aredue to maternally inherited mtDNA pointmutations (reviewed in Schon et al, 1997).Point mutations in mtDNA-encoded respirat-ory chain polypeptides are responsible, in themain, for Leber's hereditary optic neuropathy(LHON); neuropathy, ataxia, and retinitis pig-mentosa (NARP); and maternally inheritedLeigh syndrome (MILS). The great majorityof mtDNA point mutations, however, are inthe tRNA genes. The most prominent of thesedisorders are mitochondrial encephalomyopa-thy, lactic acidosis, and stroke-like episodes(MELAS); myoclonus epilepsy with ragged-red fibres (MERRF); and a mixed group ofmaternally inherited myopathies, encephalo-myopathies, and cardiomyopathies (Schonet al, 1997).

Large-scale rearrangements of mtDNA, i.e.partial deletions (AmtDNAs) and partialduplications (dup-mtDNAs), are also associ-ated with disease. The most common patholog-ies are ocular myopathy (OM) and the Kearns-Sayre syndrome (KSS), two mitochondrialdisorders associated with progressive externalophthalmoplegia (PEO) (Holt et al., 1988;Zeviani et al, 1988; Moraes et al, 1989),as well as a haematopoietic disorder calledPearson's marrow/pancreas syndrome (PS)(Rotig et al, 1989). In these disorders, the

AmtDNAs can be observed easily by Southernblot hybridization analysis as a large popula-tion (up to 80% of total mtDNA) of mtDNAmolecules migrating in electrophoretic gelsmore rapidly than full-length mtDNAs.Importantly, these three syndromes are almostnever maternally inherited, as the deletionsarise spontaneously in the patient whosemother and siblings are usually normal, bothphenotypically and genetically. In addition,the deletions are unique in each patient, withthe particular type of deletion differing amongpatients; however, about one-third of all KSS/PEO/PS patients harbour the same deletion,called the 'common deletion' (Schon et al,1989; Mita et al, 1990), which removes 4977bp of mtDNA between the ATPase8 and ND5genes (see Figure 1). Taken together, thesefindings imply that the AmtDNA populationin any one such patient is a clonal expansion ofa single spontaneous deletion event occurringearly in oogenesis or in embryogenesis.

Some mtDNA rearrangements are mater-nally inherited, but these patients do not havethe clinical features of KSS or PEO; theymore typically have maternally inherited typeII diabetes (Dunbar et al, 1993). Significantly,mtDNA duplications, but not mtDNA dele-tions, have been documented to be inherited inthis way. With but one exception (S.DiMauro,personal communication), there is no convin-cing evidence that large-scale mtDNA dele-tions can be inherited through the femalegermline, implying that high levels ofAmtDNAs (but, somewhat surprisingly, notequivalently high levels of pathogenic mtDNApoint mutations) are incapable of being trans-mitted to viable offspring.

Besides the 'clonal' mtDNA deletions foundin the sporadic rearrangement disorders, anumber of Mendelian-inherited disorders inwhich affected family members harbour largequantities of multiple species of deletions ofmtDNA in tissues (usually muscle) have beendescribed. These deletions are apparently gen-

163


E.A.Schon et al.

erated during the lifespan of the patient(Zeviani et al, 1989; Nishino et al, 1999).

Intriguingly, while most mtDNA pointmutations can be maternally transmitted (andmaternally inherited), a growing group ofmutations appear to violate this dictum. Ofthe few known sporadic mtDNA pointmutations or microdeletions (-10% of allpoint mutations), almost all are located inpolypeptide-coding genes. While they are pre-dominantly in the cytochrome b subunit ofcomplex III (Dumoulin et al, 1996; Andreuet al, 1998, 1999a), these mutations have alsobeen found in other complexes (Keightleyet al, 1996; Hanna et al, 1998; Andreu et al,1999b). The prevalence of sporadic mutationsin mtDNA-encoded polypeptides implies that,like AmtDNAs, the loss of a functional respir-atory chain somehow selects against viableprogeny in the germline.

Deletions of mtDNA in somatic tissuesduring normal ageing

In the last few years it has become clear thattissues from normal individuals, especiallyterminally differentiated (post-mitotic) tissueswith high oxidative requirements, e.g. muscleand brain, contain low amounts of AmtDNAs(Schon et al, 1996). These deleted mtDNAs,which are observable only after amplificationusing the polymerase chain reaction (PCR), arequalitatively identical to the highly abundantpopulations of AmtDNAs that have beendescribed in patients with sporadic PEO andKSS, sporadic Pearson's syndrome, andautosomal-dominant PEO.

Importantly, the AmtDNAs found in normalindividuals appear to accumulate during age-ing. Using a quantitative PCR technique tomeasure the amount of the common deletion,we have found that this species of AmtDNAaccumulates in muscle by a factor of 10 000over the course of the normal human lifespan,reaching a level of -0.1% of total muscle

mtDNA by the age of 84 years (Simonettiet al, 1992; Pallotti et al, 1996). Besidesthe common deletion, numerous other deletedspecies are also present in ageing muscle(Zhang etal, 1992; Chen et al, 1993; Pallottiet al, 1996). Thus, both the overall numberof different AmtDNA species, and the amountof each such species, increases with time, suchthat, in the aggregate, perhaps as much as 5-10% (or, according to some workers, evenmore) of total mtDNA is deleted in elderlyindividuals. These AmtDNAs are not distrib-uted homogeneously, at least in muscle.Rather, individual muscle fibres appear tocontain the majority of deleted molecules(Schon et al, 1996; Brierley et al, 1998), andit is these fibres that are respiratorily deficient.

Presence of mtDNA deletions innormal human oocytes

We currently do not know why, or how,AmtDNAs accumulate with age. A logarithmicincrease in AmtDNAs with respect to agewould be consistent with the idea thatAmtDNAs are competent for replication, andthe AmtDNAs observed at later ages are pro-geny of AmtDNAs that arose earlier in time.This exponential accumulation of AmtDNAsposes an interesting mechanistic question.AmtDNAs appear to arise afresh with eachgeneration, as our data indicate that there arefour orders of magnitude fewer AmtDNAs ininfancy as compared to old age. If this is true,how can a woman of childbearing age, whoalready harbours a low but detectable level ofAmtDNAs (at least in her somatic tissues),give birth to a baby containing almost noAmtDNAs at all? In other words, how doesthe organism 'reset' the level of mtDNAmutations in each generation? This may bedue to one of at least three possibilities: (i)AmtDNAs arise and accumulate in somatictissue, but do not arise in female germlinetissue (a hypothesis proposed by Nagley et al,

164



1992); (ii) AmtDNAs do arise in oocytes, butare actively eliminated via some unknownmechanism; or (iii) AmtDNAs do arise priorto or during oogenesis but are rarely transmit-ted to the viable embryo because they cannotpass through the narrow mitochondrial popula-tion bottleneck during oogenesis and sub-sequent embryogenesis (Hauswirth and Laipis,1985; Ashley et al, 1989; Jenuth et al, 1996,1997). The massive accumulation of a singlespecies of AmtDNA in spontaneous Kearns-Sayre syndrome (in which the population ofdeleted molecules is present in all tissues, andis presumably a clonal expansion of an initialdeletion event occurring early in oogenesis orembryogenesis) is consistent with all threepossibilities, but is perhaps most compatiblewith the last one.

In 1995, we began to address this question.Using single-oocyte PCR, we were able toquantify the total amount of mtDNA in unfer-tilized human oocytes left over from an IVFprogramme: the average oocyte contained-100 000 mitochondrial genomes (Chen et al,1995). In addition, we detected a single markerAmtDNA4977 (the 4977 bp common deletion)in about half of all the oocytes we tested. Theamount of this species of AmtDNA variedamong oocytes, but was 10-100 such molec-ules per oocyte (Chen et al., 1995). Theseresults have been confirmed by some groups(Keefe et al, 1995; Brenner et al, 1998), butnot by others (Miiller-Hocker et al, 1996).Deletions of mtDNA have also been observedin ageing ovarian tissue (Kitagawa et al, 1993;Suganuma et al, 1993) and in spermatozoa(Reynier et al, 1998).

Hypothesis

Maternal meiotic non-disjunction resulting inaneuploidy shows a very strong relationshipwith increasing maternal age, with an expo-nentially rising curve after -32 years. Thismaternal age effect holds for all chromosomal

pairs, with the exception of the very largechromosomes (where there seems to be littlematernal age effect) and for chromosomes 2and 16 (where the effect is linear) (Warburtonand Kinney, 1996). It has been estimated thatat least 30% of oocytes in women aged >40years have undergone non-disjunction.

These findings have stimulated us and others(Jansen and De Boer, 1998; Van Blerkomet al, 1998) to consider the possibility thatan age-related increase in mtDNA mutationsin oocytes could be an underlying cause ofthe maternal age effect on chromosomal non-disjunctions. Specifically, we postulate that,as a woman ages, mtDNA mutations accumu-late in her oocytes, and that the frequencyof aneuploidy increases once a threshold ofmitochondrial energy deficit is crossed. Asdescribed above, the process of meiotic inhibi-tion and resumption is the result of the produc-tion of autonomous factors by the oocyte, aswell as the response to extrinsic signalling bythe surrounding follicular cells. Consequently,mtDNA mutations that accumulate over timein the surrounding somatic cells might alsolead to aneuploidy. This 'mitochondrial ageinghypothesis' would explain the relationshipbetween maternal age and the frequency oftrisomies, e.g. Down's syndrome.

The idea that oocytes require high concen-trations of ATP (Van Blerkom and Runner,1984; Van Blerkom, 1991) and that a mito-chondrial energy deficit could result in aneu-ploidies is not a new one (Beermann andHansmann, 1986). However, the most compel-ling reason to consider a mitochondrial aeti-ology in germline aneuploidies may be thatthere are no mechanistic data that explainsuccessfully the age-related increase in humantrisomies. Specifically, no data exist to explainwhy the mere passage of time should resultin the increased frequency, of non-disjunctionevents. An age-related phenomenon can bethe result, of course, of age-related changesin nuclear DNA, or to epigenetic and environ-

165


E.A.Schon et al.

mental factors, but it is more difficult toinvoke these causes (which probably varyfrom individual to individual) to explain theconsistent and specific observation of age-related aneuploidies, which show little or novariation with geographical region, ethnicity,or social class. On the other hand, the 'mito-chondrial paradigm', with its focus on popula-tion genetics, on random but cumulativeeffects of mutations, on heteroplasmy (and, inageing, 'polyplasmy'), and on mitotic segre-gation (coupled with the mitochondrion'smonopoly in oxidative energy metabolism) isan attractive alternative hypothesis to explainthe ageing-aneuploidy correlation.

Other facts also support a mitochondrialaetiology. As noted above, there is an exponen-tial increase in the amount of the commondeletion in muscle over time (Simonetti et al,1992; Pallotti et al, 1996). The shape of thiscurve roughly parallels the curve describingthe increase in the frequency of trisomies asa function of maternal age (Penrose, 1933;Hassold and Jacobs, 1984; Hassold and Chiu,1985), that is, the curve begins to accelerateat -30-40 years.

The vast majority of all chromosomal non-disjunction events are maternal, not paternal(although for a few specific chromosomes,e.g. chromosome 2, paternal disomies maypredominate) (Zaragoza et al, 1998), andnearly 80% of maternal non-disjunctions occurduring meiosis I (Antonarakis et al., 1991,1992, 1993; Sherman et al., 1991; McFaddenetal., 1993; Abruzzo and Hassold, 1995). Thefrequency of Mil disomies in comparisonwith MI disomies varies from chromosome tochromosome (Fisher et al., 1995; Hassoldet al., 1995), but it appears that both MI andMil errors increase with advancing maternalage (Yoon et al., 1996). However, there arealso data to suggest that all non-disjunctionevents are the result of errors in MI (Lambet al., 1996). In either case, we believe it issignificant that oocytes are generated only

during fetal life and have the same age as thewoman bearing them.

Data from other organisms such as yeast andDrosophila have shown that factors causingdecreased recombination result in an increasein meiotic non-disjunction and aneuploidy.Analysis of recombination events in humantrisomic conceptions has demonstrated a con-sistent decrease in recombination for maternalnon-disjunctive events classified as meiosis I,as well as a change in the chromosomaldistribution of these recombination events(Lamb et al, 1996, 1997; Robinson et al,1998). For Mil errors, the results have beenless consistent across chromosomes (Buggeet al, 1998), but for trisomy 21, an increasein recombination in pericentromeric regionswas described.

These data have led to the hypothesis thatthe maternal age effect on non-disjunction isa 'two-hit' process. In the first hit, the under-lying distribution of recombination events perchromosome bivalent, established during fetallife, creates pairs susceptible to non-disjunc-tion, either because of too little or too muchpericentromeric recombination. While a youngovary can handle most configurations success-fully, older oocytes are likely to undergonon-disjunction, given these less than optimalrecombination patterns (Lamb et al, 1996).In the second hit, our hypothesis concerningthe role of accumulating mitochondrial DNAmutations addresses the reason for the changein oocyte competence with maternal age.

Chromosomal disjunction, whether in mei-osis I or meiosis II, is almost certainly a highlyenergy-intensive and ATP-dependent process,because the synapsed chromosome pairs (inMI) or sister chromatids (in Mil) must bepulled to the opposite ends of the dividingoocyte along the 'pulley-and-rope' machineryof the mitotic spindle apparatus (Eichenlaub-Ritter et al, 1996; Eichenlaub-Ritter, 1998).Recent studies have shown that, in vitro,matured oocytes taken from antral follicles of

166



ageing women show spindle abnormalitiesand problems with chromosome alignment(Battaglia et al, 1996a; Volarcik et al, 1998).Moreover, it has been shown that an increasedfrequency of oocyte aneuploidies in mice canbe related to a specific mitochondrial genotypein the ooplasm (Beermann et al, 1988).

We speculate that mitochondria andmtDNAs in oocytes turn over in order toprovide a certain minimum level of energy foroocyte viability. The supposition that mtDNAreplicates in the resting oocytes of primordialfollicles (currently unverified experimentally)would provide a means by which mutationsin mtDNA could arise in otherwise quiescentoocytes and then accumulate over time. Like-wise, the surrounding somatic cells, includinggranulosa precursors, must maintain viabilityover many years.

The accumulation of mtDNA mutationscould have harmful effects on mitochondrialenergy production and on overall energy levelsin the oocyte. There is only one mitochondrialgenome per organelle in oocytes (Michaelset al, 1982; Piko and Taylor, 1987), asopposed to an average of five mtDNAs perorganelle in somatic cells (Satoh and Kuroiwa,1991). Thus, oocytes should be particularlyprone to the effects of mtDNA mutations, asmutation in only a single mtDNA moleculecould theoretically impair the function of theentire organelle. We also note that the mito-chondrial respiratory chain is the singlegreatest source of free radicals in the body,and that cumulative defects in mtDNA integ-rity could be intimately associated with dam-age by reactive oxygen species. Thus, theaccumulation of mtDNA mutations might notonly affect ATP production directly, but couldimpair it indirectly and/or secondarily, throughthe downstream effects of free radicals andoxidative stress on cellular functions (Tarfn,1996; Tarin et al, 1996; 1998).

Regarding the oocyte-follicle as a singlefunctional unit may also aid in our understand-

ing of mitochondrial perturbations, in twoways. First, endocrinopathies are a major fea-ture of mitochondrial disorders. The ovarianfollicle would certainly fit the description ofan active and dynamic endocrinological unit,particularly during maturation, which is undertightly regulated, sequential hormonal control,with secretion of sex steroids as well as growthfactors and regulatory peptides. Second, thefollicular fluid contains a plethora of moleculessecreted by the somatic cells, including mito-chondrion-derived steroids. Granulosa devel-opment, in fact, is accompanied by increasesin the numbers of mitochondria and topograph-ical shifts in their subcellular location (Meidanet al, 1990; Dive et al, 1992). This supposi-tion is consistent with the suggestion thatmitochondrial mutations are particularly proneto disrupt molecular transport across cell mem-branes (Schon et al, 1997). Thus, mitochon-drial mutations could theoretically altermembrane transport and thereby alter the sig-nals perceived by the oocyte.

Problems with the hypothesis

There are at least two key pieces of evidencethat do not support this hypothesis. First,there is no evidence that levels of pathogenicmtDNA mutations high enough to cause overtdisease can cause karyotypic abnormalities inpatient cells. For example, we were unable todetect karyotypic changes in fibroblasts fromMERRF patients harbouring >80% mutatedmtDNAs (mutation in the tRNALys ^ene), eventhough it is known that high levels of theMERRF mutation can impair mitochondrialprotein synthesis (Masucci et al, 1995). Thisresult implies that either an extremely highlevel of mutant mtDNAs are required beforemitotic non-disjunction occurs (i.e. that non-disjunction is a threshold phenomenon), orthat chromosomes can divide well even withreduced ATP values. Of course, evidence

167


E.A.Schon et al.

obtained on mitotic cells may not be applicableto meiotic cells.

A potentially more serious problem for thehypothesis is the fact that women harbouringhigh levels of pathogenic mtDNA pointmutations (usually 50-70% of total mtDNA)are able to have children who are destined tohave a serious mitochondrial disorder, yetwhose cells are karyotypically normal. Thisimplies that chromosomes in oocytes withseverely reduced levels of ATP can still replic-ate and divide normally. On the other hand,the failure to transmit mtDNA deletions frommother to child, and the extremely high preval-ence of spontaneous cytochrome b mutations(and the lack of reported maternally inheritedcytochrome b mutations) implies that at leastsome types of mtDNA mutations interferewith germline or zygotic viability. Examiningreproductive histories of women bearing chil-dren with mitochondrial disorders forincreases in infertility, in spontaneous abor-tions, or in trisomic offspring might be worth-while in this context.

Predictions of the hypothesis

The hypothesis that mtDNA mutations areassociated with trisomies predicts a numberof consequences, all of which can be testedexperimentally.

First, it predicts that the proportion ofmtDNA mutations in oocytes increases as afunction of age, irrespective and independentof the presence of aneuploidies in individualgerm cells. This issue can be addressed byquantifying mtDNA deletions in oocytes ver-sus age, using methods already described(Chen et al, 1995).

Second, it predicts that the proportion ofmtDNA mutations is higher in oocytes har-bouring aneuploidies than in normal cells,irrespective of age. This issue can be addressedby performing simultaneous karyotyping andquantitative mtDNA analysis on individual

oocytes. In addition, if non-disjunction is dueto a high proportion of mtDNA mutations inthe oocyte itself, and if this high level ofmutation is transmitted to the fetus (i.e. noloss through the bottleneck in early embryo-genesis), the hypothesis predicts that theproportion of mtDNA mutations ought to begreater in the somatic cells of trisomic patientsor fetuses than in normals, and that patientswith trisomies might have mitochondrial dis-eases at a relatively high frequency.

Third, it implies that mtDNA mutations inoocytes should have detectable biochemicalconsequences, such as loss of cytochrome coxidase activity (but see Mtiller-Hocker et al.,1996). This can be tested by performing respir-atory chain histochemistry and immunohisto-chemistry (Bonilla et al., 1992) on karyotypedoocytes.

Fourth, it also implies that there is mtDNAturnover (presumably by replication) in germ-line tissue. This might be able to be verifiedexperimentally via pulse labelling of mtDNAsin vivo (Moraes and Schon, 1995).

Finally, if aneuploidy is primarily the resultof energy deficits in the follicular cells, thereought to be a correlation between the level ofmtDNA mutations in granulosa cells and thepresence of aneuploidy in the oocyte containedwithin that follicle. This can be tested, forexample, by performing in-situ hybridizationon sectioned follicles to detect AmtDNAs(Mita et al, 1989).

Positive results from experiments of thesetypes would not only support a mitochondrialaetiology for age-related aneuploidies, butmight also point the way towards rationalapproaches to prevention of these devastatingbirth defects.

AcknowledgementsThis work was supported by grants from the NationalInstitutes of Health (NS28828, NS11766, and HD32062)and the Muscular Dystrophy Association.

168



ReferencesAbruzzo, M.A. and Hassold, T.J. (1995) Etiology of

nondisjunction in humans. Environ. Mol. Mutagen.,25 (Suppl. 26), 38-47.

Anderson, S., Bankier, A.T., Barrell, B.G. et al. (1981)Sequence and organization of the humanmitochondrial genome. Nature, 290, 457^-65.

Andreu, A.L., Bruno, C , Shanske, S. et al. (1998)Missense mutation in the mtDNA cytochrome b genein a patient with myopathy. Neurology, 51,1444-1447.

Andreu, A.L., Bruno, C , Dunne, T.C. et al. (1999a) Anonsense mutation (G15059A) in the cytochromeb gene in a patient with exercise intolerance andmyoglobinuria. Ann. Neurol, 45, 127-130.

Andreu, A.L., Tanji, K., Bruno, C. etal. (1999b) Exerciseintolerance due to a nonsense mutation in the mtDNAND4 gene. Ann. Neurol, 45, 820-823.

Antonarakis, S.E. and the Down Syndrome CollaborativeGroup (1991) Origin of the extra chromosome intrisomy 21 as indicated by analysis of DNApolymorphisms. N. Engl. J. Med., 324, 872-876.

Antonarakis, S.E., Petersen, M.B., Mclnnis, M.G. et al.(1992) The meiotic stage of nondisjunction in trisomy21: determination using DNA polymorphisms. Am. J.Hum. Genet., 50, 544-550.

Antonarakis, S.E., Avramopoulos, D., Blouin, J.-L. et al.(1993) Mitotic errors in somatic cells cause trisomy21 in about 4.5% of cases and are not associated withadvanced maternal age. Nature Genet., 3, 146-150.

Ashley, M.V., Laipis, P.J. and Hauswirth, W.W. (1989)Rapid segregation of heteroplasmic bovinemitochondria. Nucleic Acids Res., 17, 7325-7331.

Battaglia, D.E., Goodwin, P., Klein, N.A. and Soules,M.R. (1996a) Influence of maternal age on meioticspindle assembly in oocytes from naturally cyclingwomen. Hum. Reprod., 11, 2217-2222.

Battaglia, D.E., Klein, N.A. and Soules, M.R. (1996b)Changes in centrosomal domains during meioticmaturation in the human oocyte. Mol. Hum. Reprod.,2, 845-851.

Beermann, F. and Hansmann, I. (1986) Follicularmaturation, luteinization and first meiotic division inoocytes after inhibiting mitochondrial function in micewith chloramphenicol. Mutat. Res., 160, 47-54.

Beermann, E, Hummler, E., Franke, U. and Hansmann,I. (1988) Maternal modulation of the inheritablemeiosis I error Dipl I in mouse oocytes is associatedwith the type of mitochondrial DNA. Hum. Genet.,79, 338-340.

Bonilla, E., Sciacco, M., Tanji, K. et al. (1992) Newmorphological approaches to the study ofmitochondrial encephalomyopathies. Brain Pathol, 2,113-119.

Brenner, C.A., Wolny, Y.M., Barritt, J.A. et al. (1998)Mitochondrial DNA deletion in human oocytes andembryos. Mol. Hum. Reprod., 4, 887-892.

Brierley, E.J., Johnson, M.A., Lightowlers, R.N. et al.(1998) Role of mitochondrial DNA mutations inhuman aging: implications for the central nervoussystem and muscle. Ann. Neurol., 43, 217-223.

Bugge, M., Collins, A., Petersen, M.B. et al. (1998) Non-disjunction of chromosome 18. Hum. Mol. Genet., 7,661-669.

Byskov, A.G., Yding Andersen, C, Hossaini, A. andGuoliang, X. (1997) Cumulus cells of oocyte-cumuluscomplexes secrete a meiosis-activating substancewhen stimulated with FSH. Mol. Reprod. Dev., 46,296-305.

Chen, X., Simonetti, S., DiMauro, S. and Schon, E.A.(1993) Accumulation of mitochondrial DNA deletionsin organisms with various lifespans. Bull. Mol. Biol.Med., 18, 57-66.

Chen, X., Prosser, R., Simonetti, S. et al. (1995)Rearranged mitochondrial genomes are present inhuman oocytes. Am. J. Hum. Genet., 57, 239-247.

Chesnel, E, Wigglesworth, K. and Eppig, J.J. (1994)Acquisition of meiotic competence by denuded mouseoocytes: participation of somatic-cell product(s) andcAMP. Dev. Biol, 161, 285-295.

Clayton, D.A. (1982) Replication of animal mito-chondrial DNA. Cell, 28, 693-705.

Clayton, D.A. (2000) Transcription and replication ofthe mitochondrial genome. Hum. Reprod., 15 (Suppl.2), 11-17.

de Pennart, H., Cibert, C , Petzelt, C. and Maro, B.(1994) Microtubule tracks can be detected in mouseoocytes with an antibody directed against a calciumtransporter. J. Cell Set, 107, 1899-1908.

Dive, C , Yoshida, T.M., Simpson, D.J. and Marrone,B.L. (1992) Flow cytometric analysis of steroidogenicorganelles in differentiating granulosa cells. Biol.Reprod., 47, 520-527.

Dumoulin, R., Sagnol, I., Ferlin, T. et al. (1996) Anovel gly290asp mitochondrial cytochrome b mutationlinked to a complex III deficiency in progressiveexercise intolerance. Mol. Cell. Probes, 10, 389-391.

Dunbar, D.R., Moonie, PA., Swingler, R.J. et al. (1993)Maternally transmitted partial direct tandemduplication of mitochondrial DNA associated withdiabetes mellitus. Hum. Mol. Genet, 2, 1619-1624.

Edwards, R.G. (1965) Maturation in vitro of humanovarian oocytes. Lancet, 2 (7419), 926-929.

Eichenlaub-Ritter, U. (1998) Genetics of oocyte ageing.Maturitas, 30, 143-169.

Eichenlaub-Ritter, U., Baart, E., Yin, H. and Betzendahl,I. (1996) Mechanisms of spontaneous and chemically-induced aneuploidy in mammalian oogenesis: basisof sex-specific differences in response to aneugensand the necessity for further tests. Mutat. Res., 372,279-294.

Eppig, J.J., Wigglesworth, K. and Chesnel, F. (1993)Secretion of cumulus expansion enabling factor by

169


E.A.Schon et al.

mouse oocytes: relationship to oocyte growth andcompetence to resume meiosis. Dev. Biol, 158,400-409.

Fisher, J.M., Harvey, J.F., Morton, N.E. and Jacobs, P.A.(1995) Trisomy 18: studies of the parent and celldivision of origin and the effect of aberrantrecombination on nondisjunction. Am. J. Hum. Genet.,56, 669-675.

Giles, R.E., Blanc, H., Cann, H.M. and Wallace, D.C.(1980) Maternal inheritance of human mitochondrialDNA. Proc. Natl Acad. Sci. USA, 77, 6715-6719.

Hanna, M.G., Nelson, I.P., Rahman, S. et al. (1998)Cytochrome c oxidase deficiency associated with thefirst stop-codon point mutation in human mtDNA.Am. J. Hum. Genet., 63, 29-36.

Hassold, T. and Chiu, D. (1985) Maternal age-specificrates of numerical chromosomal abnormalities withspecial reference to trisomy. Hum. Genet., 70, 11-17.

Hassold, T.J. and Jacobs, P.A. (1984) Trisomy in man.Ann. Rev. Genet., 18, 69-97.

Hassold, T., Merrill, M., Adkins, K. et al. (1995)Recombination and maternal age-dependentnondisjunction: molecular studies of trisomy 16. Am.J. Hum. Genet, 57, 867-874.

Hauswirth, W.W. and Laipis, P.J. (1985) Transmissiongenetics of mammalian mitochondria: a molecularmodel and experimental evidence. In Quagliarello, E.(ed.), Achievements and Perspectives of MitochondrialResearch, Elsevier, Amsterdam, The Netherlands, pp.49-59.

Hawley, R.S., Frazier, J.A. and Rasooly, R. (1994)Separation anxiety: the etiology of nondisjunction inflies and people. Hum. Mol. Genet., 3, 1521-1528.

Holt, I.J., Harding, A.E. and Morgan-Hughes, J.A. (1988)Deletions of mitochondrial DNA in patients withmitochondrial myopathies. Nature, 331, 717-719.

Hunt, P.A. and LeMaire-Adkins, R. (1998) Geneticcontrol of mammalian female meiosis. Curr. Top. Dev.Biol, 37, 359-381.

Hutchison, C.A., Newbold, J.B., Potter, S.S. and Edgell,M.H. (1974) Maternal inheritance of mammalianmitochondrial DNA. Nature, 251, 536-538.

Jansen, R.P. and De Boer, K. (1998) The bottleneck:mitochondrial imperatives in oogenesis and ovarianfollicular fate. Mol. Cell. Endocrinol, 145, 81-88.

Jenuth, J.P., Peterson, A.C., Fu, K. and Shoubridge, E.A.(1996) Random genetic drift in the female germlineexplains the rapid segregation of mammalianmitochondrial DNA. Nature Genet., 13, 146-151.

Jenuth, J.P., Peterson, A.C. and Shoubridge, E.A. (1997)Tissue-specific selection for different mtDNAgenotypes in heteroplasmic mice. Nature Genet., 16,93-95.

Keefe, D.L., Niven-Fairchild, T., Powell, S. andBuradagunta, S. (1995) Mitochondrial deoxyribo-

nucleic acid deletions in oocytes and reproductiveaging in women. Fertil. Steril, 64, 577-583.

Keightley, J.A., Hoffbuhr, K.C., Burton, M.D. et al.(1996) A microdeletion in cytochrome c oxidase(COX) subunit III associated with COX deficiency andrecurrent myoglobinuria. Nature Genet., 12, 410-416.

Kitagawa, T., Suganuma, N., Nawa, A. et al. (1993)Rapid accumulation of deleted mitochondrialdeoxyribonucleic acid in postmenopausal ovaries.Biol. Reprod., 49, 730-736.

Klein, N.A., Battaglia, D.E., Miller, P.B. et al. (1996)Ovarian follicular development and the follicular fluidhormones and growth factors in normal women ofadvanced reproductive age. J. Clin. Endocrinol.Metab., 81, 1946-1951.

Lamb, N.E., Freeman, S.B., Savage-Austin, A. et al.(1996) Susceptible chiasmate configurations ofchromosome 21 predispose to non-disjunction in bothmaternal meiosis I and meiosis II. Nature Genet., 14,400-405.

Lamb, N.E., Feingold, E., Savage, A. et al. (1997)Characterization of susceptible chiasma configurationsthat increase the risk for maternal nondisjunction ofchromosome 21. Hum. Mol. Genet., 6, 1391-1399.

McFadden, D.E., Kwong, L.C., Yam, I.Y.L. and Langlois,S. (1993) Parental origin of triploidy in human fetuses:evidence for genomic imprinting. Hum. Genet., 92,465^69.

Masucci, J., Davidson, M., Koga, Y. et al. (1995) In vitroanalysis of mutations causing myoclonus epilepsy withragged-red fibers in the mitochondrial tRNALys gene:two genotypes produce similar phenotypes. Mol. Cell.Biol, 15, 2872-2881.

Meidan, R., Girsh, E., Blum, O. and Aberdam, E. (1990)In vitro differentiation of bovine theca and granulosacells into small and large luteal-like cells:morphological and functional characteristics. Biol.Reprod., 43, 913-921.

Michaels, G.S., Hauswirth, W.W. and Laipis, P.J. (1982)Mitochondrial DNA copy number in bovine oocytesand somatic cells. Dev. Biol, 94, 246-251.

Mita, S., Schmidt, B., Schon, E.A. etal. (1989) Detectionof 'deleted' mitochondrial genomes in cytochrome coxidase-deficient muscle fibers of a patient withKearns-Sayre syndrome. Proc. Natl Acad. Sci. USA,86, 9509-9513.

Mita, S., Rizzuto, R., Moraes, C.T. et al. (1990)Recombination via flanking direct repeats is a majorcause of large-scale deletions of human mitochondrialDNA. Nucleic Acids Res., 18, 561-567.

Moraes, C.T. and Schon, E.A. (1995) Replication of aheteroplasmic population of normal and partiallydeleted human mitochondrial genomes. In Palmieri,F., Papa, S., Saccone, C. and Gadaleta, M.N. (eds),Progress in Cell Research. Vol. 5. Elsevier,Amsterdam, The Netherlands, pp. 209-215.

170



Moraes, C.T., DiMauro, S., Zeviani, M. et al. (1989)Mitochondrial DNA deletions in progressive externalophthalmoplegia and Kearns-Sayre syndrome. N.Engl. J. Med., 320, 1293-1299.

Miiller-Hocker, J., Schafer, S., Weis, S. et al. (1996)Morphological-cytochemical and molecular geneticanalyses of mitochondria in isolated human oocytes inthe reproductive age. Mol. Hum. Reprod., 2, 951-958.

Nagley, P., Mackay, I.R., Baumer, A. et al. (1992)Mitochondrial DNA mutation associated with agingand degenerative disease. Ann. N.Y. Acad. Sci., 673,92-102.

Nishino, I., Spinazzola, A. and Hirano, M. (1999) Loss-of-function mutations in thymidine phosphorylaseare associated with MNGIE, a human mitochondrialdisorder Science, 283, 689-692.

Pallotti, R, Chen, X., Bonilla, E. and Schon, E.A. (1996)Evidence that specific mtDNA point mutations maynot accumulate in skeletal muscle during normalhuman aging. Am. J. Hum. Genet., 59, 591-602.

Penrose, L.S. (1933) The relative effects of paternal andmaternal age in mongolism. J. Genet., 44, 17-28.

Peters, H., Himelstein-Braw, R. and Faber, M. (1976)The normal development of the ovary in childhood.Acta Endocrinol, 82, 617-630.

Piko, L. and Taylor, K.D. (1987) Amounts ofmitochondrial DNA and abundance of somemitochondrial gene transcripts in early mouseembryos. Dev. Biol, 123, 364-374.

Reynier, P., Chretien, M.F., Savagner, F. et al. (1998)Long PCR analysis of human gamete mtDNA suggestsdefective mitochondrial maintenance in spermatozoaand supports the bottleneck theory for oocytes.Biochem. Biophys. Res. Commun., 252, 373—377.

Robinson, W.P., Kuchinka, B.D., Bernasconi, F. et al.(1998) Maternal meiosis I non-disjunction ofchromosome 15: dependence of the maternal ageeffect on level of recombination. Hum. Mol. Genet.,7, 1011-1019.

Rotig, A., Colonna, M., Bonnefont, J.P. et al. (1989)Mitochondrial DNA deletion in Pearson's marrow/pancreas syndrome. [Letter.] Lancet, i, 902-903.

Satoh, M. and Kuroiwa, T. (1991) Organization ofmultiple nucleoids and DNA molecules inmitochondria of a human cell. Exp. Cell Res., 196,137-140.

Schon, E.A., Rizzuto, R., Moraes, C.T. et al. (1989) Adirect repeat is a hotspot for large-scale deletions ofhuman mitochondrial DNA. Science, 244, 346-349.

Schon, E.A., Sciacco, M., Pallotti, F. et al. (1996)Mitochondrial DNA mutations and aging. In CellularAging and Cell Death, Wiley-Liss, Inc., New York,USA, pp. 19-34.

Schon, E.A., Bonilla, E. and DiMauro, S. (1997)Mitochondrial DNA mutations and pathogenesis. J.Bioenerg. Biomembr., 29, 131-149.

Sherman, S.L., Takaesu, N., Freeman, S.B. et al. (1991)Trisomy 21: association between reducedrecombination and nondisjunction. Am. J. Hum.Genet., 49, 608-620.

Simonetti, S., Chen, X., DiMauro, S. and Schon, E.A.(1992) Accumulation of deletions in humanmitochondrial DNA during normal aging: analysisby quantitative PCR. Biochim. Biophys. Acta, 1180,113-122.

Sirard, M.A. and Bilodeau, S. (1990) Effects of granulosacell co-culture on in vitro meiotic resumption ofbovine oocytes. J. Reprod. Fertii, 89, 459-465.

Suganuma, N., Kitagawa, T., Nawa, A. and Tomoda, Y.(1993) Human ovarian aging and mitochondrial DNAdeletion. Horm. Res. 39 (Suppl. 1), 16-21.

Tarin, JJ. (1996) Potential effects of age-associatedoxidative stress on mammalian oocytes/embryos. Mol.Hum. Reprod., 2, 717-724.

Tarin, JJ., Vendrell, F.J., Ten, J. et al. (1996) Theoxidizing agent tertiary butyl hydroperoxide inducesdisturbances in spindle organization, c-meiosis, andaneuploidy in mouse oocytes. Mol. Hum. Reprod., 2,895-901.

Tarin, JJ., Vendrell, F.J., Ten, J. and Cano, A. (1998)Antioxidant therapy counteracts the disturbing effectsof diamide and maternal ageing on meiotic divisionand chromosomal segregation in mouse oocytes. Mol.Hum. Reprod., 4, 281-288.

Tsafriri, A. and Pomerantz, S.H. (1984) Regulation ofthe development of meiotic competence and of theresumption of oocyte maturation in the rat. Symp.Soc. Exp. Biol., 38, 25^3.

Van Blerkom, J. (1991) Microtubule mediation ofcytoplasmic and nuclear maturation during the earlystages of resumed meiosis in cultured mouse oocytes.Proc. NatlAcad. Sci. USA, 88, 5031-5035.

Van Blerkom, J. and Runner, M.N. (1984) Mitochondrialreorganization during resumption of arrested meiosisin the mouse oocyte. Am. J. Anat., 171, 335-355.

Van Blerkom, J., Sinclair, J. and Davis, P. (1998)Mitochondrial transfer between oocytes: potentialapplications of mitochondrial donation and the issueof heteroplasmy. Hum. Reprod., 13, 2857-2868.

Volarcik, K., Sheean, L., Goldfarb, J. et al. (1998)The meiotic competence of in-vitro matured humanoocytes is influenced by donor age: evidence thatfolliculogenesis is compromised in the reproductivelyaged ovary. Hum. Reprod., 13, 154-160.

Warburton, D. and Kinney, A. (1996) Chromosomaldifferences in susceptibility to meiotic aneuploidy.Environ. Mol. Mutagen., 28, 237-247.

Yoon, P.W., Freeman, S.B., Sherman, S.L. et al. (1996)Advanced maternal age and the risk of Downsyndrome characterized by the meiotic stage of thechromosomal error: a population-based study. Am. J.Hum. Genet., 58, 628-633.

171


E.A.Schon et al.

Zaragoza, M.V., Millie, E., Redline, R.W. and Hassold,T.J. (1998) Studies of non-disjunction in trisomies, 2,7, 15, and 22: does the parental origin of trisomyinfluence placental morphology? J. Med. Genet., 35,924-931.

Zeviani, M., Moraes, C.T., DiMauro, S. et al. (1988)Deletions of mitochondnal DNA in Kearns-Sayresyndrome. Neurology, 38, 1339-1346.

Zeviani, M., Servidei, S., Gellera, C. et al. (1989) Anautosomal dominant disorder with multiple deletionsof mitochondrial DNA starting at the D-loop region.Nature, 339, 309-311.

Zhang, C , Baumer, A., Maxwell, R.J. et al. (1992)Multiple mitochondrial DNA deletions in an elderlyhuman individual. FEBS Lett., 297, 34-38.

172


Date post:	04-Jan-2017
Category:	Documents
Upload:	phamthuy
View:	219 times
Download:	0 times

Chromosomal non-disjunction in human oocytes: is there a ...

Documents