Do mitochondrial mutations cause recurrent miscarriage?
Advanced Access publication on March
Do mitochondrial mutations cause recurrent miscarriage?
Milja Kaare 2
Alexandra Go¨ tz 1
Veli-Matti Ulander 0
Sarah Ariansen 5
Risto Kaaja 0
Anu Suomalainen 1 4
Kristiina Aittoma¨ ki 2 3
0 Department of Obstetrics and Gynecology, Helsinki University Central Hospital , PO Box 140, 00029 HUS Helsinki , Finland
1 Research Program of Molecular Neurology, Biomedicum-Helsinki, University of Helsinki , Helsinki , Finland
2 Folkha ̈lsan Institute of Genetics, University of Helsinki , PO Box 63, FIN-00014 Helsinki , Finland
3 Department of Clinical Genetics, Helsinki University Central Hospital , PO Box 140, 00029 HUS Helsinki , Finland
4 Department of Neurology, Helsinki University Central Hospital, University of Helsinki , PO Box 63, 00014 Helsinki , Finland
5 Pathology Clinic , Molecular Pathology , Rikshospitalet Medical Center , N-0027 Oslo , Norway
The cause of recurrent miscarriage (RM) can be identified in 50% of cases, whereas in others, unknown genetic factors are actively being sought. As mitochondrial functions, and therefore also the mitochondrial genome [mitochondrial DNA (mtDNA)], have an important role in human development, through ATP production and participation in apoptosis, we aimed to study the role of mtDNA variations in RM. We screened 48 women with RM and 48 age-matched control women for heteroplasmic mitochondrial mutations using denaturing high performance liquid chromatography, a sensitive method that can detect 5% heteroplasmy. As a result, we detected a heteroplasmic mtDNA variation in 13 RM women (27%) and in 9 control women (19%). Seven synonymous and five non-synonymous changes were detected within coding regions. In addition, seven heteroplasmic variations were detected within the non-coding control region. We were also able to show the presence of the variations in eight placental samples from three heteroplasmic women. In three of these cases, the proportion of variant mtDNA was higher in the placenta compared with that in the mother. We conclude that our sensitive methodology revealed a higher frequency of samples with heteroplasmic variations than expected in women with both RM and controls. However, no apparent increased frequency of heteroplasmic mtDNA variations or amounts of aberrant mtDNA was detected in the RM group. In addition, none of the detected variations were previously known to be pathogenic and therefore they are an unlikely cause of miscarriage.
DHPLC / mitochondrial DNA variations / recurrent miscarriage
The human mitochondrial DNA (mtDNA) is a circular
doublestranded DNA, which contains 37 genes encoding 13 proteins of
the respiratory chain, 2 rRNAs and 22 tRNAs needed for
mitochondrial protein synthesis (DiMauro, 2001; Thorburn and Dahl, 2001).
The maternally inherited mitochondria are the most abundant
organelles in the human oocyte and early embryo. The main function of
the mitochondria is cellular oxidative ATP production, but the
organelle participates in a variety of other cellular functions, e.g. apoptosis,
which is of paramount importance in early embryonic development.
Structural or genetic dysfunctions that alter the capacity of
mitochondria to produce ATP by oxidative phosphorylation are likely to affect
early human development. In addition, mitochondrial dysfunctions,
changing the activation cascade of apoptosis, have been suggested
to cause oocyte wastage and early embryo demise in humans (Van
Today over 200 mtDNA point mutations have been reported in the
Mitomap database (http://www.mitomap.org) to be associated with a
wide variety of human diseases. The phenotype of diseases caused by
mtDNA mutations shows remarkable variation even within families
due to different levels of heteroplasmy (Ghosh et al., 1996; Blok
et al., 1997; Jacobi et al., 2001). The rapid changes in segregation of
mutations in mtDNA between generations have been explained by
an ‘mtDNA bottleneck’ in the germline, which is responsible for the
random segregation of mtDNA sequence variants between
generations. When the pathogenic threshold for a given tissue is surpassed,
the disease will manifest (DiMauro, 2001; Thorburn and Dahl, 2001;
McKenzie et al., 2004). Significant enrichment of a pathogenic
mtDNA mutation can occur between a mother and her offspring,
and complete allele switching has been observed in a single generation
(Koehler et al., 1991; Jacobi et al., 2001). Homoplasmy for a severe
pathogenic mtDNA mutation, however, is rarely observed,
presumably because of embryo lethality. Although a homoplasmic tRNA
& The Author 2009. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email:
mutation has been reported in a family with six neonatal deaths and a
miscarriage (McFarland et al., 2002), mitochondrial mutations for
diseases such as mitochondrial encephalopathy with lactic acidosis and
stroke-like episodes (MELAS) and myoclonic epilepsy and ragged-red
fibers (MERRF) are invariably heteroplasmic, and homoplasmic
mutations causing these syndromes would probably cause fetal
demise (Jenuth et al., 1996; Howell et al., 2000).
Recurrent miscarriage (RM), defined as the occurrence of three or
more consecutive pregnancy failures, is estimated to occur in 1 – 2% of
all couples. Approximately 50% of RM cases remain unexplained, and
a significant proportion of these is suggested to be caused by as yet
unknown genetic mechanisms (Carrington et al., 2005; Rai and
Regan, 2006). We hypothesized that mtDNA mutations could
constitute one of the unknown factors. This study was therefore designed to
explore involvement of mtDNA mutations in RM by screening women
with RM for heteroplasmic mutations in the mitochondrial genome
and by studying the identified variations in placental tissue, when
Materials and Methods
Women with RM treated at the Helsinki University Central Hospital
(HUCH), Department of Gynaecology and Obstetrics, were chosen for
this study. The women included in this study were (i) aged 18 – 40 years
(mean age 31 years) and had (ii) a history of three or more consecutive
miscarriages, (iii) a normal result on karyotyping and (iv) no known
uterine abnormalities (examined by ultrasonography or hysterosonogram).
A total of 48 women with unexplained RM were included. In 45 women,
all abortions had taken place during the first trimester (,13 weeks),
whereas two women had experienced a second trimester and
one women a third trimester ( 24 weeks) intrauterine fetal death.
In addition, at least one placental sample from a spontaneously
aborted pregnancy was available from seven of the women included in
The control group consisted of 48 age-matched unrelated healthy
women who had at least one normal pregnancy and no known history
of miscarriage. They were recruited from the same hospital during the
same time period as the patients and were of the same ethnic origin
(Finnish Caucasian) as the patients.
Informed consent was obtained from all participants prior to enrollment
in the study. The ethics committee of the Department of Obstetrics
and Gynaecology, Helsinki University Central Hospital, approved this
Polymerase chain reaction
DNA was extracted according to the manufacturer’s instructions from
whole blood samples and placental samples using Puregene DNA isolation
kits (Gentra Systems, Minneapolis, USA).
The mtDNA genome was amplified, in 40 previously reported
overlapping amplicons (Biggin et al., 2005), by PCR performed in 25 ml reaction
mixes containing the following reagents: 50 ng of genomic DNA, 1
PCR buffer, 2 nmol each dNTP, 10 pmol forward primer, 10 pmol
reverse primer and 0.1 U of AmpliTaq Gold DNA-polymerase (Applied
Biosystems, Foster City, USA). Initial denaturation at 958C for 10 min
was followed by 35 cycles of denaturation at 958C for 30 s, annealing at
568C for 45 s and extension at 728C for 45 s. A final extension was
performed at 728C for 10 min. Amplification products were checked by
agarose gel electrophoresis before further analysis.
Denaturing high-performance liquid
chromatography sample analysis
Denaturing high-performance liquid chromatography (DHPLC) analysis
was carried out using a Transgenomic WAVE Nucleic Acid Fragment
Analysis System (Transgenomic, Omaha, USA) and the associated
Navigator software. Before analyzing the samples with DHPLC, the PCR
products were denatured for 3 min at 958C, and then gradually re-annealed
by decreasing the temperature from 958C to 508C over a period of
50 min to enable the formation of heteroduplexes.
To obtain optimal resolution of homoduplex and heteroduplex DNA
fragments, the temperature was set for partially denaturing conditions.
The melting profile of the amplicons and the optimal oven temperature
were predicted for each PCR fragment using the Navigator software.
The buffer gradients for the elution of the fragments were created
automatically with the Navigator software. The analysis temperatures of the
fragments and time shifts used for DHPLC analysis for each amplicon
are given in Table I.
Following DHPLC analysis, samples showing heterozygous peaks were
sequenced in order to determine the nature of the sequence change.
PCR products were purified using exonuclease I and shrimp alkaline
phosphatase (USB Corporation, Cleveland, USA), and the purified products
were sequenced using BigDye version 3.1 sequencing chemistry and an
ABI 3730 DNA Analyzer (Applied Biosystems) according to the
Sequence analysis of samples with low-level heteroplasmic variations
either gave no clear result or could not detect the variations. To allow
sequence analysis of low-level heteroplasmic mtDNA, fractions with a
small-sized heteroduplex peak were collected using an FCW-200 fraction
collector integrated with the WAVE system. After re-amplification with the
primers used to produce the original amplicon, the heteroplasmy level in
the sample increased allowing sequence analysis.
For quantification of the heteroplasmy levels, we utilized solid-phase
minisequencing according to a previously described protocol (Suomalainen
and Syva¨nen, 2000). Specific primers were designed for each of the
variations studied (Table II). The PCR was performed in 50 ml reaction
mixes containing the following reagents: 50 ng of genomic DNA, 1
PCR buffer, 2 nmol each dNTP, 10 pmol biotinylated primer, 50 pmol
unbiotinylated primer and 1 U of Dynazyme II DNA-polymerase
(Finnzymes, Espoo, Finland). Initial denaturation at 958C for 4 min was followed
by 30 cycles of denaturation at 958C for 30 s, annealing at 588C for 45 s
and extension at 728C for 45 s. A final extension was performed at 728C
for 10 min.
Quantitative real-time PCR
The mitochondrial cytochrome b in mtDNA and the nuclear beta-amyloid
precursor protein (APP) genes were simultaneously amplified by
quantitative TaqMan real-time PCR assay in the ABI Prism 7000 Detection
System Cycler to quantify the amount of mtDNA. The primers and
probes used have been previously described (Go¨ tz et al., 2008).
Genomic DNA (25 ng) was amplified using TaqMan Universal PCR
Master Mix (Roche, Applied Biosystems), 900 nM of each primer and
250 nM of each probe in a 30 ml volume in following conditions: an
enzyme activation step for 2 min at 508C, followed by denaturing for
10 min at 958C, and 40 amplification cycles of 15 s at 958C, and 1 min
at 608C. Fluorescence acquisition was done at the end of each annealing
step, and the relative amounts of mtDNA and the nuclear gene APP were
Table I The size of the analyzed amplicons and the
temperatures and time shifts used for DHPLC analysis
Amplicon Region Size DHPLC temperatures
(bp) (8C) [time shift (min)]
1 16 336–283 517 56, 60 (0.5)
2 16 496–452 526 58.5
3 361–921 561 58, 60 (0.5)
4 756–1425 670 59
5 1234–1769 536 58.5
6 1587–2216 630 56.5, 58 (0.5)
7 2105–2660 556 57
8 2417–3063 647 58, 59.5
9 2986–3460 474 58.2
10 3397–3957 561 58, 60.5
11 3815–4262 448 57, 58
12 4150–4682 533 57.5
13 4656–5171 516 55, 57
14 5097–5555 459 58.5
15 5468–5933 466 58.5
16 5862–6381 520 60
17 6361–6846 486 57
18 6744–7255 512 59
19 7178–7728 551 56.5, 58 (0.5)
20 7645–8215 571 59.5
21 8205–8669 465 56.5
22 8640–9102 463 56, 59 (0.5)
23 8900–9365 466 59
24 9310–9878 569 57.5, 59
25 9754–10 275 522 55, 58.5 (1.0)
26 10 127–10 556 430 54.5, 56.5 (0.5)
27 10 291–10 764 474 55, 56.5
28 10 601–11 161 561 58.5 (0.5)
29 11 142–11 713 572 58.5
30 11 622–12 212 591 58.5
31 12 174–12 755 582 57
32 12 601–13 166 566 59 (1.0)
33 13 099–13 580 482 58.5
34 13 428–13 947 520 57.5, 59
35 13 715–14 388 674 57
36 14 046–14 583 538 55, 57 (0.5)
37 14 434–14 996 563 57
38 14 924–15 423 500 59
39 15 260–15 774 515 58
40 15 733–16 355 623 56, 57.5 (0.5)
DHPLC, denaturing high performance liquid chromatography.
compared in the exponential phase of PCR, using the ABI Prism 7000 SDS
software (Applied Biosystems). The ratio of mtDNA to nuclear DNA was
used as a measure of mtDNA content (22DDCT method) (Livak and
Schmittgen, 2001), with triplicates for each sample and two individual
runs per sample.
In screening of 48 RM women and 48 age-matched control women for
heteroplasmic mtDNA variations using DHPLC, we detected
altogether 19 different variations in these samples. At least one
heteroplasmic mtDNA variation was detected in 13 RM women (27%)
and in 9 control women (19%). Of the detected 19 variations, 18
have previously been reported as polymorphic variations in the
Mitomap or mtDB database, whereas the variation 14653C.T, a
synonymous change in the ND6 gene, has not been previously
reported. The location and the predicted amino acid changes of all
the detected variations are listed in Table III. One of the detected
variations was an insertion, whereas all others were transitions of 1 bp.
No heteroplasmic variations were detected within the mitochondrial
tRNA and mitochondrial rRNA regions.
Altogether 12 different variations were detected within
mitochondrial protein coding genes. Seven of these were synonymous and
five were non-synonymous changes. A non-synonymous change was
detected in four RM women and four control women. The
heteroplasmy levels of the non-synonymous variations were studied using
solid-phase minisequencing (Table IV). The level of heteroplasmy
varied between 5% and 95% in the samples with a non-synonymous
variation. The frequencies of these variations are given in Table IV.
Seven heteroplasmic variations were detected within the
noncoding control region. To study if the variations occurring in the
control region of mtDNA were affecting the mtDNA copy number,
the mtDNA levels in samples with a variation in the control region
(n ¼ 8) were analyzed using real-time quantitative PCR and compared
to the mtDNA levels in samples with no heteroplasmic mtDNA
variations (n ¼ 6). In addition, the mtDNA level of one placental sample
with a control region variation was compared with the mtDNA levels
in placental samples without a control region variation (n ¼ 6). We did
not detect significant differences (P . 0.05) in the mtDNA copy
number between samples with a control region variation and
control samples with no variations (Fig. 1).
Placental samples were available from three women with a
heteroplasmic mtDNA variation, namely 195T.C (one sample),
11404A.G (three samples) and 14653C.T (four samples). By
analyzing these eight placental samples using DHPLC, we were able to
show the presence of the variations in the placenta. The heteroplasmy
level of these variations was studied using solid-phase minisequencing
(Table IV). In three cases, the proportion of the mtDNA variation was
increased in the placenta compared with the mother.
Recent studies suggest that mitochondrial dysfunction in the oocyte
may be a critical determinant of the developmental competence of
an early human embryo. Accordingly, disturbances in the capacity
of mitochondria to produce ATP or to activate the apoptosis
cascade have been suggested to be possible causes of early human
embryo wastage. Low mtDNA copy number or point mutations
affecting the ATP level in the oocyte or early embryo may result in
defects in chromosomal segregation or developmental arrest during
preimplantation stages. As mitochondrial replication does not occur
before implantation, the developmental competence of the early
blastocyst is dependent upon normal function and ATP production of the
Table II Primers used for minisequencing
Table III The base substitutions, predicted amino acid changes and the location of the heteroplasmic variations detected
in patients and control women
oocyte mitochondria. Mitochondrial dysfunctions affecting ATP
production during early developmental stages may be lethal for the
embryo, even if only a portion of the mitochondria are dysfunctioning.
Consequently, it is possible that some mtDNA mutations cause
developmental arrest already before the pregnancy is clinically
recognized (Van Blerkom et al., 1998; Van Blerkom, 2004).
However, if development is arrested after the pregnancy is
recognized, then the mitochondrial genes could be considered as candidate
genes for miscarriage. To study the role of mtDNA mutations in RM,
we screened women with RM for possible pathogenic mtDNA
variations. We hypothesized that a generational shift in the heteroplasmy
toward a higher level of variant mtDNA could explain a proportion
We detected altogether 19 different variations when screening 48
RM women and 48 age-matched control women for heteroplasmic
mtDNA variations using DHPLC. At least one heteroplasmic
mtDNA variation was detected in 13 RM women (27%) and in 9
control women (19%). The percentage of individuals with
heteroplasmic variations was unexpectedly high as the presence of two different
mtDNA populations in an individual is reported to be rare, and in
general, heteroplasmic variations are expected to be pathogenic
(Jenuth et al., 1996; DiMauro and Schon, 2001; McKenzie et al.,
2004). However, Grzybowski (2000) has suggested that the high
level of homoplasmy might be over-estimated and that low level
heteroplasmy may be more common than suspected. This could be due
to the fact that many studies have used sequencing, which cannot
Table IV Heteroplasmy levels of mtDNA variations
and the reported frequency of the variation
*According to mtDB.
detect low levels of variant mtDNA sequences. The fact that we
detected heteroplasmy in so many control women is most likely
due to the sensitivity of the DHPLC method. In 8 out of 19
samples (42%), in which the level of heteroplasmy was tested, the
ratio of variant sequences was ,10% or .90%, and this could not
be detected by direct sequencing. However, even though the
sensitivity and specificity of the DHPLC method are reported to exceed
96% (Xiao and Oefner, 2001), and the method is able to detect
heteroplasmy levels under 5% (Biggin et al., 2005; Lim et al., 2008), there
are also instances in which mutations may not be detected. Mutations
located in GC-rich areas in fragments with otherwise normal
nucleotide ratios may in some cases be difficult to detect, and some
variations can only be detected at specific temperatures (Xiao and
Oefner, 2001). Consequently, it is possible that some heteroplasmic
mutations may have gone undetected in our sample series.
Non-synonymous changes, predicted to change an amino acid,
were detected in four RM women. These variations have been,
however, previously reported as homoplasmic polymorphisms with
a frequency of 0.037 – 22.6% according to the mtDNA sequences
reported in mtDB and are therefore likely to be neutral. One of the
detected coding region variations is novel, not previously reported
in the Mitomap or in mtDB database. This variation, 14653C.T, in
the ND6 gene is, however, unlikely to be pathogenic as it is not
changing an amino acid. In addition, this variation was detected at a
heteroplasmy level of 85% in one otherwise healthy RM patient.
Figure 1 mtDNA copy number analysis by quantitative PCR
(Q-PCR) in cases with mtDNA control region variations. Lymphocyte
samples of women with RM (n ¼ 5) and controls (n ¼ 3) and one
placental sample from an aborted pregnancy with at least one variation
in the mtDNA control region were compared with control
lymphocyte (n ¼ 6) and placental samples (n ¼ 6) without mtDNA
variations. The mtDNA copy number of one of the controls in each
pool was set as 100% reference point. The results are average
values of two Q-PCR experiments, each with triplicate samples.
Consequently, we conclude this novel variation to be neutral. The
remaining synonymous variations in the coding region are all previously
reported to be polymorphisms and are therefore not likely to cause
Analysis of the mtDNA control region revealed seven
heteroplasmic variations among the studied samples. These have previously
been reported as polymorphisms. However, point mutations in the
control region may affect replication of the mitochondrial genome,
which could be an important factor for the outcome of a pregnancy.
It has been reported that the localization of mitochondria in the
cleavage state embryo is strictly regulated and that the mitochondrial
distribution may play a role in defining the long-term viability of a
blastomere and embryonic patterning. Consequently, mtDNA copy
number may be an important determinant of embryo survival (Van
Blerkom et al., 2000; Van Blerkom, 2004; Dumollard et al., 2007).
To test if the control region variations affected mtDNA replication,
we analyzed the mtDNA copy number in these samples. No
significant differences in the mtDNA copy number were shown.
The inheritance of heteroplasmic mtDNA variations could be
studied in three women from whom eight placental samples were
available. In three cases, these variations (Table IV) were present in
a slightly larger proportion in the placenta compared with maternal
blood, and in the remaining cases, the proportion was lower. The
result is similar to previous studies showing variable levels of
heteroplasmic variations in the offspring (Ghosh et al., 1996; Blok et al.,
1997). The mtDNA bottleneck enables dramatic changes in the
level of mutant mtDNA in the fetus as the mtDNA genotype in the
embryo is largely determined by the genotype of the few mitochondria
in the primordial germ cell (Jenuth et al., 1996; Shoubridge, 2000).
In conclusion, we studied the role of mtDNA mutations in miscarriages
by screening women with RM for heteroplasmic mtDNA variations. As a
result, we detected altogether 19 different variations in these samples.
The percentage of samples with heteroplasmic variations was
unexpectedly high as the presence of two different mtDNA populations in an
individual has previously been reported to be rare. A non-synonymous
change was detected in four RM patients, and a heteroplasmic variation
within the control region was detected in seven RM patients, but no
evidence for accumulation of pathogenic mtDNA variations in RM patients
was found. The detected variations were judged to be polymorphisms,
and not likely to contribute to the miscarriages experienced by the
studied women. This, however, does not exclude the role of other
heteroplasmic mtDNA variations in miscarriage.
We thank all couples with RM for participating in this study.
This work was supported by the Sigrid Juselius Foundation and by a
Finnish State Grant.
Biggin A , Henke R , Bennetts B , Thorburn DR , Christodoulou J. Mutation screening of the mitochondrial genome using denaturing highperformance liquid chromatography . Mol Genet Metab 2005 ; 84 : 61 - 74 .
Blok RB , Gook DA , Thorburn DR , Dahl HH . Skewed segregation of the mtDNA nt 8993 (T - .G) mutation in human oocytes . Am J Hum Genet 1997 ; 60 : 1495 - 1501 .
Carrington B , Sacks G , Regan L. Recurrent miscarriage: pathophysiology and outcome . Curr Opin Obstet Gynecol 2005 ; 17 : 591 - 597 .
DiMauro S. Lessons from mitochondrial DNA mutations . Semin Cell Dev Biol 2001 ; 12 : 397 - 405 .
DiMauro S , Schon EA . Mitochondrial DNA mutations in human disease . Am J Med Genet 2001 ; 106 : 18 - 26 .
Dumollard R , Duchen M , Carroll J. The role of mitochondrial function in the oocyte and embryo . Curr Top Dev Biol 2007 ; 77 : 21 - 49 .
Ghosh SS , Fahy E , Bodis-Wollner I , Sherman J , Howell N. Longitudinal study of a heteroplasmic 3460 Leber hereditary optic neuropathy family by multiplexed primer-extension analysis and nucleotide sequencing . Am J Hum Genet 1996 ; 58 : 325 - 334 .
Go¨ tz A, Isohanni P , Pihko H , Paetau A , Herva R , Saarenpa¨a¨-Heikkila¨ O, Valanne L , Marjavaara S , Suomalainen A. Thymidine kinase 2 defects can cause multi-tissue mtDNA depletion syndrome . Brain 2008 ; 131 : 2841 - 2850 .
Grzybowski T. Extremely high levels of human mitochondrial DNA heteroplasmy in single hair roots . Electrophoresis 2000 ; 21 : 548 - 553 .
Howell N , Chinnery PF , Ghosh SS , Fahy E , Turnbull DM. Transmission of the human mitochondrial genome . Hum Reprod 2000 ; 15 (Suppl. 2): 235 - 245 .
Jacobi FK , Leo-Kottler B , Mittelviefhaus K , Zrenner E , Meyer J , Pusch CM , Wissinger B. Segregation patterns and heteroplasmy prevalence in Leber's hereditary optic neuropathy . Invest Ophthalmol Vis Sci 2001 ; 42 : 1208 - 1214 .
Jenuth JP , Peterson AC , Fu K , Shoubridge EA . Random genetic drift in the female germline explains the rapid segregation of mammalian mitochondrial DNA . Nat Genet 1996 ; 14 : 146 - 151 .
Koehler CM , Lindberg GL , Brown DR , Beitz DC , Freeman AE , Mayfield JE , Myers AM . Replacement of bovine mitochondrial DNA by a sequence variant within one generation . Genetics 1991 ; 129 : 247 - 255 .
Lim KS , Naviaux RK , Wong S , Haas RH. Pitfalls in the denaturing high-performance liquid chromatography analysis of mitochondrial DNA mutation . J Mol Diagn 2008 ; 10 : 102 - 108 .
Livak KJ , Schmittgen TD . Analysis of relative gene expression data using real-time quantitative PCR and the 2(- Delta Delta C(T)) method. Methods 2001 ; 25 : 402 - 408 .
McFarland R , Clark KM , Morris AA , Taylor RW , Macphail S , Lightowlers RN , Turnbull DM . Multiple neonatal deaths due to a homoplasmic mitochondrial DNA mutation . Nat Genet 2002 ; 30 : 145 - 146 .
McKenzie M , Liolitsa D , Hanna MG . Mitochondrial disease: mutations and mechanisms . Neurochem Res 2004 ; 29 : 589 - 600 .
Rai R , Regan L. Recurrent miscarriage . Lancet 2006 ; 368 : 601 - 611 .
Shoubridge EA . Mitochondrial DNA segregation in the developing embryo . Hum Reprod 2000 ; 15 (Suppl. 2): 229 - 234 .
Suomalainen A , Syva¨nen AC. Quantitative analysis of human DNA sequences by PCR and solid-phase minisequencing . Mol Biotechnol 2000 ; 15 : 123 - 131 .
Thorburn DR , Dahl HH . Mitochondrial disorders: genetics, counseling, prenatal diagnosis and reproductive options . Am J Med Genet 2001 ; 106 : 102 - 114 .
Van Blerkom J. Mitochondria in human oogenesis and preimplantation embryogenesis: engines of metabolism, ionic regulation and developmental competence . Reproduction 2004 ; 128 : 269 - 280 .
Van Blerkom J , Sinclair J , Davis P. Mitochondrial transfer between oocytes: potential applications of mitochondrial donation and the issue of heteroplasmy . Hum Reprod 1998 ; 13 : 2857 - 2868 .
Van Blerkom J , Davis P , Alexander S. Differential mitochondrial distribution in human pronuclear embryos leads to disproportionate inheritance between blastomeres: relationship to microtubular organization , ATP content and competence . Hum Reprod 2000 ; 15 : 2621 - 2633 .
Xiao W , Oefner PJ . Denaturing high-performance liquid chromatography . Hum Mutat 2001 ; 17 : 439 - 474 .