Mitochondrial point mutations do not limit the natural
lifespan of mice
Marc Vermulst1, Jason H Bielas1, Gregory C Kujoth2, Warren C Ladiges3, Peter S Rabinovitch1,
Tomas A Prolla2& Lawrence A Loeb1
Whether mitochondrial mutations cause mammalian aging,
or are merely correlated with it, is an area of intense debate1.
Here, we use a new, highly sensitive assay2to redefine the
relationship between mitochondrial mutations and age. We
measured the in vivo rate of change of the mitochondrial
genome at a single–base pair level in mice, and we
demonstrate that the mutation frequency in mouse
mitochondria is more than ten times lower than previously
reported. Although we observed an 11-fold increase in
mitochondrial point mutations with age, we report that a
mitochondrial mutator mouse3was able to sustain a 500-fold
higher mutation burden than normal mice, without any
obvious features of rapidly accelerated aging. Thus, our results
strongly indicate that mitochondrial mutations do not limit the
lifespan of wild-type mice.
Mitochondria are functionally diverse organelles with a central role in
both oxidative phosphorylation4and apoptosis5. Accordingly, post-
mitotic tissues with high rates of oxygen consumption, such as brain
and heart, are exceptionally sensitive to mitochondrial dysfunction6.
The mitochondrial theory of aging postulates that a lifelong accumu-
lation of mitochondrial DNA (mtDNA) mutations in multiple tissues
eventually results in mitochondrial failure, which, in synergy with
downstream processes such as apoptosis, results in loss of cellularity
and the progressive decline of tissue functioning known as aging3,7.
These mutations may compromise the integrity of the electron
transport chain and increase the formation of reactive oxygen species
(ROS), creating a vicious cycle of mutagenesis that continuously
amplifies the production of cytotoxic oxygen radicals8,9.
Although several studies have quantified either tissue-specific muta-
tion frequencies10or mutation accumulation as a function of age11,
most methods rely heavily on PCR and cloning-based strategies12,13,
techniques that are limited in throughput and that may be con-
founded by polymerase infidelity on damaged templates and by
cloning artifacts. Therefore, we explored the relationship between
age and mutation accumulation in mtDNA with an adaptation
of the random mutation capture (RMC) assay2,14, a quantitative
PCR-based approach that relies on PCR amplification of single
molecules for mutation detection (Supplementary Fig. 1 online)
but is not limited by polymerase fidelity2,14. This methodology allows
for exact determination of mutation frequencies in high-throughput
screens that interrogate millions of base pairs simultaneously2,14.
Approximately 150 million bp were screened for mutation detection
in this study.
ATaqI restriction site (TCGA) located in the gene encoding the 12S
rRNA subunit (bp 634–637) was selected for mutation frequency
determination. Examination of mtDNA derived from brain tissue of
four young mice (1–3 months) yielded an average mutation frequency
of 6.0 ? 10?7± 0.9 ? 10?7per bp (Fig. 1a). A gradual increase in
mutation burden was observed with advancing age, with mutations
accumulating to a mean frequency of 1.1 ? 10?5± 0.3 ? 10?5per bp
in old animals (24–33 months). Mutations started accumulating
rapidly after 16 months of age, resulting in a mutation frequency
that was ten times higher in old animals (24–33 months) compared
with young animals (1–10 months) (P ¼ 0.0026, two-tailed t-test,
Fig. 1b). This accumulation was fitted best along an exponential line.
We obtained similar results in heart tissue from the same group
of animals, with frequencies ranging from 8.5 ? 10?7per bp in a
2.5-month-old animal to 1.3 ? 10?5per bp in a 33-month-old animal
(Fig. 1a). However, subtle fluctuations from these averages may occur
between cell types within tissues. Mutation frequencies determined via
the RMC assay in young mice are approximately one to two orders
of magnitude lower than previously documented by conventional
methods7,13. Although the RMC assay has an increased level of
sensitivity over these methods, it registers mutational events at the
interrogated restriction site only; thus, it is possible that local sequence
context or negative selection at this locus accounts for this discre-
pancy. However, interrogation of a second TaqI restriction site, located
at bp 15253–15256 in the mitochondrial cytochrome b gene, yielded
similar results: the mutation frequency was 1.2 ? 10?6in 2.5-month-
old animals but rose to 7.9 ? 10?6in a 33-month-old animal, sug-
gesting that our results can be extended beyond the local genetic
environment of any one site (data not shown). Furthermore,
in parallel experiments, we observed mutation frequencies of up to
Received 19 October 2006; accepted 25 January 2007; published online 4 March 2007; doi:10.1038/ng1988
1Department of Pathology, University of Washington, Seattle, Washington 91895, USA.2Departments of Genetics and Medical Genetics, University of Wisconsin,
Madison, Wisconsin 53706, USA.3Department of Comparative Medicine, University of Washington, Seattle, Washington 98195, USA. Correspondence should be
addressed to L.A.L. (firstname.lastname@example.org).
540VOLUME 39 [ NUMBER 4 [ APRIL 2007 NATURE GENETICS
© 2007 Nature Publishing Group http://www.nature.com/naturegenetics
1.5 ? 10?3at both restrictionsites in brain tissue and mouse
embryonic fibroblasts (MEFs) derived from 2.5-month-old animals
completely deficient in the proofreading activity of DNA polymerase g
(Polg), the mitochondrial replicative enzyme. Because these animals
are healthy at this age and cell lines are viable despite a B2,500-fold
increase in mutation burden, these data strongly suggest that negative
selection is not an influential factor in determining mutation
load at either site. Nevertheless, we cannot exclude the possibility
that mutational hotspots that lie outside of our mutational target
could raise the overall mutation frequency per bp reported
here. However, the contribution of hotspots would be mitigated
by their rareness and diluted by the presence of large regions of
uniform mtDNA. Thus, we expect their influence to the mutation
frequency per bp to be modest, and we conclude that our results
indicate that the mutation burden in mitochondria of wild-type mice
is more than ten times lower than previously reported. This discre-
pancy is most likely to be the result of mutations introduced ex vivo
on (damaged) DNA templates during PCR before cloning steps in
conventional assays, as PCR amplification, prior to application
of the RMC assay, increased the mutation frequency at least 32-fold
(Supplementary Fig. 2 online). In contrast, the RMC assay does not
seem to be substantially influenced by DNA damage (Supplementary
Fig. 3 online).
mtDNA is located in the vicinity of the electron transport chain, the
primary site of ROS production. To investigate whether oxidative
damage is responsible for mutation acquisition, we measured the
mutation load in MEFs and heart tissue of transgenic animals carrying
an extra gene that targets human catalase (mCAT), an ROS scavenger,
to the mitochondrion15. In the hearts of three 26- to 28-month old
mice that showed stable expression of the mCAT gene, we measured
an average mutation frequency of 1.4 ? 10?6± 0.1 ? 10?6, whereas
five age-matched wild-type animals showed an average mutation
frequency of 4.0 ? 10?6± 1.0 ? 10?6(P ¼ 0.036, Mann-Whitney
test, Supplementary Fig. 4 online). In addition, we measured a
mutation frequency of 1.3 ? 10?6in primary MEFs of a wild-type
animal but did not find any mutations in a screen of 6.3 million bp
from a similar primary cell line derived from an mCAT littermate. We
observed a marked difference in the mutation spectrum of wild-type
and mCAT-expressing animals (Supplementary Fig. 5 online). Col-
lectively, these findings are in agreement with extensive literature that
proposes a substantial role for oxidative
damage in mitochondrial mutagenesis16.
A compilation of sequencing data collected
from wild-type animals identified 81% of all
mutation events as GC to AT transitions
(Fig. 2), suggesting that a limited number
of sources is responsible for mutation acqui-
sition. The detection of five deletions, ran-
ging from 1 to 80 bp, and one insertion
(Fig. 2 and Supplementary Fig. 5), attests
to the versatility of the RMC assay. GC to AT
transitions are the most commonly observed
mutations after oxidative stress17and are
thought to be produced predominantly
through cytosine oxidation followed by
rapid decomposition of the destabilized base
into uracil glycol and 5-hydroxyuracil, both
of which mispair with adenine18. Repair of
these lesions creates abasic sites, repair inter-
mediates that are preferentially paired with
adenine during Polg-mediated synthesis19.
However, an altered pattern of mutagenesis in proofreading-deficient
Polg mice argues against a substantial role for Polg misinsertions in
the absence of DNA damage (Supplementary Fig. 5). The muta-
tion spectrum remained constant between three restriction sites
(bp 634–637, 7667–7680 and 15253–15256) and was independent of
age or tissue type (brain or heart). However, even though the TaqI site
is a palindrome, the mutations were not symmetrically divided over all
4 bp (Supplementary Fig. 5); this asymmetry may occur if the
mutation rate of the two DNA strands is not equal20. Because DNA
oxidation and deamination occur faster on single-stranded DNA than
on double-stranded DNA21, it is possible that temporary single-
strandedness of mtDNA during replication is a driving force in
mitochondrial mutagenesis20. Finally, we did not find any evidence
for clonal expansions during spectrum analysis, consistent with the
post-mitotic nature of the tissues surveyed.
To directly evaluate the hypothesis that the accumulation of muta-
tions in mtDNA is a causal factor in the aging process, we compared
the mutation burden of wild-type mice with that of mice that contain
Mutation frequency/bp (× 10–6)
Mutation frequency (× 10–6)
R2 = 0.6997
R2 = 0.8377
Figure 1 Frequency of mitochondrial mutations as a function of age. (a) Mutation frequency was
determined at TaqI restriction site 634–637 in brain and heart. Each data point represents one animal.
On average, 3 ? 106bp were screened per animal. (b) Mutation burden (mean ± s.e.m.) of young
animals (1–10 months, n ¼ 7) and old animals (24–33 months, n ¼ 6). ** P o 0.01, two-tailed
Figure 2 Mitochondrial mutation spectrum in wild-type animals. Mutation
spectrum was composed of mutations captured at three restriction sites
(634–637, 7667–7670 and 15253–15256) in both brain and heart.
Absolute numbers of mutations captured are listed inside the bars. Indels:
inversions and deletions.
NATURE GENETICS VOLUME 39 [ NUMBER 4 [ APRIL 2007541
© 2007 Nature Publishing Group http://www.nature.com/naturegenetics
an allelic substitution in the exonuclease domain of Polg. This
substitution abolishes the proofreading activity of Polg, causing it to
be error prone. We3and others22have recently demonstrated that
mice homozygous for this substitution (Polgmut/mut) show a marked
reduction in lifespan and several features of premature aging. This
phenotype could be attributed to an increased rate of mtDNA
mutation, which we initially reported to be three- to eightfold higher
in homozygous mutant mice compared with wild-type mice,
as measured by a standard DNA sequencing approach. Here, we
correct this to be B2,500-fold. Because our measurements in
young Polgmut/mutmice (1.5 ? 10?3± 0.2 ? 10?3) are equivalent to
what was previously reported3,22with conventional assays, this differ-
ence can be attributed solely to the increased sensitivity of the RMC
assay, which allowed us to correctly determine the very low mutation
frequency in wild-type mice. Thus, a key finding of this study is
that the exonuclease domain of Polg is a principal caretaker of
Heterozygous animals (Polg+/mut) did not show a statistically
significant reduction in mean lifespan (P ¼ 0.875, Fig. 3), consistent
with previous reports3,22. Additionally, no significant increase in age-
related pathology has thus far been detected in these animals22(T.A.P.,
unpublished data). We re-examined the mutation load of the Polg+/mut
mice with the RMC assay. Notably, we recorded an average muta-
tion frequency of 3.3 ? 10?4± 0.9 ? 10?4per bp (Fig. 4a) in brain
tissue of young (2- to 3-month-old) Polg+/mutanimals at the 12S
rRNA locus—approximately 500 times higher than in age-matched
wild-type animals (P ¼ 0.008). Notably, this mutation burdenwas also
29 times higher than the burden in old wild-type animals (24–33
months, Fig. 4a, P ¼ 0.001). The lack of a classic mismatch repair
mechanism in mouse mitochondria23probably allows for this vast
increase in mitochondrial mutagenesis in an exonuclease-deficient
background. Because the frequency of intracellular expansions of
mtDNA mutations depends directly on the mutation rate24, these
data suggest that both heteroplasmic and homoplasmic mutations are
markedly higher in Polg+/mutcells than in wild-type cells. We con-
firmed these results in heart tissue of three young Polg+/mutmice,
which showed an average mutation frequency of 1.6 ? 10?4± 0.5 ?
10?4, approximately 220 times higher than in four young wild-type
animals, which carried a mutation burden of 7.1 ? 10?7± 0.1 ? 10?7
(P ¼ 0.0175) and 30 times higher than in six old wild-type animals
with a frequency of 5.4 ? 10?6± 1.7 ? 10?6(Fig. 4b, P ¼ 0.004). This
observation was strengthened by similar results from a probe of three
additional TaqI restriction sites, randomly dispersed throughout the
mitochondrial genome in both brain (Supplementary Fig. 6 online)
and heart (data not shown).
As the mutation frequency is determined at the same loci in
all genotypes, these measurements unambiguously describe the
relative differences in mutation rate between them. Because hetero-
zygous mice are born with a 30-fold higher mutation burden than the
oldest wild-type animals without suffering a phenotype that resembles
premature aging, we conclude that the threshold at which mitochon-
drial mutations become limiting for lifespan is unlikely to be reached
in wild-type mice. Notably, heterozygous carriers of human variants
of polymerase g (POLG) are asymptomatic as well, suggesting
that there may be many healthy human carriers of POLG alleles that
harbor large numbers of undetected mtDNA mutations. Although
our data strongly argue against a causal role for mitochon-
drial mutations in natural aging, it should be noted that despite our
ability to detect small deletions, large mtDNA deletions are not
detected by our methods. Large mtDNA deletions have been
correlated with the demise of certain specialized tissues such
as the substantia nigra25,26. However, twinkle transgenic mice,
which have an organism-wide increase in the accumulation of large
deletions, do not show a decrease in lifespan or a premature
The mutation burden rapidly grew after 16 months of age, around
the time when reproduction has ceased and the average life expectancy
of these animals in the wild has expired. As ROS seem to be the
primary source of mutagenesis in mitochondria, this study is con-
sistent with the hypothesis that the production of ROS increases with
age28and that mtDNA is a primary target of ROS, but that the most
prevalent result of oxidative lesions, point mutations, are extremely
rare and do not determine the rate of aging of wild-type mice.
Tissue homogenization and organelle separation. All tissues were harvested
within 5 min of death. All animals were cared for according to approved
guidelines at the University of Washington. Tissues were sliced in pieces with a
scalpel and rinsed in 1? PBS before homogenization in a Dounce-type glass
Figure 3 Kaplan-Meier survival curves. Median survival is 423 d for
Polgmut/mutmice, 758 d for Polg+/mutmice and 864 d for wild-type mice.
The curves for wild-type and Polg+/mutmice do not differ statistically
significantly (log rank test, P ¼ 0.875).
Figure 4 Mutation burden in wild-type and Polg exonuclease–deficient
mice. (a,b) Mutation frequency (mean ± s.e.m.) was determined at TaqI
restriction site 634–637 in brain (gray) and heart (red) and is plotted on
a logarithmic scale. For brain tissue, young wild-type (WT) animals are
1–3 months of age (n ¼ 4); old wild-type animals, 24–33 months of age
(n ¼ 6); young Polg+/mutanimals, 2.5 months of age (n ¼ 3) and young
Polgmut/mutanimals, 2.5 months of age (n ¼ 2). * P o 0.05. ** P o 0.01.
For heart tissue, young wild-type animals are 1–16 months of age (n ¼ 4),
old wild-type animals are 24–33 months of age (n ¼ 6) and young Polg+/mut
animals are 2.5 months of age (n ¼ 3).
542VOLUME 39 [ NUMBER 4 [ APRIL 2007 NATURE GENETICS
© 2007 Nature Publishing Group http://www.nature.com/naturegenetics
homogenizer with 25 firm strokes of a hand-driven glass pestle. Homogeniza-
tion buffer contained 0.075 M sucrose, 0.225 M sorbitol, 1 mM EGTA, 0.1%
fatty acid–free BSA and 10 mM Tris HCl (pH 7.8). Differential centrifugation
was performed as described previously12to obtain a crudely purified mito-
DNA extraction and restriction digest. Mitochondrial pellets were digested for
1 h in a buffer containing 0.2 mg/ml proteinase K (Sigma), 0.75% SDS, 0.01 M
Tris HCl, 0.15 M NaCl and 0.005 M EDTA at pH 7.8. mtDNAwas subsequently
isolated using phenol-chloroform extraction (1:1, vol/vol) followed by ethanol
precipitation. mtDNAwas then diluted and digested in a specialized TaqI buffer
(New England Biolabs) in the presence of 100 units TaqaI (New England
Biolabs) and 1? BSA (New England Biolabs) for 10 h, with the addition of
100 units of TaqaI per h.
Mutation detection. PCR was performed using a DNA Engine Opticon
Monitor 2, a continuous fluorescence detection system (BioRad), and
amplicons were visualized with Stratagene’s Brilliant SYBR Green qPCR
Master Mix. Primers used for amplification are listed in Supplementary
Table 1 online.
All real-time qPCR reactions were performed in 25-ml reactions containing
1? Brilliant SYBR Green qPCR Master Mix from Stratagene, 20 pmol forward
and reverse primers and 2 units uracil DNA glycosylase (UDG). The samples
were amplified as follows: UDG incubation at 37 1C for 10 min and 95 1C for
10 min followed by 45 cycles of 95 1C for 30 s, 60 1C for 1 min and 72 1C for
1.5 min. Samples were held at 72 1C for 5 min and then immediately stored
at –20 1C. All PCR products were either (i) incubated with TaqaI and then
verified by agarose gel electrophoresis to be insensitive to digestion or (ii)
isolated with the QiaQuick PCR Purification Kit (Qiagen) and sequenced to
identify the mutation at the TaqI recognition site.
Cell culture. Mouse embryonic fibroblasts were cultured in DMEM
(Gibco-BRL) containing 10% (vol/vol) FBS (HyClone), 1% L-glutamine and
1% penicillin-streptomycin (Gibco). An atmosphere of 5% CO2was main-
tained in a humidified incubator at 37 1C. mCATand corresponding wild-type
cell lines were grown with 2% oxygen.
Note: Supplementary information is available on the Nature Genetics website.
This work was supported by US National Institutes of Health grants AG001751
(L.A.L., P.S.R.), CA102029 (L.A.L.), ES11045 (L.A.L., W.C.L.) and AG021905
(T.A.P., G.C.K.). J.H.B. was supported by a research fellowship from the Canadian
Institutes of Health. The authors thank G.M. Martin, R.S. Mangalindan,
R.N. Venkatesan and C.-Y. Chen for editing this manuscript, technical assistance
M.V. carried out all the experiments described and wrote the paper. M.V., J.H.B.
and L.A.L. conceived the project. G.C.K. and T.A.P. provided Kaplan-Meier curves
and statistical analysis of mouse cohorts. J.H.B., W.C.L., G.C.K., T.A.P. and
P.S.R. provided technical assistance, animal care and tissues. L.A.L. supervised the
experimental work and interpretation of data. All authors commented on and
discussed the paper.
COMPETING INTERESTS STATEMENT
The authors declare no competing financial interests.
Published online at http://www.nature.com/naturegenetics
Reprints and permissions information is available online at http://npg.nature.com/
1. Khrapko, K., Kraytsberg, Y., de Grey, A.D., Vijg, J. & Schon, E.A. Does premature aging
of the mtDNA mutator mouse prove that mtDNA mutations are involved in natural
aging? Aging Cell 5, 279–282 (2006).
2. Bielas, J.H. & Loeb, L.A. Quantification of random genomic mutations. Nat. Methods
2, 285–290 (2005).
3. Kujoth, G.C. et al. Mitochondrial DNA mutations, oxidative stress, and apoptosis in
mammalian aging. Science 309, 481–484 (2005).
4. Saraste, M. Oxidative phosphorylation at the fin de siecle. Science 283, 1488–1493
5. Newmeyer, D.D. & Ferguson-Miller, S. Mitochondria: releasing power for life and
unleashing the machineries of death. Cell 112, 481–490 (2003).
6. Wallace, D.C. A mitochondrial paradigm of metabolic and degenerative diseases, aging,
and cancer: a dawn for evolutionary medicine. Annu. Rev. Genet. 39, 359–407 (2005).
7. Trifunovic, A. Mitochondrial DNA and ageing. Biochim. Biophys. Acta 1757, 611–617
8. Balaban, R.S., Nemoto, S. & Finkel, T. Mitochondria, oxidants, and aging. Cell 120,
9. Miquel, J., Economos, A.C., Fleming, J. & Johnson, J.E., Jr. Mitochondrial role in cell
aging. Exp. Gerontol. 15, 575–591 (1980).
10. Khrapko, K. et al. Mitochondrial mutational spectra in human cells and tissues. Proc.
Natl. Acad. Sci. USA 94, 13798–13803 (1997).
11. Michikawa, Y., Mazzucchelli, F., Bresolin, N., Scarlato, G. & Attardi, G. Aging-
dependent large accumulation of point mutations in the human mtDNA control region
for replication. Science 286, 774–779 (1999).
12. Copeland, W.C. Mitochondrial DNA: methods and protocols. Methods Mol. Biol. 197,
13. Zhang, D. et al. Construction of transgenic mice with tissue-specific acceleration of
mitochondrial DNA mutagenesis. Genomics 69, 151–161 (2000).
14. Bielas, J.H., Loeb, K.R., Rubin, B.P., True, L.D. & Loeb, L.A. Human cancers express a
mutator phenotype. Proc. Natl. Acad. Sci. USA 103, 18238–18242 (2006).
15. Schriner, S.E. et al. Extension of murine life span by overexpression of catalase
targeted to mitochondria. Science 308, 1909–1911 (2005).
16. Mandavilli, B.S., Santos, J.H. & Van Houten, B. Mitochondrial DNA repair and aging.
Mutat. Res. 509, 127–151 (2002).
17. Wang, D., Kreutzer, D.A. & Essigmann, J.M. Mutagenicity and repair of oxidative DNA
damage: insights from studies using defined lesions. Mutat. Res. 400, 99–115
18. Kreutzer, D.A. & Essigmann, J.M. Oxidized, deaminated cytosines are a source of C -
T transitions in vivo. Proc. Natl. Acad. Sci. USA 95, 3578–3582 (1998).
19. Pinz, K.G., Shibutani, S. & Bogenhagen, D.F. Action of mitochondrial DNA polymerase
gamma at sites of base loss or oxidative damage.J.Biol. Chem.270,9202–9206 (1995).
20. Tanaka, M. & Ozawa, T. Strand asymmetry in human mitochondrial DNA mutations.
Genomics 22, 327–335 (1994).
21. Frederico, L.A., Kunkel, T.A. & Shaw, B.R. A sensitive genetic assay for the detection of
cytosine deamination: determination of rate constants and the activation energy.
Biochemistry 29, 2532–2537 (1990).
22. Trifunovic, A. et al. Premature ageing in mice expressing defective mitochondrial DNA
polymerase. Nature 429, 417–423 (2004).
23. Larsen, N.B., Rasmussen, M. & Rasmussen, L.J. Nuclear and mitochondrial DNA
repair: similar pathways? Mitochondrion 5, 89–108 (2005).
24. Elson, J.L., Samuels, D.C., Turnbull, D.M. & Chinnery, P.F. Random intracellular drift
explains the clonal expansion of mitochondrial DNA mutations with age. Am. J. Hum.
Genet. 68, 802–806 (2001).
25. Bender, A. et al. High levels of mitochondrial DNA deletions in substantia nigra
neurons in aging and Parkinson disease. Nat. Genet. 38, 515–517 (2006).
26. Kraytsberg, Y. et al. Mitochondrial DNA deletions are abundant and cause functional
impairment in aged human substantia nigra neurons. Nat. Genet. 38, 518–520 (2006).
27. Tyynismaa, H. et al. Mutant mitochondrial helicase Twinkle causes multiple mtDNA
deletions and a late-onset mitochondrial disease in mice. Proc. Natl. Acad. Sci. USA
102, 17687–17692 (2005).
28. Hamilton, M.L. et al. Does oxidative damage to DNA increase with age? Proc. Natl.
Acad. Sci. USA 98, 10469–10474 (2001).
NATURE GENETICS VOLUME 39 [ NUMBER 4 [ APRIL 2007543
© 2007 Nature Publishing Group http://www.nature.com/naturegenetics