Copyright ? 2009 by the Genetics Society of America
Evidence That Msh1p Plays Multiple Roles in Mitochondrial Base
Leah Pogorzala, Shona Mookerjee and Elaine A. Sia1
Department of Biology, University of Rochester, Rochester, New York 14627
Manuscript received April 13, 2009
Accepted for publication April 23, 2009
Mitochondrial DNA is thought to be especially prone to oxidative damage by reactive oxygen species
generated through electron transport during cellular respiration. This damage is mitigated primarily by
the base excision repair (BER) pathway, one of the few DNA repair pathways with confirmed activity on
mitochondrial DNA. Through genetic epistasis analysis of the yeast Saccharomyces cerevisiae, we examined
the genetic interaction between each of the BER proteins previously shown to localize to the mitochon-
dria. In addition, we describe a series of genetic interactions between BER components and the MutS
homolog MSH1, a respiration-essential gene. We show that, in addition to their variable effects on mito-
chondrial function, mutant msh1 alleles conferring partial function interact genetically at different points
in mitochondrial BER. In addition to this separation of function, we also found that the role of Msh1p in
BER is unlikely to be involved in the avoidance of large-scale deletions and rearrangements.
be attributed to mitochondrial genome instability, as the
respiratory capacity of the mitochondria is dependent on
an intact genome. Since respiration is essential for the
survival of eukaryotic obligate aerobes, the facultative an-
aerobe Saccharomyces cerevisiae is an ideal model system for
mitochondrial studies. Despite the difference in size bet-
encoded components are required for the same process,
studying how S. cerevisiae maintain mitochondrial DNA
(mtDNA) could lend valuable insight into mitochondrial
genome maintenance in higher eukaryotes (Perocchi
et al. 2008).
The necessary process of electron transport during
respiration can cause damage to proteins, lipids, and
nucleic acids through the formation of reactive oxygen
species (ROS) (Longo et al. 1996). Because mtDNA
exists in this harsh environment, it is thought that it is
especially prone to oxidative damage (Bohr 2002).
Damaged bases can be mutagenic by misincorporation
translesion synthesis beyond the damaged base. There-
fore, the repair of oxidative lesions is essential for the
stability of the mitochondrial genome.
EPLETION of mitochondrial function has been
implicated in the human aging process as well as
Animportant mechanismfor repair ofoxidative DNA
damage is the base excision repair (BER) pathway
(Croteau and Bohr 1997; Nilsen and Krokan 2001;
Bohr 2002). This pathway is well studied in the nucleus
of many organisms, and isoforms of several key compo-
nents have been shown to localize to the mitochondrial
compartment (Rosenquist et al. 1997; You et al. 1999;
Vongsamphanh et al. 2001). However, despite their
role of these isoforms in the repair of mtDNA is poorly
BER is initiated when an N-glycosylase recognizes a
damaged base and cleaves the glycosidic bond between
it and the sugar-phosphate backbone, creating an
apurinic/apyrimidinic (AP) site that can be repaired
by one of two BER pathways. In short patch BER, the AP
site is processed by an AP endonuclease on the 59 side
of the damaged base and by the AP lyase activity of
a glycosylase, or polymerase b, on the 39 side of the
damage, to create a single-strand gap (Wilson et al.
1998). This gap is filled by a DNA polymerase and then
of long-patch BER, the DNA is again cleaved by an AP
endonuclease to generate an available 39-end for syn-
thesis by a DNA polymerase at the nick, displacing the
existing sequence containing the abasic site and creat-
ing a 59 flap. This flap is cleaved by a flap endonuclease,
and the resulting nick is sealed by DNA ligase, complet-
ing the repair. Biochemical studies suggest that both
short-patch and long-patch pathways are active in
mitochondria (Akbari et al. 2008; Liu et al. 2008;
Szczesny et al. 2008).
1Corresponding author: University of Rochester, RC Box 270211,
Rochester, NY 14627.E-mail: firstname.lastname@example.org
Genetics 182: 699–709 (July 2009)
In this study, we examine the mitochondrial roles of
Apn1p, Ntg1p, and Ogg1p, three well-studied BER
components. The N-glycosylase Ogg1p is important for
the repair of oxidatively damaged DNA, and studies of
ogg1-D strains have found an increase in point mutations
in both nuclear and mitochondrial DNA (Thomaset al.
1997; Singh et al. 2001). In yeast, it was previously dem-
onstrated that a deletion of the N-glycosylase NTG1, or
the AP endonuclease APN1, leads to a decrease in mito-
chondrial mutations as measured by rates of erythromy-
cin resistance, suggesting that the actions of Ntg1p and
Apn1p create mutagenic intermediates in mtDNA dur-
the increases seen for nuclear DNA mutationrates in the
presenceofthese deletionalleles, indicatingthat it isnot
of BER proteins on the basis of their nuclear functions,
thus making mitochondrial-specific studies necessary
(Ramotar et al. 1991, 1993; Alseth et al. 1999; Bennett
1999). In addition, there are likely to be mitochondrial-
specific players in the pathway. Here we show that the
mismatch repair homolog Msh1p plays multiple roles in
Msh1p is the only one of six yeast homologs of MutS,
the bacterial mismatch repair protein, which has been
found localized to the mitochondria (Reenan and
Kolodner 1992; Chi and Kolodner 1994). Msh1p is
essential for mitochondrial function and maintenance
of mtDNA, necessitating the use of partial function
mutants to study the role of Msh1p in mtDNA mainte-
nance (Mookerjee et al. 2005). Although the effects of
its disruption have been examined in multiple studies,
the mechanism by which Msh1p acts to carry out its
essential functions remains unclear (Reenan and
Kolodner 1992; Koprowski et al. 2002; Mookerjee
et al. 2005; Mookerjee and Sia 2006). Its role as a mito-
chondrial mismatch repair protein has been disputed,
particularly since there are no other mismatch repair
proteins that localize to the mitochondria. However,
since mtDNA has such a high potential requirement for
BER, it is possible that this pathway in the mitochondria
may utilize Msh1p. Previous studies have shown genetic
interactions between Msh1p and the BER proteins
Ogg1p, Apn1p, and Ntg1p (Dzierzbicki et al. 2004;
Kaniak et al. 2009). Using msh1 alleles disrupted in
conserved DNA binding and ATPase domains, we have
examined the frequency and spectrum of mutations
responsible for the different mutation rates seen with
MATERIALS AND METHODS
Strains and media: Rich dextrose medium (YPD) contains
1% yeast extract, 2% Bacto peptone, and 2% dextrose. Rich
glycerol medium (YPG) contains 1% yeast extract, 2% Bacto
peptone, and 2% glycerol. YPD geneticin medium is YPD
supplemented with 200 mg/liter geneticin (Invitrogen). YG
erythromycin medium contains 1% yeast extract and 2%
glycerol supplemented with 50 mm sodium phosphate buffer
(pH 6.5) and 4 g/liter of erythromycin (MP Biomedicals).
Synthetic dextrose media contain 0.17% yeast nitrogen base,
0.5% ammonium sulfate, 2% dextrose, and necessary amino
acids. Synthetic glycerol (SGly) media contain 0.17% yeast
nitrogen base, 0.5% ammonium sulfate, 2% glycerol, and
necessary amino acids.
All S. cerevisiae strains used in this study are isogenic to the
his3 arg8ThisG; r1) (Sia et al. 2000) (Table 1). Primers located
?200 bp upstream and downstream of the gene of interest
were used to amplify the kanMX cassette from the S. cerevisiae
deletion collection (Open Biosystems). The ogg1-D strain was
created using primers 59-AGGCATTTGAAGCGTCCTGATT
TGAGAGTGCACC-39 and 59-CGCCTTTTCGGTCGCGTGCT
TATTTTCACACCGC-39 to amplify the HIS3 gene from the
pRS413 plasmid (Stratagene, La Jolla, CA). Because msh1 mu-
tant strains are difficult to manipulate while maintaining res-
piratory competence, multiple methods of strain construction
were utilized. Strains containing multiple deletions of BER
genes or deletion of a BER gene in an msh1 mutant back-
ground were constructed by mating otherwise isogenic single
mutants of opposite mating type and then sporulating and
ntg1-D, msh1-R813Wapn1-D, and msh1-R813W ntg1-D were con-
structed by deleting APN1 or NTG1 in the appropriate msh1
mutant background. The resulting strains were nonrespiring,
the cells were mated with NPY3 (DFS160 r1), which contains a
kar1-1 mutation preventing nuclear fusion, and haploid re-
spiring cytoductants with the DFS188 nuclear background
were screened for a Lys?and respiration-proficient pheno-
type. To construct the msh1-F105A ogg1-D strain, the msh1-
F105A strain was transformed with the Ura-selectable pRS416-
MSH1plasmid (Mookerjeeetal. 2005) tomaintain awild-type
copy ofMsh1p.Theresulting strainwasmatedtoDFS188 ogg1-
D, dissected to isolate the double mutant, and then grown on
5-FOA toselectfor lossoftheplasmid.The resultingstrain was
nonrespiring and was cytoduced with wild-type mtDNA as
For construction of the msh1-G776D 1 pRK2 and msh1-
G776D 1 pRK2 APN1 strains, YSM84 was transformed with the
appropriate plasmid. The resulting strains were nonrespiring,
so functional mtDNA was reintroduced by cytoduction as
Mitochondrial point mutation rates: Point mutation rates
(mutations/cell division) were estimated by performing
fluctuation analysis using spontaneous resistance to the drug
erythromycin (EryR) as a reporter. Independent colonies were
isolated on YPG plates and grown at 30? for 3 days. Twenty
independent colonies were used to inoculate 5 ml each of
YPG, which were grown at 30? with agitation for 2 days. For the
Apn1p overexpression studies, independent colonies were
isolated on SGly–Leu media, and singles were used to in-
oculate SGly–Leu media and grown for 3 days. Appropriate
dilutions were plated on YPG, allowing growth of all respiring
cells in a culture and YG plates supplemented with 4 g/liter
erythromycin to select EryRcolonies. Plates were incubated at
30? for 7 days. Rates of erythromycin resistance per cell
division were determined using the method of the median
(Lea and Coulson 1949). Each rate was determined from at
least two independent assays.
Determining the mutational spectra: The spectra of muta-
tions causing erythromycin resistance were determined in
mutant strains by amplifying the region of the 21S rRNA gene
700L. Pogorzala, S. Mookerjee and E. A. Sia
previously shown to give rise to EryRstrains from indepen-
dently derived EryRcolonies (Sor and Fukuhara 1982). This
region, which is located at 1900–2000 on the mitochondrial
genome, was amplified using primers 59-GGTAAATAG-
TTACAGTAAAGC-39. The amplified product was sequenced
at the Cornell University sequencing facility.
Respiration loss assay: To determine the proportion of
nonrespiring cells in each strain, cultures were grown over-
night in YPG medium and colonies were isolated on YPG
plates. Ten to 20 independent YPD cultures were inoculated
with a single colony each and grown overnight, and appropri-
tests were performed in duplicate, and the average median
values were used for analysis.
Statistical analysis: Fluctuation analyses were performed
two or more times. Statistical analysis was performed using
Instat 3 for Macintosh (GraphPad Software, San Diego). Error
is represented as 6SEM. Two-tailed P-values were calculated
using unpaired t-tests to compare average rates and frequen-
cies. Mutational spectra were compared by chi-square analysis.
Differential functions of BER on mitochondrial and
nuclear DNA: In the nucleus, various effects have been
reported for deletion of NTG1 and APN1. Deletion of
APN1 has been reported to increase nuclear mutation
range from insignificant to 40-fold over wild-type levels
(Ramotar et al. 1991, 1993; Bennett 1999; Hanna et al.
2004; Doudican et al. 2005). To determine how mtDNA
is affected by these deletions, we assayed for spontane-
ous resistance to EryR. In agreement with previously
published findings (Phadnis et al. 2006), we found that
the mitochondrial point mutation rates in the ntg1-D
strain (0.13 3 10?7mutations/cell division) and the
significantly lower than the wild-type rate of 1.2 3 10?7
2). These findings suggest that the context in which
these BER enzymes function in the mitochondria may
be substantially different from that in the nucleus,
despite identical biochemistry. Furthermore, we show
that the rates of mitochondrial point mutations in an
apn1-D ntg1-D strain (0.75 3 10?7mutations/cell di-
vision) are not significantly different from rates of
mutation in a strain containing a deletion of APN1 or
NTG1 alone (P ¼ 0.25 and 0.18, respectively), which is
consistent with their action in a single genetic pathway
for prevention of EryRmutations.
In contrast to the decreased rates in apn1-D and ntg1-
D, we show that deletion of OGG1 causes a significant,
(4.9 3 10?7mutations/cell division P ¼ 0.025), which is
consistent with previous results (Singh et al. 2001).
AdditionaldeletionofAPN1 orNTG1 inanogg1-Dstrain
fold relative to the rate of apn1-D and ntg1-D strains
Strains used in this study
Strain Relevant genotypeReference
MATa ura3-52 leu2-3, 112 lys2 his3 arg8ThisG
DFS188 apn1-DTkanMX ntg1-DTkanMX
DFS188 apn1-DTkanMX ntg1-DTkanMX ogg1-DTHIS3
DFS188 apn1-DTkanMX ogg1-DTHIS3
DFS188 ntg1-DTkanMX ogg1-DTHIS3
DFS188 msh1-F105A apn1-DTkanMX
DFS188 msh1-F105A ntg1-DTkanMX
DFS188 msh1-F105A ogg1-DTHIS3
DFS188 msh1-G776D apn1-DTkanMX
DFS188 msh1-G776D ntg1-DTkanMX
DFS188 msh1-G776D ogg1-DTHIS3
DFS188 msh1-G776D apn1-DTkanMX ogg1-DTHIS3
DFS188 1 pRK2
DFS188 1 pRK2 APN1
DFS188 msh1-G776D 1 pRK2
DFS188 msh1-G776D 1 pRK2 APN1
DFS188 msh1-R813W apn1-DTkanMX
DFS188 msh1-R813W ntg1-DTkanMX
DFS188 msh1-R813W ogg1-DTHIS3
Sia et al. (2000)
Phadnis et al. (2006)
Phadnis et al. (2006)
Mookerjee and Sia (2006)
Mookerjee and Sia (2006)
Phadnis et al. (2006)
Phadnis et al. (2006)
Mookerjee and Sia (2006)
Msh1p Plays Multiple Roles701
alone (P ¼ 0.032 and 0.009, respectively). With a
mitochondrial point mutation rate of 3.5 3 10?7, the
apn1-D ntg1-D ogg1-D strain is not significantly different
from either the apn1-D ogg1-D or the ntg1-D ogg1-D strain
(P ¼ 0.33 and 0.62, respectively). In contrast to their
effects on mitochondrial point mutations, we have
found that deletion of APN1, NTG1, or OGG1 does not
significantly affect the frequency of respiration loss as
compared to the wild-type strain (P ¼ 0.40, 0.60, and
0.37, respectively) (Table 3).
The difference between mitochondrial and nuclear
suggests a role for unique mitochondrial components.
As demonstrated by epistasis analysis, we reveal specific
interactions between MSH1 and the BER genes using
mitochondrial point mutation accumulation as an out-
put parameter, consistent with a role for Msh1p in BER
(Table 2). Moreover, these genetic relationships are
dependent on defined mutations to the conserved
functional domains of MSH1, suggesting that specific
activities of Msh1p contribute differently to BER as
DNA binding by Msh1p occurs early in mitochon-
drial BER: Mutation at a conserved phenylalanine
residue within the MutS homolog mismatch recogni-
tion domain disrupts both the mismatch recognition
and the DNA-binding capacities of MutS homologs
(Bowers et al. 1999; Dufner et al. 2000; Yamamoto
et al. 2000; Drotschmann et al. 2001; Schofield et al.
confers a point mutation rate of 6.3 3 10?7mutations/
cell division, about fivefold over wild type (P , 0.001).
This relatively mild defect is contrasted by a respiration
competence that is only 10% of the wild-type strain
(Tables 2 and 3). The levels of Msh1p-F105A are
comparable to steady-state levels of Msh1p, demonstrat-
ing that the associated phenotypes are specific to
disruption of the DNA-binding domain (Mookerjee
et al. 2005). Additional deletion of OGG1 does not
significantly alter the mitochondrial point mutation
phenotype (P ¼ 0.08), consistent with Msh1p-binding
activity functioning upstream of Ogg1p. This finding
suggests that Msh1p may bind to damaged bases or
mismatches and interact with or even preferentially
bind particular downstream components of BER.
Rates of erythromycin resistance
Rate of erythromycin
of the mean)
Wild type 1.2 (0.01)
Wild type 1 pRK2
Wild type 1 pRK2APN1
msh1-G776D 1 pRK2
msh1-G776D 1 pRK2APN1
apn1-D ntg1-D ogg1-D
Relevant genotype% nonrespiring cells
msh1-G776D apn1-D ogg1-D
Wild type 1 pRK2
Wild type 1 pRK2APN1
msh1-G776D 1 pRK2
msh1-G776D 1 pRK2APN1
apn1-D ntg1-D ogg1-D
702L. Pogorzala, S. Mookerjee and E. A. Sia
In contrast to theinteraction observed withthe OGG1
deletion, msh1-F105A, in conjunction with the APN1 or
NTG1 deletion, synergistically increases point mutation
accumulation over levels observed with msh1-F105A
alone (25 3 10?7and 29 3 10?7mutations/cell division;
P , 0.001 and P ¼ 0.026, respectively), suggesting that
the repair steps following Msh1p binding are at least
partially independent of pathways involving Apn1p and
Ntg1p. Unlike point mutation rates, deletion of APN1,
NTG1, or OGG1 does not alter the respiration loss seen
in the msh1-F105A strain, indicating that the mutator
phenotype conferred by the msh1-F105A allele is specif-
ically due to a defect in BER, while respiration loss
occurs through gross instability that is likely a separate
feature of MSH1 disruption (Table 3).
ATP-bound Msh1p functions in downstream BER
events: The msh1-G776D substitution within the con-
served ATPase domain is analogous to mutations that
disrupt the ability of MutS homologs to bind ATP
(Alani et al. 1997; Studamire et al. 1998). The steady-
Msh1p, demonstrating that the associated phenotypes
et al. 2005). It is the most deleterious of the three
mutations studied here, conferring a 27-fold increase
in point mutation rates relative to wild-type rates (32 3
10?7mutations/cell division; P ¼ 0.036), as well as an
almost complete loss of respiratory competence (Table
2, Table 3). However, in conjunction with deletion of
APN1 or NTG1, we find a complete rescue of the msh1-
G776D EryRphenotype to wild-type levels (P ¼ 0.83
and 0.19, respectively). In contrast, loss of Ogg1p,
which, like Ntg1p, can initiate BER by glycosidic bond
cleavage, does not significantly alter the rate conferred
by msh1-G776D alone (P ¼ 0.17). Since the activities of
Apn1p and Ntg1p appear to be important for the msh1-
G776D phenotype, these data are consistent with a
model in which Msh1p bound to ATP functions
downstream of endonucleolytic cleavage. One predic-
tion of this model is that an excess of Apn1p would
create more damage and increase the mitochondrial
point mutation rate. Under these experimental con-
ditions, we find that an extra copy of APN1 under a
relative to the wild-type rate (P ¼ 0.04), but does not
affect the mutation rate relative to msh1-G776D alone
(P ¼ 0.35).
As with msh1-F105A strains, respiration loss in msh1-
G776D strains is not affected by deletion of the BER
genes or overexpression of Apn1p (Table 3). However,
the msh1-G776D apn1-D ogg1-D strain has an even more
dramatic respiration loss than msh1-G776D alone, and it
is difficult to maintain respiring cells, indicating that
these proteins do play a partially redundant role in
maintaining respiratory competence. With this level of
measure mitochondrial point mutation rates.
Msh1p ATPase activity is epistatic to APN1 or NTG1
deletion: In contrast to the msh1-G776D allele, an
R813W substitution in Msh1p presumably allows ATP
binding and a binding-dependent conformational
change, as in other MutS homologs, but disrupts ATP
hydrolysis (Studamire et al. 1999). Therefore, compar-
ing these mutants should allow us to differentiate
between those functions that require ATPase activity
and those that require only ATP binding. The msh1-
R813W allele confers a 3.5-fold increase in point
mutation accumulation compared to wild type, which
under the conditions of these experiments is not quite
significant (4.2 3 10?7mutations/cell division; P ¼
3). Unlike the msh1-G776D allele, the rate of EryRin
msh1-R813W strains is not significantly altered by addi-
tional deletion of APN1 or NTG1 (P ¼ 0.14 and 0.24,
increase over the rate of msh1-R813W alone (41 3 10?7
mutations/cell division; P , 0.001). As with msh1-F105A
and msh1-G776D, respiration loss conferred bythe msh1-
R813W allele is unaffected by the additional deletion of
APN1, NTG1, or OGG1 individually.
Disruption of BER shifts the mutational spectra of
mtDNA: An important consideration of these studies is
notonlytherate ofmutationaccumulation, but also the
mutational spectrum, which can indicate a specific
mutational mechanism. Because a different biochemi-
cal activity is disrupted in each of the msh1-F105A, msh1-
G776D, and msh1-R813W mutants, they may display
different mutator effects. To determine the frequencies
of mutational events responsible for EryRwith each of
the msh1 alleles, the region from 1900 to 2000 of the
mitochondrial 21S rRNA gene was sequenced from
independently isolated EryRcolonies to determine the
types of substitutions that give rise to EryRin individual
colonies. Contained within this 100-bp region is a
stretch of four nucleotides whose mutation commonly
confers EryR(Figure 1A) (Vanderstraeten et al. 1998;
Kalifa and Sia 2007). All sequenced EryRclones in this
study resulted from mutations in this region. In wild-
type EryRstrains, the majority of mutations found are A-
to-G transitions and A-to-T transversions at frequencies
¼ 1.0). However, deletion of NTG1 causes a significant
shift toward A-to-G transitions as compared to the wild-
type strain (P ¼ 0.047) (Table 4).
Ogg1p recognizes and excises multiple types of
damaged bases from DNA. Its most well-studied sub-
strate is 8-oxo-7,8-dihydroguanine (8-oxo-G), but it acts
with similar efficiencies on 7,8-dihydro-8-oxoadenine,
2,6-diamino-4-hydroxy-5-formamidopyrimidine, and 4,6-
et al. 1996; Klungland et al. 1999; Krishnamurthy et al.
2008). It has been shown that 8-oxo-G mispairs with
adenine, and in subsequent rounds of replication, causes
a G-to-T transversion (Grollman and Moriya 1993). We
Msh1p Plays Multiple Roles 703
indicating that, while detectable, it is not a highly
represented mutation. We expected that deletion of
OGG1 would increase G-to-T events through failure of
8-oxo-G repair. Surprisingly, we found that, in an ogg1-D
strain, none of 30 EryRcolonies sequenced contained this
mutation. However, 67% of these contained an A-to-G
colonies in wild-type cells (P ¼ 0.013). These data suggest
by loss of Ogg1p. Instead, it suggests that, in the mito-
Reyes et al. 1998), we predict that adenine is more likely
to be the damaged base. Both FapyA and hypoxanthine
in A-to-G mutations in subsequent rounds of replication,
and these represent significant lesions in mtDNA (Reyes
et al. 1998; Tudek 2003).
Erythromycin resistance in msh1-F105A strains results
from A-to-G substitutions in 95% of the mutants
sequenced (Table 4). This mutational spectrum is also
significantly shifted toward A-to-G transitions in the
msh1-G776D strain (P ¼ 0.048), although not in the
msh1-R813W strain (P ¼ 0.77). Using the data presented
in Tables 2 and 4, we can estimate the frequency of each
B–E, the frequency of mutational changes at adenine
residues increases in ogg1-D and msh1 mutant strains.
These data indicate that each msh1 allele affects both
the rate and the spectrum of mutation differently. In
Figure 1.—(A) The sequence from 1945 to 1957 of the 21S rRNA gene encoded in the mitochondrial genome, the most com-
mon site of nucleotide substitutions conferring erythromycin resistance. (B–E) Frequency of base changes in EryRmutant strains.
The spectrum of mutations for each strain was determined by sequencing from independent erythromycin-resistant strains, as
described in materials and methods. Graphs represent data from 15 to 50 independent colonies for each strain. The frequency
of each type of substitution was estimated by multiplying the fraction of each type by the average frequency of EryRfor each strain.
704 L. Pogorzala, S. Mookerjee and E. A. Sia
addition, we find that, when the mutation rates are
reduced to wild-type levels in the msh1-G776D apn1-D
and msh1-G776D ntg1-D strains, the mutational spectra
are indistinguishable from apn1-D and ntg1-D strains
with respect to the number of A-to-G substitutions (P ¼
0.37 and 0.14, respectively).
The importance of maintaining the mitochondrial
genome for respiratory functioncannot be overstated,as
is demonstrated by the variety and severity of inherited
mtDNA mutation accumulation, ROS production, and
aging (Loeb et al. 2005; Kai et al. 2006; Vermulst et al.
2007). To understand how accumulating mtDNA muta-
tions affect cellular functions over time, we need to
elucidate the mechanisms that contribute to mtDNA
maintenance. Base excision repair was the earliest con-
firmed pathway of mtDNA repair, and its presence is
particularly intriguing, given the prevalent assumption
that mtDNA is highly susceptible to oxidative damage,
the main substrate for BER activity.
In the nuclear model of both short-patch and long-
patch BER, AP endonuclease cleavage is a required
intermediate step following base removal. The similar-
ities between apn1-D and ntg1-D, and their disparity to
ogg1-D with respect to point mutation accumulation,
connected in function while Ogg1p function is at least
partially independent. The finding that accumulation
of mitochondrial point mutations in the apn1-D ntg1-D
ogg1-D strain is similar to that in both the apn1-D ogg1-D
and ntg1-D ogg1-D strains further supports the connec-
tion between Apn1p and Ntg1p. We hypothesize that
Apn1p and Ntg1p form a functional complex in mito-
chondria. Such an Apn1p–Ntg1p complex may func-
tion in a nucleotide incision repair (NIR) role where
glycosylase activity is absent and endonucleolytic cleav-
age initiates repair, while Ogg1p plays no role in this
transitions in both ogg1-D and ntg1-D strains, but not in
apn1-D, however, may suggest that NIR activity is not
primarily acting on damaged adenine nucleotides and
that Apn1p and Ntg1p activities are not completely
interdependent. Unfortunately, the protein compo-
nents of short-patch BER, long-patch BER, and NIR
overlap, making it difficult to separate the contribution
of each pathway by genetic analysis.
Msh1p appears to constitute an additional compo-
nent of the mitochondrial BER pathways. Each of the
mutant alleles results in a unique set of phenotypes
Spectra of mutations conferring erythromycin resistance
% with nucleotide substitutions
Relevant genotypeA to CA to G A to T G to AG to T G insertion
Wild type3 34131469 35
Wild type 1 pRK2
Wild type 1 pRK2APN1
msh1-G776D 1 pRK2
msh1-G776D 1 pRK2APN1
Msh1p Plays Multiple Roles705
Figure 2.—Proposed model for Msh1p in mitochondrial BER. Distorted solid boxes (A and B) represent damaged bases, while
the box with an ‘‘X’’ (A) represents a mismatch across from a damaged base after replication by a translesion polymerase. The
double half-moon structures represent Msh1p dimers.
706L. Pogorzala, S. Mookerjee and E. A. Sia
consistent with multiple roles for Msh1p in BER, as well
as with additional roles in genome stability. Previously,
we attributed the mutator phenotypes resulting from
the msh1-G776D, msh1-R813W, and msh1-F105A alleles to
defects in the same Msh1p mechanism (Mookerjee
et al. 2005). However, we observe here that the delete-
rious effects of these alleles on point mutation are
differentially dependent on the activities of BER en-
zymes. For example, the increased point mutation rates
in msh1-G776D mutant strains seem to require func-
tional Apn1p and Ntg1p, presumably due to the normal
role of Msh1p in an Apn1p- and Ntg1p-dependent base
excision repair pathway. In contrast, point mutation
accumulation in msh1-R813W strains is unaffected by
deletion of APN1 or NTG1, and the msh1-F105A allele
appears defective independently from an Apn1p- and
These data support a model for Msh1p function in
mitochondrial BER in which at least two partially over-
the mismatch recognition/DNA-binding domain in the
msh1-F105A allele confers similar phenotypes with
regards to point mutation rates and mutational spectra
as the deletion of OGG1 does, we propose that Msh1p
may aid in the function of Ogg1p through its DNA-
binding activity (Figure 2A). It is possible that Msh1p
recognizes aspecific DNAaberrationanddirectsOgg1p
to the damaged site or stabilizes an existing association.
In Figure 2A, we have depicted the Msh1p substrate as
an oxidative lesion opposite a mispair. Such binding is
consistent with the known binding specificity of the
human MutSa complex (Mazurek et al. 2002; Larson
et al. 2003), but for Msh1p we do not exclude recogni-
tion of other lesions. Because deletion of APN1 and
NTG1 results in similar phenotypes, we propose that
they act in a complex that does not require Msh1p
binding to initiate repair (Figure 2B) but does require
that Msh1p bind ATP to complete the repair (Figure
2D). In our model, we have depicted Msh1p bound to a
flap intermediate since Msh2p–Msh6p complexes have
been demonstrated to bind such structures (Surtees
and Alani 2006). Studamire et al. have shown that the
nuclear Msh2p–Msh3p complex is important for single-
strand annealing, presumably by binding flapped DNA
structures. This activity is disrupted by a mutation in
analogous to msh1-R813W. These data suggest that the
ability of the Msh2p–Msh3p complex to bind ATP, but
not necessarily to hydrolyze it, is necessary for this
structure recognition (Studamire et al. 1998). Our data
suggest a similar separation of function for Msh1p,
where the mutator phenotype of msh1-R813W is only
slightly enhanced while it is dramatically increased in
The disparities between the point mutation rates and
respiration loss in the msh1 mutants suggest that the
participation of Msh1p in BER is likely unrelated to its
role in maintaining larger-scale mtDNA stability. Respi-
rationloss isa commonmarker for the overall stability of
the mitochondrial genome. Cells can lose respiration
both through accumulation of point mutations in the
mitochondrial genome and through large-scale rear-
rangements and deletions of mtDNA. In fact, such rear-
rangements, resulting in r?petites, are the primary
mechanisms that generate nonrespiring cells in labora-
these events occur is high and may result in respiration
loss in several percentages of cells in a culture after
overnight growth. Point mutation rates, which occur on
the order of one in 107cells for EryR, would have to
increase to phenomenal levels before they would begin
to be observed as increases in respiration loss over this
background of r?cells. Consequently, increases in point
respiration loss. So, while the mitochondrial point mutati-
ons in the msh1-G776D mutant strain can be suppressed by
not similarly reduced. Therefore, the essential function of
Msh1p is to stabilize the mitochondrial genome through a
broader mechanism than point mutation prevention.
Most of the proteins functioning within the mitochon-
dria are encoded in the nucleus, and many have addi-
tional, well-studied roles throughout the cell. This study
exemplifies the necessity for examination of mitochon-
drial-specific roles of proteins that also function else-
where. We have shown that Apn1p, Ntg1p, and Ogg1p
have unexpected mitochondrial phenotypes and that
their function is closely tied to the mitochondrial specific
protein, Msh1p. We have also demonstrated that Msh1p
plays multiple roles in mitochondrial base excision repair
but that this function is separate from a broader role in
We thank Linnell Randall and Shweta Krishnan for help with the
fluctuation analysis. We also thank Lidza Kalifa for critical reading of
the manuscript. This work was supported by grants from the National
Science Foundation (MCB0543084) and the National Institutes of
Mitochondrial base excision repair of uracil and AP sites
takes place by single-nucleotide insertion and long-patch DNA
synthesis. DNA Repair (Amst.) 7: 605–616.
Alani, E., T. Sokolsky, B. Studamire, J. J. Miret and R. S. Lahue,
1997Genetic and biochemical analysis of Msh2p-Msh6p: role
of ATP hydrolysis and Msh2p-Msh6p subunit interactions
in mismatch base pair recognition. Mol. Cell. Biol. 17: 2436–
Alseth, I., L. Eide, M. Pirovano, T. Rognes, E. Seeberg et al.,
1999 The Saccharomyces cerevisiae homologues of endonuclease
III from Escherichia coli Ntg1 and Ntg2 are both required for ef-
ficient repair of spontaneous and induced oxidative DNA dam-
age in yeast. Mol. Cell. Biol. 19: 3779–3787.
Bennett, R. A., 1999The Saccharomyces cerevisiae ETH1 gene, an in-
ducible homolog of exonuclease III that provides resistance to
DNA-damaging agents and limits spontaneous mutagenesis.
Mol. Cell. Biol. 19: 1800–1809.
T.Visnes, H. E.Krokan andM.Otterlei,
Msh1p Plays Multiple Roles 707
Bohr, V. A., 2002
mitochondrial DNA, and some changes with aging in mamma-
lian cells. Free Radic. Biol. Med. 32: 804–812.
Bowers, J., T. Sokolsky, T. Quach and E. Alani, 1999
in the MSH6 subunit of the Saccharomyces cerevisiae MSH2–MSH6
complex disrupts mismatch recognition. J. Biol. Chem. 274:
Chi, N.-W., and R. D. Kolodner, 1994
ization of MSH1, a yeast mitochondrial protein that binds to
DNA mismatches. J. Biol. Chem. 269: 29984–29992.
Croteau, D. L., and V. A. Bohr, 1997
nuclear and mitochondrial DNA in mammalian cells. J. Biol.
Chem. 272: 25409–25412.
Doudican, N. A., B. Song, G. S. Shadel and P. W. Doetsch,
2005Oxidative DNA damage causes mitochondrial genomic
instability in Saccharomyces cerevisiae. Mol. Cell. Biol. 25: 5196–5204.
Drotschmann, K., W. Yang, F. E. Brownewell, E. T. Kool and T. A.
Kunkel, 2001 Asymmetric recognition of DNA local distortion.
Structure-based functional studies of eukaryotic Msh2-Msh6. J.
Biol. Chem. 276: 46225–46229.
Dufner, P., G. Marra, M. Raschle and J. Jiricny, 2000
recognition and DNA-dependent stimulation of the ATPase activ-
ity of hMutS-alpha is abolished by a single mutation in the
hMSH6 subunit. J. Biol. Chem. 275: 36550–36555.
Dujon, B., 1981Mitochondrial genetics and functions, pp. 505–635
in The Molecular Biology of the Yeast Saccharomyces: Life Cycle and
Inheritance. Cold Spring Harbor Laboratory Press, Cold Spring
Dzierzbicki, P., P. Koprowski, M. U. Fikus, E. Malc and Z. Ciesla,
2004Repair of oxidative damage in mitochondrial DNA of Sac-
charomyces cerevisiae: involvement of the MSH1-dependent path-
way. DNA Repair (Amst) 3: 403–411.
Foury, F., T. Roganti, N. Lecrenier and B. Purnell, 1998
complete sequence of the mitochondrial genome of Saccharomy-
ces cerevisiae. FEBS Lett. 440: 325–331.
Grollman, A. P., and M. Moriya, 1993
nine: an enemy within. Trends Genet. 9: 246–249.
Hanna, M., B. L. Chow, N. J. Morey, S. Jinks-Robertson, P. W.
Doetsch et al., 2004 Involvement of two endonuclease III ho-
mologs in the base excision repair pathway for the processing of
DNA alkylation damage in Saccharomyces cerevisiae. DNA Repair
(Amst.) 3: 51–59.
Kai, Y., C. Takamatsu, K. Tokuda, M. Okamoto, K. Irita et al.,
2006 Rapid and random turnover of mitochondrial DNA
in rat hepatocytes of primary culture. Mitochondrion 6:
Kalifa, L., and E. A. Sia, 2007 Analysis of Rev1p and Polz in mito-
chondrial mutagenesis suggests an alternative pathway of dam-
age tolerance. DNA Repair (Amst.) 6: 1732–1739.
Kaniak, A., P. Dzierzbicki, A. T. Rogowska, E. Malc, M. Fikus et al.,
2009Msh1p counteracts oxidative lesion-induced instability of
mtDNA and stimulates mitochondrial recombination in Saccharo-
myces cerevisiae. DNA Repair (Amst.) 8: 318–329.
Klungland, A., I. Rosewell, S. Hollenbach, E. Larsen, G. Daly
et al., 1999Accumulation of premutagenic DNA lesions in mice
defective in removal of oxidative base damage. Proc. Natl. Acad.
Sci. USA 96: 13300–13305.
Koprowski, P., M. U. Fikus, P. Mieczkowski and Z. Ciesla,
2002A dominant mitochondrial mutator phenotype of Sac-
charomyces cerevisiae conferred by msh1 alleles in the sequence
encoding the ATP-binding domain. Mol. Genet. Genomics
glycosylases. Biochemistry 47: 1043–1050.
Larson, E. D., K. Iams and J. T. Drummond, 2003
processing of 8-oxoguanine by the human mismatch repair path-
way: inefficient removal of 8-oxoguanine paired with adenine or
cytosine. DNA Repair (Amst.) 2: 1199–1210.
Lea, D. E., and C. A. Coulson, 1949
bers of mutants in bacterial populations. J. Genet. 49: 264–285.
Liu, P., L. Qian, J. S. Sung, N. C. de Souza-Pinto, L. Zheng et al.,
2008Removal of oxidative DNA damage via FEN1-dependent
long-patch base excision repair in human cell mitochondria.
Mol. Cell. Biol. 28: 4975–4987.
Repair of oxidative DNA damage in nuclear and
Purification and character-
Repair of oxidative damage to
Mutagenesis by 8-oxogua-
The distribution of the num-
Loeb, L. A., D. C. Wallace and G. M. Martin, 2005
drial theory of aging and its relationship to reactive oxygen spe-
cies damage and somatic mtDNA mutations. Proc. Natl. Acad.
Sci. USA 102: 18769–18770.
Longo, V. D., E. B. Gralla and J. S. Valentine, 1996
dismutase activity is essential for stationary phase survival in Sac-
charomyces cerevisiae. Mitochondrial production of toxic oxygen
species in vivo. J. Biol. Chem. 271: 12275–12280.
Mazurek, A., M. Berardini and R. Fishel, 2002
man MutS homologs by 8-oxo-guanine DNA damage. J. Biol.
Chem. 277: 8260–8266.
Mookerjee, S. A., and E. A. Sia, 2006
Msh1p and putative recombination proteins Cce1p, Din7p, and
Mhr1p in large-scale recombination and genome sorting events
in the mitochondrial genome of Saccharomyces cerevisiae. Mutat.
Res. 595: 91–106.
Mookerjee, S. A., H. D. Lyon and E. A. Sia, 2005
tional domains of the mismatch repair homologue Msh1p and its
Nilsen, H., and H. E. Krokan, 2001
work of defence and tolerance. Carcinogenesis 22: 987–998.
Perocchi, F., E. Mancera and L. M. Steinmetz, 2008
screens for human disease genes, from yeast to human and back.
Mol. Biosyst. 4: 18–29.
Phadnis, N., R. Mehta, N. Meednu and E. A. Sia, 2006
base excision repair protein, generates mutagenic intermediates
in yeast mitochondrial DNA. DNA Repair (Amst.) 5: 829–839.
Ramotar, D., S. C. Popoff, E. B. Gralla and B. Demple,
1991Cellular role of yeast Apn1 apurinic endonuclease/
39diesterase: repair of oxidative and alkylation DNA damage and
control of spontaneous mutation. Mol. Cell. Biol. 11: 4537–4544.
Ramotar, D., C. Kim, R. Lillis and B. Demple, 1993
localization of the Apn1 DNA repair enzyme of Saccharomyces cer-
evisiae. Nuclear transport signals and biological role. J. Biol.
Chem. 268: 20533–20539.
Reenan, R. A. G., and R. D. Kolodner, 1992
sertion mutations in the Saccharomyces cerevisiae MSH1 and MSH2
genes: evidence for separate mitochondrial and nuclear func-
tions. Genetics 132: 975–985.
Reyes, A., C. Gissi, G. Pesole and C. Saccone, 1998
directional mutation pressure in the mitochondrial genome of
mammals. Mol. Biol. Evol. 15: 957–966.
Rosenquist, T. A., D. O. Zharkov and A. P. Grollman,
1997Cloning and characterization of a mammalian 8-oxogua-
nine DNA glycosylase. Proc. Natl. Acad. Sci. USA 94: 7429–7434.
Schofield, M. J., F. E. Brownewell, S. Nayak, C. Du, E. T. Kool
et al., 2001 The Phe-X-Glu DNA binding motif of MutS. The
role of hydrogen bonding in mismatch recognition. J. Biol.
Chem. 276: 45505–45508.
Sia, E. A., C. A. Butler, M. Dominska, P. Greenwell, T. D. Fox et al.,
2000Analysis of microsatellite mutations in the mitochondrial
DNA of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 97:
Singh, K. D., B. Sigala, H. A. Sikder and C. Schwimmer,
2001 Inactivation of Saccharomyces cerevisiae OGG1 DNA repair
gene leads to an increased frequency of mitochondrial mutants.
Nucleic Acids Res. 29: 1381–1388.
Sor, F., and H. Fukuhara, 1982Identification of two erythromycin
resistance mutations in the mitochondrial gene coding for the
large ribosomal RNA in yeast. Nucleic Acids Res. 10: 6571–6577.
Studamire, B., T. Quach and E. Alani, 1998
Msh2p and Msh6p ATPase activities are both required during
mismatch repair. Mol. Cell. Biol. 18: 7590–7601.
Studamire, B., G. Price, N. Sugawara, J. E. Haber and E. Alani,
1999Separation-of-function mutations in Saccharomyces cerevi-
siae MSH2 that confer mismatch repair defects but do not affect
nonhomologous-tail removal during recombination. Mol. Cell.
Biol. 19: 7558–7567.
Surtees, J. A., and E. Alani, 2006
MSH3 binds and alters the conformation of branchedDNA struc-
tures predicted to form during genetic recombination. J. Mol.
Biol. 360: 523–536.
2008 Long patch base excision repair in mammalian mitochon-
drial genomes. J. Biol. Chem. 283: 26349–26356.
Activation of hu-
Overlapping contributions of
Analysis of the func-
Base excision repair in a net-
Characterization of in-
Mismatch repair factor MSH2–
708L. Pogorzala, S. Mookerjee and E. A. Sia
Tanaka, M., and T. Ozawa, 1994
tochondrial DNA mutations. Genomics 22: 327–335.
Thomas, D., A. D. Scot, R. Barbey, M. Padula and S. Boiteux,
1997Inactivation of OGG1 increases the incidence of G ? C
/ T ? A transversions in Saccharomyces cerevisiae: evidence for en-
dogenous oxidative damage to DNA in eukaryotic cells. Mol.
Gen. Genet. 254: 171–178.
Tudek, B., 2003Imidazole ring-opened DNA purines and their bi-
ological significance. J. Biochem. Mol. Biol. 36: 12–19.
van der Kemp, P. A., D. Thomas, R. Barbey, R. de Oliveira and S.
Boiteux, 1996Cloning and expression in Escherichia coli of the
OGG1 gene of Saccharomyces cerevisiae, which codes for a DNA gly-
cosylase that excises 7,8-dihydro-8-oxoguanine and 2,6-diamino-
4-hydroxy-5-N-methylformamidopyrimidine. Proc. Natl. Acad.
Sci. USA 93: 5197–5202.
Vanderstraeten, S., S. Van den Brule, J. Hu and F. Foury,
1998The role of 39-59 exonucleolytic proofreading and mis-
matchrepairinyeastmitochondrial DNAerroravoidance.J. Biol.
Chem. 273: 23690–23697.
Vermulst, M., J. H. Bielas, G. C. Kujoth, W. C. Ladiges, P. S.
Rabinovitch et al., 2007
do not limit the natural lifespan of mice. Nat. Genet. 39: 540–
Vongsamphanh, R., P.-K. Fortier and D. Ramotar, 2001
diates translocation of the yeast Apn1pendonuclease into the mi-
Strand asymmetry in human mi-
Mitochondrial point mutations
tochondria to maintain genomic stability. Mol. Cell. Biol. 21:
Wallace, D. C., 2005 A mitochondrial paradigm of metabolic and
degenerative diseases, aging, and cancer: a dawn for evolutionary
medicine. Annu. Rev. Genet. 39: 359–407.
Weissman, L., N. C. de Souza-Pinto, T. Stevnsner and V. A. Bohr,
2007 DNA repair, mitochondria, and neurodegeneration. Neu-
roscience 145: 1318–1329.
Wilson, D. M., B. P. Engelward and L. Samson, 1998
base excision repair, pp. 29–64 in DNA Damage and Repair. Vol. I:
DNA Repair in Prokaryotes and Lower Eukaryotes, edited by J. A.
Nickoloff and M. F. Hoekstra. Humana Press, Totowa, NJ.
Yamamoto, A., M. J. Schofield, I. Biswas and P. Hsieh,
2000 Requirement for Phe36 for DNA binding and mismatch
repair by Escherichia coli MutS protein. Nucleic Acids Res. 28:
You, H. J., R. L. Swanson, C. Harrington, A. H. Corbett, S. Jinks-
Robertson et al., 1999
Saccharomyces cerevisiae Ntg1p and
Ntg2p: broad specificity N-glycosylases for the repair of oxidative
DNA damage in the nucleus and mitochondria. Biochemistry 38:
Communicating editor: E. Alani
Msh1p Plays Multiple Roles 709