MutS and MutL are dispensable for maintenance of the genomic mutation rate in the halophilic archaeon Halobacterium salinarum NRC-1.

Abstract

The genome of the halophilic archaeon Halobacterium salinarum NRC-1 encodes for homologs of MutS and MutL, which are key proteins of a DNA mismatch repair pathway conserved in Bacteria and Eukarya. Mismatch repair is essential for retaining the fidelity of genetic information and defects in this pathway result in the deleterious accumulation of mutations and in hereditary diseases in humans.
We calculated the spontaneous genomic mutation rate of H. salinarum NRC-1 using fluctuation tests targeting genes of the uracil monophosphate biosynthesis pathway. We found that H. salinarum NRC-1 has a low incidence of mutation suggesting the presence of active mechanisms to control spontaneous mutations during replication. The spectrum of mutational changes found in H. salinarum NRC-1, and in other archaea, appears to be unique to this domain of life and might be a consequence of their adaption to extreme environmental conditions. In-frame targeted gene deletions of H. salinarum NRC-1 mismatch repair genes and phenotypic characterization of the mutants demonstrated that the mutS and mutL genes are not required for maintenance of the observed mutation rate.
We established that H. salinarum NRC-1 mutS and mutL genes are redundant to an alternative system that limits spontaneous mutation in this organism. This finding leads to the puzzling question of what mechanism is responsible for maintenance of the low genomic mutation rates observed in the Archaea, which for the most part do not have MutS and MutL homologs.

Full text

Page 1

MutS and MutL Are Dispensable for Maintenance of the
Genomic Mutation Rate in the Halophilic Archaeon
Halobacterium salinarum NRC-1
Courtney R. Busch., Jocelyne DiRuggiero*.
Department of Biology and Molecular Genetics, University of Maryland, College Park, Maryland, United States of America
Abstract
Background: The genome of the halophilic archaeon Halobacterium salinarum NRC-1 encodes for homologs of MutS and
MutL, which are key proteins of a DNA mismatch repair pathway conserved in Bacteria and Eukarya. Mismatch repair is
essential for retaining the fidelity of genetic information and defects in this pathway result in the deleterious accumulation
of mutations and in hereditary diseases in humans.
Methodology/Principal Findings: We calculated the spontaneous genomic mutation rate of H. salinarum NRC-1 using
fluctuation tests targeting genes of the uracil monophosphate biosynthesis pathway. We found that H. salinarum NRC-1 has
a low incidence of mutation suggesting the presence of active mechanisms to control spontaneous mutations during
replication. The spectrum of mutational changes found in H. salinarum NRC-1, and in other archaea, appears to be unique to
this domain of life and might be a consequence of their adaption to extreme environmental conditions. In-frame targeted
gene deletions of H. salinarum NRC-1 mismatch repair genes and phenotypic characterization of the mutants demonstrated
that the mutS and mutL genes are not required for maintenance of the observed mutation rate.
Conclusions/Significance: We established that H. salinarum NRC-1 mutS and mutL genes are redundant to an alternative
system that limits spontaneous mutation in this organism. This finding leads to the puzzling question of what mechanism is
responsible for maintenance of the low genomic mutation rates observed in the Archaea, which for the most part do not
have MutS and MutL homologs.
Citation: Busch CR, DiRuggiero J (2010) MutS and MutL Are Dispensable for Maintenance of the Genomic Mutation Rate in the Halophilic Archaeon
Halobacterium salinarum NRC-1. PLoS ONE 5(2): e9045. doi:10.1371/journal.pone.0009045
Editor: Rodolfo Aramayo, Texas A&M University, United States of America
Received October 16, 2009; Accepted January 5, 2010; Published February 4, 2010
Copyright: ? 2010 Busch, DiRuggiero. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by grants from NASA (EXB04-0000-0055), NSF (MCB-0425825), and AFOSR ((FA95500710158) to J.D.R. The funders had no role
in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: jdiruggiero@jhu.edu
. These authors contributed equally to this work
Introduction
DNA mismatch repair (MMR) is the major pathway for the
repair of DNA replication errors such as nucleotide mismatches,
insertions, and deletions [1]. Defects in the MMR pathway lead to
genomic instability that can cause a 10 to 1000-fold increase in
spontaneousmutability, meiotic defects ineukaryotes,andtolerance
to DNA alkylating agents [1,2,3]. In humans, inactivation of the
MMR pathway leads to a predisposition to hereditary nonpolyposis
colon cancer and other types of tumors [2,3]. The MMR pathway
also plays an important role in preventing recombination events
between divergent sequences [1,2].
The key proteins of the MMR pathway, MutS and MutL, are
highly conserved between Bacteria and Eukarya. The pathway has
been characterized in bacterial and eukaryal systems and
comprises three basic steps: (1) MutS/L recognition of mismatch,
(2) excision of the mismatched base and surrounding DNA, and (3)
repair synthesis [1,2,4,5]. The three-dimensional structure of
MutS has been resolved for the Escherichia coli and Thermus aquaticus
proteins [6,7]. MutS is a 95kDa protein that functions as a dimer
in vivo [3,8]. MutS has ATPase activity with Walker A/B sequence
motifs and a highly conserved Phe-X-Glu motif responsible for
binding DNA [9]. MutL is a 68kDa protein that exists as a dimer
in solution and is a member of the Bergerat-fold ATPase/kinase
family [3,10]. Eukaryotes have multiple homologs of the MutS and
MutL proteins that form heterodimers suggesting a more complex
system than in bacteria with multiple interactions [1,2]. In contrast
to E. coli and several other gram-negative bacteria, eukaryotes and
most bacteria do not have a methylation-directed MMR system
for strand discrimination or a MutH homolog. Studies suggest a
nick-directed mechanism using Okazaki fragments produced
during replication of the lagging strand or a strand discrimination
mechanism directed by the proliferating cell nuclear antigen
(PCNA), thus coupling replication and MMR [2,11].
MMR has not been investigated in the Archaea but studies of
the genomic mutation rate in the thermophilic acidophile,
Sulfolobus acidocaldarius, and the halophile, Haloferax volcanii, revealed
rates of spontaneous mutation very similar to rates previously
reported for DNA-based microorganisms with 3.461023sponta-
neous mutations per genome, per replication, suggesting that DNA
PLoS ONE | www.plosone.org1 February 2010 | Volume 5 | Issue 2 | e9045

Page 2

mismatches resulting from DNA replication errors are actively
corrected in those organisms [12]. Surprisingly, only 11 out of the
54 archaeal genomes sequenced so far encode for homologs of the
conserved MutS1 protein subfamily found in Bacteria and
Eukarya [13]. Archaea with MutS1 homologs include halophiles
and methanogens, all members of the domain Euryarchaeota. These
archaeal MutS1 proteins share identical domain structure with
their bacterial counterparts and are likely the result of a lateral
gene transfer event [13]. Also detected in the Archaea are MutS2-
like proteins [13]. In the hyperthermophilic archaeon Pyrococcus
furiosus, MutS2 has been shown to have ATPase and DNA binding
activity but no specific DNA mismatch binding activity [14].
Proteins from the MutS2 subfamily, which are not thought to be
involved in MMR, are different in structure and sequence to the
MutS1 subfamily proteins except for the MutSAc domain essential
for dimerization, ATPase and DNA binding activities [13].
Halobacterium salinarum NRC-1 is an extremely halophilic
archaeon growing optimally in 4M NaCl [15,16,17]. The high
osmotic pressure from its environment is counterbalanced by a 4M
intracellular concentration of KCl [18]. Previous studies revealed
the exceptional resistance of H. salinarum NRC-1 to desiccation,
UV and ionizing radiation, which was attributed to efficient DNA
repair and detoxification systems and to its adaptation to
hypersaline environments, characterized by high levels of solar
radiation and periodic desiccation [19,20,21]. The genome of H.
salinarum NRC-1 has been sequenced [22] and encodes for
proteins of conserved DNA repair pathways that include damage
reversal, base excision repair (BER), nucleotide excision repair
(NER), homologous recombination, and the bacterial-like MMR
proteins MutS and MutL [22]. We demonstrated that the
eukaryal-like homologous recombination protein, Mre11, is
essential for the repair of DNA double strand breaks in H.
salinarum NRC-1, whereas Rad50 is dispensable, representing a
shift from the eukaryotic model of recombinational repair [23].
Crowley et al. [24] showed that the bacterial NER homologs
UvrA/B/C encoded in the genome of H. salinarum NRC-1 were
essential for the survival of the organism to UV irradiation. These
studies demonstrate the mosaic nature of the DNA repair
pathways in H. salinarum NRC-1, and in the Archaea in general,
and raise questions about the nature of DNA MMR in this
organism. Through computational analysis we found that H.
salinarum NRC-1 has three bacterial-like mutS genes, a bacterial-
like mutL gene, 4 bacterial-like recJ exonuclease genes, 1
eukaryotic-like rad2 59-39 exonuclease gene, and a bacterial-like
uvrD helicase gene, all potentially involved in MMR. Two of the
MutS proteins in H. salinarum NRC-1, MutS1 and MutS2, are
homologous to the MutS1 protein subfamily, and have been
renamed MutS1A and MutS1B in this study, while the third MutS
protein, MutS3, is homologous to proteins found in the MutS2
subfamily [13]. Whole-genome transcriptomic studies conducted
on H. salinarum cells exposed to UV and gamma radiation, and to
oxidative stress revealed no significant changes in mRNA level for
mutS1A, mutS1B, and mutL when compared to untreated cells
[20,21, unpublished].
Here, we used a genetic approach to determine the spontaneous
genomic mutation rate in H. salinarum NRC-1 and to determine
the cellular role of the bacterial-like MMR proteins MutS and
MutL encoded in its genome. Our analysis, using fluctuation tests
targeting genes of the uridine monophosphate (UMP) biosynthesis
pathway, revealed a genomic mutation rate similar to that of other
DNA-based microorganisms and a markedly different spectrum of
mutational changes. The phenotypic analysis of deletion mutants
for the mutL, mutS1A, mutS1B, and uvrD genes and a mutS1A/
mutS1B double mutant showed little difference between the mutant
and background strains indicating that the MutS and MutL
protein homologs found in H. salinarum NRC-1 are not essential
for maintaining the low incidence of spontaneous mutations
observed in this organism.
Results
Genomic Mutation Rate
We calculated the spontaneous genomic mutation rate of H.
salinarum NRC-1 to determine the replication fidelity in this
mesophilic archaeon. We performed six independent fluctuation
tests [12], targeting forward mutations in genes of the UMP
biosynthetic pathway producing 5-fluoroorotic acid (5-FOA) resis-
tant mutants. The mutation rate was calculated using the equation
m=ln(m/Nav) [25] and was found to be 3.7361027+/21.4461027
mutations per replication. Sequencing of purified 5-FOA-resistant
mutants revealed that only 13 out of 55 sequenced mutants had a
mutation in the pyrF gene (orotate decarboxylase), only 42 out of 69
had a mutation in the pyrE2 gene, and none out of 61 had mutation
in the pyrE1 gene (orotate phosphoribosyl transferases). We therefore
adjusted the gene mutation rates for the pyrF and pyrE2 genes by
factorsof0.24(13/55)and0.60(42/69),respectively,anddidnotuse
the pyrE1 gene in our calculation. The resulting spontaneous
mutation rates at the gene level were 8.9561028and 2.2461027
mutations per gene per replication, for the pyrF and pyrE2 genes,
respectively. To correct this rate for the fraction of undetected
mutations producing no phenotypic effect, we adjusted the total
number of base pair substitutions (BPS) using published information
on BPS detection efficiency (approximately 0.2) [4]. The resulting
rate estimate per gene was calculated as follows:
gene rate x indelsz BPS=0:2
½
indelszBPS
ðÞ?
Thisratewasthenconvertedinto agenomicratebydividingbygene
size (pyrF=803bp, and pyrE2=527bp) and multiplying by genome
size (2,571,010bp), resulting in an average genomic mutation rate,
corrected for undetected mutations, for H. salinarum NRC-1 of
1.676102361.461023mutation per replication (0.6261023for
pyrF and 2.761023for pyrE2).
Mutational Spectrum
One hundred forty-nine 5-FOA-resistant mutants were recov-
ered from two fluctuation tests and of those 55 were sequenced
using primers for pyrF, 61 using primers for pyrE1, and 65 using
primers for pyrE2. Mutations were only found in the coding
regions of the pyrF and pyrE2 genes and a number of mutants had
mutations in more than one gene (Table 1). No mutation was
found within a 100-bp region upstream of either gene. A
disproportionate number of deletions were found in the pyrE2
gene when compared to the pyrF gene (Figure 1, Table 2). All of
the pyrE2 mutations occurred at a single hotspot in the gene at
positions 382–397, with a 7-nucleotide (nt) deletion (GTCGACG)
found in 42 mutants; two 1-nt deletions and two single BPS also
contributed to the changes observed in this gene (Table 2,
Figure 2A). A 9-nt sequence found at the pyrE2 gene mutational
hotspot was the direct repeat of a sequence located immediately
upstream (Figure 2A). Mutations were distributed throughout the
pyrF gene with a concentration of insertions at position 354–365
(Table 2, Figure 2B). None of the mutations were the result of a
transposon element insertion. Indels out numbered BPS and
approximately 80% of the BPS resulted in non-synonymous amino
acid changes. BPS found in the pyrF and pyrE2 genes were mostly
transversions (80%) (Table 2).
Halobacterium Mismatch Repair
PLoS ONE | www.plosone.org2 February 2010 | Volume 5 | Issue 2 | e9045

Page 3

MutS and MutL Are Not Essential for the Low Incidence
of Mutation Observed in H. salinarum NRC-1
To determine whether the bacterial-like MMR proteins
encoded in the genome of H. salinarum NRC-1 were essential in
maintaining the low genomic mutation rate we observed, we
carried out targeted gene deletions of the mutS1A, mutS1B, mutL,
and uvrD genes, with a double deletion of the mutS1A and mutS1B
genes, using the background strains AK07 (Dura3) and CB08
(Dura3Dzim) [26,27]. Genotypes
(DmutS1A), CB072 (DmutS1B), CB073 (DmutS1ADmutS1B), CB074
(DmutL), and CB081 (DuvrD) were confirmed by Southern blot
hybridization after initial screening by PCR (Figure 3). Phenotypic
characterization of the mutant strains revealed no growth defects
at 37, 42, and 45uC when compared to the background strain
AK07 (Dura3). While tolerance to alkylating agents is a hallmark of
bacterial and eukaryal MMR systems, we found no increased
tolerance to alkylation when the mutants were exposed to 600mg/
L N-Methyl-N9-Nitro-N-Nitrosoguanidine (MNNG) (Figure 4).
Mismatch repair deletion mutants were also constructed in
another background strain, CB08 (Dura3Dzim), to confirm the
of mutant strainsCB071
validity of our gene deletions. All mismatch deletion mutants in the
Dura3Dzim background showed the same phenotypes as the
mutants constructed in the Dura3 background, which together
with PCR and Southern blot analyses established that we
constructed true gene deletions.
Significant increases in mutation rate following deletion of
mismatch repair genes has been demonstrated in both Bacteria
and Eukarya [1,2,3]. Here we designed a 5-FOA mutation
frequency assay to test whether our H. salinarum NRC-1 MMR
deficient strains showed increased mutation rates. This assay,
targeting forward mutations in a plasmid copy of the pyrF (ura3)
gene, allowed us to compare the mutation frequencies for each of
the mutant strains tested with that of the background strain. The
use of a plasmid assay was born out of necessity because our MMR
deletions were constructed in a pyrF deletion background; the
limited genetic markers available for H. salinarum NRC-1
constrained our ability to restore the pyrF gene into the genome
of H. salinarum NRC-1. Using fluctuation tests, we obtained
comparable mutation frequencies (ratio of 5-FOA-resistant
colonies to the average number of colonies plated) for the
background and deletion mutant strains (Table 3). Sequencing
of 20 mutants for strain AK07 (Dura3), 22 mutants for strain
CB071 (DmutS1A), 19 mutants for strain CB073 (DmutS1AD-
mutS1B), and 33 mutants for strain CB074 (DmutL) showed that
approximately 70% of the mutants had changes in the plasmid
copy of the pyrF gene and that the distribution of BPS, insertions
and deletions in the pyrF gene of the MMR mutants was similar to
that found for the control strain AK07 (Table 3). We found that all
the changes in the plasmid copies of the pyrF gene were BPS, with
the exception of the insertion of a T in one of the CB071
(DmutS1A) mutants, and that most BPS were C/G or G/C
transversions (55 to 75%).
Figure 1. Distribution of mutations in 5-FOA-resistant mutants.
Insertions, deletions and base pair substitutions (BPS) in the pyrF and
pyrE2 genes were obtained by sequencing 5-FOA-resistant uracil
auxotrophs of H. salinarum NRC-1.
doi:10.1371/journal.pone.0009045.g001
Table 1. 5-FOA-resistant uracil auxotrophs of H. salinarum
NRC-1 with mutations in multiple UMP biosynthetic genes.
Mutations(1)
Genes analyzed# of clones sequenced pyrFpyrE2 none
pyrF4211
pyrE1
pyrE2
pyrF195 nd 14
pyrE1
pyrF 161(2)
151
pyrE2
pyrE18 nd44
pyrE2
(1)no mutation in pyrE1.
(2)mutation in both pyrF and pyrE2.
nd not determined.
doi:10.1371/journal.pone.0009045.t001
Table 2. Types and positions of spontaneous mutations in
the pyrF and pyrE2 genes.
Gene
position (bp)
Number of independent
isolates with this mutationType of mutation(1)
pyrF
3544Insertion (C)
3564Insertion (C)
3574Insertion (A)
3594 Insertion (CG)
3601Insertion (A)
3614GRC(2)
3624 Insertion (TC)
3654Insertion (G)
3978GRT(2)
4631CRA(2)
7983Insertion (C)
pyrE2
38242Deletion (GTCGACG)
39042CRT
39242Deletion (T)
394 42CRG(2)
39742 Deletion (G)
(1)insertion indicated was found prior to the stated base pair position.
(2)non-synonymous base pair changes.
doi:10.1371/journal.pone.0009045.t002
Halobacterium Mismatch Repair
PLoS ONE | www.plosone.org3February 2010 | Volume 5 | Issue 2 | e9045

Page 4

Figure 2. Location of mutations in the pyrE2 and pyrF genes. Mutations were identified by sequencing 5-FOA-resistant mutants. One-letter
code for amino acid is under the gene nucleotide sequence; start and stop codons are boxed with solid lines; putative TATA box is boxed with a
dotted line; , indicates insertion of the base(s) specified in parenthesis next to the symbol; n indicates deletion of bases directly located below the
symbol; BPS changes are indicated above the sequence in bold; highlighted in green and yellow in pyrE2 are the two 9-nt direct repeats. (A) pyrE2
gene and (B) pyrF gene.
doi:10.1371/journal.pone.0009045.g002
Halobacterium Mismatch Repair
PLoS ONE | www.plosone.org4February 2010 | Volume 5 | Issue 2 | e9045

Page 5

Discussion
The MMR pathway is essential for maintaining genome
stability by correcting errors introduced by DNA polymerases
during DNA replication [1,2,3]. We determined the genomic
mutation rate of the halophilic mesophile, H. salinarum NRC-1,
one of the few archaea to encode for a bacterial-like version of the
conserved MutS and MutL proteins. We found that H. salinarum
NRC-1 mutation rate was similar to that previously calculated for
DNA-based microorganisms, with 1.6761023mutations per
genome per replication, suggesting the presence of high fidelity
replication machinery in this organism [4,12,28,29,30,31,32]. The
DNA-based organisms that formed the basis for this genomic
mutation rate comparison are all mesophiles, belong to the
Bacteria and the Eukarya domains, and include several bacterio-
phages. When we compared the mutation rate of H. salinarum
NRC-1 with that of organisms adapted to high temperature, we
found that it was almost twice that of the acidophilic archaeon, S.
acidocaldarius, and more than an order of magnitude higher than
that of the bacterium Thermus thermophilus, with 1.861023and
9.761024mutations per genome per replication, respectively
[4,33]. While the low genomic mutation rate found in thermo-
philes might be an adaption to extreme temperature conditions
[33], the mutation rate of the mesophilic halophile, H. volcanii was
also found to be extremely low with 4.561024mutations per
genome per replication [5]. It is possible that this extremely low
genomic mutation rate is the result of phenotypic lag or the effects
of H. volcanii polyploidy [5]. However, H. salinarum NRC-1,
similarly to H. volcanii, has approximately 20 copies of its
chromosome per cell [5,34] suggesting that high gene redundancy
does not necessarily result in mutations not being efficiently
detected if they occur at low copy number.
H. volcanii has an unusual mutational spectrum with a
prevalence of in-frame indels flanked by direct repeats [5]. In H.
salinarum NRC-1, a 9-nt direct repeat in the pyrE2 gene resulted in
a mutational hotspot that strongly biased the mutational spectrum
of this gene toward deletions. Short direct repeats, tandem repeats
and monotonic runs are known to promote strand misalignments
during DNA replication and might explain the increased deletion
frequency we observed in the H. salinarum NRC-1 pyrE2 gene [35].
An alternative explanation might be that this 7-nt deletion and the
4 other mutations found in the pyrE2 gene – shared by 42 of the
mutants - were a pre-existing set of mutations present in the initial
cultures used for the fluctuation tests, which were a subset of the 6
fluctuation tests performed to calculate the genomic mutation rate
of H. salinarum NRC-1. With the exception of the pyrE2 gene
Figure 3. Analysis of deletions in the mutL, mutS1(mutS1A), mutS2 (mutS1B), double mutS, and uvrD genes. Probes for Southern blot
analysis were designed to hybridize to regions 500 nt downstream of the target genes coding region. PCR analysis primers were located 500 nt
upstream of the start codon of the targeted gene and 1000 nt downstream of the stop codon. (A) Southern hybridizations. (B) PCR analysis: positive
lanes template was wildtype H. salinarum NRC-1 DNA, negative lanes had no template DNA.
doi:10.1371/journal.pone.0009045.g003
Figure 4. Survival of H. salinarum NRC-1 background and
mutant strains to MNNG. H. salinarum NRC-1 background strain
Dura3 (AK07) and mutant strains DmutL (CB074), DmutS1A (CB071),
DmutS1B (CB072), DmutS1ADmutS1B (CB073), and DuvrD (CB081) were
exposed to 50, 100, and 600mg/L of MNNG. Survival was calculated as
the average ratio (N/No) of surviving CFU from treated cultures (N) and
untreated (No) cultures. Data are the average of a least three
independent experiments, with standard errors shown.
doi:10.1371/journal.pone.0009045.g004
Halobacterium Mismatch Repair
PLoS ONE | www.plosone.org5February 2010 | Volume 5 | Issue 2 | e9045