Download A type III-like restriction endonuclease functions as a major barrier to

A type III-like restriction endonuclease functions as
a major barrier to horizontal gene transfer in
clinical Staphylococcus aureus strains
Anna R. Corvagliaa, Patrice Françoisb, David Hernandezb, Karl Perrona,1, Patrick Lindera,2, and Jacques Schrenzelb
a
Department of Microbiology and Molecular Medicine, University Medical Center, University of Geneva, 1211 Geneva 4, Switzerland; and bGenomic Research
Laboratory, Department of Internal Medicine, Geneva University Hospitals, 1211 Geneva 14, Switzerland
Edited* by Alan M. Lambowitz, University of Texas, Austin, TX, and approved May 19, 2010 (received for review January 14, 2010)
Staphylococcus aureus is an versatile pathogen that can cause lifethreatening infections. Depending on the clinical setting, up to
50% of S. aureus infections are caused by methicillin-resistant
strains (MRSA) that in most cases are resistant to many other antibiotics, making treatment difficult. The emergence of communityacquired MRSA drastically changed the picture by increasing the
risk of MRSA infections. Horizontal transfer of genes encoding
for antibiotic resistance or virulence factors is a major concern of
multidrug-resistant S. aureus infections and epidemiology. We identified and characterized a type III-like restriction system present in
clinical S. aureus strains that prevents transformation with DNA
from other bacterial species. Interestingly, our analysis revealed
that some clinical MRSA strains are deficient in this restriction system, and thus are hypersusceptible to the horizontal transfer of
DNA from other species, such as Escherichia coli, and could easily
acquire a vancomycin-resistance gene from enterococci. Inactivation
of this restriction system dramatically increases the transformation
efficiency of clinical S. aureus strains, opening the field of molecular
genetic manipulation of these strains using DNA of exogenous origin.
antibiotic resistance
| transformation | targetron
T
he opportunistic pathogen Staphylococcus aureus colonizes
20–50% of the human population (1, 2) and is responsible for
a broad spectrum of human and animal diseases, ranging from
benign skin infections to such severe diseases as osteomyelitis,
endocarditis, and sepsis (3, 4). A major issue in the treatment of
S. aureus infections is the presence of methicillin-resistant strains
(MRSA) responsible for hospital-acquired infections and, more
recently, highly virulent community-acquired infections. Methicillin resistance is encoded by the mecA gene present on a mobile
genetic element, termed the staphylococcal cassette chromosome (SCC). The origin of this element, as well as of many other
virulence and resistance determinants, remains unclear. Nevertheless, the acquisition of plasmids, transposons, and lysogenic
bacteriophages definitely contributes to the virulence capacity of
this pathogen. Furthermore, the acquisition by MRSA of part of
the vancomycin-resistance encoding operon harbored by Enterococcus strains has been extensively described and represents
a real threat (5, 6).
Acquisition of resistance genes and virulence factors occurs by
horizontal gene transfer (HGT) through the uptake of exogenous
DNA by transduction, transformation, or conjugation. In some
cases, the virulence genes are encoded by a bacteriophage genome (7). An efficient restriction system drastically reduces
the frequency of interspecies and even intraspecies HGT (8, 9).
Three types of restriction systems have been characterized in
detail. The type I restriction system is composed of three subunits:
the specificity subunit (HsdS), which is responsible for the recognition of a specific sequence; the modification subunit (HsdM),
which methylates host DNA to protect it from self-digestion; and
the endonuclease (HsdR), which cleaves the nonprotected DNA
(10). For restriction by the type I restriction system, two inverted
recognition sites are required on the same DNA molecule, and
11954–11958 | PNAS | June 29, 2010 | vol. 107 | no. 26
cleavage occurs on collision of two translocating restriction
complexes away from the recognition site (11). The type II restriction system is well known by all scientists who perform molecular biology experiments (12, 13). It consists of a site-specific
methylase and a site-specific restriction endonuclease that cleaves
DNA at the recognition site. The type III restriction system was
originally described as a phage- or plasmid-encoded restriction
system (EcoP1 and EcoP15) (14, 15). This system encodes a
methylase that is also responsible for sequence specificity and an
endonuclease responsible for cleavage of unmethylated DNA.
The type I and type III restriction endonucleases harbor a superfamily 2 helicase domain (16–18), which is required for translocation on dsDNA (17, 19). Finally, a fourth restriction system
cleaves DNA when it contains modified cytosine residues (20).
Interestingly, the acquisition of the SCCmec element by S.
aureus yielding to MRSA is observed in only a limited subset of
genetic clones, presumably corresponding to different restriction
system specificities (21, 22). Recently, Lindsay et al. (23) reported that the laboratory strain RN4220, a derivative of NCTC
8325, is mutated in the hsdR gene encoding the type I restriction
endonuclease. This strain is widely used for propagation of plasmid DNA before transformation into clinical strains, and thus can
serve as a very important tool for molecular genetic analyses in
S. aureus. More recently, a bovine pathogenic S. aureus strain that
is hypersusceptible to the acquisition of vancomycin resistance
from Enterococcus faecalis has been described (6). This strain is
naturally mutated in both hsdS genes and thus devoid of a functional type I restriction system.
Based on the observation that plasmids prepared in Escherichia
coli fail to efficiently transform S. aureus strains devoid of a functional type I restriction system, we report the identification of
a type III-like restriction system present in most S. aureus strains.
We show that this restriction system represents a major barrier for
transformation in many clinical strains. Inactivation of this type
III-like restriction endonuclease allows for efficient transformation, with broad implications for the epidemiology of HGT of resistance and virulence genes, as well as for the molecular genetic
analysis of clinical strains.
Results
Inactivation of the Type I Restriction Endonuclease HsdR in Clinical S.
Aureus Strains. The type I restriction system of S. aureus, com-
posed of the HsdR endonuclease, the specificity subunit HsdS,
Author contributions: P.L. and J.S. designed research; A.R.C., P.F., and K.P. performed
research; A.R.C., P.F., and D.H. analyzed data; and P.F., P.L., and J.S. wrote the paper.
The authors declare no conflict of interest.
*This Direct Submission article had a prearranged editor.
1
Present address: Department of Botany and Plant Biology, University of Geneva, 1211
Geneva 4, Switzerland.
2
To whom correspondence should be addressed. E-mail: [email protected].
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1000489107/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1000489107
Identification of a New Barrier. In contrast to the laboratory strain
RN4220, in both hsdR-deficient strains (UAMS-1 and SA564),
Table 1. Transformation efficiency (number of transformants/
μg of plasmid DNA)
Source of plasmid DNA
Recipient strain
RN4220
SA564
SA564 hsdR−
SA564 type III−
SA564 hsdR−, type III−
UAMS-1
UAMS-1, hsdR−
UAMS-1, type III−
UAMS-1, hsdR−, type III−
E. coli DH5α
RN4220
SA564
× 10
×
×
×
×
×
×
×
×
×
×
5.8
<1
17
7.6
7.1
<1
<1
35
5.0
3
× 104
× 104
× 102
1.2
5.1
8.9
7.2
3.0
1
15
4
2.9
3
10
104
104
104
104
× 102
3.4
1.3
1.6
1.4
3.0
<1
<1
<1
96
UAMS-1
2
10
104
104
104
103
8.2
2.1
6.9
3.0
6.0
6
<1
10
8
×
×
×
×
×
102
104
104
104
104
Transformation efficiencies were established using 500 ng of plasmid
DNA for SA564 and its derivatives and 1 μg for UAMS-1 and its derivatives.
To compare efficiencies, the same plasmid preparation was used for transformation of the strain collections, and the different competent cell preparations were used for the different plasmid preparations. Strains were
grown to an OD600 = 0.6, and an equivalent of 5–7 mL of culture per transformation was used. The values represent two independent experiments for
plasmid preparations from RN4220 or E. coli and are representative of many
more transformation experiments. Plasmids used: pCN38 or pCN33. <1, no
transformants detected.
Corvaglia et al.
transformation with plasmid DNA prepared from E. coli was not
possible. This inability to transform the hsdR-disrupted strains
with plasmid DNA prepared in E. coli suggests the presence of an
additional barrier for HGT. Thus, we UV-mutagenized the hsdR−
mutant of SA564 and selected for a clone that could be transformed with plasmid DNA purified from E. coli. For this, we
transformed the mutagenized SA564 with the plasmid pNL9162
(ErmR) (26) prepared in E. coli. Eighteen transformant colonies
were obtained. For a second round of enrichment, we then
pooled the candidates in two batches for transformation using
plasmid pCN38 (CmR). Transformation of pool 10–18 resulted in
very high transformation efficiency; thus, clones 10–18 were individually transformed, and clone 17, the one with the best
transformation efficiency, was selected for further investigation.
Whole genome sequencing of the parental and mutant strains
revealed a single base-pair substitution and a frameshift mutation
(additional A at coordinate 865 in a row of six As) in a gene
demonstrating 98% identity to ORF2790 in strain NCTC 8325.
To exclude the possibility of an undetected second-site mutation, we confirmed the phenotype of the UV-induced mutation
by directed inactivation of the candidate gene using the targetron
disruption method in both SA564 hsdR− and UAMS-1 hsdR−. As
expected, the resulting double-mutant strains could be efficiently
transformed directly with plasmid DNA from E. coli (Table 1).
Because the phenotype of our double-mutant strains is similar to
that of RN4220 in terms of transformation efficiency, we sequenced the corresponding gene in RN4220. The ORF of this
gene in RN4220 is mutated (Fig. S2), explaining why this widely
used laboratory strain can be efficiently transformed even with
foreign DNA and indicating that the identified gene represents
an important barrier for HGT between different bacterial species. Moreover, the transformation efficiency with plasmid prepared from RN4220 was higher in the UAMS-1 double mutant
compared with the UAMS-1 hsdR− single mutant (Table 1),
indicating that the discovered gene also may function as a barrier
between different S. aureus lineages.
To confirm that the increased transformation efficiency observed in the mutant strains is indeed due to the disruption of the
identified gene, we cloned the ORF, including 500 nucleotides
of the upstream and downstream flanking regions, into pCN38.
The resulting plasmid was transformed into the SA564 strain
disrupted for the identified gene and into RN4220. Transformation into the complemented strains using pCN33 was strongly
reduced (by a factor of 100), whereas transformation into the
strains carrying the empty vector remained highly efficient. As a
control, we transformed the complemented strains with plasmid
DNA prepared from RN4220, which showed the same transformation efficiency for the complemented and noncomplemented
strains. The low level of transformation in the complemented
strains is most likely due to inefficient expression, because the
complementing plasmid carries only the second cistron, not the
entire operon.
Identified Gene Resembles a Restriction Endonuclease. The candidate protein displays a superfamily 2 helicase domain and long
N- and C-terminal domains (Fig. 1). Such a primary structure is
typical for type I or type III restriction endonucleases that
translocate in an ATP-dependent manner on the dsDNA before
cleavage of the target DNA (28). The known type I restriction
endonuclease HsdR in S. aureus strains is encoded at a locus
distant from the hsdM-hsdS operons; thus, the identified gene
also could represent a second copy of a HsdR restriction endonuclease. Nevertheless, a blast analysis of the protein sequence
revealed high conservation of this protein with sequences automatically annotated in trEMBL as type III restriction endonuclease. The “classical” type III restriction endonucleases in other
bacterial species generally are encoded by an operon that includes a methylase (29, 30); however, the locus encoding the
PNAS | June 29, 2010 | vol. 107 | no. 26 | 11955
MICROBIOLOGY
and the HsdM methylase, is an important barrier to HGT (23).
Most S. aureus strains contain two operons encoding for HsdS
and HsdM. Alignment of the HsdS protein sequences from
NCTC 8325 (parent of RN4220, clonal complex CC8) with those
of the CC30 strain UAMS-1 (24) and CC5 strain SA564 (25)
reveals important differences in their respective target recognition domains (TRDs) (Fig. S1). Each TRD binds to part of the
recognition sequence, and identical sequences indicate identical
recognition sequences, whereas differences in the TRD indicate
different target sequences. Based on our sequence alignment,
three of the four TRDs from NCTC 8325 are different from those
of UAMS-1, but only one is different from SA564. These differences explain the type I restriction barrier and the impossibility
to transform UAMS-1 with plasmid DNA prepared from RN4220
(23). Consequently, we inactivated the hsdR gene in the clinical
strains UAMS-1, which could not be transformed at all, and
SA564, which could be transformed only with plasmid DNA from
RN4220. The site-directed disruption of the hsdR genes in these
strains was performed using the targetron system (mobile group
II introns) adapted for S. aureus (26) through introduction of the
targetron plasmid by phage transduction. As expected, the resulting UAMS-1 hsdR− mutant strain could be transformed with
plasmid DNA (pCN33, pCN38, and pNL9162) (26, 27), albeit at
low frequency (Table 1). The transformation efficiency with
plasmid DNA prepared from RN4220 into SA564 was similar in
the mutant and parent strains. In general, we observed a lower
transformation frequency with plasmids prepared from S. aureus
compared with those prepared from E. coli, most likely due the
different protocols for plasmid preparation.
Importantly, because the nucleotide sequence of the target
DNA for integration of the targetron in different S. aureus
genomes (e.g., UAMS-1, SA564, USA300, Mu3, Newman, JH1,
JH9 Mu50, COL, NCTC 8325, MSSA476, MW2, N315, MRS252)
is highly conserved, the targeting of the intron in the hsdR gene
should be feasible in most strains. Thus, it should be possible to
render many clinical strains easily transformable with plasmid
DNA prepared from RN4220, provided that transducing phages
are available.
MSRLLNDFNQSLKKGFIDKDISHKGNYTPKLLVNNKNEKVLSTIIDELQK
CETFYFSVAFITESGLASLKAQLLDLSNKGVKGKILTSNYLGFNSPKMYG
ELLKLKNVEVRLTDIAGFHAKGYIFEHKDYSSMVIGSSNLTSNALKVNYE
HNVLLSTMKNGDLVDSVKNEFELLWQKSTPLTEQWINSYKESFEYRSLEK
LAEVEQTQMLLADKVKKSVEIVPNLMQAEALRSLKAIRDKTKDKALIISA
TGTGKTILCALDVREVNPNKFLFIVHNEGILNRAKEEFKKVLPIKNDSDF
active, or that the plasmids used for transformation do not
contain any suitable recognition site.
Similarly, our sequence alignment of the type III-like restriction endonuclease also revealed a nonsense codon at coordinate
444 in strains N315, Mu3, and Mu50 (Fig. S2). Transformation of
Mu3 and Mu50 with plasmid pCN38 was indeed as efficient with
plasmid DNA prepared from E. coli as with plasmid originating
from RN4220.
GLLTGKHRDVDAKYLFATIQTLSRDDNFKQFDENEFDYIVFDEAHRSAAS
TYQRVFNYFKPKFMLGMTATPERSDELSIFELFDYNIAYEIRLQAALESD
ILCPFHYFGVTDYVHQGIKEDDVTKLRYLTSDERVNYIIQKTDYYGYSGE
ILQGLIFVSSKKEAYDLADKLSSKGIKSVALTGDDSVNYRQIVIEKLKEG
KINYIITVDLFNEGIDIPEVNQVVMLRPTESSIIFIQQLGRGLRKSANKE
YVTVIDFIGNYKTNYLIPIALSGDQSQNKDNYKKFLTNNDSINGVSTINF
EEVAKKQIYNSLDAVSLNQNKLILKAYEEVENRLGHMPLLMDFIQQHSID
PSVIFSKFSNYYEFLLRYKKIDALLTENESKNLVFFSRQIAPGLKRIDSL
VLEELLKNELTYDELKNKMLNEVKDITEDDIDTSLRILDFSFYNAGIEKI
YGSPIIECNERMIRLSDAFTNALSNQTFKIFLEDLIELSKYNNEKYQKGK
NGLILYNKYSREDFSKIFNWNKNGSSVIMGYMIRSQEMPIFITYDKHEDI
SDSTKYEDEFLSQDELKWFTKSNRTLKSKEVQKILSHRAKGIKMYIFVQK
KDDDGIYFYYLGTAGYIEGSEKQDKMPNGSNVVTMDLALDKAVRDDIYRY
ITN
Fig. 1. Sequence of the type III-like restriction endonuclease from strain
SA564. The sequence corresponding to the SF2 helicases is indicated in
purple, and some conserved residues of the typical helicase motifs I, II, III,
and VI are indicated in yellow.
novel restriction endonuclease is consistently preceded by a short
ORF encoding a MutT homolog [encoding a putative dGTP
pyrophosphohydrolase (31)] but, surprisingly, not by a methylase. Because the genetic organization of the potential restriction
endonuclease does not include a methylase, it is formally possible that this gene is an alternative endonuclease of the type I
restriction system.
Consequently, we deemed it important to examine whether the
identified gene could be a type I restriction endonuclease. In the
absence of the two specificity genes hsdS1 and hsdS2, the HsdR
endonuclease is not active, as observed in a bovine S. aureus
pathogenic isolate from ST151 lineage (6). Thus, we disrupted the
two hsdS genes in the WT SA564 strain using the targetron
method. But despite disruption of the two hsdS genes, this mutant
SA564 was still resistant to transformation with plasmid DNA
from E. coli. Therefore, we conclude that the identified endonuclease does not require hsdS genes and thus is not a type I
restriction endonuclease. Rather, we suggest that this gene encodes a type III–like restriction endonuclease whose cognate
methylase remains to be identified.
Some Clinical MRSA Strains Are Deficient in the Type III–Like
Restriction Endonuclease. Our analysis of susceptibility to acquire
exogenous DNA in major endemic MRSA clones identified two
strains (HUG-A and HUG-B) of sequence type 228 (ST228) as
being easily transformable with plasmid DNA prepared from
E. coli. This suggests that these strains are missing a restriction
barrier and thus might be highly susceptible to the acquisition of
additional resistance or virulence determinant genes. Thus, we
sequenced the type I restriction system (hsdS1, hsdS2, hsdR) and
the type III-like restriction system of these strains. Whereas the
nucleotide sequences of hsd genes are identical to those of the
WT, the locus of the type III-like restriction endonuclease is
partially deleted in these strains. As in the case of SA564, this
suggests that the Hsd restriction system in these strains is not
11956 | www.pnas.org/cgi/doi/10.1073/pnas.1000489107
Efficient Transfer of DNA from E. faecalis into S. aureus. It is interesting to note that the bovine pathogen lineage ST151 is highly
susceptible to acquisition of the vanA-encoding transposon naturally harbored by enterococci, and that this clone shows mutations
in both hsdS1 and hsdS2 genes (6). Our results suggested that
mutations in hsdS1 and hsdS2 are not sufficient, and thus we analyzed the RF122 genome of the ST151 lineage for the type III–
like restriction endonuclease (32). Interestingly, this genome does
not contain any homolog of the newly discovered type III–like
restriction endonuclease system.
To test whether the type III–like restriction system is indeed
a barrier for the transfer of DNA from enterococci to S. aureus,
we prepared pMW401 (33) plasmid DNA from E. faecalis (34)
and transformed the parent and mutant SA564 strains. Whereas
the parent and the hsdR mutant strain could not be transformed
or could be only barely transformed (i.e., no colonies detected for
1 μg of DNA), the mutant strain with a disrupted type III–like
restriction endonuclease and the double mutant (type I– and type
III–like) were both readily transformable, albeit at low frequency
(95 and 330 colonies per μg, respectively). This shows that the
type III–like endonuclease gene functions as an important barrier
for the transfer of DNA from E. faecalis to S. aureus.
Discussion
We report the identification of a genetic barrier to HGT in
S. aureus. The transfer of genetic information among bacteria of
the same species, as well as among phylogenetically distant species,
is important in the evolution of these organisms. In bacteria of
medical importance, such as the opportunistic pathogen S. aureus,
the transfer of antibiotic resistance and virulence genes represents
a major concern. The discovery of the transfer of the vancomycinresistance gene vanA on transposon Tn1546 harbored by enterococci to S. aureus demonstrates the importance of HGT (5, 35).
Similarly, several decades after the identification of the genetic
basis of methicillin resistance in S. aureus, the origin of the
SCCmec element and its transfer are neither established nor
fully understood.
Nevertheless, the efficient transfer of genetic information between bacteria of different species, even those closely related
from a phylogenetic standpoint, is often limited by one or more
restriction systems. The type I restriction system in S. aureus has
been described as an efficient barrier for genetic transfer by
Waldron and Lindsay (23). Those authors showed that the hsdS1
and hsdS2 genes can differ among S. aureus isolates, thus limiting
intraspecies transfer of DNA and shaping the highly clonal
structure of MRSA (21). The difference in the type I recognition
sequences explains our inability to transform the clinical strain
UAMS-1, even with DNA propagated in strain RN4220. Accordingly, disruption of the hsdR gene in UAMS-1 permits its
transformation with plasmid prepared from RN4220 (Fig. 2).
Sung and Lindsay (6) described a ST151 strain that is hypersusceptible to the transfer of resistance genes from enterococci. This
strain is deficient in the two hsdS genes and thus is unable to use
the HsdR restriction endonuclease.
Despite inactivation of the type I restriction system in UAMS-1,
transformation of this strain with plasmid DNA prepared from
RN4220 remains very inefficient. This suggests that UAMS-1 has
other barriers that reduce HGT. One of these barriers could be
the presence of a capsule that reduces transformation efficiency.
Corvaglia et al.
ty
pe
III
RN4 2 2 0
h sd
R
t yp
eIII
UAMS-1
SA5 6 4
hsdR
t ypeIII
Fig. 2. Summary of intrastrain transfer of plasmid DNA and its barriers.
Black arrows indicate compatibility for transformation, and red arrows denote compatibilities generated by mutations, indicated by (|—). For the introduction of plasmid DNA by transformation into clinical S. aureus strains,
the DNA is first transformed into the laboratory strain RN4220 (black arrow).
From there, plasmid DNA may be transformed into some clinical strains, like
SA564 (black arrow), but not into others (UAMS-1). Only after disruption of
the hsdR gene in UAMS-1 can plasmid DNA prepared from RN4220 be
transformed into UAMS-1 (red arrow). Finally, disruption of the gene
encoding the type III-like restriction endonuclease improves transformation
from RN4220 into UAMS-1 and, most importantly, allows direct transformation from E. coli (or E. faecalis) into the clinical strains SA564 and
UAMS-1 (red arrows).
Indeed, agglutination assays have shown that the UAMS-1 strain
is heavily capsulated (36). Moreover, preparation of competent
cells requires the use of B2 medium supplemented with histidine,
indicating that the growth conditions affect the constitution of the
bacterium. In accordance with this observation, transformation of
UAMS-1 WT or hsdR− with “self”-DNA is possible, but is inefficient compared with that of RN4220 or SA564.
The additional barrier also could be due to the presence of
another restriction system. A survey of S. aureus genomes at rebase
(rebase.neb.com) indeed shows the presence of type II or type IV
restriction systems in some strains. Our findings identify an additional system (Figs. 1 and 2). Based on the protein sequence with
the presence of a superfamily 2 helicase domain, the identified
gene may encode a type I or type III restriction enzyme. Because
the inactivation of both hsdS genes does not abolish the transformation barrier, we propose that this gene is a type III–like restriction endonuclease. Nevertheless, further analysis of this
protein in vitro and the identification of its dedicated methylase is
needed to verify it as a bona fide type III restriction endonuclease.
Interestingly, this type III–like restriction system has escaped
previous identification. Our sequence analysis of the corresponding gene in RN4220 reveals the presence of a nonsense
codon, explaining the possibility of transforming this strain directly with plasmid DNA prepared from E. coli. Similarly, analysis
of the genome of ST151 revealed a deletion of this locus, again
explaining the hypersusceptibility of strains from ST151 lineage to
accept foreign DNA (6). The high transformation efficiency of
the local MRSA strains also can be explained by partial deletion
of the corresponding gene (Fig. S2).
Surprisingly, the SA564 strain with a disruption of the type III–
like gene can be directly transformed with plasmid DNA prepared from E. coli (Fig. 2). This might be due to the absence of
hsd expression or the presence of an unidentified antirestriction
system; however, we cannot exclude the possibility that our various plasmids do not contain the adequate recognition sites, even
though the likelihood of this is very low. Nevertheless, it should
be kept in mind that even a genome as large as that of phage T7 is
insensitive to the type III restriction system, because all 36 recCorvaglia et al.
ognition sites are present in the same orientation and efficient
restriction requires a head-to-head orientation of two sites for
cleavage to occur (37).
In the present work, we mainly used transformation assays to
test for HGT. Although transduction and conjugation play a very
important role in HGT (23, 38), we believe that our tests accurately measure the barrier effects of the different restriction
systems. Moreover, S. aureus is often present in biofilms with the
presence of extracellular DNA, and transformation might occur
under such circumstances. In light of the coexistence of S. aureus
with other bacteria, such as enterococci, transformation in particular (and HGT in general) appears to be a crucial mechanism
for exchanging genetic material.
In accordance with the foregoing results, we have shown that
disruption of the hsdR and type III–like genes in a nontransformable clinical S. aureus strains render them accessible to
transformation with plasmid DNA prepared in bacteria of the
same species. This opens the possibility of performing molecular
genetic screens using plasmid-based libraries. Because the targetron integration sequences in the hsdR and type III–like genes are
highly conserved, this strategy should be applicable to almost any
clinical strain. Therefore, the only limiting factor is the availability
of a transducing phage to introduce the targetron plasmid in the
first instance. Additional restriction systems, such as the type II or
type IV restriction systems, may reduce the frequency of transformation in certain clinical strains, but because none of these
barriers is absolutely tight, these systems should not impede the
molecular genetic analysis of these strains.
In conclusion, we have identified a hitherto undescribed S. aureus gene that encodes a type III–like restriction endonuclease.
The inactivation of this gene renders clinical strains hypersusceptible to the acquisition of foreign DNA, such as the vancomycinresistance gene from E. faecalis, a remotely related bacterial species. Moreover, the inactivation of this gene opens the possibility of
performing simple molecular genetic experiments, such as transformation with multicopy or expression libraries. Because the hsdR
and type III–like restriction endonuclease genes are highly conserved among the clinical strains of S. aureus, the targetron disruption method used in this study should be easily applicable to
most of these strains, providing a valuable tool for addressing the
challenging topic of adaptability and virulence in clinical strains.
Materials and Methods
Strains and Media. E. coli DH5α was used for all cloning experiments and for
plasmid preparations. The Staphylococcus and Enterococcus strains are listed
in Table S1. Strains HUG-A and -B are prevalent isolates encountered in our
local hospital settings, responsible for >90% of our MRSA infections. These
strains, belonging to clonal complex 5 and showing sequence type 228, have
been identified in various European countries for decades. They harbor a
SCCmec type I element encoding for methicillin resistance and display numerous other resistance determinants, suggesting HGT ability.
The media used included Mueller-Hinton broth (MHB; Difco), TSB (bioMérieux), and B2 (39), supplemented with 60 μg/mL of histidine. Erythromycin (10 mg/L) and chloramphenicol (10 mg/L) were added as appropriate.
Transformation and Bacteriophage Transduction. Transformation was carried
out as described previously (40). Strains were grown in MHB medium before
preparation of competent cells, except for UAMS-1, which was grown in B2
medium in the presence of histidine (39). Phage transduction was carried out
as described previously (41).
Targetron Disruptions. Targetron insertions were carried out according to the
procedure described by Yao et al. (26) and the Sigma-Aldrich targetron site
(http://www.sigmaaldrich.com/life-science/functional-genomics-and-rnai/
targetron.html). Plasmid-carrying cells were screened by PCR using internal
(intron-specific) and external primers for integration of the intron in the
target gene (Table S2). After loss of the targetron plasmid, the insertion was
reconfirmed by PCR analysis.
The hsdR disruption plasmid was constructed to insert the group II intron
of plasmid pNL9162 at position 739 in the hsdR gene of UAMS-1. Table S2
PNAS | June 29, 2010 | vol. 107 | no. 26 | 11957
MICROBIOLOGY
E. coli
lists the oligonucleotides used. After transformation into RN4220, the plasmid was transduced using phage Φ11 into the UAMS-1 strain. For introduction of plasmids into SA564, plasmid DNA was prepared in RN4220
and then electroporated into SA564.
For disruption of the type III–like endonuclease using the targetron
method, the intron was inserted at coordinate 735. For the construction of
the double mutants and the type III–like mutation in SA564 parental strain,
plasmid DNA was prepared from RN4220. For the disruption of the type III–
like gene in UAMS-1, the plasmid was introduced by phage transduction, as
described above. Disruptions of the hsdS1 and hsdS2 genes in SA564 were
performed using the same procedure.
Cloning the Type III-Like Restriction Endonuclease. For complementation of
the type III–like restriction endonuclease mutant, we amplified the ORF with
500 nucleotides upstream and downstream by PCR and cloned the resulting
fragment into the BamHI and EcoRI sites of pCN38.
Mutagenesis. UV-induced mutagenesis was carried out on bacteria grown
overnight in MHB and resuspended in 0.9% NaCl. The survival rate was
measured by plating after mutagenesis and found to be ~80%. Cells were
then diluted into MH medium and grown to OD600 = 0.6 for preparation of
electrocompetent cells. The transformation was done with the temperaturesensitive plasmid pNL9162 (26) prepared in E. coli DH5α. In a second round of
transformation, the pools of candidate clones were transformed with pCN38
(CmR) (27).
in MHB (10 mL) for 5 h. Bacterial cells were rinsed twice in 10 mL of TrisEDTA (10 mM and 1 mM, respectively), suspended in 2 mL of Tris-EDTA
containing 100 μg/mL of lysostaphin (Ambicin; Applied Microbiology), and
then incubated for 10 min at 37 °C. DNA was then extracted and purified
using the DNeasy Kit (Qiagen). DNA purity and quantity were assessed with
a NanoDrop 1000 spectrophotometer (Thermo Scientific). Whole genome
sequencing was performed using an Illumina Genome Analyzer GAII. In brief,
5 μg of genomic DNA was physically fragmented by nebulization into 50- to
500-bp fragments. After end repair and ligation of the bar-coded paired-end
adaptors, the products were purified on agarose gel to recover products with
inserts of ~200 bp. Quality control was performed by cloning an aliquot of the
library into a TOPO plasmid and capillary sequencing eight clones per library.
The samples were then used to generate DNA colonies (or DNA clusters) using
one channel of a paired-end flow cell at dilutions of 4 pM. The flow cell was
then submitted to 2 × 76 cycles of sequencing on the genome analyzer. Base
calling was performed using the GAPipeline 1.3.2 software; a total of 11.1
million reads (pass filter) were obtained. After bar code selection, between
2.4 and 3.4 million reads of 2 × 71 bases were obtained for each library.
Assembling was performed using the Edena assembler (42), and comparison
were performed using the MUMmer software package (43).
Genomic Sequencing. Genomic DNA from the hsdR− mutant of SA564 and
from the mutant clone 17 was prepared according to the method of Hernandez et al. (42). The DNA was sequenced by Fasteris. Strains were grown
ACKNOWLEDGMENTS. We thank M. Tangomo-Bento and S. Luethi for
technical support, D. Belin and P. Viollier for helpful discussions, and J.
Casadesus and P. Redder for comments on the manuscript. This work was
supported by funding from Infectigen (to J.S.), the Bonninchi Foundation (to
P.L.), the Ernst et Lucie Schmidheiny Foundation (to P.L.), the Swiss National
Science Foundation (Grants 31003A-118309 to P.L., 3100A0-116075 to P.F., and
3100A0-112370/1 to J.S.), and the Canton de Genève.
1. Maslow JN, et al. (1995) Variation and persistence of methicillin-resistant Staphylococcus
aureus strains among individual patients over extended periods of time. Eur J Clin
Microbiol Infect Dis 14:282–290.
2. Kluytmans J, van Belkum A, Verbrugh H (1997) Nasal carriage of Staphylococcus
aureus: Epidemiology, underlying mechanisms, and associated risks. Clin Microbiol
Rev 10:505–520.
3. Leonard FC, Markey BK (2008) Mehticillin-resistant Staphylococcus aureus in animals:
A review. Vet J 175:27–36.
4. Lowy FD (1998) Staphylococcus aureus infections. N Engl J Med 339:520–532.
5. Weigel LM, et al. (2003) Genetic analysis of a high-level vancomycin-resistant isolate
of Staphylococcus aureus. Science 302:1569–1571.
6. Sung JM, Lindsay JA (2007) Staphylococcus aureus strains that are hypersusceptible to
resistance gene transfer from enterococci. Antimicrob Agents Chemother 51:2189–2191.
7. Kaneko J, Kimura T, Narita S, Tomita T, Kamio Y (1998) Complete nucleotide sequence
and molecular characterization of the temperate staphylococcal bacteriophage
phiPVL carrying Panton-Valentine leukocidin genes. Gene 215:57–67.
8. Hoskisson PA, Smith MC (2007) Hypervariation and phase variation in the
bacteriophage “resistome.” Curr Opin Microbiol 10:396–400.
9. Tock MR, Dryden DT (2005) The biology of restriction and anti-restriction. Curr Opin
Microbiol 8:466–472.
10. Murray NE (2000) Type I restriction systems: Sophisticated molecular machines (a
legacy of Bertani and Weigle). Microbiol Mol Biol Rev 64:412–434.
11. Janscak P, MacWilliams MP, Sandmeier U, Nagaraja V, Bickle TA (1999) DNA
translocation blockage, a general mechanism of cleavage site selection by type I
restriction enzymes. EMBO J 18:2638–2647.
12. Sussenbach JS, Steenbergh PH, Rost JA, van Leeuwen WJ, van Embden JD (1978) A
second site-specific restriction endonuclease from Staphylococcus aureus. Nucleic Acids
Res 5:1153–1163.
13. Pingoud A, Fuxreiter M, Pingoud V, Wende W (2005) Type II restriction endonucleases: Structure and mechanism. Cell Mol Life Sci 62:685–707.
14. Hadi SM, Bächi B, Iida S, Bickle TA (1983) DNA restriction-modification enzymes of
phage P1 and plasmid p15B: Subunit functions and structural homologies. J Mol Biol
165:19–34.
15. Arber W, Wauters-Willems D (1970) Host specificity of DNA produced by Escherichia coli,
XII: The two restriction and modification systems of strain 15T-. Mol Gen Genet 108:
203–217.
16. Bourniquel AA, Bickle TA (2002) Complex restriction enzymes: NTP-driven molecular
motors. Biochimie 84:1047–1059.
17. Webb JL, King G, Ternent D, Titheradge AJ, Murray NE (1996) Restriction by EcoKI is
enhanced by co-operative interactions between target sequences and is dependent
on DEAD box motifs. EMBO J 15:2003–2009.
18. Murray NE (2002) 2001 Fred Griffith review lecture. Immigration control of DNA in
bacteria: Self versus non-self. Microbiology 148:3–20.
19. Bianco PR, Hurley EM (2005) The type I restriction endonuclease EcoR124I couples ATP
hydrolysis to bidirectional DNA translocation. J Mol Biol 352:837–859.
20. Raleigh EA, et al. (1988) McrA and McrB restriction phenotypes of some E. coli strains
and implications for gene cloning. Nucleic Acids Res 16:1563–1575.
21. Lindsay JA (2010) Genomic variation and evolution of Staphylococcus aureus. Int J
Med Microbiol 300:98–103.
22. Robinson DA, Enright MC (2003) Evolutionary models of the emergence of methicillinresistant Staphylococcus aureus. Antimicrob Agents Chemother 47:3926–3934.
23. Waldron DE, Lindsay JA (2006) Sau1: A novel lineage-specific type I restrictionmodification system that blocks horizontal gene transfer into Staphylococcus aureus
and between S. aureus isolates of different lineages. J Bacteriol 188:5578–5585.
24. Gillaspy AF, et al. (1995) Role of the accessory gene regulator (agr) in pathogenesis of
staphylococcal osteomyelitis. Infect Immun 63:3373–3380.
25. Somerville GA, et al. (2002) In vitro serial passage of Staphylococcus aureus: Changes
in physiology, virulence factor production, and agr nucleotide sequence. J Bacteriol
184:1430–1437.
26. Yao J, et al. (2006) Use of targetrons to disrupt essential and nonessential genes in
Staphylococcus aureus reveals temperature sensitivity of Ll.LtrB group II intron
splicing. RNA 12:1271–1281.
27. Charpentier E, et al. (2004) Novel cassette-based shuttle vector system for grampositive bacteria. Appl Environ Microbiol 70:6076–6085.
28. Janscak P, Sandmeier U, Szczelkun MD, Bickle TA (2001) Subunit assembly and mode
of DNA cleavage of the type III restriction endonucleases EcoP1I and EcoP15I. J Mol
Biol 306:417–431.
29. Iida S, et al. (1983) DNA restriction-modification genes of phage P1 and plasmid p15B:
Structure and in vitro transcription. J Mol Biol 165:1–18.
30. Su P, Im H, Hsieh H, Kang’A S, Dunn NW (1999) LlaFI, a type III restriction and
modification system in Lactococcus lactis. Appl Environ Microbiol 65:686–693.
31. Bessman MJ, Frick DN, O’Handley SF (1996) The MutT proteins or “Nudix” hydrolases,
a family of versatile, widely distributed, “housecleaning” enzymes. J Biol Chem 271:
25059–25062.
32. Herron-Olson L, Fitzgerald JR, Musser JM, Kapur V (2007) Molecular correlates of host
specialization in Staphylococcus aureus. PLoS ONE 2:e1120.
33. Wirth R, An F, Clewell DB (1987) Highly efficient cloning system for Streptococcus
faecalis: Protoplast transformation, shuttle vectors, and applications. Streptococcal
Genetics, eds Ferretti JJ, Curtiss RI (American Society for Microbiology, Washington, DC).
34. Perreten V, Schwarz FV, Teuber M, Levy SB (2001) Mdt(A), a new efflux protein
conferring multiple antibiotic resistance in Lactococcus lactis and Escherichia coli.
Antimicrob Agents Chemother 45:1109–1114.
35. Zhu W, et al. (2008) Vancomycin-resistant Staphylococcus aureus isolates associated
with Inc18-like vanA plasmids in Michigan. Antimicrob Agents Chemother 52:452–457.
36. Francois P, et al. (2003) Rapid detection of methicillin-resistant Staphylococcus aureus
directly from sterile or nonsterile clinical samples by a new molecular assay. J Clin
Microbiol 41:254–260.
37. Meisel A, Bickle TA, Krüger DH, Schroeder C (1992) Type III restriction enzymes need
two inversely oriented recognition sites for DNA cleavage. Nature 355:467–469.
38. Ubeda C, et al. (2005) Antibiotic-induced SOS response promotes horizontal
dissemination of pathogenicity island-encoded virulence factors in staphylococci. Mol
Microbiol 56:836–844.
39. Schenk S, Laddaga RA (1992) Improved method for electroporation of Staphylococcus
aureus. FEMS Microbiol Lett 73:133–138.
40. Lee JC (1995) Electrotransformation of staphylococci. Methods Mol Biol 47:209–216.
41. Dyer DW, Rock MI, Lee CY, Iandolo JJ (1985) Generation of transducing particles in
Staphylococcus aureus. J Bacteriol 161:91–95.
42. Hernandez D, François P, Farinelli L, Osterås M, Schrenzel J (2008) De novo bacterial
genome sequencing: Millions of very short reads assembled on a desktop computer.
Genome Res 18:802–809.
43. Kurtz S, et al. (2004) Versatile and open software for comparing large genomes.
Genome Biol 5:R12.
11958 | www.pnas.org/cgi/doi/10.1073/pnas.1000489107
Corvaglia et al.