Previous Article | Next Article ![]()
Journal of Clinical Microbiology, October 2008, p. 3246-3258, Vol. 46, No. 10
0095-1137/08/$08.00+0 doi:10.1128/JCM.00136-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Department of Bacteriology, Juntendo University, Tokyo, Japan,1 Department of Infection Control Science, Graduate School of Juntendo University, Tokyo, Japan,2 Radioisotope Research Facility, The Jikei University School of Medicine, Tokyo, Japan,3 Laboratory of Applied Microbiology, Department of Microbial Biotechnology, Graduate School of Agricultural Science, Tohoku University, Sendai, Japan,4 Division of Laboratory Medicine, Kinan General Hospital, Wakayama, Japan,5 Kotobiken Medical Laboratories, Inc., Ibaragi, Japan,6 Department of Medical Microbiology and Parasitology, China Medical University, Shenyang, China7
Received 24 January 2008/ Returned for modification 27 April 2008/ Accepted 14 July 2008
|
|
|---|
Sa2958-carrying Panton-Valentine leukocidin (PVL) gene, which was lysogenized in a sequence type 5 staphylococcal cassette chromosome mec (SCCmec) type II strain of methicillin-resistant Staphylococcus aureus (MRSA). Based on the nucleotide sequences of PVL phages, we developed PCRs to discriminate among five PVL phages, with a preliminary classification into two morphological groups (elongated-head type and icosahedral-head type) with four PCRs, including two PCRs for identifying the gene lineage between lukS-PV and the tail gene. The phages were then classified into five types by four PCRs identifying each phage-specific structure. With these PCRs, we examined the PVL phage types of 67 MRSA strains isolated in Japan from 1979 through 1985 and since 2000 and found that two morphologically distinct phages were predominant in Japan. The icosahedral-head-type phage, represented by the
108PVL type, was identified for 39 of 53 strains isolated from 1979 through 1985. Of 26 other Japanese isolates, 25 belonged either definitively or presumably to elongated-head types as follows: 3 belonged to the
Sa2958 type; 8 were determined to belong to an elongated-head type, but a determination of greater specificity was not made; and 14 belonged to a
Sa2958-like phage of unknown type. We induced prophages by treatment with mitomycin C from six strains of the
Sa2958 type or of
Sa2958-like unknown-type phages; five of six strains carried intact PVL-carrying phages, which can infect other S. aureus strains and might generate novel PVL-positive strains of S. aureus. That various SCCmec elements were carried by different strains of the same phage type suggests that S. aureus strains might independently acquire PVL phages before they acquire various SCCmec elements. |
|
|---|
PVL was first reported in 1932 (35). It is a two-component toxin composed of LukF-PV, with a molecular mass of 34 kDa, and LukS-PV, with a molecular mass of 32 kDa (36). These two proteins, LukF-PV and LukS-PV, act synergistically and cause damage on cell membranes by forming pores, resulting in the lysis of polymorphonuclear leukocytes and macrophages (6, 13).
The genes for these two proteins, lukS-PV and lukF-PV, respectively, are carried by prophages that are integrated into the S. aureus chromosome. To our knowledge, five PVL phages,
PVL,
SLT,
Sa2mw,
108PVL, and
Sa2usa, have been reported to date (1, 9, 20, 21, 28, 32). These phages carry the lukS-PV and lukF-PV genes, which show more than 99% nucleotide identity and are integrated at the same position in the S. aureus chromosome. However, the morphologies of the phage particles are not identical. The phage morphologies can be classified into two types, namely, the icosahedral-head type and the elongated-head type (4). We have examined PVL-positive MRSA strains isolated from 1979 through 1985 and have determined the nucleotide sequence of
108PVL, carried by SCCmec type IV.3 (SCCmec types are designated hereafter by a capital roman numeral [e.g., type IV] or by a capital roman numeral, a decimal point, and an Arabic numeral [e.g., type IV.3]) sequence type 30 (ST30) MRSA strain 81/108 (28). When we examined the carriage of
108PVL phage in representative MRSA strains isolated from 1979 through 1985, phage
108PVL was most often identified in ST30 MRSA strains, whereas the SCCmec type II.1 ST5 MRSA strain JCSC2958 was negative by five of six PCRs using different sets of primers to identify
108PVL, indicating that the strain might carry another new PVL phage. Both PVL-positive methicillin-sensitive S. aureus (MSSA) strains and MRSA strains have been isolated worldwide, including in the United States (2, 3, 10, 37), France (39), Australia (7), England (15), Canada (30), Singapore (16), Belgium (8), and Uruguay (27). We wondered how these PVL-positive strains evolved by acquiring each PVL phage and how each PVL phage evolved by acquiring the lukS-PV and lukF-PV genes. In an attempt to answer these questions, we determined the nucleotide sequence of a new PVL phage carried by JCSC2958 and developed a PCR system to identify the lysogenized PVL phages. By applying the newly developed PCRs, we have found that Japanese PVL-positive S. aureus strains carry two morphologically distinct PVL phages, namely, the icosahedral-head type, represented by
108PVL phage, and the elongated-head type, represented by
Sa2958, and that these types differ from those for strains disseminated in the United States and France.
|
|
|---|
SLT was also used as a control for the induction experiment for PVL-positive phages. To identify phages induced from lysogenized bacteria, two MSSA strains, RN4220 and 1039, were used as indicator strains. These mutant strains, which lack prophages and restriction systems, were obtained from the NCTC8325 and Terashima strains, respectively (41).
SCCmec typing and identification of virulence genes. SCCmec typing and identification of virulence genes were performed with PCRs as described previously (22, 27, 34).
Determination of the entire nucleotide sequence of
Sa2958.
The fosmid library from the genomic DNA of JCSC2958 was constructed by using the CopyControl fosmid library production kit (Epicentre Biotechnologies, Madison, WI) according to the protocols recommended by manufacturer. Briefly, genomic DNA was prepared from cells of JCSC2958 cells with Isoplant (Nippon Gene Co., Tokyo, Japan). Approximately 400 µg of the genomic DNAs was sheared into small fragments by aspiration and expulsion through an injection needle. Subsequent size selection of 40-kb DNA fragments was performed by running samples in agarose gel (SeaKem gold agarose; Cambrex Bio Science Rockland, Inc., Rockland, ME). To recover the size-fractionated DNA after electrophoresis, agar slices containing the corresponding sizes of DNA were cut and then melted in GElase solution. The recovered DNA was purified and treated with the end repair enzyme mix that came with the kit. The purified insert DNA was ligated into CopyControl pCC1FOS cloning-ready vector with DNA ligase and then packaged according to the manufacturer's instructions. Each 10 µl of the fosmid library was mixed with 100 µl of EPI300-T1R cells and incubated for 20 min to promote phage infection. Each reaction mixture was spread on an L-agar plate containing 12.5 µg/ml of chloramphenicol and incubated at 37°C overnight. Five hundred colonies were screened with two primer pairs for the PVL gene and the integrase gene (Table 1), which are conserved in nearly all extant PVL phages. Plasmids were extracted from positive clones and used as templates for nucleotide sequencing. In some cases, DNA fragments amplified by long-range PCR were used as the templates. Primers used for long-range PCR are listed in Table 1. Nucleotide sequences were subsequently determined with the primer-walking method.
|
View this table: [in a new window] |
TABLE 1. List of primers used in this study
|
Induction of prophages from S. aureus cells. A 0.3-ml portion of overnight culture was inoculated to 3 ml of BHI broth. After the culture had been shaken for 2 h at 37°C, mitomycin C (Kyowa Hakko Kogyo Co., Ltd., Tokyo, Japan) was added to a final concentration of 1 µg/ml, and the culture was cultivated at 37°C with shaking for 50 min. Cells were then precipitated by centrifugation and resuspended with 300 µl of L broth. A 100-µl portion was added to 3 ml of L broth and incubated further at 37°C until the cell suspension became transparent. The supernatant was sterilized by means of a membrane filter with 0.45-µm-diameter pores (Whatman Corp., Clifton, NJ).
Identification of PVL-carrying phages by plaque hybridization. Phage solutions were diluted 10-fold serially, and a 0.1-ml portion of each solution was mixed with a 0.3-ml overnight culture of the RN4220 or 1039 strain and kept at room temperature for 15 min. Three milliliters of L broth containing 0.6% agar was added to each mixture, which was then poured onto heart infusion agar plates. The plates were incubated overnight at 30°C to form plaques on the lawns. Plates with an appropriate numbers of plaques were selected, and the plaques were transferred onto a piece of Biodyne membrane (Pall Biodyne A, pore size, 1.2 µm; Pall Life Sciences, Ann Arbor, MI). The filters were denatured by submersion for 5 min in 0.5 M NaOH and 1.5 M NaCl and then neutralized by submersion twice for 5 min in a solution of 0.5 M Tris-HCl (pH 7.3), 1.5 M NaCl, and 1 mM EDTA.
The DNAs were cross-linked to the nylon membrane by use of the Stratalinker UV cross-linker (Stratagene Japan K. K., Tokyo, Japan). Probes to identify PVL-carrying phages were prepared by labeling DNA fragments with digoxigenin by use of the digoxigenin DNA labeling and detection kit (Roche Applied Science, Penzberg, Germany). The DNA fragments used for plaque hybridization were amplified with PCR and genomic DNA of the JCSC2958 strain as a template with a pair of primers to identify the lukS-PV and lukF-PV genes (Table 1). The subsequent experiment was performed as described by the manufacturer.
PCRs to classify PVL phages.
We developed eight PCRs to classify the PVL-encoding prophages by use of chromosomal DNA preparations from PVL-positive S. aureus strains as templates. Primers used for identifying PVL-encoding phages are listed in Table 1. These primers were designed on the bases of nucleotide sequences of five PVL phages, namely,
108PVL,
Sa2mw,
PVL,
SLT, and
Sa2958, which are deposited in DDBJ/EMBL/GenBank databases under accession nos. AB009866, BA000033, AB009866, NC_002661, and AP009363, respectively. The locations of the primers are shown below (see Fig. 3).
![]() View larger version (43K): [in a new window] |
FIG. 3. Structural comparison of 5 PVL phages. Structures of 108PVL, PVL, SLT, Sa2mw, and Sa2958 are illustrated on the bases of the nucleotide sequences as follows: 108PVL (DDBJ/EMBL/GenBank databases under accession no. AB243556), PVL (DDBJ/EMBL/GenBank accession no. AB009866), SLT (DDBJ/EMBL/GenBank accession no. NC_002661), Sa2mw (DDBJ/EMBL/GenBank accession no. BA000033), and Sa2958 (DDBJ/EMBL/GenBank accession no. AP009363). Genes having nucleotide identities of more than 90% are linked by blocks filled with colors as follows: pink, genes conserved among five PVL phages; light purple, genes conserved among Sa2958, Sa2mw, and SLT; blue, genes conserved between PVL and 108PVL. The positions of primers for M-PCR 1 to 7 are indicated by green arrowheads. Blue arrows indicate ORFs well conserved in all PVL phages other than lukS and lukF genes; black arrows indicate lukS and lukF genes; light green and dark green arrows indicate genes encoding major tail proteins of the elongated-head type and the icosahedral-head type, respectively; and yellow arrows and ocher arrows indicate genes encoding terminase subunits (large and small), portal proteins, prohead proteases, and capsid proteins of the elongated-head type and the icosahedral-head type, respectively.
|
PVL and
108PVL. PCR-2 was designed to identify the carriage of phages with elongated heads by amplifying the portal gene and the head gene, which are conserved in
Sa2mw,
Sa2958, and
SLT. If two DNA fragments could be amplified with PCR-1, we then proceeded to PCR-3, which was designed to identify the gene lineage between the tail gene and the lukS gene and to verify that the tail gene identified with PCR-1 belonged to a PVL-carrying phage with a hexagonal head.
If two DNA fragments were amplified with PCR-2, we proceed to PCR-4, which was designed to identify the gene lineage between the tail gene and the lukS gene to verify that the tail gene of the elongated shape is located in relation to the lukS gene with a primer pair commonly conserved among
Sa2958,
Sa2mw, and
SLT.
PCR-5 to -8 were designed to identify five PVL phages by amplifying the gene lineage between the integrase gene, which is commonly carried by all reported PVL-carrying phages, and genes located on the region related to the lysogeny or recombination of each PVL phage.
PCR-5 was designed to identify
PVL and
108PVL by integrase and two open reading frames (ORFs), the repressor for
PVL and the antirepressor for
108PVL. PCR-6, -7, and -8 were designed to identify
Sa2958,
Sa2mw, and
SLT by detecting the gene lineage between the integrase gene and the genes located downstream of the gene: those for a hypothetical protein (JCP004) for
Sa2958, cro repressor protein for
Sa2mw, and a single-stranded binding protein for
SLT.
Chromosomal DNAs were prepared with the small-scale phenol extraction method and used as PCR templates. The PCRs were performed using a Gene Amp 9600 thermal cycler (Perkin-Elmer Cetus Instruments, Emeryville, CA). The reaction mixtures for PCR-1, -2, and -5 to -8 contained 50 ng of template DNA, each oligonucleotide primer (0.2 mM), 400 mM of each deoxynucleoside triphosphate, 1x Ex Taq buffer with magnesium, and 4 U of Ex Taq polymerase (Takara Shuzo Co. Ltd., Kyoto, Japan) in final volumes of 50 µl. PCRs for five PCRs (PCR-1, -2, -5, -6, and -7) consisted of 30 cycles of denaturation (95°C, 60 s), annealing (50°C, 60 s), and extension (72°C, 2 min).
PCR-3, -4, and -8 were carried out with long-range PCR using an Expand high-fidelity PCR kit adhering to the protocols recommended by manufacturer. The 6-µl PCR mixture was subjected to agarose gel electrophoresis to detect amplified DNA fragments.
Nucleotide sequence accession number.
The entire nucleotide sequence of
Sa2958 has been deposited in the DDBJ/EMBL/GenBank database under accession no. AP009363.
|
|
|---|
Sa2958.
The
Sa2958 phage is 46,046 bp in length, which is a size comparable to those of other PVL phages identified to date. A total of 59 predicted ORFs of larger than 99 bp were identified in
Sa2958 (Fig. 1 and Table 2). The whole G+C content of the novel PVL phage was 33.1%, which is comparable to those of S. aureus genomes. The gene coding potential for
Sa2958 was 90.6%, with approximately 1.28 genes/kbp of nucleotide sequence.
![]() View larger version (17K): [in a new window] |
FIG. 1. (A) The structure of Sa2958. Black arrowheads indicate the locations of primers used to amplify the entire Sa2958 genome. The two red arrowheads flanking the core sequence indicate the att sites on the phage element. (B) The ORFs in Sa2958. The ORFs are shown as squares in six possible reading frames. The direction of the arrow indicates the transcriptional direction for each ORF. Color codes are as follows: dark green, ORFs (or the parts of ORFs) that are well conserved among four other PVL-carrying phages, namely, Sa2mw, SLT, PVL, and 108PVL; light green, ORFs that are highly homologous to Sa2mw and SLT; light purple, ORFs that are highly homologous to Sa2mw; orange, an ORF that is homologous to SLT and PVL; yellow, an ORF that is homologous to SLT and 108PVL; white, ORFs that are unique to Sa2958; light blue, ORFs that are homologous to Sa2mw; dark blue, ORFs that are homologous to SLT.
|
|
View this table: [in a new window] |
TABLE 2. ORFs in and around 2958PVL and their similarities to Sa2mw and 108PVLa
|
Sa2958 exactly matched the corresponding core sequence conserved in
Sa2mw but differed from those of
SLT,
PVL, and
108PVL by 1 bp. The 25-bp sequences of the attB site located on the chromosome side were identified flanking both ends of
Sa2958. The rightmost (attB-R) sequence was identical with that of
Sa2mw, whereas the leftmost (attB-L) sequence differed by 2 bp from that of
Sa2mw and differed significantly from those of
PVL and
108PVL. The 25-bp attachment sites (attP) at both ends of the phage were well conserved among all five PVL phages.
The organization of the
Sa2958 genome was similar to those of other extant PVL phages, indicating that the following order of modules on the phage genome is well conserved: regions related to lysogeny, DNA replication/transcriptional regulation regions, the packaging/head, the tail, and the lysis module as well as lukS-PV and lukF-PV. Fifty-nine ORFs were roughly classified into five groups on the basis of similarities with extant PVL phages.
Group 1 contained 10 genes (JP013, JP014, JP023, JP024, JP027, JP056, JP057, JP058, JP059, and JP060) conserved among five extant PVL phages, with nucleotide identities of more than 90% (Fig. 1). The integrase gene (int) was located at the leftmost side of
Sa2958. It showed high similarity to int genes carried by extant PVL phages, with more than 98% identity in amino acids. ORFs JP056, JP057, JP059, and JP060 encoded holin, amidase, and LukS-PV and LukF-PV proteins, respectively. Both the lukF-PV and lukS-PV genes showed nucleotide identities of more than 99% with those of other PVL phages. In particular, the lukF-PV gene in
Sa2958 exhibited 100% homology with that of
PVL. ORF JP058 encodes 48 amino acids with no assigned function identified by means of a BLASTP search. This ORF was also identified in
PVL, although it was disrupted and its function could not be assigned because of the insertion of a "T" nucleotide.
Group 2 contained 27 ORFs (JP011, JP012, JP028, and JP032 to -055) conserved among
Sa2958,
Sa2mw, and
SLT (Fig. 1). Because eight ORFs that might be associated with phage DNA packaging and head and tail formation are involved (JP034 [terminase small subunit], JP035 [terminase large subunit], JP036 [portal protein], JP037 [prohead protease], JP038 [capsid protein], JP043 and JP044 [major tail protein], and JP047 [phage tail tape measure protein]),
Sa2958 could be determined on the basis of taxonomy criteria as belonging to the same subgroup as
Sa2mw and
SLT.
Group 3 contained 12 ORFs (JP015 to JP022, JP025, JP029, JP030, and JP031) that showed high similarity only with
Sa2mw (Fig. 1). Three genes were associated with phage replication during the phage lyric process. Representative genes in group 3 encode DNA polymerase (JP019), DNA-binding protein (JP021), helices (JP031), and virulence-associated protein E (JP029).
Group 4 contained three ORFs that were homologous only to one or two other phages. JP026, JP003, and JP010 (Fig. 1) were homologous only to
108PVL and
SLT, only to
PVL and
SLT, and only to
108PVL, respectively.
Group 5 contained six ORFs (JP004 to -009) that were uniquely conserved in
Sa2958 (Fig. 1). Interestingly, JP006, containing the transcription repressor gene, was not homologous to the corresponding genes in other extant PVL phages. A search with the BLASTP program revealed that JP006 showed 100% identity with the corresponding gene conserved in lytic phage 47 and showed 65% identity with that in lytic phage 29, whose entire genomic sequences have been determined by Kwan et al. (23).
The JP006 protein contains a helix-turn-helix DNA-binding motif that is present in a large family of transcriptional regulators. Another unique ORF, JP008, which is conserved in
Sa2958 and was identified as an antirepressor gene, was shown to have 93% similarity with the corresponding gene in phage 47. The remaining four unique ORFs in
Sa2958 could not be assigned any functions through searches with the BLASTP program.
Characterization of PVL-carrying prophages in Japanese MRSA strains.
We conducted PCR experiments to identify PVL phages carried by 53 MRSA strains isolated in from 1979 through 1985 and 12 MRSA strains isolated in the 2000s with PCRs as described in Materials and Methods. The representative results of these PCRs are shown in Fig. 2 and the results of PCR experimentation are summarized in Table 3. When PCR-1 was performed with the chromosomal DNAs of these same five strains carrying each PVL phage, DNA fragments of two expected sizes, 489 bp for the tail gene and 569 bp for the head gene, were amplified with chromosomal DNAs of 81/108 (
108PVL), ATCC 49775 (
PVL), and MW2 (
Sa2mw). When PCR-2 was performed with chromosomal DNAs of these same five strains, two DNA fragments of expected sizes, 466 bp for the tail gene and 655 bp for the head gene, were amplified with chromosomal DNAs of JCSC2958 (
Sa2958), RN4220 (
SLT), MW2 (
Sa2mw), and 81/108 (
108PVL).
![]() View larger version (59K): [in a new window] |
FIG. 2. Representative results of seven multiplex PCRs (M-PCRs) for identifying PVL phages. A 1-kb molecular weight marker (MWM) was run on both sides of the gel. Lanes: 1, PCR-1 using chromosomal DNA of 81/108; 2, PCR-2 using chromosomal DNA of JCSC2958; 3, PCR-3 using chromosomal DNA of 81/108; 4, PCR-4 using chromosomal DNA of JCSC2958; 5, PCR-5 using chromosomal DNA of 81/108; 6, PCR-5 using chromosomal DNA of ATCC 49775; 7, PCR-6 using chromosomal DNA of JCSC2958; 8, PCR-7 using chromosomal DNA of MW2; and 9, PCR-8 using chromosomal DNA of RN4220 ( SLT).
|
|
View this table: [in a new window] |
TABLE 3. Results of PCR experimentationa
|
Sa3mw, has the same morphogenesis region as does
PVL (1, 19). Strain 81/108 was also positive with PCR-2 and seemed to carry other phages that did not carry lukS-PV and lukF-PV genes or have the structure to react with primers for PCR-2. Most of the PCR-1-positive strains we had tested up to this point in our study were also PCR-2 positive. Therefore, to confirm whether the structures related to morphology belonged to PVL-carrying phages, we performed PCR-3 and PCR-4 to verify whether the PVL-carrying prophage belonged to either the icosahedral-head-type phages or the elongated-head-type phages by amplifying gene lineages between lukS and mtp that belong to either the icosahedral-head type or the elongated-head type.
DNA fragments of the expected sizes were amplified using PCR-3 and chromosomal DNAs of 81/108 (
108PVL) and ATCC 49775 (
PVL) and using PCR-4 and chromosomal DNAs of MW2 (
Sa2mw), JCSC2958 (
Sa2958), and RN4220 (
SLT).
Up to that point, 38 of 39 PCR-1-positive strains were positive by PCR-2. However, PCR-3 was performed with chromosomal DNAs of 39 strains, and a DNA fragment of 10.5 kb was successfully amplified from 35 of 39 PCR-1-positive strains, indicating that these strains carried PVL phages of the icosahedral-head type. Sixty-six of 67 strains were positive by PCR-2, whereas DNA fragments of different sizes were amplified with chromosomal DNA of JCSC4274. When we performed PCR-4 with chromosomal DNAs with 26 strains for which no or only one DNA fragment was amplified by PCR-1, DNA fragments were amplified with chromosomal DNAs of 11 of these 26 strains.
Furthermore, we performed PCR-5 to -8 to classify these PVL phages into five different types. We conducted PCR-5 to determine whether these phages belonged to either the
PVL type or the
108PVL type. A DNA fragment of 4,340 bp, indicating the carriage of a
108PVL-specific gene (fragment PCR-1 A), was amplified with the DNAs of 33 of 35 strains, whereas a DNA fragment of 1,411 bp, indicating the carriage of
PVL (fragment PCR-2 A), was amplified in two cases, together with a DNA fragment of 4,340 bp. Because two PCR experiments to identify the other component of
PVL yielded negative results and four PCR experiments to identify the
108PVL-specific regions yielded positive results (data not shown), we conclude that these 33 strains that were positive by PCR-1 amplified fragment PCR-1 A, which carried the phage belonging
108PVL type. Four strains that were negative by PCR-4 were confirmed to belong to the
108PVL type, with four PCR experiments to identify
108PVL-specific regions and two long-range PCRs to amplify the DNA fragments covering the region from mtp to lukS. Therefore we regarded these four strains as being of the
108PVL-like type. Two strains, with the gene lineage between lukS and mtp common to
108PVL and
PVL, could not be classified into one of the five extant PVL phage types tested for up to that point.
In contrast, of the 26 strains for which no or only one DNA fragment was amplified by PCR-1, 25 were positive by PCR-2. In the case of strain JCSC7247, only one DNA fragment was amplified, suggesting the carriage of mtp in common with the elongated-head type.
PCR-4 was performed with DNAs of 26 strains to verify the gene lineage between the tail gene and the lukS gene by use of a primer pair designed for the tail gene that is commonly conserved in
Sa2958,
Sa2mw, and
SLT and for lukS. Of the 14 strains isolated from 1979 through 1985, only 1 was positive by PCR; in contrast, 10 of 20 strains isolated in the 2000s were positive by PCR. In addition, we performed PCR-6, -7, and -8 to classify elongated-head-type phages as of either the
Sa2958 type, the
Sa2mw type, or the
SLT type. Interestingly, neither the
Sa2mw type nor the
SLT type, both of which are carried by S. aureus strains isolated in the United States or France, was identified, whereas three strains carrying
Sa2958-type phage and eight strains carrying the elongated-head-type phage were identified. We determined the nucleotide sequence of a prophage lysogenized in MRSA strain JCSC2958, which was isolated in Japan in 1981. However, 13 of 14 strains isolated from 1979 through 1985 were definitively determined to be of the
Sa2958 type, because we could not amplify the region between mtp and lukS with long-range PCR. Because these 13 strains as well as a strain isolated in the 2000s were positive by PCR-6, identifying a specific ORF in
Sa2958, and by several types of long-range PCR, amplifying the gene lineage between mtp to some ORFs of
Sa2958 located between the tail gene and lukS, e.g., JP052 or JP053, we consider them to be
Sa2958-like phage of an unknown type. One strain carried a prophage that could not be classified into either one of the two types; this strain was exceptional, because it was negative by the seven PCRs other than PCR-6.
Characterization of PVL-positive MRSA clones.
We next characterized these strains by determining the coagulase isotypes and SCCmec types of all 65 strains and the MLST types of selected strains (Table 4). Fifty-two of the 53 MRSA strains isolated from 1979 through 1985 and 11 of the 12 MRSA strains isolated in the 2000s carried type 4 coagulase. Only JCSC2958, which we used for sequencing the
2958PVL genome, carried type 2 coagulase, and only JCSC7247 carried type 7 coagulase. When we performed MLST of the chosen isolates, we found that all coagulase type 4 strains belonged to ST30, whereas a coagulase type 2 strain belonged to ST5 and a coagulase type 7 strain belonged to ST59. The data showed that the genotypes of MRSA strains carrying PVL phages isolated in Japan were extremely homogeneous and were represented by ST30 coagulase type 4 strains.
|
View this table: [in a new window] |
TABLE 4. Characterization of PVL-positive MRSA strains isolated in 1979 to 1985, 2002, and 2004
|
108PVL type, SCCmec type IV.3 strains (21 of 33) were the most frequent and were followed by SCCmec type IV.1 strains (8 of 33), SCCmec type II strains (2 of 33), and SCCmec type I strains (1 of 33). Three strains identified as being of the
Sa2958 type carried type II SCCmec and type IV.1 SCCmec elements. Eight strains classified as the elongated-head type carried type IV.3 SCCmec and nontypeable SCCmec elements, which were similar to those of type IV.3 because they carried type 2 ccr and the J1 region of type IV.3 SCCmec and IS1272. Since we could not amplify the gene lineage between mecA and IS1272, we tentatively considered them as untypeable. Other strains also carried several SCCmec elements.
Interestingly, no strains with icosahedral-head-type phages were identified in isolates from the 2000s. The prophages of three strains were judged to be of the
Sa2958 type. In contrast to JCSC2958, two other strains isolated in the 2000s belonged to the coagulase type 4 ST30 genotype and carried a type IV.3 SCCmec element. A
Sa2958-like phage of unknown type was identified in the 2000s as belonging to the coagulase type 4 ST765 (CC30) group, carried type I SCCmec, and was similar to the isolates from 1979 through 1985. A strain carrying an untypeable PVL-carrying phage belonged to coagulase type 7 ST59 and carried the type V SCCmec element.
Induction of infective PVL phages.
To date, the
SLT is the only PVL phage that has been induced from cells and found capable of infecting other cells. Through a structural comparison of PVL phages, we found that both
Sa2958 and
Sa2mw carried possibly intact genes encoding head and tail proteins, and we inferred that these PVL phages might also be infective. To confirm this hypothesis, we induced prophages by treating these phages with mitomycin C and mixed them with cells of restriction-negative strains RN4220 and 1039. Because plaque formation does not indicate whether these plaques were generated by induced PVL phages, we conducted plaque hybridization experiments with a probe for the lukS-PV and lukF-PV genes to examine whether these plaques were of PVL-carrying phages.
The results are summarized in Table 5. We tested four MRSA strains (JCSC2973, JCSC4465, M11, and JCSC2958) isolated from 1979 through 1985, along with MRSA strains isolated in 2002 (JCSC4625) and 2004 (JCSC6605). Three strains carried
Sa2958-type PVL phages, and three strains carried
Sa2958-like PVL phages of unknown type. In addition, prophages
Sa2mw and
108PVL, whose genomes have been sequenced, were induced from strains MW2 and 81/108, as was lysogenized
SLT from strain RN4220 (
SLT). The cell suspensions of eight of nine tested strains became transparent after the addition of mitomycin C, indicating that phages might be generated upon the stress of DNA synthesis inhibition by mitomycin C. When these phage lysates were diluted and mixed with cells of indicator strains RN4220 and 1039, we found that most of the induced phages could propagate in 1039, whereas only some of them could propagate in RN4220.
|
View this table: [in a new window] |
TABLE 5. Identification of PVL phages induced by mitomycin C treatment
|
Sa2mw and
Sa2958 can infect other S. aureus strains, as can the previously reported
SLT. The phage lysate from 81/108 generated plaques, but no PVL-positive plaque was observed, indicating that the
108PVL phage did not induce or cannot infect the strains tested so far. This result was consistent with our speculation that
108PVL might be defective, because the tail gene of
108PVL was disrupted in a manner similar to that seen for
PVL (21).
Phage lysates of five of six MRSA strains carrying either the
Sa2958-type PVL phage or the
Sa2958-like PVL phage of unknown type generated PVL-positive plaques on indicator strains; however, the ratios of PVL-positive plaques ranged from 2% to 100%. In the case of JCSC6055, whose cell suspension did not become transparent after the addition of mitomycin C, no plaques were formed on the lawns of 1039 and RN4220 after the addition of the filtrate.
|
|
|---|
Sa2958.
Canchaya et al. have classified the Staphylococcal phages into five groups based on similarities in genes related to morphogenesis (4). All extant PVL phages are Sfi21-like cos-site phages of the Siphoviridae family, which can be further subdivided into two groups based on morphology: those with isomeric hexagonal heads and those with elongated heads. The PVL phages
Sa2mw and
Sa2usa, which are carried by MRSA strains isolated in the United States, have elongated heads. The
SLT phage was isolated from a French strain of S. aureus and has been shown by electron microscopy to have an elongated head measuring 100 by 50 nm and a flexible tail that is 400 nm long (32). Interestingly, the
Sa2958 phage we described here belongs to the elongated-head group, whereas
108PVL, identified in Japan, has an isometric hexagonal head, as does
PVL (20).
When we performed a detailed comparison of the genomic structure of
Sa2958 with those of other extant PVL phages, represented by
Sa2mw and
108PVL (Fig. 3 and Table 2), we found that most of the region between
Sa2958 and
Sa2mw is conserved. Similarly, this region is also conserved among
Sa2usa and
SLT. In contrast, the regions containing ORFs related to lysogeny (ORFs JP004 to -009) were unique to each phage. The average G+C content of the region from JP004 to JP009 was approximately 29.2%, which was lower than the average of the whole
Sa2958 genome (33.1%). These data indicated that these genes might be acquired from organisms with lower G+C contents.
The theory of modular exchange in phages, which is known as the modular theory, has become the basis of a popular hypothesis for phage evolution (26, 33). According to this theory,
Sa2958 and
Sa2mw must share a common ancestor and must have evolved as close relatives, while the module exchange was likely a recent event. We note here that functionally important structures of the repressor and antirepressor genes are located in this module (25). We now know that in most known prophages, immunity is elicited by a repressor protein that prevents transcription initiation at promoters controlling the expression of lytic function (14, 24). Furthermore, the immunity of the phage is sometimes complicated by the presence of an antirepressor gene that can prevent the expression of the repressor activity. The nucleotide sequences of the repressor genes conserved in
Sa2958 and
Sa2mw exhibited differences from each other at the DNA level; however, when we investigated the protein structure encoded by these repressor genes, we found that the essential helix-turn-helix domain was well conserved, indicating that both repressor genes execute functional activities. This finding suggests that
Sa2958 and
Sa2mw might utilize different mechanisms for maintaining their immunity when they are integrated into bacterial genomes; to achieve this goal, these phages had to adopt the most beneficial module during the process when they interacted with host bacteria or with other phages.
Two lineages of PVL phage are predominant in MRSA and MSSA strains in Japan.
We designed several PCRs to identify PVL phages with the notion that the type of PVL-carrying phage should be determined on the bases of the gene lineage between lukS-PV and lukF-PV and other phage components. We first developed PCR-1 and -2 to identify portal or head genes, but we soon noticed that these PCRs might also be used to identify components of phages other than PVL-carrying phages. Therefore, we developed PCRs to identify the gene lineage between lukS and mtp, which is located rather far from the lukS-PV and lukF-PV genes. In addition, we developed PCRs to classify individual phages by identifying the genes in lysogeny-related regions or recombination-related regions in combination with the integrase gene, which is carried by all reported PVL-carrying phages. During this step, we noticed that some modules are commonly shared among phages. As shown in Fig. 3,
Sa2958 was similar to
Sa2mw, and
Sa2958 was similar to
SLT (data not shown). The primer pairs we reported here are only those selected pairs that react only with one of five phages.
We have found that two PVL lineages, the icosahedral-head type and the elongated-head type, are present in Japan. Phages of the
108PVL type were identified mostly in MRSA strains isolated from 1979 through 1985. Of the 53 MRSA strains examined, 33 isolates were identified as the carrier of PVL-carrying phage of the
108PVL type, and 4 isolates were identified as carriers of the
108PVL-like type; however, no phages of this type were identified in the isolates from the 2000s. Furthermore, no
PVL-type strain was identified among the MRSA strains tested. However, when 13 PVL-positive MSSA strains isolated in the 2000s were examined, 2 were found to belong to the icosahedral-head type and were likely to be of the
PVL type. Because the tail gene of
108PVL was truncated, we assume that the phage had already lost its infectivity but carried an integrated truncated prophage, which might explain why
108PVL-type strains have not been identified recently.
In contrast, phages with an elongated head, e.g.,
Sa2958 as well as
Sa2mw and
SLT, carried intact tail and head genes.
Sa2958 was identified from an ST5 SCCmec type II MRSA strain identified in 1981, whereas two
Sa2958 type strains of ST30 SCCmec type IV.3 strains were identified in the 2000s, suggesting that the phage might infect other strains and become integrated into their chromosomes as a prophage. We considered the prophages carried by 14 strains (13 isolates from 1975 to 1985 and 1 isolate from the 2000s) as
Sa2958-like phages of unknown type, because we could not clarify the gene lineage between mtp and lukS-PV. The strains carrying the prophage might be derived from the same clone because they belonged to ST30 and carry a type I SCCmec element and because the prophages of all three tested strains could be induced, proving their infectivity; these findings suggest that the recent isolate might be a descendant or that the phage might infect other strains. Interestingly, eight strains with elongated-head-type PVL-carrying prophages were identified. Because these strains were identified not only from seven isolates in Wakayama prefecture but also from an isolate from a 27-year-old man in Kanagawa prefecture, which is far from Wakayama, in 2002 (strain EB00449), we speculate that the prophage might have been induced from the cells of S. aureus and then infected other S. aureus cells to generate the novel PVL phage-carrying strain. A coagulase type 7 ST59 strain carried an as-yet-untypeable PVL phage. Vandenesch et al. reported that five major STs, namely, ST1, -8, -30, -59, and -80, carry PVL phages and that ST59 PVL-positive strains have been isolated in the United States (39). Because the ST59 strain is not common in Japan, the strains of other countries might have been introduced into Japan. Whether the strain carries a novel PVL phage is the next problem that should be clarified. It might carry a PVL phage that does not belong to the two extant groups.
When we characterized these PVL-positive strains by means of SCCmec typing and genotyping, we found that the most of them produced type 4 coagulase and belonged to ST30 or CC30 and carried SCCmec elements of various types, including type I, type II, type IV.1, and type IV.3 (17, 18, 29). We speculate that MSSA strains of coagulase type 4 ST30 independently acquired two phages,
108PVL and
Sa2958, or other elongated-head phages before they acquired SCCmec elements. The characterization of PVL phages carried by MSSA strains isolated in the 1960s has shown that MSSA strains carrying either of the PVL phages were present in the Japanese community. Therefore, we presume that the MSSA strains evolved into PVL-positive MRSA strains by acquiring SCCmec elements of various types. When we examined the types of SCCmec elements carried by PVL-positive MRSA strains, we found that 20 of 33 MRSA strains of the
108PVL type carried type IV.3 SCCmec and that 8 of 33 carried type IV.1 SCCmec. It was noteworthy that most of the PVL-positive SCCmec type IV.1 strains had been isolated at the Tokyo Geriatric Hospital, whereas, and most of the PVL-positive SCCmec type IV.3 strains had been isolated at Gunma University Hospital. The data indicated that MRSA clones disseminating in different Japanese hospitals were not identical and that the spread of MRSA strains is regional.
Sa2958 and
Sa2958-like phages of unknown type are intact prophages.
When we studied the structure of
Sa2958, identified in this study, we realized that the amino acid sequence of the tail-length tape measure protein designated JP047 showed 98% similarity in with that of
SLT, which has been shown to be the first PVL phage that could infect S. aureus strains experimentally (32). Because the tail-length tape measure protein of
Sa2mw carried by MW2, a CA-MRSA strain isolated in North Dakota, was similar to that of
SLT, we presumed that the newly identified
Sa2958 as well as
Sa2mw might also be able to infect to S. aureus strains.
To determine whether these two phages as well as phages of the
Sa2958 type were intact and could infect S. aureus strains, we tested MW2 and six MRSA strains carrying
Sa2958-type phage or
Sa2958-like PVL phage of unknown type. The PVL phages carried by five of six MRSA strains could infect indicator cells, although the ratios of PVL phages in the phage lysates were not similar. In addition, the sizes of plaques generated on the lawn were larger with 1039 than with RN4220. With three strains, namely, M11, JCSC2958, and JCSC4625, all phages induced by treatment with mitomycin C were PVL positive. In contrast, with two strains, JCSC2973 and JCSC4625, the ratios of PVL-positive phages were very low. In the case of JCSC6055, which did not generate any plaques, the apparent reason that phages were not induced was that the cell suspension did not turn transparent after mitomycin C was added. These discrepancies may be due to many host cell factors, e.g., the presence of other prophages and differences in gene expression that confer the SOS response. Although we did not test a large number of strains, we chose
Sa2958-type MRSA strains of a representative genotype identified in this study. Except for JCSC6055, the five strains (derived from four different clones as indicated), namely, JCSC2973 (ST30 SCCmec type I), M11 and JCSC4625 (ST30 SCCmec type IV.3), JCSC4465 (ST30 SCCmec type IV.n), and JCSC2958 (ST5 SCCmec type II), generated intact PVL phages. This result indicates that
Sa2958-type and
Sa2958-like phages of unknown type can be frequently induced from many S. aureus strains with diverse genetic backgrounds. These PVL phages might be induced with or without any stress from S. aureus and infect other appropriate recipient S. aureus strains, resulting in the phage conversion of nonvirulent bacteria to virulent bacteria. The facts that two recently isolated strains carried
Sa2958 and that one of them generated infective PVL phages suggest that the
Sa2958 type might be the cause of the recent emergence of PVL-positive MRSA strains in Japan. As shown in Table 5,
Sa2mw carried by MW2 was infectious for two indicator strains, RN4220 and 1039. The USA300 strain has recently been the predominant CA-MRSA clone in the United States.
Sa2usa carries PVL and belongs to the same family as
Sa2mw and
Sa2958, which have elongated heads. Although we could not identify any plaques of
Sa2usa, we have tested FPR3757, whose genome has been sequenced; these
Sa2mw-type or
Sa2usa-type PVL phages might be the cause of the recent appearance of PVL-positive MRSA strains in the United States. Our data also indicate that ST30 MRSA strains carry two different PVL phages. Further study will clarify whether ST30 PVL-positive MRSA clones can be grouped with Western Samoan clones.
In conclusion, we have developed a method of combining several PCRs to identify PVL phage carried by an organism. Our results indicate that two lineages of PVL phages have existed in Japan for many years and that intact phages with elongated heads, such as
Sa2958 and
Sa2958-like phage of unknown type or other phages known simply as elongated-head types, might disseminate among MRSA strains in Japan and confer greater virulence.
This work was supported by a grant-in-aid for 21st-century COE research, a grant-in-aid for scientific research on priority areas, and a grant-in-aid for scientific research (C19590456) from the Ministry of Education, Science, Sports, Culture and Technology of Japan.
Published ahead of print on 6 August 2008. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2010 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»