Efficient transfer of antibiotic resistance plasmids by transduction within methicillin-resistant Staphylococcus aureus USA300 clone
Correspondence: Jiri Doskar, Department of Experimental Biology, Faculty of Science, Masaryk University, Kotlarska 2, 611 37 Brno, Czech Republic. Tel.: +420 549493557; fax: +420 549492570; e-mail: email@example.com
The epidemic community-associated methicillin-resistant clone Staphylococcus aureus USA300 is a major source of skin and soft tissue infections and involves strains with a diverse set of resistance genes. In this study, we report efficient transduction of penicillinase and tetracycline resistance plasmids by bacteriophages φ80α and φJB between clinical isolates belonging to the USA300 clone. High transduction frequencies (10−5–10−6 CFU/PFU) were observed using phages propagated on donor strains as well as prophages induced from donors by ultraviolet light. Quantitative real-time PCR was employed to detect penicillinase plasmids in transducing phage particles and determine the ratio of transducing particles in phage lysates to infectious phage particles (determined as approximately 1 : 1700). Successful transfer of plasmids between strains in USA300 clone proves transduction is an effective mechanism for spreading plasmids within the clone. Such events contribute to its evolution and to emergence of new multiple drug-resistant strains of this successful clone.
Staphylococcus aureus is an important human pathogen causing both nosocomial and community-acquired infections ranging from minor superficial skin infections to life-threatening systemic diseases. Staphylococcus aureus USA300 is one of the S. aureus clones most widespread worldwide. Typically, USA300 strains are associated with infections occurring in the community, but, more recently, these strains have been reported to cause infection among patients in health care facilities (Tenover & Goering, 2009). The most noticeable feature of the USA300 genome is its rapid diversification and acquisition of different mobile genetic elements, including plasmids (Kennedy et al., 2008; Li et al., 2009). USA300 strains harbor a diverse set of plasmids with a broad spectrum of antibiotic resistance genes (Kennedy et al., 2010; Carpaij et al., 2011).
The most common mechanism of horizontal gene transfer in S. aureus is apparently transduction, because there is a little evidence that transformation occurs and conjugative plasmids or transposons are not widespread in S. aureus (Lindsay, 2008). Many transduction experiments have been conducted intending either to test the transduction ability of staphylococcal phages (Dowell & Rosenblum, 1962; Novick, 1990) or prove the mobility of variable genetic elements with genes encoding antibiotic resistance or toxins (Cohen & Sweeney, 1970; Ruzin et al., 2001; Nakaminami et al., 2007; Chen & Novick, 2009). Most clinically important human strains of S. aureus harbor at least one prophage, the presence of which may affect the strain's capability for gene transfer (Lindsay, 2008; Goerke et al., 2009). To date, however, only limited knowledge is available whether naturally occurring phages are able to mediate effective transfer of plasmids in vivo within the population of clinical S. aureus strains.
The main aim of this study was to prove that the penicillinase and tetracycline resistance plasmids are efficiently transferred within the USA300 clone by transduction. According to our best knowledge, this is also the first work providing quantitative real-time PCR (qPCR) estimation of functional plasmids packaged in transducing particles in S. aureus.
Materials and methods
Bacterial strains and bacteriophages
Five strains from the USA300 clone, designated 07/235, 07/759, 08/019, 08/629, and 08/986 (all isolated in Czech hospitals), were obtained from the National Reference Laboratory for Staphylococci, National Institute of Public Health, Prague. Their assignment to the USA300 clone was based on PCR screening for the arginine catabolic mobile element and lukF-PV and lukS-PV genes (Diep et al., 2006), spa typing (Shopsin et al., 1999), and multi locus sequence typing (Enright et al., 2000). Diverse plasmid content of these strains has been identified previously (Kuntová et al., 2012). Other strains used in transduction experiments were methicillin-resistant S. aureus (MRSA) isolate Jevons B obtained from Dr. G. Pulverer (Hygiene-Institut, Köln, Germany) and laboratory strain RN4220 kindly provided by Prof. F. Götz (University of Tübingen, Germany). For transduction purposes with induced phage lysate, the lysogen of 07/759 was constructed by inserting φJB (see below) into its chromosome as previously reported (Borecká et al., 1996). All clinical strains and their characteristics are listed in Table 1.
Table 1. Clinical strains used in this study as donor or recipient strains
|07/235||Recipient||Czech Rep.||2007||ST8/t211||IV||Sa2, Sa3||A, FA||3||None|
|07/235 (φ80α+), pPENa||Donor||Czech Rep.||2007||ST8/t008||IV||Sa2, Sa3, Sa7||A, B, FA||3; 27||blaZ, cadD|
|07/759||Donor and recipient||Czech Rep.||2007||ST8/t008||IV||Sa2, Sa3||A, FA||31||blaZ, cadD|
|07/759 (φJB+)||Donor||Czech Rep.||2007||ST8/t008||IV||Sa2, Sa3, Sa6||A, B, FA||31||blaZ, cadD|
|08/019||Donor and recipient||Czech Rep.||2008||ST8/t008||IV||Sa2, Sa3||A, FA||27||blaZ, cadD|
|08/019, pT181||Donor||Czech Rep.||2008||ST8/t008||IV||Sa2, Sa3||A, FA||4.4; 27||tetK, blaZ, cadD|
|08/629||Donor and recipient||Czech Rep.||2008||ST8/t008||IV||Sa2, Sa3||A, FA||3; 27||blaZ, cadD|
|08/986||Donor and recipient||Czech Rep.||2008||ST8/t008||IV||Sa2, Sa3||A, FA||3; 27||blaZ, cadD|
|Jevons B||Donor||UK||1961||ST250/t008||I||Sa6b||Bb||4.4; 28||tetK, blaZ, cadD|
Two transducing bacteriophages, φ80α (Novick, 1963) obtained from Dr C. Wolz (University of Tübingen) and φJB induced by UV light from the MRSA strain Jevons B in this work, both of serological group B from the Siphoviridae family, were employed as mediators for plasmid transfer. Bacteriophage φJB has been deposited in the Czech Collection of Microorganisms under designation CCM 7872.
The recipient strain was grown overnight to the titer of approximately 3 × 108 CFU mL−1 in meat peptone broth prepared from 13.0 g of nutrient broth CM1 (Oxoid, Basingstoke, UK), 3.0 g of yeast extract powder L21 (Oxoid), and 5.0 g of peptone L37 (Oxoid) dissolved in distilled water to 1000 mL (pH 7.4). Calcium chloride was added to final concentration 2 mM, and the culture was mixed with stock phage lysate so the multiplicity of infection was below 1. The mixture was shaken moderately at 37 °C for 25 min. Sodium citrate was then added to the mixture to final concentration 15 mM, and cells were centrifuged at 1100 g for 15 min. The cell pellet was resuspended in 1 mL of sodium citrate (17 mM), and the cells were spread onto agar plates (nutrient agar CM3; Oxoid) supplemented with sodium citrate (20 mM) and Cd(NO3)2·4H2O (final concentration 50 μM) or tetracycline (5 μg mL−1). The plates were incubated at 37 °C for 48 h. The colonies of transductants were picked up and passaged successively through selective medium with sodium citrate (as described above), nonselective medium with sodium citrate and nonselective medium without sodium citrate. Finally, cells were resuspended in Hogness modified freezing medium (3.6 mM K2HPO4, 1.3 mM KH2PO4, 2 mM sodium citrate, 1 mM MgSO4·7H2O, 4.4% (v/v) glycerol) and stored at −80 °C. The transduction frequency was calculated as the ratio of the number of transductants (CFU) obtained to the number of plaque-forming units (PFU).
To obtain induced phage lysates for transduction purposes, the lysogenic strains were precultivated in meat peptone broth at 37 °C with aeration after cells had reached logarithmic growth phase. Twice-washed cells resuspended in 10 mL saline solution (0.85% NaCl) to optical density OD600 nm = 0.15 were irradiated using a 15-W ultraviolet lamp at distance 60 cm for 30 s. The following steps were performed as described previously (Duval-Iflah, 1972).
Isolation and characterization of plasmids
The plasmid DNA from donors and transductants was isolated using the High Pure Plasmid Isolation Kit (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's protocol with modifications of cell lysis. Briefly, 8 mL of overnight culture was washed twice in phosphate-buffered saline and resuspended in 235 μL of Suspension Buffer with RNase A + 15 μL of lysostaphin (Dr. Petry Genmedics, Reutlingen, Germany) (0.5 mg mL−1). Then, it was left incubating at 37 °C for 15 min. After the treatment, 250 μL of lysis buffer was added.
The restriction endonuclease HindIII (Roche Diagnostics) was used to digest plasmid DNA according to the manufacturer's protocol. The digested DNA was analyzed by electrophoresis in 1.8% agarose gel (Serva, Heidelberg, Germany) in 1× TAE buffer at 5 V cm−1. 2-Log DNA Ladder (New England Biolabs, Ipswich, MA) was used as DNA molecular weight marker. Ethidium bromide staining and UV irradiation were employed for DNA visualization.
Sequencing of plasmid DNA
The complete nucleotide sequence of the 3 kb cryptic plasmid present in strain 07/235 was determined by Sanger capillary sequencing. All sequencing steps were performed by Eurofins MWG Operon (Ebersberg, Germany).
Plasmid-borne resistance genes were detected by PCR using primers for the β-lactamase gene blaZ (Martineau et al., 2000), tetracycline resistance gene tetK (Ng et al., 2001), and cadmium resistance gene cadD (primers cadD-F GGATATTAGGTTTATTGGGTT and cadD-R CGCCACAACTTGCTATCGTA). Each reaction mixture (25 μL) contained 1× PCR buffer, 0.2 mM dNTP, 1.5 mM MgCl2, 0.2 mM of each primer, 1 U Taq DNA polymerase (Invitrogen Life Technologies, Carlsbad, CA), and 10 ng of template plasmid DNA. Initial denaturation of DNA at 94 °C for 5 min was followed by 30 amplification cycles (94 °C for 30 s, 55 °C for 30 s, 72 °C for 45 s), ending with a final extension phase at 72 °C for 4 min. PCR products were separated by electrophoresis as was plasmid DNA.
Bacteriophage integrase types and morphogenesis gene types corresponding to serological groups of prophages in the genomes of the strains were identified by multiplex PCR as described previously (Kahánková et al., 2010).
The test for β-lactamase production was made using nitrocefin disk assay according to the manufacturer's recommendations (Erba Lachema, Brno, Czech Republic).
DNA from phage particles was isolated as described previously (Doškař et al., 2000). RNase A (Serva) and DNase I (Sigma, St Louis, MO) were added to the samples to final concentration 1 and 5 μg mL−1, respectively, to remove contaminating exogenous bacterial DNA.
qPCR experiments were performed on the Applied Biosystems 7300 Real-Time PCR System (Applied Biosystems, Foster City, CA). Each reaction mixture (25 μL) contained 12.5 μL 2× FastStart Universal SYBR® Green Master (Rox) (Roche Diagnostics), 900 nM of each primer, and 10 ng of template DNA. For the standard, the amount of template DNA ranged from 10 ng to 0.1 pg in 10-fold fashion. Reaction mixtures were prepared in triplicates in MicroAmp® Optical 96-Well Reaction Plate (catalog number 4316813) sealed with optical adhesive cover (Applied Biosystems). Initial denaturation of DNA at 95 °C for 10 min was followed by 40 cycles of amplification (95 °C for 15 s and 60 °C for 45 s), ending with a dissociation phase at 95 °C for 15 s, 60 °C for 60 s, 95 °C for 15 s, and 60 °C for 15 s. Primers were as follow: blaZ-F (ACGAAATCGGTGGAATCAAA) and blaZ-R (AGCAGCAGGCGTTGAAGTAT) for blaZ (product length, 115 bp); mecA1575 (AGGTTACGGACAAGGTGAAATACTG) and mecA1657 (TGTCTTTTAATAAGTGAGGTGCGTTAA) for mecA (product length, 106 bp); and 53D-F (CGACAAAAGGCATTCAACAA) and 53D-R (ACGTTCAAAAATCGCTTGCT) for the 4867-bp HindIII DNA fragment of bacteriophage φ53 cloned in the pUC18 vector (GenBank accession number, AF513856; product length, 139 bp) that is specific for all serogroup B phages (Doškař et al., 2000). The basis for calculating the unknown quantity of PCR product is that the 10-fold difference in the amount of DNA in two samples will manifest in the difference in their quantification cycles with a value of 3.22 (Lee et al., 2006).
By comparing quantities of the blaZ plasmid gene and mecA single-copy chromosome gene, the plasmid copy number (PCN) in the donor strain was determined, which is necessary to determine the number of transducing particles carrying the penicillinase plasmid in comparison with the standard consisting of genomic DNA. PCN was determined according to the equation PCN = [size of chromosomal DNA (bp) × amount of plasmid DNA (pg)]/[size of plasmid DNA (bp) × amount of genomic DNA (pg)] (Lee et al., 2006).
To calculate the standard copy number (SCN), the following formula was used: SCN = (amount of DNA per reaction in ng)/(Mr/NA), where Mr = size of S. aureus USA300 genomic DNA × normalized weight of nucleotide base (650 Da), and NA is the Avogadro constant. The standard curves were repeatable, and amplification efficiency of PCR reactions ranged between 90% and 100%.
Results and discussion
Genotypic characteristics and plasmid content of clinical strains used as donors and/or recipients in transduction experiments are listed in Table 1. Ability to transduce plasmids from the USA300 strains was first confirmed using the φ80α phage. The prophageless, plasmidless, and restriction-deficient RN4220 strain was used as control recipient strain for the transductions. Transduction frequency in this system ranged from 3.9 × 10−6 CFU/PFU to 5.1 × 10−6 CFU/PFU for penicillinase plasmids and from 2.7 × 10−6 CFU/PFU to 3.4 × 10−6 CFU/PFU for tetracycline resistance plasmid pT181. Based upon successful transduction of penicillinase plasmids and the tetracycline resistance plasmid into RN4220, plasmids were transduced between the isolates from the USA300 clone by the φ80α phage and subsequently by the naturally occurring φJB prophage which in a number of our experiments demonstrated excellent transducing abilities (unpublished results).
Transduction of plasmids by propagated lysates
Transduction of penicillinase plasmids from all four USA300 donor strains into the recipient 07/235 by φ80α and φJB are summarized in Table 2. High transduction frequency was observed in all transduction mixtures, ranging around 10−5 CFU/PFU. The highest frequency was during transmission of the 31 kb plasmid from the 07/759 donor strain.
Table 2. Transduction frequencies obtained with phages φ80α and φJB
|07/759||φ80α||31 kb; blaZ, cadD||4.7 × 107||07/235||1.5 × 10−5|
|08/019||φ80α||27 kb; blaZ, cadD||8.5 × 107||07/235||9.2 × 10−6|
|08/629||φ80α||27 kb; blaZ, cadD||6.2 × 107||07/235||1.1 × 10−5|
|08/986||φ80α||27 kb; blaZ, cadD||8.2 × 107||07/235||7.9 × 10−6|
|08/019, pT181||φ80α||4.4 kb; tetK||6.3 × 107||07/759||4.6 × 10−6|
|07/759||φJB||31 kb; blaZ, cadD||4.2 × 105||07/235||5.0 × 10−6|
|08/019||φJB||27 kb; blaZ, cadD||5.1 × 108||07/235||1.1 × 10−6|
|08/629||φJB||27 kb; blaZ, cadD||3.0 × 106||07/235||2.7 × 10−6|
|08/986||φJB||27 kb; blaZ, cadD||3.4 × 108||07/235||9.0 × 10−7|
|08/019, pT181||φJB||4.4 kb; tetK||2.3 × 107||07/759||2.8 × 10−6|
Testing for β-lactamase production, growth on selection medium, PCR for detecting the blaZ and cadD genes, and cleaving of plasmids by HindIII restriction endonuclease confirmed that plasmids were transferred into all transductants with functioning genes and without structural rearrangements. Sporadic lysogenization of transductants 07/235 by the φ80α bacteriophage was discovered by PCR for detecting prophage genes. We then used these lysogenic transductants as donor strains for the penicillinase plasmid in transductions mediated by the induced prophage.As none of the USA300 donor strains naturally contain the pT181 tetracycline resistance plasmid, it was first necessary to prepare such a strain. For this purpose, the pT181 plasmid was transduced from the Jevons B strain by means of φ80α to the 08/019 strain. Subsequently, transductions of pT181 from such prepared strain were made using φ80α and φJB into other strains of the USA300 clone. However, pT181 was only transduced into 07/759 and transfer of the plasmid did not occur in other strains. As all these strains contain a 3-kb cryptic plasmid (Table 1), we hypothesized this plasmid is incompatible with pT181. To test this hypothesis, the complete nucleotide sequence of the cryptic plasmid present in strain 07/235 was determined. Bioinformatic analysis revealed that this plasmid is in fact identical to plasmid pUSA01 (GenBank accession number NC_007790) from S. aureus USA300_FPR3757. Based upon Kennedy et al. (2010) who found out that pUSA01 shows almost no similarity with the tetracycline resistance plasmid pT181, we concluded that it is highly unlikely these two plasmids could be mutually incompatible. The reason why pT181 was not transduced into strains possessing cryptic plasmid pUSA01 remains unresolved.
In our study, we reached significantly higher transduction frequency values for the penicillinase plasmids and the pT181 in the USA300 clone than did Asheshov (1969) using PS80 strain as donor and 17855 as recipient and Kayser et al. (1972) using E142 as donor and various recipients. It is therefore probable the transfer of plasmids between strains of USA300 originating from the same clonal complex 8 (CC8) is not affected by activity of the Sau1 restriction-modification system, which seems to be the main barrier to transfer of mobile genetic elements between various clonal lineages (Waldron & Lindsay, 2006).
Detection of plasmid DNA in transducing phage particles and their quantification in transducing lysates
To indentify transducing particles containing the penicillinase plasmid and determine the number of infectious phage particles in lysates, respectively, qPCR assay targeting the blaZ gene and a part of the conservative gene encoding the long tail fibers of serological group B phages was introduced. To quantify the blaZ gene, a standard using genomic DNA of transductant 07/759/φ80α/07/235 was prepared. The standard for determining the number of infectious particles was the pUC18 construct with the 4867 bp DNA fragment of bacteriophage φ53.
It was determined that the penicillinase plasmid occurs on average in three copies per cell (exact value deduced by qPCR is 2.98). This value correlates with the expected copy number for such a large plasmid per cell (Novick, 1990). Based upon absolute quantification of the blaZ gene by qPCR, the number of copies of this gene in 1 mL of transducing phage lysate was determined as 1216. An analogous approach enabled determining the number 2.108 × 106 infectious phage particles in the lysate. Comparing the aforementioned values, we determined the approximate ratio of transducing particles to number of infectious phages to be 1 : 1700.
The number of transducing virions carrying blaZ (2.71 × 104) deduced from the qPCR data and from the titer of infectious phages in the lysate (4.7 × 107 PFU mL−1), as well as the number of acquired transductants (720 CFU mL−1), enabled determining the effectiveness of transduction as 2.7%. Transduction effectiveness was determined on the multiplicity of infection level of 0.16, when the probability of introducing more plasmids into a single recipient cell and superinfection of the created transductants followed by their elimination is very low.
Transduction of plasmids within USA300 clone by induced prophages
In further experiments, we explored the possibility for disseminating antibiotic resistance genes by the φJB prophage induced from donor lysogenic cells prepared by lysogenization of strain 07/759. Using UV radiation, a transducing lysate with titer 8.6 × 105 PFU mL−1 was prepared from the lysogenic strain 07/759 (φJB+) and was used successfully to transfer the 31 kb penicillinase plasmid into the strain 07/235 with frequency 2.3 × 10−6 CFU/PFU. The genotype of transductants was determined in the same way as of the transductants obtained by propagated phage lysates.
Another donor strain used for these experiments was the lysogenic transductant 07/235 (φ80α+) containing the φ80α prophage and 27 kb penicillinase plasmid of the 08/986 strain described above. The objective was to clarify the hypothesis whether by receiving a plasmid and integration of φ80α phage 07/235 became a new potential donor capable of transferring the plasmid into other strains after induction of the φ80α prophage. The induced lysate with titer of 1.6 × 106 PFU mL−1 was used for transducing plasmid into the RN4220 strain, which successfully received it with a frequency of 3.1 × 10−6 CFU/PFU. This shows that if the transductant is lysogenized, the plasmid can be very effectively mobilized.
Transduction experiments with induced lysates proved that prophages that abundantly occur in a number of clinical strains can play an important role in transferring plasmids. Transduction of a resistance plasmid from one strain into others may be a quite efficient way of spreading antibiotic resistance. A more efficient way, however, seems to be spreading associated with lysogenizations of transductants or generally of plasmid-positive strains by transducing bacteriophages. Such events increase the risk of emerging multiresistant strains with transducing bacteriophages able to transfer resistance determinants into other strains. Effective transfer of resistance plasmids between strains of the USA300 clone intermediated by transduction contributes to this clone's faster evolution.
Comparative DNA analysis of the φJB phage and prophages of several clinical S. aureus strains demonstrated substantial match in their DNA profiles. The highest similarity rate was for the prophage of recent MRSA isolate E53 (ST624/t211/SCCmec IA) related to the Iberian clone and of the φNM4 prophage of S. aureus strain Newman (Bae et al., 2006), both members of the CC8 lineage. This evidences that in naturally occurring S. aureus strains some prophages of serological group B are very closely related to the φJB prophage, for which similar transduction abilities can be expected.
We would like to thank P. Petráš from the National Reference Laboratory for Staphylococci, National Institute of Public Health, Prague for providing MRSA isolates. We gratefully acknowledge financial support of the Czech Science Foundation (310/09/0459), and Ministry of Education, Youth and Sports of the Czech Republic (MSM 0021622415).