The most significant events in the history of S. aureus antibiotic resistance are the emergence of resistance to penicillin (which is now almost ubiquitous), the emergence of and epidemic rise in methicillin resistance, the recognition of strains with intermediate resistance to vancomycin and the recent emergence of S. aureus that is fully resistant to vancomycin. The mechanisms of resistance to other antimicrobial agents are summarized in Table 5.3.
Penicillin-resistant S. aureus strains emerged in the early 1940s, shortly after the introduction of penicillin into clinical practice. Resistance to methicillin and other fi-lactamase-resistant penicillins was likewise observed soon after methicillin was introduced into clinical use in Britain (Jevons 1961). At this time the methicillin-resistant strains isolated in Britain demonstrated heterogeneous resistance to methicillin (i.e. affecting only a minority of the cell population), were multiply antibiotic resistant and were isolated from hospitalized patients (Barber 1961). After the mid-1970s, large outbreaks of infection caused by MRSA were reported in many hospitals in Britain (Shanson, Kensit and Duke 1976; Cookson and Phillips 1988), Ireland (Cafferkey etal. 1985), the United States (Schaefler etal. 1981) and Australia (Pavillard etal. 1982). Many of these early MRSA epidemics were caused by a single epidemic strain that was transferred between hospitals (Duckworth, Lothian and Williams 1988). Since then, many clones of MRSA associated with epidemic spread or sporadic infections have been described throughout the world.
Approximately one-third of serious S. aureus infections in the United Kingdom are now caused by MRSA, although the figure varies considerably worldwide. Until recently, MRSA was mostly confined to the hospital setting and MRSA colonization of those discharged from the community rarely persisted long term except when associated with defects in the skin integrity or the presence of prosthetic material. However, community-acquired MRSA associated with both colonization and infection is being increasingly recognized (Daum etal. 2002; Okuma etal. 2002). These strains are resistant to fewer non-fi-lactam antibiotics than most of the previously defined MRSA strains.
The gene encoding methicillin resistance (mecA) is carried by the chromosome of MRSA and methicillin-resistant S. epidermidis (MRSE), the mechanism for which is the synthesis of an altered low-affinity PBP termed PBP2a. mecA is part of a mobile genetic element termed staphylococcal cassette chromosome mec (SCCmec) (Katayama, Ito and Hiramatsu 2000). Four types of SCCmec have been defined based on sequence analysis. Comparison of SCCmec types I—III has demonstrated substantial differences in both size and nucleotide sequence but shared conserved terminal inverted repeats and direct repeats at the integration junction points, conserved genetic organization around the mecA gene and the presence of ccr genes, which are responsible for the movement of the island (Ito etal. 2001). Evaluation of 38 epidemic MRSA strains isolated in 20 countries showed that the majority possessed one of the three typical SCCmec elements on the chromosome (Ito etal. 2001). SCCmec type IV was subsequently defined in community-acquired MRSA (Daum etal. 2002). This element is smaller and lacks non-fi-lactam genetic-resistance determinants. Multiple MRSA clones carrying type IV SCCmec have been identified in community-acquired MRSA strains in the United States and Australia (Okuma etal. 2002). An evaluation of SCCmec types (as defined by the PCR) present in 202 MRSA isolates from two hospitals in Florida reported that four isolates failed to give an amplification product, indicating the possible existence of additional types (Chung etal. 2004).
Genetic relationships of 254 MRSA isolates recovered between 1961 and 1992 from nine countries on four continents have been analysed using electrophoretic mobility of enzymes (multilocus enzyme electrophoresis) (Musser and Kapur 1992). Fifteen distinctive electrophoretic types (clones) were identified, and the mec gene was found in divergent phylogenetic lineages. This was interpreted as evidence that multiple episodes of horizontal transfer and recombination have contributed to the spread of methicillin resistance. MLST has subsequently been used to define the population genetic structure of MRSA and methicillin-susceptible strains (Enright etal. 2002). Eleven major MRSA clones were identified within five groups of related genotypes. The major MRSA clones appeared to have arisen repeatedly from successful epidemic MSSA strains, and isolates with decreased susceptibility to vancomycin have arisen from some of these major MRSA clones (Enright etal. 2002).
Possible precursors and reservoirs of the genetic determinants of methicillin resistance include Staphylococcus sciuri (Wu, de Lencastre and Tomasz 1998) and Staphylococcus hominis subsp. novobiosepticus (Kloos etal. 1998b). A genetic element that is structurally very similar to SCCmec except for the absence of mecA has been defined in a methicillin-susceptible strain of S. hominis (Katayama etal. 2003). An SCC element that lacks mecA has also been defined during sequencing of S. aureus MSSA252 (Holden etal. 2004). The role of this element in horizontal transfer of genes encoding antibiotic resistance and other determinants that may endow a biological fitness requires further investigation.
Some mecA S. aureus strains exhibit borderline susceptibility to methicillin (oxacillin MIC, 1-2 |lg/ml) and other fi-lactamase-resistant penicillins (McMurray, Kernodle and Barg 1990; Barg, Chambers and Kernodle 1991). This is partly due to the hyperproduction of fi-lactamase type A (McDougal and Thornsberry 1986; Chambers, Archer and Matsuhashi 1989).
The glycopeptide antibiotics vancomycin and teicoplanin prevent the transglycosylation and transpeptidation steps of cell-wall peptidog-lycan synthesis by binding to the peptidyl-D-alanyl-D-alanine termini
Table 5.3 Antibiotic resistance mechanisms for S. aureus (see text for P-lactams and glycopeptides)
Major activity or mechanism
Macrolide, lincosamide and streptogramin B (MLS)
23S rRNA methylases A-C Methylases cause a conformational change in the ermA ribosome, resulting in a reduced binding of MLS antibiotics to target site in 50s subunit Require induction by erythromycin or are made ermB constitutively ermC
Macrolide and streptogramin B (MS)
ATP-dependent efflux ATP-dependent efflux pump to remove protein MsrA MS antibiotics
Acetyltransferase Inactivation of drag by enzymatic modification msrA
ATP-binding proteins Mechanism not fully elucidated
Streptogramin B Hydrolase
Hydrolysis of drug vgaA vgaB vgaAv
(variant of A)
3-Lincomycin, 4-clindamycin Inactivates lincosamides linA'
Tetracycline export proteins Energy-dependent efflux pump removes tetK
Tetracycline and minocycline
Ribosomal protection protein tetM
Aminoglycoside Acetylation of an amino group: tobramycin, acetyltransferase (AAC) netilmicin, amikacin and gentamicin aac( 6')-aph( 2")
Chromosome: on transposon Tn554 which prefers a single att site
Plasmids and chromosome: on transposon Tn557 which can transpose to many sites Plasmids (mainly class I) > chromosome
Murphy (1985a), Murphy, Huwyler and de Freire Bastos Mdo (1985), Thakker-Varia etal. (1987), Lim etal. (2002), Strommenger etal. (2003)
Khan and Novick (1980), Luchansky and Pattee (1984)
Horinouchi and Weisblum (1982), Weisblum (1985)
Plasmid vgaAv is carried by Tn5406 located on plasmid and/or chromosome
Allignet etal. (1993, 1996), Allignet, Loncle and el Sohl (1992), Allignet and el Solh (1995), Allignet and El Solh (1997), Haroche etal. (2000, 2003), Haroche, Allignet and El Solh (2002), Strommenger etal. (2003)
Located on plasmids conferring resistance to A compounds
Allignet etal. (1988), Allignet, Liassine and el Solh (1998), Haroche etal. (2003)
Khan and Novick (1983), Mojumdar and Khan (1988), Guay, Khan and Rothstein (1993), Strommenger etal. (2003)
Chromosome: Tn916- and Tn916-like Nesin etal. (1990), Burdett (1993), Strommenger etal. transposons (2003)
Plasmids (mainly class III) and Townsend, Grabb and Ashdown (1983), Townsend etal. (1984), chromosome on transposons (Tn4001 Rouch etal. (1987) and Tn4001-like, Tn4031, Tn3851)
Aminoglycoside phosphotransferase (APH)
Aminoglycoside adenylyltransferase (ANT)
Trimethoprim Dihydrofolate reductase (DHFR)
Chloramphenicol Chloramphenicol acetyltransferase (CAT)
Fluoroquinolones DNA gyrase subunits GyrA and GyrB
DNA topoisomerase IV
Multidrug efflux protein NorA
DNA-dependent RNA polymerase ß-subunit
Phosphorylation of a hydroxyl group: kanamycin, neomycin, paromomycin, amikacin and gentamicin
Adenylation of a hydroxyl group: tobramycin, amikacin, paromomycin, kanamycin, neomycin, gentamicin and dibekacin
Single amino acid substitution of dhfr gene leads to marked reduced affinity for trimethoprim
Acquisition of second gene via plasmid that encodes trimethoprim-resistant DHFR
Acetylates chloramphenicol via acetyl coenzyme A to yield derivatives that are unable to bind to the ribosome
Mutation in genes encoding two subunits of DNA gyrase
Mutation in genes encoding two subunits of DNA gyrase
Reduced target enzyme expression (ParE)
Energy-dependent efflux pump removes hydrophilic fluoroquinolones; associated with alterations in gene or increased transcription of norA
Single base pair change in rpoB resulting in an amino acid substitution in the ß subunit of RNA polymerase
Low-level mupirocin resistance (MIC 8-256 mg/1) because of mutation in gene for target enzyme
High-level resistant (MIC >512 mg/1) mediated by ileS-2 gene, which encodes an additional isoleucyl-tRNA synthetase aph(3')-IIIa (aphA) Plasmids and chromosome ant(4')-Ia (aadD) Plasmids
Gray and Fitch (1983), Coleman etal. (1985), el Solh, Moreau and Ehrlich (1986)
Schwotzer, Kayser and Schwotzer (1978), Thomas and Archer (1989)
Chromosome: on transposon Tn554 Murphy (1985b)
grlB (also termed parC and parE) norA
Regulation of norA affected by flqB
Plasmid > chromosome
Plasmids (class I)
Chromosome: with cat plasmid integrated
Val-to-Phe changes at either residue 588 (V5 88F) or residue 631 (V63IF) Plasmid
Coughter, Johnston and Archer (1987), Tennent etal. (1988), Rouch etal. (1989), Dale, Then and Stuber (1993)
Shaw (1983), Projan and Novick (1988), Cardoso and Schwarz (1992)
Sreedharan, Peterson and Fisher (1991), Goswitz etal. (1992), Margerrison, Hopewell and Fisher (1992), Brockbank and Barth (1993), Ito etal. (1994)
Tracksis, Wolfson and Hooper (1991), Ferrero etal. (1994), Ferrero, Cameron and Crouzet (1995), Hooper (2001), Ince and Hooper (2003)
Ohshita, Hiramatsu and Yokota (1990), Yoshidaetal. (1990), Kaatz, Seo and Ruble (1993), Ng, Tracksis and Hooper (1994)
Aubry-Damon, Soussy and Courvalin (1998)
Gilbart, Perry and Slocombe (1993), Hodgson etal. (1994), Cookson (1998), Antonio, McFerran and Pallen (2002)
\o of peptidoglycan precursors. Glycopeptides are important in clinical practice because they are used to treat severe infections caused by MRSA. Three categories of S. aureus resistances to vancomycin have been described since 1997:
1. S. aureus with intermediate-level resistance to vancomycin (VISA);
2. S. aureus with heteroresistance to vancomycin (hVISA);
3. vancomycin-resistant S. aureus (VRSA).
Criteria have been developed by the CDC for the identification of VISA. These are the following:
1. growth within 24 h on commercial brain-heart infusion agar screen plates containing 6 |lg/ml of vancomycin;
3. broth microdilution MIC of 8-16|g/ml (Tenover etal. 1998).
A variety of screening methods has been described to detect hVISA, but the optimal method is still under debate (Liu and Chambers 2003). VRSA is defined by an MIC >32 |g/ml (National Committee for Clinical Laboratory Standards 2002a).
Staphylococcus aureus with intermediate-level resistance to vancomycin (VISA) was first detected in Japan in 1996 (Hiramatsu etal. 1997b). The strain (Mu50) was isolated from an infant who developed a sternal wound infection that was refractory to treatment following surgery to correct a congenital cardiac defect. This strain was reported to have an MIC of 8 mg/l (Hiramatsu et al. 1997b). VISA strains have since been identified worldwide, although they are currently rare in clinical practice, and most appear to evolve from MRSA strains in patients who have received prolonged vancomycin treatment. The first hVISA (Mu3) was identified in Japan from the sputum of a patient with MRSA pneumonia following surgery (Hiramatsu etal. 1997a). This strain was reported to have an MIC of 3 |g/ml (Hiramatsu etal. 1997a). Serial passage of Mu5 in increasing concentrations of vancomycin gave rise to subpopulations with levels of resistance comparable to those of Mu50, and typing showed that Mu3 and Mu50 had the same PFGE pattern (Hiramatsu etal. 1997a). hVISA has since been reported from around the world and appears to be more common than VISA (Hiramatsu etal. 1997a; Hiramatsu 2001; Liu and Chambers 2003). The first VRSA was reported in Michigan in July 2002, with a second apparently unrelated case in Pennsylvania 2 months later (Centers for Disease Control and Prevention 2002a, 2002b).
The mechanisms of vancomycin resistance for VISA and hVISA are not fully elucidated. These strains do not carry the enterococcal vancomycin-resistance gene vanA, vanB or vanC 1-3 (Hanaki etal. 1998a). Findings at the time of writing indicate that changes in the cell wall are important in this process. Strain Mu50 has increased amounts of glutamine-non-amidated muropeptides and decreased cross-linking of peptidoglycan compared with Mu3, together with a decreased dimer-to-monomer ratio of muropeptides (Hanaki etal. 1998b). The peptidoglycan of Mu50 binds 1.4 times more vancomycin than that of Mu3 (Hanaki etal. 1998b). Both strains produce three to five times the amount of PBPs when compared with vancomycin-susceptible S. aureus control strains with or without methicillin resistance, and transmission electron microscopy has shown a doubling in the cellwall thickness in Mu50 compared with control strains (Hanaki etal. 1998a). An increase in the consumption of vancomycin by the cell wall and reduction in the amount of vancomycin reaching the cyto-plasmic membrane may both contribute to vancomycin resistance. Microarray transcription analysis of two clinical VISA isolates was compared with two derivatives with MIC of 32 | g/ml after passage in the presence of vancomycin (Mongodin etal. 2003). Of 35 genes with increased transcription, 15 involved purine biosynthesis or transport, and the regulator of the major purine biosynthetic operon was mutant (Mongodin etal. 2003). These genes may be involved in the metabolic processes required for the production of the thicker cell wall. The genome sequences of Mu50 and vancomycin-susceptible MRSA strains N315, EMRSA-16 and COL have been compared (Avison etal. 2002). Several mutations affecting important cell-wall biosynthesis and intermediary metabolism genes were identified in Mu50.
Vancomycin-resistant enterococci were reported in 1988. Transfer from Enterococcus faecalis to S. aureus of high-level resistance to both vancomycin and teicoplanin was performed in the laboratory in the early 1990s through the interstrain transfer of vanA (Noble, Virani and Cree 1992). This gene encodes a ligase that causes an alteration in the composition of the terminal dipeptide in muramyl pentapeptide cell-wall precursors, leading to decreased binding affinity to glyco-peptides (Fraimow and Courvalin 2000). This represents the mechanism of resistance in the two VRSA isolated in the United States that contain the vanA gene (Weigel etal. 2003; Tenover etal. 2004).
Resistance to Other Antibiotics
Table 5.3 summarizes the antibiotic-resistant mechanisms for S. aureus to the other major antibiotic groups.
Staphylococcus aureus plasmids have been classified into three general classes. Class I plasmids are small (1-5 kbp), have a high copy number (10-55 copies per cell) and are either cryptic or encode a single antibiotic resistance. These plasmids are the most widespread throughout the genus Staphylococcus. For example, the pT181 family comprises a group of small (4-4.6 kbp) plasmids that usually encode tetracycline or chloramphenicol resistance, and the pSN2 family plasmids often encode erythromycin resistance. Class II plasmids, commonly called penicillinase or P-lactamase plasmids, are relatively large (15-33 kbp), have a low copy number (4-6 per cell) and carry several combinations of antibiotic and heavy-metal resistance genes, many of which are located on transposons (Shalita, Murphy and Novick 1980; Lyon and Skurray 1987). Class III plasmids appear to be assemblages of transposons and transposon remnants (Gillespie etal. 1987).
Most strains of S. aureus are multiply lysogenic (Verhoef, Winkler and van Boven 1971; Pulverer, Pillich and Haklova 1976). The temperate phages of S. aureus may be subdivided into three main serological groups termed A, B and C. The relevance of bacteriophage is threefold:
1. these are commonly used in experimental transduction;
2. bacterial susceptibility to phage has been used for many years as the basis for a typing scheme;
3. phage may influence the gene complement or gene expression of an organism.
Alterations in gene complement or expression may occur either via carriage of genes into the organism [lysogenic conversion by prophage that carries genes encoding staphylokinase and staphylo-coccal enterotoxin A (SEA)] or via negative lysogenic conversion in which genes that contain phage-attachment sites within their sequence are disrupted (e.g. hlb encoding P-toxin; Coleman etal. 1986).
The genome of S. aureus is composed of a single chromosome of around 2.8 Mb, which is predicted to encode approximately 2500 genes. The genome sequence of two related S. aureus strains (N315 and Mu50) was published in 2001 (Kuroda etal. 2001). N315 is an MRSA strain isolated in Japan in 1982, and Mu50 is an MRSA strain with intermediate-level resistance to vancomycin, isolated in Japan in 1997. It was observed that most of the antibiotic resistance genes were carried either by plasmids or by mobile genetic elements including a unique resistance island and that many putative virulence genes seem to have been acquired by lateral gene transfer. Three classes of new pathogenicity islands were identified: a TSST island family, an exotoxin island and an enterotoxin island. Sequencing of strain MW2 was subsequently undertaken (Baba etal. 2002). This strain is a community-acquired MRSA isolated from a 16-month-old girl in the United States with fatal septicaemia and septic arthritis. The genome of this strain was compared with that of N315 and Mu50, including a comparison between the staphylococcal cassette chromosome mec (SCCmec) that carries the mec gene encoding methicillin resistance. Nineteen additional virulence genes were found in the MW2 genome, all but two of which were carried by one of the seven genomic islands of MW2. Sequencing of two further isolates by The Wellcome Trust Sanger Institute, UK, has provided an opportunity to undertake comparative genomics for five genomes (Holden etal. 2004). The strains sequenced were a methicillin-susceptible S. aureus (MSSA) strain (MSSA476) that is phylogenetically close to MW2 and a hospital-acquired representative of the epidemic methicillin-resistant EMRSA-16 clone (MRSA252) that is phylogenetically divergent from any other sequenced strain. Schematic circular diagrams of the S. aureus MRSA252 and MSSA476 genomes are shown in Plate 2. Comparison of the five S. aureus whole-genome sequences showed that 166 genes (6%) in the phylogenetically divergent MRSA252 genome were not found in the other published genomes. Around 80% of the genome was highly conserved between strains, the remaining 20% of each consisting of highly variable genetic elements that spread horizontally between strains at high frequency. This is consistent with the findings of a study that used a DNA microarray representing more than 90% coverage of the S. aureus genome (Fitzgerald etal. 2001). Comparison of the genomes of 36 strains from divergent clonal lineages showed that genetic variation was very extensive, with approximately 22% of the genome being comprised of dispensable genetic material.
Publication of the genome sequence of two further strains (COL and 8325) is imminent, with genome sequence of additional strains reportedly in progress. Comparison of the S. aureus genome with the genome sequence of S. epidermidis (ATCC 12228) has demonstrated that S. epidermidis contains a lower number of putative virulence determinants (Zhang etal. 2003).
Sequencing of smaller fragments of the genome in populations of isolates has provided interesting insights into rates of genetic mutation. Examination of the sequence changes at multilocus sequence typing (MLST) loci during clonal diversification has shown that point mutations give rise to new alleles at least 15-fold more frequently than by recombination (Feil etal. 2003). This contrasts with the naturally transformable species Neisseria meningitidis and Streptococcus pneumoniae, in which alleles change between five- and tenfold more frequently by recombination than by mutation.
Was this article helpful?