Tweet
The Lancet Infectious Diseases 2002; 2:180-189
DOI:10.1016/S1473-3099(02)00227-X
Secrets of success of a human pathogen: molecular evolution of pandemic clones of meticillin-resistant Staphylococcus aureus
Duarte C Oliveira a, Alexander Tomasz a and Dr Herm?nia de Lencastre
a 
See Errata
Summary
The adaptive power of Staphylococcus aureus: emergence of antibiotic-resistant strains
Antibiotic resistance in an evolutionary context
Evolutionary zero time in the emergence of MRSA
Molecular typing of MRSA
The nature of early MRSA isolates
Identification of five pandemic MRSA clones
Identification of two distinct ancestral lineages among the five pandemic MRSA clones
The mec element as an evolutionary marker
MRSA lineages and the acquisition of mec elements
The origin of the mec element
New trends in MRSA epidemiology
Search strategies and selection criteria
References
Summary
The first European isolate of meticillin-resistant Staphylococcus aureus (MRSA) was detected in 1960. Since then MRSA has become a leading cause of nosocomial infections worldwide. Using molecular typing techniques?primarily pulsed-field gel electrophoresis (PFGE)?we identified five major MRSA clones that accounted for almost 70% of the over 3000 MRSA isolates recovered in hospitals mainly in southern and eastern Europe, South America, and the USA. Most of our surveillance studies were done in these areas. Multilocus sequencing typing (MLST) of representative isolates of this collection showed that these five pandemic MRSA clones have evolved from only two distinct ancestral genetic backgrounds, one of which can be traced back to the very first European MRSA isolates and also to meticillin susceptible S aureus strains circulating in Danish hospitals during the mid to late 1950s?ie, shortly before the introduction of meticillin into therapy. The second lineage with a completely different MLST profile included MRSA frequently recovered in the USA, Japan, and among paediatric isolates from several parts of the world. A few isolates with a third distinct MLST type corresponding to that of EMRSA-16 were also detected in the early Danish isolates. The four structural types of mec element, the heterologous DNA segment containing the meticillin resistance determinant mecA, were present in unique combinations with the MRSA clonal types. Our findings establish evolutionary associations in the most widely spread pandemic clones of MRSA. The epidemiological factors that contributed to the massive dissemination of a few MRSA clones are not well understood. We suggest, however, that the secrets of effectiveness of MRSA could be hidden in the unique genetic background of a surprisingly few lineages of S aureus particularly well able to cope with the contemporary clinical environment.
Back to top
The adaptive power of Staphylococcus aureus: emergence of antibiotic-resistant strains
Staphylococcus aureus has long been recognised as one of the major human pathogens responsible for a wide range of afflictions from minor infections of the skin to wound infections, bacteraemia, infections of the central nervous system, respiratory and urinary tracts, and infections associated with intravascular devices and foreign bodies.1,2 Most S aureus strains are opportunistic pathogens that can colonise individuals, without symptoms, for either short or extended periods of time, causing disease when the immune system becomes compromised. Indications are that this bacterial pathogen has affected human beings since prehistoric times. Before the antibiotic era S aureus diseases had high mortality rates. In 1941 the mortality rate of S aureus bacteraemia at the Boston City hospital was reported to be 82%.3 In versatility of pathogenic strategies, numbers of virulence factors, and capacity to survive and multiply in a wide range of environments, S aureus is unsurpassed by any other human pathogen. The immense genetic repertoire of this bacterium for adapting to rapidly changing and uniformly hostile environments was repeatedly shown by the emergence of S aureus strains that acquired resistance mechanisms to virtually all antimicrobial agents shortly after the introduction of these drugs into clinical practice.
The introduction of benzylpenicillin into chemotherapy in the early 1940s found S aureus fully susceptible and several of the first successes of penicillin therapy were related to the cure of formerly untreatable staphylococcal diseases.4 But by the mid 1950s the number of S aureus clinical isolates showing high-level resistance to penicillin increased rapidly, to such an extent that penicillin ceased to be a useful therapeutic agent against staphylococcal infections. The mechanism of penicillin resistance involved the acquisition of a plasmid-borne penicillinase capable of degrading the antibiotic before it had reached its cellular targets. The effects of this ?plasmid epidemic? were first seen on hospital isolates of S aureus but soon afterwards the penicillinase-based mechanism found its way into community isolates as well.5 Currently, most S aureus isolates that cause disease or simply colonise healthy individuals are resistant to penicillin. A study of the S aureus colonising flora of 1001 healthy volunteers showed that 97% of the S aureus isolates recovered carried the resistant trait to penicillin.6
The power of S aureus to mount counter-strategies to additional products of the antibiotic era besides penicillin is also well documented. Records of the Danish Health Board registered the years of introduction of various antimicrobials into clinical practice, beginning with penicillin in 1945?1946, streptomycin in 1948, tetracycline in 1950, and erythromycin in 1953.5 The same records indicate that S aureus bloodstream isolates resistant to penicillin, streptomycin, tetracycline, and erythromycin were recovered as early as 1957. The introduction of meticillin in clinical practice in 1960 was followed by the appearance of the first bloodstream isolate of S aureus that was resistant not only to penicillin, streptomycin, and tetracycline (and occasionally to erythromycin), but to meticillin as well.
Meticillin, originally called celbenine, was the first mechanism-based antimicrobial agent: it is a semisynthetic derivative of penicillin chemically modified to withstand the degradative action of penicillinase.7 The drug was introduced into therapy in Europe in 1959?1960. 1 year later, the first meticillin-resistant S aureus (MRSA) strains were detected,8 and the first clinical failure by an MRSA strain was described,9 followed by a report on the first MRSA outbreak in 1963.10 Since the 1960s, MRSA strains have spread among hospital isolates in several waves, which eventually disseminated these strains worldwide. Recent surveillance studies in hospitals in various parts of the world indicate a varying incidence of MRSA strains depending on the country and the hospital. In the USA, the National Nosocomial Infections Surveillance System (NNISS) recorded an increase of MRSA in large US hospitals, from 4% in the 1980s to 50% in the late 1990s. In some hospitals resistance frequencies as high as 80% have been recorded.11
Back to top
Antibiotic resistance in an evolutionary context
The introduction of vast quantities of structurally diverse antimicrobial agents into the human environment during the past 60 years has presented a new set of challenges to bacterial pathogens such as S aureus. Effective lineages of contemporary pathogens must excel in several capacities: they must be able to acquire resistant genes and to construct regulatory mechanisms that can adjust resistance levels to increasing concentrations of the antimicrobial agent. In the environment, the resistance-related determinants must find their way into genetic backgrounds that assure the capacity to compete with other bacteria. Pathogens must be able to spread, establish ecological reservoirs, colonise, and cause disease.
Detection and elucidation of the mechanisms of drug resistance requires the collection of S aureus isolates from the clinical environment. Strain collection in increasingly sophisticated epidemiological contexts and analysis of the strains by molecular typing techniques has led to the emergence of the new discipline, molecular epidemiology, which has greatly improved the capacity of hospitals to track the source and transmission routes of bacterial pathogens during hospital outbreaks. Beyond producing sophisticated tools for epidemiological tracking, molecular epidemiology has become an alliance of clinical and molecular microbiologists that began to tackle issues of the population structure of microbial pathogens in their native environment.12?17
Back to top
Evolutionary zero time in the emergence of MRSA
The emergence and dissemination of meticillin-resistant S aureus is a case of accelerated evolution in which these bacteria are propelled by the selective pressure of vast quantities of antimicrobial-related agents in the planetary environment.18 The evolutionary zero time of this process is the acquisition of the central genetic determinant of MRSA, the mecA gene, a 2?1 kb stretch of DNA that is not native to the species S aureus and which is embedded in a larger block?up to 60kb?of additional ?foreign? DNA called the mec element19 or staphylococcal chromosomal cassette (SCCmec),20 which together with the mecA is incorporated into the S aureus chromosome at a site-specific location.21,22 The mecA gene encodes for a 78 kDa penicillin-binding protein (PBP2A), which has very low affinity for beta-lactam antibiotics.23?25 Recent evidence indicates that PBP2A is a transpeptidase26 that, assisted by the transglycosidase domain of the native PBP2 of S aureus,27 takes over the function of cell wall biosynthesis in the presence of beta-lactam antibiotics in the medium.
Back to top
Molecular typing of MRSA
The purpose of this overview is not to detail the molecular mechanism of meticillin resistance but to survey recent information that has allowed insights into the evolutionary process that has led to the emergence of a relatively few pandemic clones of MRSA. Critical for obtaining this information were the combination of several molecular typing techniques, the elucidation of the structure of the mec elements,22 and the availability of large collections of MRSA in our laboratories. These collections, from various parts of the world, include the strain collection of the Danish surveillance system that preserved every S aureus bloodstream isolate recovered in Denmark since the mid 1950s.5
A particularly successful strategy for MRSA typing has been the combination of three molecular techniques with different discriminative characteristics:28 (i) macrorestriction pattern of chromosomal DNA after Smal digestion and separation of the fragments by pulsed-field gel electrophoresis (PFGE); (ii) polymorphisms in the vicinity of the mecA gene detected by probing Clal- digested DNAs with a mecA probe;29 and (iii) transposon Tn554 insertion patterns detected by probing C/al-digested DNAs with a specific probe.29 The individual and combined resolving ability of these methods is shown in table 1, where analysis of a group of MRSA recovered in a surveillance study in 12 hospitals in New York showed that the highest resolution was achieved by the PFGE patterns.30 PFGE has been recognised as a highly discriminatory technique and suggested to be the gold standard for outbreak investigations.31 Two sequencing-based techniques, multilocus-sequencing type (MLST) and spaA typing, have been developed recently, with great potential for global epidemiology studies.13,15,32 MLST is based on the DNA sequencing of the internal fragments of seven unlinked housekeeping genes (arcC, aroE, glpF, gmk, pta, tpi, and yqiL) and spaA typing is based on the DNA sequencing of the polymorphic region of protein A, a superantigen of S aureus.
Table 1. Resolution of three molecular typing methods (adapted from reference 30)
Back to top
The nature of early MRSA isolates
The preservation of all S aureus bacteraemic isolates in Denmark since the mid 1950s has allowed us to look at the epidemiological context of the appearance of MRSA in that country. Among 3704 S aureus bacteraemic isolates recovered between 1957 and 1970, 646 MRSA isolates were identified and all but nine of these were resistant not only to meticillin but to penicillin, streptomycin, tetracycline, and often to erythromycin. Furthermore, 616 of these 646 isolates (95%) belonged to unique phage groups called phage group III or the related 83A complex.5 Data from the Danish surveillance system indicate that almost half of all S aureus bacteraemic isolates?both MRSA and meticillin-susceptible S aureus (MSSA)?recovered in Denmark during this period belonged to these phage groups. The very first European MRSA isolates identified in the UK also belonged to the same phage group and were resistant to penicillin, streptomycin, tetracycline (PST), and occasionally to erythromycin (PSTE) as well.8,10 These observations strongly suggested that we might be able to identify among these Danish MSSA isolates one of the first S aureus lineages that received the mecA gene. Additional evidence pointing in this direction was the sequential appearance of MSSA strains belonging to phage group III and/or 83A with the PST antibiotype, and the appearance of MRSA strains belonging to the same phage group and sharing the same PST or PSTE multiresistance pattern in addition to being meticillin resistant?PSTM and PSTEM (figure 1). Molecular typing with MLST, PFGE, and spaA typing fully confirmed this point.33 The overwhelming majority of historically early MSSA and MRSA isolates from Denmark and the earliest MRSA isolates from the UK shared identical or very similar molecular profiles (table 2). Characterisation of these earliest, archaic, strains of MRSA showed remarkably uniform properties in individual isolates:34 low meticillin minimum inhibitory concentration (MIC) value (6?25 μg/ml) for most cells; heterogeneous expression of resistance; a single, common Clal-mecA polymorph II; lack of the regulatory gene medI; lack of Tn554 in most isolates; and a closely related PFGE profile. The epidemic capacity of these archaic MRSA strains was shown by their recovery from as many as 18 hospitals in Denmark.
Click to enlarge image
Figure 1. Sequential appearance of meticillin-susceptible and meticillin-resistant blood isolates of S aureus belonging to phage group III and the related 83A complex in Denmark. The numbers plotted represent all S aureus blood isolates identified in Denmark during the particular period (adapted from reference 33).
Table 2. Properties of historically early strains of MSSA recovered in Denmark (adapted from reference 33)
These results identify the progeny that must have been one of the first S aureus receiving the heterologous mec element from an unknown bacterial donor?ie, the first stage in the evolutionary process of emergence of MRSA. The genetic background of early MSSA and MRSA isolates was very similar or identical to the genetic background of one of the most widely spread contemporary multidrug-resistant clones of MRSA, the Iberian clone.33,34 Although the contemporary Iberian clone shares the same genetic background as the ?archaic? MRSA isolates, this clone seems to be a further step ahead in MRSA evolution because it has acquired extra resistance determinants, some of them resident in mobile elements such as plasmids (pUB110) and transposons (Tn554). Thus, by contrast with the limited antibiotic resistance pattern of the archaic MRSA (PSTEM), most isolates of the Iberian clone are resistant to most commonly used antimicrobial agents, with the exception of the co-trimoxazole (sulfamethoxazole-trimethoprim) and the glycopeptides.35 The integration of the linearised plasmid pUB110 in the downstream vicinity of the mecA gene explains the differences in the mecA polymorphism detected between the archaic strains (mecA polymorph II) and the contemporary Iberian MRSA strains (mecA polymorph I).36
Back to top
Identification of five pandemic MRSA clones
In recent years, the combination of mecA polymorphisms, Tn554 insertion patterns, and PFGE profiles was applied to characterise more than 3000 MRSA isolates collected in surveillance studies and outbreak investigations in southern and eastern Europe, Latin America, and the USA between 1994 and 2000, through the international CEMNET initiative.17 Five major clonal lineages were defined, which are widely disseminated in those areas and accounted for a large proportion?68%?of the isolates, indicating that they represented successful lineages in terms of ability to cause infection, to persist, and to spread from one geographic site to another, including across continents (table 3). The names assigned to these five MRSA?Iberian, Brazilian, Hungarian, New York/Japan, and paediatric pandemic MRSA clones?reflect the geographic area in which they were first identified and/or indicate some unique epidemiological property. In addition to their distinct PFGE patterns, the prototypes of these five clonal types had distinct properties. The Iberian clone carried a ClaI::mecA polymorph I and ClaI::Tn554 pattern E (clonal type I::E). The corresponding clonal types in the other MRSA clones were as follows: Brazilian clone XI::B; Hungarian clone III::B; New York/Japan clone I::A; and paediatric clone II::NH. All strains were tested for susceptibility to nine antibiotics (oxacillin, penicillin, clindamycin, erythromycin, gentamicin, spectinomycin, co-trimoxazole, tetracycline, and vancomycin). The Iberian, Hungarian, and New York/Japan clones were only susceptible to co-trimoxazole and the Brazilian clone was only susceptible to spectinomycin. The paediatric clone was only resistant to oxacillin, penicillin, gentamycin, and occasionally erythromycin. All isolates were susceptible to vancomycin.
Table 3. Pandemic MRSA clones prevalence (adapted from reference 37)
Figure 2 documents the spread of these five pandemic MRSA clones. The Iberian clone was first reported in Spain in 1989 and since then has been reported in Portugal, Italy, the UK, Germany, Belgium, Switzerland, France, Czech Republic, Poland, and the USA. The Brazilian clone was first described in 1992 in Brazil and then in Portugal, Argentina, Uruguay, Chile, and the Czech Republic. The Hungarian clone was extensively disseminated among hospitals in that country and was most recently also identified in Taiwan. The New York/Japan clone was identified as the dominant MRSA in hospitals in metropolitan New York, New Jersey, Pennsylvania, and Connecticut, and also in one hospital in Tokyo, Japan. The paediatric clone was first reported in a paediatric hospital in Portugal in 1992 and was also reported in Poland, USA, Argentina, and Colombia (for a list of references concerning the spread of the MSRA pandemic clones see table 1 of reference 37).
Click to enlarge image
Figure 2. International spread of the pandemic MSRA clones.
Molecular typing techniques have not only allowed the identification of pandemic clones of MRSA, but have also enabled us to monitor MRSA clones circulating in different hospitals and at different time intervals in a country providing evidence for the stability of epidemic clones over time. Figure 3 shows the MRSA clonal distribution in ten hospitals in six Portuguese cities during 9 years. Figure 4 shows that in the largest Portuguese teaching hospital located in Coimbra, during three distinct surveillance periods, the Iberian and Brazilian clones were present in different proportions but together accounted for virtually all the MRSA isolates. In this case, molecular typing techniques were able to document the introduction of the Brazilian clone during the first surveillance period. During the following surveillance periods, the prevalence of this clone increased from 5% to 38%, whereas the previous endemic Iberian MRSA clone decreased from 89% to 55%.38
Click to enlarge image
Figure 3. MRSA clonal distribution in hospitals in six cities in Portugal.
Click to enlarge image
Figure 4. Variation of the incidence of two pandemic MRSA clones during three consecutive surveillance periods in the largest Portuguese hospital, located in Coimbra (adapted from reference 38).
Back to top
Identification of two distinct ancestral lineages among the five pandemic MRSA clones
The primary criterion used for the initial classification of MRSA isolates into the five pandemic clonal types was their PFGE pattern. PFGE is a high-resolution technique capable of registering DNA rearrangements and acquisition or loss of genetic determinants?ie, genetic events that can happen on a short time scale. While providing a high degree of resolution, this technique could blur long-range evolutionary associations in MRSA strains. By contrast, MLST variations accumulate slower than variations in PFGE, since they arise from neutral mutations in housekeeping genes. Thus, the same MLST allelic profile could include temporally distant strains that have accumulated differences in PFGE patterns.13
Using MLST we identified two major genetic backgrounds, A and B, among representative isolates of the five pandemic MRSA clones.37 Background A has four closely related variants (single-locus variants). Backgrounds A1 and A2 were both identified among members of the ?archaic? and Iberian clones, background A3 was detected in a minor USA clone (clone V), and background A4 characterises both the Brazilian and Hungarian clones (table 4). The variant alleles in backgrounds A2 and A3 each differ from Al in a single nucleotide at one of the alleles, suggesting evolution through single point mutations.14 Backgrounds A1, A2, and A3 also share the same spaA type motif. Background A4 is a single locus variant of background A3, but it differs in three nucleotides in the arcC allele, suggesting evolution by recombinational events.14 Background A4 is also characterised by a distinct spaA type. Together, these findings suggest that the Iberian, Brazilian and Hungarian clones, and clone V, sharing the closely related genetic backgrounds A1?A4, have evolved from a common ancestor with a genotype very similar to that of the archaic MRSA. A second and completely different chromosomal background B was identified in MRSA strains originally recognised in the New York/Japan and paediatric MRSA clones.
Table 4. Genetic backgrounds and SCCmec types of the pandemic MRSA clones.
Back to top
The mec element as an evolutionary marker
The initial genetic event marking the emergence of an MRSA lineage is the acquisition of the mec element, introducing the meticillin-resistance determinant mecA into a MSSA lineage. Four types of mec elements or SCCmec have been identified so far. Type I (34 kb) was detected in the first MRSA strain isolated in 1961 in the UK (strain NCTC10442); type II (52 kb) was identified in an MRSA strain isolated in 1982 in Japan (strain N315); type III (66 kb) was identified in an MRSA strain isolated in 1985 in New Zealand (strain 82/2082);22 and type IV (20?24 kb) was identified in two community-acquired MRSA strains,39 as well as in representatives of the paediatric clone from Poland and from Portugal.37
The mec element structures resident in representative isolates of the five pandemic MRSA clones were recently characterised, and specific patterns of association between clonal types and SCCmec types were recorded38 (table 4). MRSA strains with the genetic background A1 and A2 (archaic and Iberian clone) carried SCCmec types I and IA; strains with genetic background A4 (Brazilian and Hungarian clone) carried SCCmec types III and IIIA. SCCmec type IA (Iberian clone) differs from type I by the presence of the linearised plasmid pUB110 in the mecA downstream vicinity, whereas SCCmec type IIIA (Brazilian clone) varies from type III by the absence of the linearised plasmid pT181 in the mecA downstream vicinity. Strains representing the New York/Japan clone (genetic background B) carried SCCmec type II. The SCCmec type IV was found in strains representing the paediatric clone (genetic background B) and also in the minor clone V (background A3) (figure 5).
Click to enlarge image
Figure 5. SCCmec element structures and variants. The basic structures of SCCmec types I, II III and IV are depicted, as well as its variants SCCmec IA and IIIA.21,22,38,39 LJ=chromosomal left junction, RJ= chromosomal right junction.
A search of the MLST database identified two additional important genetic backgrounds characteristic of E-MRSA 15 (allelic profile 7-6-1-5-8-8-6, background D) and E-MRSA 16 (allelic profile 2-2-2-2-3-3-2, background C), two MRSA clones that currently represent the most frequent MRSA in UK hospitals.40,41 E-MRSA 15 was also detected in Germany.42 An E-MRSA 16 representative strain is being sequenced. Examination of its preliminary genome sequence data indicates that it harbours a SCCmec type II, the same mec element shown to be resident in a completely different genetic background: background B characteristic of the New York/Japan clone. The presence of SCCmec type II in both of these distinct genetic backgrounds (B and C) could represent either an independent acquisition event or horizontal transfer of the mec element from one S aureus strain to another.
MLST studies of MSSA strains uncovered a great diversity of allelic profiles,12,13 which contrasts with the clonal structure of the highly epidemic MRSA. The worldwide spread of MRSA seems to be primarily driven by the dissemination of a few pandemic clones in which the mec element found genetic backgrounds that provided appropriate fitness for virulence and epidemic spread of the bacteria. One of these genetic backgrounds?A1?has recently been identified in MSSA strains that were highly prevalent in the late 1950s in Denmark.33 MSSA with genetic backgrounds A3 (clone V), B (Brazilian and Hungarian clones), and C (E-MRSA 16) were also detected in a few Danish isolates from this era (table 2).33 In a recent communication by Enright,43 the same backgrounds were detected among an independent collection of 394 epidemic MRSA isolates from 22 different countries characterised by MLST. The most frequent clonal complex I described by this author appears to correspond to genetic background A and its variants described in this communication. While MRSA with either ST22 or ST36 were not detected among pandemic isolates in our collection, a few MSSA isolates with ST36?corresponding to our genetic background C?were detected among historically early MSSA from Denmark.33 What Enright43 describes as a fourth type of MRSA detected among isolates from the USA and Europe presumably corresponds to our genetic background B (MLST 1-4-1-4-12-1-10) characteristic of both the New York/Japan and the paediatric clones of MRSA, which share a common MLST but may be distinguished by the distinct SCCmec type they carry (see table 4).
Back to top
MRSA lineages and the acquisition of mec elements
Putting together all data one might propose important inferences as to the evolutionary pathways through which MRSA lineages could have emerged. MRSA strains have a strong clonal structure, by contrast with MSSA strains, probably due to their shorter existence (four decades). The mec element, containing the meticillin-resistance determinant, seems to have been imported into the species of S aureus at multiple, yet restricted and independent, occasions, as shown by the specific pattern of association between clonal types and SCCmec types. Horizontal transfer of the mec element between different S aureus lineages could also have happened as suggested by the presence of SCCmec type IV in backgrounds A, B, and D, and SCCmec type II in backgrounds B and C. However, horizontal transfer seems to be a rare event and the spread of meticillin-resistance in S aureus seems to be predominantly due to the clonal expansion of very few lineages. A tentative scheme outlining steps in the evolution of MRSA clonal families and the acquisition of SCCmec elements is presented in figure 6. The proposed evolutionary associations are primarily based on the properties of MRSA strains analysed and, in our case, these came mainly from sources in southern and eastern Europe, Latin America, and the USA. Important findings were also made possible by the availability of the Danish MRSA collection and by the information available in the MLST database. Analysis of isolates from other geographic areas would be of great interest.
Click to enlarge image
Figure 6. Proposed evolutionary pathways of the contemporary pandemic MRSA clones. Letter colors represent different genetic backgrounds (MLST), box colours represent different SCCmec types and dash lines represent possible horizontal transfers of the mec element.
Back to top
The origin of the mec element
The origin of the mec element is not known. The assembly of the several mec element structures may have involved multiple hosts, possibly among coagulase-negative staphylococci.44 A close homologue of the S aureus mecA gene was recently identified as a gene native to the animal commensal species Staphylococcus sciuri, most strains of which are fully susceptible to meticillin and all beta-lactam antibiotics.45,46 The insertion sequence IS1272 present in SCCmec types I and IV appears in several copies in the Staphylococcus haemotyticus genome.47
The distribution of SCCmec type IV provides some interesting clues concerning possible evolutionary associations. In MRSA this mec element structure is seen in three different genetic backgrounds, and SSCmec type IV was also seen in community-acquired MRSA6,39 and also in coagulase-negative staphylococci (D C Oliveira, unpublished data).39 These findings suggest an enhanced mobility of this mec element, perhaps explained by its smaller size. SCCmec type IV has the same downstream structure as the SCCmec type I. A high degree of similarity is also apparent upstream of mecA where both SCCmec I and IV have a deleted mecR1, an IS1272, and ccrAB flanked by two homologous sequences. However, the region between the chromosomal left junction and the upstream region of ccrAB?ie, the gene pls and its flanking regions, close to 18 kb in size?is missing in SCCmec IV and was replaced by a variable DNA segment.37 The close association of SCCmec type IV to SCCmec type I, which is present in the earliest MRSA isolates, together with its presence in hospital-acquired and community-acquired MRSA strains as well as in coagulase-negative staphylococci, suggests that this mec element structure might be closely related to a putative precursor of the mec element.
Back to top
New trends in MRSA epidemiology
The specific pattern of association of SCCmec types and MRSA clones opens the door to some speculations about possible contributions of the SCCmec itself to the pathogenic potential of MRSA strains. It has been proposed that the mec element might be a pathogenic and/or antibiotic-resistance island,22,48 capable of autonomous excision provided by the chromosomal cassette recombinase genes ccrA and ccrB present in every mec element.22 The mec element, besides being the carrier of mecA and its regulatory genes (mecI and mecR1,) also contains additional resistance determinants in the form of integrated mobile elements (pUB110, pT181, pI258, Tn554, and ψTn554). In this respect the SCCmec structure clearly represents an antibiotic-resistance island. Other regions of the SCCmec might be more directly contributing to the pathogenic potential of the S aureus cell. Such regions could be the pls gene present in SCCmec type I and the kdp operon present in SCCmec type II. The pls gene, which codes for a large surface protein, could suppress adhesion of the bacteria to their host at some stages during infection.49 The kdp operon is an extra copy of the potassium-transporter operon present elsewhere in the genome,48 which could improve the colonisation and invasion processes of the host tissues. A possible association should be considered between the SCCmec structure and the epidemiology of some MRSA clones. For instance, SCCmec type IV, which differs mainly from SCCmec type I by the absence of the pls gene and its flanking regions, characterises both the paediatric MRSA clone and the USA clone V, which have completely different genetic backgrounds. However, both clones have low-level and heterogeneous meticillin resistance and are frequently isolated from patients with immature or compromised immune systems (children and AIDS patients).30,50?52 In a recent study, 22 of 28 patients with AIDS had MRSA infections that were seen to belong to the USA Clone V.30 It is tempting to consider the possibility that the pls deletion contributes to the affinity of these strains for patients with weak immune systems.
Several recent findings suggest that MRSA might be emerging as a community pathogen53?57 following perhaps the historic precedent of penicillin-resistant S aureus.58 A screening of 1001 S aureus isolates recovered from colonisation sites of young and healthy people in a community showed that more than 95% of the isolates were resistant to penicillin but only seven were MRSA.6 However, five of these seven MRSA isolates were representatives of the pandemic Iberian, Brazilian, and paediatric clones, giving another reminder of the evolutionary process of these staphylococcal lineages.
Back to top
Search strategies and selection criteria
Four types of references are quoted to provide context and background information to the observations and hypotheses described in this communication: historical references to MRSA, MRSA in disease, and the changing epidemiology of MRSA; references to novel molecular typing techniques such as MLST and spaA typing; references to the origin, structure and putative functions of mec elements (SCCmec types); and references related to the origins, epidemiological features and initial molecular characterisation of the large strain collection which was the source of epidemic MRSA clones described in our paper.
Back to top
Acknowledgments
Some findings were made possible by the information available at the MLST database (www.mlst.net) and the E-MSRA 16 genomesequencing project (www.sanger.ac.uk). Partial support for this study was provided by project POCTI/1999/ESP/34872 from Funda??o para a Ci?ncia e Tecnologia, Lisbon, Portugal, awarded to HL, and by a grant from the US Public Health Service awarded to AT, project RO1 AI37275. DCO was supported by a grant from Funda??o para a Ci?ncia e Tecnologia, Lisbon, Portugal, and also from July 2001 by a doctoral grant from Funda??o Calouste Gulbenkian, Lisbon, Portugal.
Back to top
DOI:10.1016/S1473-3099(02)00227-X
Secrets of success of a human pathogen: molecular evolution of pandemic clones of meticillin-resistant Staphylococcus aureus
Duarte C Oliveira a, Alexander Tomasz a and Dr Herm?nia de Lencastre
a See Errata
Summary
The adaptive power of Staphylococcus aureus: emergence of antibiotic-resistant strains
Antibiotic resistance in an evolutionary context
Evolutionary zero time in the emergence of MRSA
Molecular typing of MRSA
The nature of early MRSA isolates
Identification of five pandemic MRSA clones
Identification of two distinct ancestral lineages among the five pandemic MRSA clones
The mec element as an evolutionary marker
MRSA lineages and the acquisition of mec elements
The origin of the mec element
New trends in MRSA epidemiology
Search strategies and selection criteria
References
Summary
The first European isolate of meticillin-resistant Staphylococcus aureus (MRSA) was detected in 1960. Since then MRSA has become a leading cause of nosocomial infections worldwide. Using molecular typing techniques?primarily pulsed-field gel electrophoresis (PFGE)?we identified five major MRSA clones that accounted for almost 70% of the over 3000 MRSA isolates recovered in hospitals mainly in southern and eastern Europe, South America, and the USA. Most of our surveillance studies were done in these areas. Multilocus sequencing typing (MLST) of representative isolates of this collection showed that these five pandemic MRSA clones have evolved from only two distinct ancestral genetic backgrounds, one of which can be traced back to the very first European MRSA isolates and also to meticillin susceptible S aureus strains circulating in Danish hospitals during the mid to late 1950s?ie, shortly before the introduction of meticillin into therapy. The second lineage with a completely different MLST profile included MRSA frequently recovered in the USA, Japan, and among paediatric isolates from several parts of the world. A few isolates with a third distinct MLST type corresponding to that of EMRSA-16 were also detected in the early Danish isolates. The four structural types of mec element, the heterologous DNA segment containing the meticillin resistance determinant mecA, were present in unique combinations with the MRSA clonal types. Our findings establish evolutionary associations in the most widely spread pandemic clones of MRSA. The epidemiological factors that contributed to the massive dissemination of a few MRSA clones are not well understood. We suggest, however, that the secrets of effectiveness of MRSA could be hidden in the unique genetic background of a surprisingly few lineages of S aureus particularly well able to cope with the contemporary clinical environment.
Back to top
The adaptive power of Staphylococcus aureus: emergence of antibiotic-resistant strains
Staphylococcus aureus has long been recognised as one of the major human pathogens responsible for a wide range of afflictions from minor infections of the skin to wound infections, bacteraemia, infections of the central nervous system, respiratory and urinary tracts, and infections associated with intravascular devices and foreign bodies.1,2 Most S aureus strains are opportunistic pathogens that can colonise individuals, without symptoms, for either short or extended periods of time, causing disease when the immune system becomes compromised. Indications are that this bacterial pathogen has affected human beings since prehistoric times. Before the antibiotic era S aureus diseases had high mortality rates. In 1941 the mortality rate of S aureus bacteraemia at the Boston City hospital was reported to be 82%.3 In versatility of pathogenic strategies, numbers of virulence factors, and capacity to survive and multiply in a wide range of environments, S aureus is unsurpassed by any other human pathogen. The immense genetic repertoire of this bacterium for adapting to rapidly changing and uniformly hostile environments was repeatedly shown by the emergence of S aureus strains that acquired resistance mechanisms to virtually all antimicrobial agents shortly after the introduction of these drugs into clinical practice.
The introduction of benzylpenicillin into chemotherapy in the early 1940s found S aureus fully susceptible and several of the first successes of penicillin therapy were related to the cure of formerly untreatable staphylococcal diseases.4 But by the mid 1950s the number of S aureus clinical isolates showing high-level resistance to penicillin increased rapidly, to such an extent that penicillin ceased to be a useful therapeutic agent against staphylococcal infections. The mechanism of penicillin resistance involved the acquisition of a plasmid-borne penicillinase capable of degrading the antibiotic before it had reached its cellular targets. The effects of this ?plasmid epidemic? were first seen on hospital isolates of S aureus but soon afterwards the penicillinase-based mechanism found its way into community isolates as well.5 Currently, most S aureus isolates that cause disease or simply colonise healthy individuals are resistant to penicillin. A study of the S aureus colonising flora of 1001 healthy volunteers showed that 97% of the S aureus isolates recovered carried the resistant trait to penicillin.6
The power of S aureus to mount counter-strategies to additional products of the antibiotic era besides penicillin is also well documented. Records of the Danish Health Board registered the years of introduction of various antimicrobials into clinical practice, beginning with penicillin in 1945?1946, streptomycin in 1948, tetracycline in 1950, and erythromycin in 1953.5 The same records indicate that S aureus bloodstream isolates resistant to penicillin, streptomycin, tetracycline, and erythromycin were recovered as early as 1957. The introduction of meticillin in clinical practice in 1960 was followed by the appearance of the first bloodstream isolate of S aureus that was resistant not only to penicillin, streptomycin, and tetracycline (and occasionally to erythromycin), but to meticillin as well.
Meticillin, originally called celbenine, was the first mechanism-based antimicrobial agent: it is a semisynthetic derivative of penicillin chemically modified to withstand the degradative action of penicillinase.7 The drug was introduced into therapy in Europe in 1959?1960. 1 year later, the first meticillin-resistant S aureus (MRSA) strains were detected,8 and the first clinical failure by an MRSA strain was described,9 followed by a report on the first MRSA outbreak in 1963.10 Since the 1960s, MRSA strains have spread among hospital isolates in several waves, which eventually disseminated these strains worldwide. Recent surveillance studies in hospitals in various parts of the world indicate a varying incidence of MRSA strains depending on the country and the hospital. In the USA, the National Nosocomial Infections Surveillance System (NNISS) recorded an increase of MRSA in large US hospitals, from 4% in the 1980s to 50% in the late 1990s. In some hospitals resistance frequencies as high as 80% have been recorded.11
Back to top
Antibiotic resistance in an evolutionary context
The introduction of vast quantities of structurally diverse antimicrobial agents into the human environment during the past 60 years has presented a new set of challenges to bacterial pathogens such as S aureus. Effective lineages of contemporary pathogens must excel in several capacities: they must be able to acquire resistant genes and to construct regulatory mechanisms that can adjust resistance levels to increasing concentrations of the antimicrobial agent. In the environment, the resistance-related determinants must find their way into genetic backgrounds that assure the capacity to compete with other bacteria. Pathogens must be able to spread, establish ecological reservoirs, colonise, and cause disease.
Detection and elucidation of the mechanisms of drug resistance requires the collection of S aureus isolates from the clinical environment. Strain collection in increasingly sophisticated epidemiological contexts and analysis of the strains by molecular typing techniques has led to the emergence of the new discipline, molecular epidemiology, which has greatly improved the capacity of hospitals to track the source and transmission routes of bacterial pathogens during hospital outbreaks. Beyond producing sophisticated tools for epidemiological tracking, molecular epidemiology has become an alliance of clinical and molecular microbiologists that began to tackle issues of the population structure of microbial pathogens in their native environment.12?17
Back to top
Evolutionary zero time in the emergence of MRSA
The emergence and dissemination of meticillin-resistant S aureus is a case of accelerated evolution in which these bacteria are propelled by the selective pressure of vast quantities of antimicrobial-related agents in the planetary environment.18 The evolutionary zero time of this process is the acquisition of the central genetic determinant of MRSA, the mecA gene, a 2?1 kb stretch of DNA that is not native to the species S aureus and which is embedded in a larger block?up to 60kb?of additional ?foreign? DNA called the mec element19 or staphylococcal chromosomal cassette (SCCmec),20 which together with the mecA is incorporated into the S aureus chromosome at a site-specific location.21,22 The mecA gene encodes for a 78 kDa penicillin-binding protein (PBP2A), which has very low affinity for beta-lactam antibiotics.23?25 Recent evidence indicates that PBP2A is a transpeptidase26 that, assisted by the transglycosidase domain of the native PBP2 of S aureus,27 takes over the function of cell wall biosynthesis in the presence of beta-lactam antibiotics in the medium.
Back to top
Molecular typing of MRSA
The purpose of this overview is not to detail the molecular mechanism of meticillin resistance but to survey recent information that has allowed insights into the evolutionary process that has led to the emergence of a relatively few pandemic clones of MRSA. Critical for obtaining this information were the combination of several molecular typing techniques, the elucidation of the structure of the mec elements,22 and the availability of large collections of MRSA in our laboratories. These collections, from various parts of the world, include the strain collection of the Danish surveillance system that preserved every S aureus bloodstream isolate recovered in Denmark since the mid 1950s.5
A particularly successful strategy for MRSA typing has been the combination of three molecular techniques with different discriminative characteristics:28 (i) macrorestriction pattern of chromosomal DNA after Smal digestion and separation of the fragments by pulsed-field gel electrophoresis (PFGE); (ii) polymorphisms in the vicinity of the mecA gene detected by probing Clal- digested DNAs with a mecA probe;29 and (iii) transposon Tn554 insertion patterns detected by probing C/al-digested DNAs with a specific probe.29 The individual and combined resolving ability of these methods is shown in table 1, where analysis of a group of MRSA recovered in a surveillance study in 12 hospitals in New York showed that the highest resolution was achieved by the PFGE patterns.30 PFGE has been recognised as a highly discriminatory technique and suggested to be the gold standard for outbreak investigations.31 Two sequencing-based techniques, multilocus-sequencing type (MLST) and spaA typing, have been developed recently, with great potential for global epidemiology studies.13,15,32 MLST is based on the DNA sequencing of the internal fragments of seven unlinked housekeeping genes (arcC, aroE, glpF, gmk, pta, tpi, and yqiL) and spaA typing is based on the DNA sequencing of the polymorphic region of protein A, a superantigen of S aureus.
Click to view table
Table 1. Resolution of three molecular typing methods (adapted from reference 30)
Back to top
The nature of early MRSA isolates
The preservation of all S aureus bacteraemic isolates in Denmark since the mid 1950s has allowed us to look at the epidemiological context of the appearance of MRSA in that country. Among 3704 S aureus bacteraemic isolates recovered between 1957 and 1970, 646 MRSA isolates were identified and all but nine of these were resistant not only to meticillin but to penicillin, streptomycin, tetracycline, and often to erythromycin. Furthermore, 616 of these 646 isolates (95%) belonged to unique phage groups called phage group III or the related 83A complex.5 Data from the Danish surveillance system indicate that almost half of all S aureus bacteraemic isolates?both MRSA and meticillin-susceptible S aureus (MSSA)?recovered in Denmark during this period belonged to these phage groups. The very first European MRSA isolates identified in the UK also belonged to the same phage group and were resistant to penicillin, streptomycin, tetracycline (PST), and occasionally to erythromycin (PSTE) as well.8,10 These observations strongly suggested that we might be able to identify among these Danish MSSA isolates one of the first S aureus lineages that received the mecA gene. Additional evidence pointing in this direction was the sequential appearance of MSSA strains belonging to phage group III and/or 83A with the PST antibiotype, and the appearance of MRSA strains belonging to the same phage group and sharing the same PST or PSTE multiresistance pattern in addition to being meticillin resistant?PSTM and PSTEM (figure 1). Molecular typing with MLST, PFGE, and spaA typing fully confirmed this point.33 The overwhelming majority of historically early MSSA and MRSA isolates from Denmark and the earliest MRSA isolates from the UK shared identical or very similar molecular profiles (table 2). Characterisation of these earliest, archaic, strains of MRSA showed remarkably uniform properties in individual isolates:34 low meticillin minimum inhibitory concentration (MIC) value (6?25 μg/ml) for most cells; heterogeneous expression of resistance; a single, common Clal-mecA polymorph II; lack of the regulatory gene medI; lack of Tn554 in most isolates; and a closely related PFGE profile. The epidemic capacity of these archaic MRSA strains was shown by their recovery from as many as 18 hospitals in Denmark.
Click to enlarge image
Figure 1. Sequential appearance of meticillin-susceptible and meticillin-resistant blood isolates of S aureus belonging to phage group III and the related 83A complex in Denmark. The numbers plotted represent all S aureus blood isolates identified in Denmark during the particular period (adapted from reference 33).
Click to view table
Table 2. Properties of historically early strains of MSSA recovered in Denmark (adapted from reference 33)
These results identify the progeny that must have been one of the first S aureus receiving the heterologous mec element from an unknown bacterial donor?ie, the first stage in the evolutionary process of emergence of MRSA. The genetic background of early MSSA and MRSA isolates was very similar or identical to the genetic background of one of the most widely spread contemporary multidrug-resistant clones of MRSA, the Iberian clone.33,34 Although the contemporary Iberian clone shares the same genetic background as the ?archaic? MRSA isolates, this clone seems to be a further step ahead in MRSA evolution because it has acquired extra resistance determinants, some of them resident in mobile elements such as plasmids (pUB110) and transposons (Tn554). Thus, by contrast with the limited antibiotic resistance pattern of the archaic MRSA (PSTEM), most isolates of the Iberian clone are resistant to most commonly used antimicrobial agents, with the exception of the co-trimoxazole (sulfamethoxazole-trimethoprim) and the glycopeptides.35 The integration of the linearised plasmid pUB110 in the downstream vicinity of the mecA gene explains the differences in the mecA polymorphism detected between the archaic strains (mecA polymorph II) and the contemporary Iberian MRSA strains (mecA polymorph I).36
Back to top
Identification of five pandemic MRSA clones
In recent years, the combination of mecA polymorphisms, Tn554 insertion patterns, and PFGE profiles was applied to characterise more than 3000 MRSA isolates collected in surveillance studies and outbreak investigations in southern and eastern Europe, Latin America, and the USA between 1994 and 2000, through the international CEMNET initiative.17 Five major clonal lineages were defined, which are widely disseminated in those areas and accounted for a large proportion?68%?of the isolates, indicating that they represented successful lineages in terms of ability to cause infection, to persist, and to spread from one geographic site to another, including across continents (table 3). The names assigned to these five MRSA?Iberian, Brazilian, Hungarian, New York/Japan, and paediatric pandemic MRSA clones?reflect the geographic area in which they were first identified and/or indicate some unique epidemiological property. In addition to their distinct PFGE patterns, the prototypes of these five clonal types had distinct properties. The Iberian clone carried a ClaI::mecA polymorph I and ClaI::Tn554 pattern E (clonal type I::E). The corresponding clonal types in the other MRSA clones were as follows: Brazilian clone XI::B; Hungarian clone III::B; New York/Japan clone I::A; and paediatric clone II::NH. All strains were tested for susceptibility to nine antibiotics (oxacillin, penicillin, clindamycin, erythromycin, gentamicin, spectinomycin, co-trimoxazole, tetracycline, and vancomycin). The Iberian, Hungarian, and New York/Japan clones were only susceptible to co-trimoxazole and the Brazilian clone was only susceptible to spectinomycin. The paediatric clone was only resistant to oxacillin, penicillin, gentamycin, and occasionally erythromycin. All isolates were susceptible to vancomycin.
Click to view table
Table 3. Pandemic MRSA clones prevalence (adapted from reference 37)
Figure 2 documents the spread of these five pandemic MRSA clones. The Iberian clone was first reported in Spain in 1989 and since then has been reported in Portugal, Italy, the UK, Germany, Belgium, Switzerland, France, Czech Republic, Poland, and the USA. The Brazilian clone was first described in 1992 in Brazil and then in Portugal, Argentina, Uruguay, Chile, and the Czech Republic. The Hungarian clone was extensively disseminated among hospitals in that country and was most recently also identified in Taiwan. The New York/Japan clone was identified as the dominant MRSA in hospitals in metropolitan New York, New Jersey, Pennsylvania, and Connecticut, and also in one hospital in Tokyo, Japan. The paediatric clone was first reported in a paediatric hospital in Portugal in 1992 and was also reported in Poland, USA, Argentina, and Colombia (for a list of references concerning the spread of the MSRA pandemic clones see table 1 of reference 37).
Click to enlarge image
Figure 2. International spread of the pandemic MSRA clones.
Molecular typing techniques have not only allowed the identification of pandemic clones of MRSA, but have also enabled us to monitor MRSA clones circulating in different hospitals and at different time intervals in a country providing evidence for the stability of epidemic clones over time. Figure 3 shows the MRSA clonal distribution in ten hospitals in six Portuguese cities during 9 years. Figure 4 shows that in the largest Portuguese teaching hospital located in Coimbra, during three distinct surveillance periods, the Iberian and Brazilian clones were present in different proportions but together accounted for virtually all the MRSA isolates. In this case, molecular typing techniques were able to document the introduction of the Brazilian clone during the first surveillance period. During the following surveillance periods, the prevalence of this clone increased from 5% to 38%, whereas the previous endemic Iberian MRSA clone decreased from 89% to 55%.38
Click to enlarge image
Figure 3. MRSA clonal distribution in hospitals in six cities in Portugal.
Click to enlarge image
Figure 4. Variation of the incidence of two pandemic MRSA clones during three consecutive surveillance periods in the largest Portuguese hospital, located in Coimbra (adapted from reference 38).
Back to top
Identification of two distinct ancestral lineages among the five pandemic MRSA clones
The primary criterion used for the initial classification of MRSA isolates into the five pandemic clonal types was their PFGE pattern. PFGE is a high-resolution technique capable of registering DNA rearrangements and acquisition or loss of genetic determinants?ie, genetic events that can happen on a short time scale. While providing a high degree of resolution, this technique could blur long-range evolutionary associations in MRSA strains. By contrast, MLST variations accumulate slower than variations in PFGE, since they arise from neutral mutations in housekeeping genes. Thus, the same MLST allelic profile could include temporally distant strains that have accumulated differences in PFGE patterns.13
Using MLST we identified two major genetic backgrounds, A and B, among representative isolates of the five pandemic MRSA clones.37 Background A has four closely related variants (single-locus variants). Backgrounds A1 and A2 were both identified among members of the ?archaic? and Iberian clones, background A3 was detected in a minor USA clone (clone V), and background A4 characterises both the Brazilian and Hungarian clones (table 4). The variant alleles in backgrounds A2 and A3 each differ from Al in a single nucleotide at one of the alleles, suggesting evolution through single point mutations.14 Backgrounds A1, A2, and A3 also share the same spaA type motif. Background A4 is a single locus variant of background A3, but it differs in three nucleotides in the arcC allele, suggesting evolution by recombinational events.14 Background A4 is also characterised by a distinct spaA type. Together, these findings suggest that the Iberian, Brazilian and Hungarian clones, and clone V, sharing the closely related genetic backgrounds A1?A4, have evolved from a common ancestor with a genotype very similar to that of the archaic MRSA. A second and completely different chromosomal background B was identified in MRSA strains originally recognised in the New York/Japan and paediatric MRSA clones.
Click to view table
Table 4. Genetic backgrounds and SCCmec types of the pandemic MRSA clones.
Back to top
The mec element as an evolutionary marker
The initial genetic event marking the emergence of an MRSA lineage is the acquisition of the mec element, introducing the meticillin-resistance determinant mecA into a MSSA lineage. Four types of mec elements or SCCmec have been identified so far. Type I (34 kb) was detected in the first MRSA strain isolated in 1961 in the UK (strain NCTC10442); type II (52 kb) was identified in an MRSA strain isolated in 1982 in Japan (strain N315); type III (66 kb) was identified in an MRSA strain isolated in 1985 in New Zealand (strain 82/2082);22 and type IV (20?24 kb) was identified in two community-acquired MRSA strains,39 as well as in representatives of the paediatric clone from Poland and from Portugal.37
The mec element structures resident in representative isolates of the five pandemic MRSA clones were recently characterised, and specific patterns of association between clonal types and SCCmec types were recorded38 (table 4). MRSA strains with the genetic background A1 and A2 (archaic and Iberian clone) carried SCCmec types I and IA; strains with genetic background A4 (Brazilian and Hungarian clone) carried SCCmec types III and IIIA. SCCmec type IA (Iberian clone) differs from type I by the presence of the linearised plasmid pUB110 in the mecA downstream vicinity, whereas SCCmec type IIIA (Brazilian clone) varies from type III by the absence of the linearised plasmid pT181 in the mecA downstream vicinity. Strains representing the New York/Japan clone (genetic background B) carried SCCmec type II. The SCCmec type IV was found in strains representing the paediatric clone (genetic background B) and also in the minor clone V (background A3) (figure 5).
Click to enlarge image
Figure 5. SCCmec element structures and variants. The basic structures of SCCmec types I, II III and IV are depicted, as well as its variants SCCmec IA and IIIA.21,22,38,39 LJ=chromosomal left junction, RJ= chromosomal right junction.
A search of the MLST database identified two additional important genetic backgrounds characteristic of E-MRSA 15 (allelic profile 7-6-1-5-8-8-6, background D) and E-MRSA 16 (allelic profile 2-2-2-2-3-3-2, background C), two MRSA clones that currently represent the most frequent MRSA in UK hospitals.40,41 E-MRSA 15 was also detected in Germany.42 An E-MRSA 16 representative strain is being sequenced. Examination of its preliminary genome sequence data indicates that it harbours a SCCmec type II, the same mec element shown to be resident in a completely different genetic background: background B characteristic of the New York/Japan clone. The presence of SCCmec type II in both of these distinct genetic backgrounds (B and C) could represent either an independent acquisition event or horizontal transfer of the mec element from one S aureus strain to another.
MLST studies of MSSA strains uncovered a great diversity of allelic profiles,12,13 which contrasts with the clonal structure of the highly epidemic MRSA. The worldwide spread of MRSA seems to be primarily driven by the dissemination of a few pandemic clones in which the mec element found genetic backgrounds that provided appropriate fitness for virulence and epidemic spread of the bacteria. One of these genetic backgrounds?A1?has recently been identified in MSSA strains that were highly prevalent in the late 1950s in Denmark.33 MSSA with genetic backgrounds A3 (clone V), B (Brazilian and Hungarian clones), and C (E-MRSA 16) were also detected in a few Danish isolates from this era (table 2).33 In a recent communication by Enright,43 the same backgrounds were detected among an independent collection of 394 epidemic MRSA isolates from 22 different countries characterised by MLST. The most frequent clonal complex I described by this author appears to correspond to genetic background A and its variants described in this communication. While MRSA with either ST22 or ST36 were not detected among pandemic isolates in our collection, a few MSSA isolates with ST36?corresponding to our genetic background C?were detected among historically early MSSA from Denmark.33 What Enright43 describes as a fourth type of MRSA detected among isolates from the USA and Europe presumably corresponds to our genetic background B (MLST 1-4-1-4-12-1-10) characteristic of both the New York/Japan and the paediatric clones of MRSA, which share a common MLST but may be distinguished by the distinct SCCmec type they carry (see table 4).
Back to top
MRSA lineages and the acquisition of mec elements
Putting together all data one might propose important inferences as to the evolutionary pathways through which MRSA lineages could have emerged. MRSA strains have a strong clonal structure, by contrast with MSSA strains, probably due to their shorter existence (four decades). The mec element, containing the meticillin-resistance determinant, seems to have been imported into the species of S aureus at multiple, yet restricted and independent, occasions, as shown by the specific pattern of association between clonal types and SCCmec types. Horizontal transfer of the mec element between different S aureus lineages could also have happened as suggested by the presence of SCCmec type IV in backgrounds A, B, and D, and SCCmec type II in backgrounds B and C. However, horizontal transfer seems to be a rare event and the spread of meticillin-resistance in S aureus seems to be predominantly due to the clonal expansion of very few lineages. A tentative scheme outlining steps in the evolution of MRSA clonal families and the acquisition of SCCmec elements is presented in figure 6. The proposed evolutionary associations are primarily based on the properties of MRSA strains analysed and, in our case, these came mainly from sources in southern and eastern Europe, Latin America, and the USA. Important findings were also made possible by the availability of the Danish MRSA collection and by the information available in the MLST database. Analysis of isolates from other geographic areas would be of great interest.
Click to enlarge image
Figure 6. Proposed evolutionary pathways of the contemporary pandemic MRSA clones. Letter colors represent different genetic backgrounds (MLST), box colours represent different SCCmec types and dash lines represent possible horizontal transfers of the mec element.
Back to top
The origin of the mec element
The origin of the mec element is not known. The assembly of the several mec element structures may have involved multiple hosts, possibly among coagulase-negative staphylococci.44 A close homologue of the S aureus mecA gene was recently identified as a gene native to the animal commensal species Staphylococcus sciuri, most strains of which are fully susceptible to meticillin and all beta-lactam antibiotics.45,46 The insertion sequence IS1272 present in SCCmec types I and IV appears in several copies in the Staphylococcus haemotyticus genome.47
The distribution of SCCmec type IV provides some interesting clues concerning possible evolutionary associations. In MRSA this mec element structure is seen in three different genetic backgrounds, and SSCmec type IV was also seen in community-acquired MRSA6,39 and also in coagulase-negative staphylococci (D C Oliveira, unpublished data).39 These findings suggest an enhanced mobility of this mec element, perhaps explained by its smaller size. SCCmec type IV has the same downstream structure as the SCCmec type I. A high degree of similarity is also apparent upstream of mecA where both SCCmec I and IV have a deleted mecR1, an IS1272, and ccrAB flanked by two homologous sequences. However, the region between the chromosomal left junction and the upstream region of ccrAB?ie, the gene pls and its flanking regions, close to 18 kb in size?is missing in SCCmec IV and was replaced by a variable DNA segment.37 The close association of SCCmec type IV to SCCmec type I, which is present in the earliest MRSA isolates, together with its presence in hospital-acquired and community-acquired MRSA strains as well as in coagulase-negative staphylococci, suggests that this mec element structure might be closely related to a putative precursor of the mec element.
Back to top
New trends in MRSA epidemiology
The specific pattern of association of SCCmec types and MRSA clones opens the door to some speculations about possible contributions of the SCCmec itself to the pathogenic potential of MRSA strains. It has been proposed that the mec element might be a pathogenic and/or antibiotic-resistance island,22,48 capable of autonomous excision provided by the chromosomal cassette recombinase genes ccrA and ccrB present in every mec element.22 The mec element, besides being the carrier of mecA and its regulatory genes (mecI and mecR1,) also contains additional resistance determinants in the form of integrated mobile elements (pUB110, pT181, pI258, Tn554, and ψTn554). In this respect the SCCmec structure clearly represents an antibiotic-resistance island. Other regions of the SCCmec might be more directly contributing to the pathogenic potential of the S aureus cell. Such regions could be the pls gene present in SCCmec type I and the kdp operon present in SCCmec type II. The pls gene, which codes for a large surface protein, could suppress adhesion of the bacteria to their host at some stages during infection.49 The kdp operon is an extra copy of the potassium-transporter operon present elsewhere in the genome,48 which could improve the colonisation and invasion processes of the host tissues. A possible association should be considered between the SCCmec structure and the epidemiology of some MRSA clones. For instance, SCCmec type IV, which differs mainly from SCCmec type I by the absence of the pls gene and its flanking regions, characterises both the paediatric MRSA clone and the USA clone V, which have completely different genetic backgrounds. However, both clones have low-level and heterogeneous meticillin resistance and are frequently isolated from patients with immature or compromised immune systems (children and AIDS patients).30,50?52 In a recent study, 22 of 28 patients with AIDS had MRSA infections that were seen to belong to the USA Clone V.30 It is tempting to consider the possibility that the pls deletion contributes to the affinity of these strains for patients with weak immune systems.
Several recent findings suggest that MRSA might be emerging as a community pathogen53?57 following perhaps the historic precedent of penicillin-resistant S aureus.58 A screening of 1001 S aureus isolates recovered from colonisation sites of young and healthy people in a community showed that more than 95% of the isolates were resistant to penicillin but only seven were MRSA.6 However, five of these seven MRSA isolates were representatives of the pandemic Iberian, Brazilian, and paediatric clones, giving another reminder of the evolutionary process of these staphylococcal lineages.
Back to top
Search strategies and selection criteria
Four types of references are quoted to provide context and background information to the observations and hypotheses described in this communication: historical references to MRSA, MRSA in disease, and the changing epidemiology of MRSA; references to novel molecular typing techniques such as MLST and spaA typing; references to the origin, structure and putative functions of mec elements (SCCmec types); and references related to the origins, epidemiological features and initial molecular characterisation of the large strain collection which was the source of epidemic MRSA clones described in our paper.
Back to top
Acknowledgments
Some findings were made possible by the information available at the MLST database (www.mlst.net) and the E-MSRA 16 genomesequencing project (www.sanger.ac.uk). Partial support for this study was provided by project POCTI/1999/ESP/34872 from Funda??o para a Ci?ncia e Tecnologia, Lisbon, Portugal, awarded to HL, and by a grant from the US Public Health Service awarded to AT, project RO1 AI37275. DCO was supported by a grant from Funda??o para a Ci?ncia e Tecnologia, Lisbon, Portugal, and also from July 2001 by a doctoral grant from Funda??o Calouste Gulbenkian, Lisbon, Portugal.
Back to top
Comment