This historic collection provides unique insights into of the evolution of the first MRSA lineage. Preserved for decades in their original freeze-dried state, this large collection of strains representing the earliest MRSA clone has allowed us to reconstruct the evolutionary events leading to the emergence of MRSA. Using whole genome sequencing we have been able to shed light on the time when SCCmec first entered into S. aureus, and also to estimate how many times this is likely to have happened in the archaic MRSA population.

The origins of SCCmec almost certainly lie in coagulase-negative staphylococci (CoNS) [16]. S. aureus belonging to the ST250 background appear to have been the first recipient in the transfer from CoNS, but whether the element entered the ST250 population on multiple occasions, or as a single isolated event with subsequent propagation through the population, has never been definitively resolved. A single entry of mecA followed by its evolution within the recipient background has been suggested [17]. In order to clarify this we examined the variation present within the SCCmec elements in isolates throughout the population. The variation seen within SCCmec is predominantly in the pls gene, which has been described before [18]. Functionality of this 230 kDa cell wall-anchored (CWA) protein remains unclear, but its expression has been shown to reduce adhesion to host proteins as well as decrease invasiveness [19]. This LPxTG surface protein has a highly repetitive D/S-rich structure, making it a target for homologous recombination. As noted in other lineages the CWA proteins are subject to diversifying selection and exhibit diversity between and within S. aureus lineages [20, 21]. Removal of this variation reveals that the evolutionary history of the SCCmec elements was congruent with that of the strains carrying them, which points towards a single acquisition, rather than multiple or recurrent horizontal transmissions. Supporting this hypothesis is the observation of a mutation in ccrB1 gene of the SCCmec type I element. The recombinase genes are required for both integration and excision from the chromosome. Specifically, CcrB is required for excision and the mutation present within this NCTCT10442 type I SCCmec element is believed to produce a non-functioning recombinase [22, 23]. Given that all the isolates in this collection have this frameshift mutation, this strongly supports the conclusions of the phylogenetic analysis, namely that a type I SCCmec was acquired once in the ST250 background, and then became fixed in the population due to defective recombinase apparatus that precluded excision.

One of the questions we sought to address in this study was what were the temporal events surrounding the emergence of MRSA. With the first reports of MRSA occurring only after introduction of methicillin in the UK in 1959 and Denmark in 1964 it seemed reasonable to conclude that resistance arose after the first clinical use of the drug, and resistance therefore developed in S. aureus as an adaptive response following exposure to the antibiotic. However, the results presented in this communication are not consistent with this conclusion, since the gene bestowing methicillin resistance was likely to have been acquired in the mid-1940s. It was during this period that β-lactamase-mediated penicillin resistance was becoming widespread among clinical isolates of S. aureus. Within 4 years of the introduction of penicillin for the treatment of staphylococcal infections, the first penicillin-resistant S. aureus were being described in 1944 [1]. In the years that followed the frequency of resistance in clinical isolates climbed steadily, such that by the time methicillin was introduced into clinical practice in 1960, resistance rates of 80% were common [24, 25].

Whilst the main genetic determinant associated with penicillin resistance in S. aureus is blaZ, mecA also encodes penicillin resistance via a different mechanism involving an alternative penicillin-binding protein, PBP2a [6, 26]. In the sequenced collection blaZ is widely distributed, albeit at a lower frequency than mecA (85.2% of isolates carry the blaZ gene in comparison to 95.2% for mecA), suggesting a selective advantage to possessing two distinct β-lactam resistance mechanisms. Based on the temporal calibration of the acquisition of mecA, it appears likely that methicillin resistance in S. aureus evolved long before this new β-lactam antibiotic was introduced. Thus, it was the widespread use of penicillin, rather than methicillin, that was the driver for the emergence in the archaic MRSA clone.

Beyond β-lactams our analysis uncovered evidence for the strong selective impact that a number of different antibiotics have had on the evolution of the archaic MRSA clone. Several of the antibiotics, such as tetracycline, are prescribed in far lower amounts today in human medicine than in the 1950s and 1960s, and resistance to these antibiotics in contemporary S. aureus from humans is relatively rare, which contrasts with the archaic MRSA population, in which the distribution of tetracycline resistance determinants was widespread (Fig. 2; 96% of isolates contained tetK or/and tetM) [27]. In a prescient study examining the antibiotic consumption and rates of resistance in a hospital in the US in the 1950s, Bauer et al. provided evidence of a correlation between the two, where increasing usage of tetracycline was associated with an increase in the frequency of tetracycline resistance in isolates from inpatients [25].

In addition to methicillin and tetracycline resistances, a key phenotypic marker of the archaic MRSA clone was non-susceptibility to streptomycin. In our analysis we identified a mutation predicted to confer streptomycin resistance occurring on the same branch of the tree in which we mapped the acquisition of the SCCmec element. This finding suggests that methicillin and streptomycin resistance both emerged in the archetypal MRSA progenitor population around the same time. Discovered in the early 1940s, streptomycin was demonstrated to have activity against Gram-positive pathogens, and was used in the UK in 1947 during the first ever randomized clinical trials studying the efficacy of streptomycin in the treatment of pulmonary tuberculosis [28, 29]. It therefore appears that the first MRSA clone emerged, and developed resistance to two of the earliest antibiotics—streptomycin and penicillin—almost immediately after the S. aureus population would have been first exposed to them.

At the time of its discovery, the incidence of MRSA in the general population is likely to have been very low. This is demonstrated by the fact that screening of over 5000 samples at Public Health England yielded only three methicillin-resistant isolates. Therefore, it is likely that when methicillin was introduced to circumvent penicillin resistance in S. aureus, it did not select for emergence of MRSA at that time, but instead provided the selective pressure, which drove the nosocomial spread of a pre-existing variant, at a time when infection control measures in UK hospitals were limited.