1. Findings
1.1. Among the 92 S. aureus strains,
1.1.1. 14 rep families
1.1.1.1. Up to 5 rep families were found insomestrains
1.1.2. 2 USs were detected
1.1.3. 9 strains (8.7%) could not be classified in any of the rep families or USs established
1.2. most predominant rep families were
1.2.1. rep7 (n 46)
1.2.1.1. The rep gene of pGO1 (rep21) with ant(4=)-Ia was detected in one strain. pVGA harbored vga(A), and this resistance gene was identified in all our rep7b-positive
1.2.2. rep10 (n 35)
1.2.2.1. Ten of our eleven rep10-positive strains also carried erm(C), as is the case for pDLK1
1.2.2.1.1. A specific combination of resistance genes and plasmid families was already described for some strains in the previous study
1.2.2.2. Two strains contained the gene lnu(A) and a rep gene similar to that of plasmid pBMSa1
1.2.3. rep21 (n 32)
1.2.4. rep22 (n 28)
1.2.4.1. Three strains belonging to the rep22 family contained the genes ant(4=)-Ia, tet(L), and dfrK, similar to pKKS825
1.3. A high number of plasmids (from three to eight) were detected among the 18 strains examined by S1-PFGE.
1.3.1. These plasmids differed in sizes from 500 to 2.5 kb
1.3.2. In the rep families rep5, rep7, rep20, and rep22, plasmid sizes were different from those of the plasmids in which the rep genes were originally detected
1.3.3. In some cases, two rep genes hybridized with the same plasmid size. There are two plausible explanations for this:
1.3.3.1. the strain contained two plasmids of the same size, or
1.3.3.2. both rep genes were located in the same plasmid.
1.3.3.3. Mosaic plasmids have previously been detected
1.3.3.3.1. Freitas AR, et al. 2011. Human and swine hosts share vancomycinresistant Enterococcus faecium CC17 and CC5 and Enterococcus faecalis CC2 clonal clusters harboring Tn1546 on indistinguishable plasmids. J. Clin. Microbiol. 49:925–931.
1.3.4. All rep10-positive strains harbored a rep gene identical to that of pDLK1, and a similar size was revealed by hybridization
1.3.4.1. Similar results were obtained for rep21, which in all cases hybridized to small plasmids of 2.5 kb with rep sequences similar to those of pKH21, pBMSa1, and pS0385-3
1.3.5. It is interesting to remark that small plasmids seem to remain more similar in molecular weight than larger plasmids, in which changes in size, probably due to genomic plasticity, appear to occur more frequently
1.4. A relationship between some clonal complexes (CC) and some rep genes was detected.
1.4.1. Rep genes of pS0385-2 and pS0385-3, plasmids previously found in CC398, were identified in our study in strains belonged to this CC
1.4.1.1. Moreover, pTW20_1 (rep20) was previously isolated from a MRSA clone belonging to a highly transmissible and multiresistant CC8 isolate (10), and the rep gene of this plasmid was detected in MRSA CC8
1.4.2. Similarities in some plasmid patterns of Enterococcus strains belonging to the same CC have been discovered before
1.4.2.1. Moreover, in recent studies of plasmids in staphylococci, the pUSA300-like plasmids were found only in the USA300 clone strains
1.4.2.1.1. Nonetheless, other rep genes seem to be more widespread.
1.5. MRSA CC398 usually displays a multiresistance phenotype.
1.5.1. In this study, most strains belonging to this CC showed an elevated number of rep families, indicating that this CC has a high ability to capture different plasmids.
1.5.1.1. Moreover, MRSA appears to be more likely to harbor plasmids than MSSA (2), so these mobile genetic elements could have played an indirect role in the evolution from MSSA to MRSA or the subsequent emergence of MRSA
2. Our Classification system
2.1. Our classification system showed some similarities with those presented by others.
2.1.1. For example, some rep families (rep7 and rep10) contained plasmids with sizes in accordance with the subgroups pT181 and pSN2, or the plasmids pSAS and pMW2 were grouped in the rep5 family
2.1.2. However, as occurs in Novick classification, plasmids of the same family can belong to different incompatibility groups (19), and it is not possible based on a genotype, as in our study, to determine the exact incompatibility group.
2.1.3. No chromosomal rep gene location was detected in our study.
2.1.4. Interestingly, rep families (rep2, rep6, and rep9) previously found in other Gram-positive species were detected in S. aureus strains for the first time in our study, although at a very low percentage.
2.1.4.1. This indicates plasmid transfer between genera and/or species
3. classification systems
3.1. Novick et al. divided plasmids into three classes:
3.1.1. (i) small rolling-circle,
3.1.2. (ii) multiresistance
3.1.3. (iii) conjugative
3.1.4. a fourth class (pSK639 family) has been proposed
3.1.4.1. Fischetti VA, Novick RP, Ferretti JJ, Portnoy DA, Rood JI (ed). 2006. Gram-positive pathogens, p 413–415. American Society for Microbiology, Washington, DC
3.2. A system, based on incompatibility groups, identified 15 groups within the Staphylococcus genus
3.2.1. Udo EE, Grubb WB. 1991. A new incompatibility group plasmid in Staphylococcus aureus. FEMS Microbiol. Lett. 78:33–36
3.3. a classification system that uses restriction fragment length polymorphism (RFLP) patterns was suggested for large plasmids
3.3.1. Shearer JE, et al. 2011. Major families of multiresistant plasmids from geographically and epidemiologically diverse staphylococci. G3 (Bethesda) 1:581–591
3.4. plasmid classification based on PCR amplification of conserved areas of the replication initiating genes (rep) from 103 plasmids belonging to different Gram-positive bacteria was described
3.4.1. Jensen LB, et al. 2010. A classification system for plasmids from enterococci and other Gram-positive bacteria. J. Microbiol. Methods 80:25–43
4. Aim
4.1. expand the previous classification system
4.2. to implement it with a collection of S. aureus strains of different origins.
5. Method
5.1. All S. aureus plasmid sequences published in the GenBank database until November 2010 were retrieved, and only one plasmid per rep gene was included
5.1.1. They comprised 112 sequences from 108 S. aureus plasmids (81 not defined in the previous system).
5.2. The BioNumerics software program (Applied Maths, Belgium) was used to generate nucleotide- and amino acid-based phylogenetic trees by neighbor-joining analysis
5.2.1. a cutoff value of 75% identity was chosen
5.2.2. a total of 15 rep families were defined.
5.2.2.1. 6 newly defined
5.2.2.1.1. rep7b, rep20, rep21, rep22, rep23, and rep24
5.2.2.2. 9 from the previous study
5.2.2.3. 10 unique sequences (US)
5.3. system was tested with 92strains from Spain and Denmark
5.3.1. different origins (human, animal, and food)
5.3.2. 70 methicillin-resistant S. aureus (MRSA) strains
5.3.3. 22 methicillin-susceptible (MSSA) strains
5.3.4. Based on their antimicrobial susceptibility profiles, 18 resistance genes were tested for the strains of the Spanish collection
5.4. Multiplex PCRs from the previous study, in addition to the new simplex PCRs, were performed for the 92 strains
5.4.1. obtained amplicons from 18 selected strains were sequenced (corresponding to all rep families detected for the Spanish collection)
5.5. 18 representative strains were submitted to S1 pulsed-field gel electrophoresis (S1-PFGE) nuclease digestion
5.5.1. to assess the number and size of plasmids from each of them
5.5.2. S1-PFGE and gel DNA plasmid extraction (for small plasmids) hybridizations were performed to determine the sizes of the plasmids in which the different rep genes were located