Supplementary Information: Correlation of Microarray Typing with MLST

Supplementary Information: Correlation of microarray typing with MLST

Following some suggestions from the reviewers that some more explanation is required as to the correlation of microarray typing with multi locus sequence typing (MLST), we decided to add this supplementary note in explanation of some of the classifications derived from microarray vis-à-vis the assignation of strains to clonal clusters via MLST. The information was kindly provided by Dr. Stefan Monecke who in this study is the author with most experience in the field.

MLST genes reflect phylogeny as single point mutations accumulate with time. The concept of clonal complexes has been introduced to reflect that very different sequence type designations might hide that the sequence types might still be related. For instance, ST22 and ST1117 differ in a single nucleotide (T vs. G at position 340 of the pta allele). Thus, both belong to Clonal Complex (CC) 22. Because other genes contain the same phylogenetic information, the phylogeny can also be derived from alleles of adhesion markers, agr genes etc. Several studies [1-3] have shown that the similarity of hybridisation profiles translates into similarity of MLST profiles as both reflect the core genome of a given strain. Practically, when an isolate is tested with the array, its image will be compared to a database based on about 15,000 array experiments and about 800 MLS datasets. If it yields a hit, it will be regarded as identified; if its profile is unknown, MLST is performed in order to assign a designation to the image. For example, approximately 250 CC80 isolates were typed using the microarray technique, while 20 of these underwent MLST. Thus, one can be sure that, for example, a strain displaying anabsenceof agrI, agrII and agrIV alleles, capsule 5 genes of the enterotoxin H gene, the egc cluster, ORF CM14 and the collagen adhesion gene can, as well as the presence of agrIII, capsule 8, sasG, etD and edinB belongs to CC80.

Clones or strains within a given clonal complex would be defined mainly on the absence or presence of mecA, on the type of SCC elements (if present) and of the absence or presence of PVL genes. Thus, a CC80 isolate with mecA as a part of an SCCmec IV element and lukF/S-PV would be assigned to “CC80-MRSA-IV [PVL+], European caMRSA Clone”. However, we would not differentiate between ST80, ST583 and ST728, since they differ only in single nucleotide exchanges that cant be demonstrated by array hybridization; hence, the designation as “CC80” . Some STs can be discerned, if they originate not from single nucleotide exchanges but from genomic replacements introducing foreign core genomic markers. Examples are ST239 (CC8) or ST34 (CC30).

CC6 or ST6 is unique in its characteristics. Monecke and colleagues [3] previously described that: “According to MLST data, ST6 is a double locus variant (DLV) of ST5 (arcC-12 and yqiL-3 rather than arcC-1 and yqiL-10, as in ST5). However, ST6 isolates differ in agr group, capsule and spa types. They harbour different alleles of hsdS-, set/ssl- and several MSCRAMM genes (bbp, fnbB, sdrC, sdrD, clfA, clfB, sasG). In addition, isolates of this ST carry cna but lack egc. These observations may be explained by a large scale chromosomal replacement with one parental strain belonging to CC5. The origin of the inserted region has not yet been determined. The size and location of the insert can be estimated by analysing the known positions of the probe sequences within the published CC5 genome sequences. Thus, the insert is localised around oriC, and ranges from hsdS3 downstream of oriC to the ssl/set-locus upstream of oriC. This equals ca. 1.500.000 bp (i.e., about half of the genome).” Identification of this strain type was based on a dataset of about 120 isolates out of which 5 were MLST´ed. The actual isolate found in this study was not classified by MLST but it was identical to the WA51 type strain 07-15545 from Perth, WA with the sole exception of an additional carriage of ermC.

The CC50 isolate shared basic characteristics with other CC50 isolates which had been previously characterized (e.g., absence of agrI-III alleles. Capsule 5, she and ORF CM14 and presence of agrIV, capsule 8, egc and cna). However, since this CC50-MRSA isolate was encountered for the first time, MLST was performed. It proved to be a double locus variant of ST50 (16-16-12-2-39-13-2).

The use of "clonal clusters" related to MLST designations might be problematic as they are defined by the changing content of the MLST database. For example, CC1 was once a defined clonal cluster, but it has been subsumed into CC15 as more isolates have been added to the MLST database over the years.When going back in time, there was certainly one common ancestor for all S. aureus lineages, and possibly missing links still can be found. To force constantly evolving and interbreeding subjects into a rigid framework of species/subspecies/clonal complexes is a constant problem in taxonomy, and always this is always arbitrary to a certain degree. Here we opted to retain the designation CC1 for practical purposes. CC1 is a distinct group and very different from CC15 sensu strictu (for instance, the presence of agrIII, rather than agrII, or the presence of seh in CC1).

[1]  Lindsay JA, Moore CE, Day NP, Peacock SJ, Witney AA, Stabler RA, et al. (2006) Microarrays reveal that each of the ten dominant lineages ofStaphylococcus aureushas a unique combination of surface-associated and regulatory genes

[2]  Monecke S, Slickers P, Ehricht R (2008) Assignment of Staphylococcus aureus isolates to clonal complexes based on microarray analysis and pattern recognition. FEMS Immunol Med Microbiol 53(2):237-251. DOI:10.1111/j.1574-695X.2008.00426.x

[3]  Monecke S,Coombs G,Shore AC,Coleman DC,Akpaka P,Borg M, et al.(2011)A field guide to pandemic, epidemic and sporadic clones of methicillin-resistant Staphylococcus aureus.PLoS ONE6(4):e17936.doi:10.1371/journal.pone.0017936