Evaluation of spa and DiversiLab rep-PCR typing in characterising Western Australian community MRSA clones as defined by MLST/SCCmec

  • Geoffrey Coombs, PathWest Laboratory Medicine - WA, Royal Perth Hospital, Australia
  • Julie Pearson, PathWest Laboratory Medicine - WA, Royal Perth Hospital, Australia
  • Lynne Wilson, PathWest Laboratory Medicine - WA, Royal Perth Hospital, Australia
  • Hui-Lee Tan, PathWest Laboratory Medicine - WA, Royal Perth Hospital, Australia
  • Samanth Cramer, PathWest Laboratory Medicine - WA, Royal Perth Hospital, Australia
  • Frances O'Brien, Curtin University of Technology, Australia
  • Stacie Frye, Bacterial Barcodes, United States
  • Mimi Healy, Bacterial Barcodes, United States

    • Background: In Western Australia all clinical isolates of MRSA are referred to the Gram-positive Bacteria Typing and Research Unit for molecular characterization. Using pulsed-field gel electrophoresis (PFGE) 57 pulsotypes of community associated MRSA have been identified from which 38 multilocus sequence types (MLST) belonging to 15 clonal clusters (CC) and three singletons have been described. Combining MLST and SCCmec typing 57 MRSA clones have been characterised including five sequence types (ST) with typeIV and type V SCCmec elements and four ST with non typable SCCmec elements. Five MRSA clones have more than one pulsotype. In this study the clonal designations and CCs based on MLST were compared to those made by sequencing the variable repeats in the protein A gene spa and the DiversiLab rep-PCR assay which is based on the PCR amplification of specific regions between non-coding repetitive sequences in the bacterial genome.

    Methods: spa sequences obtained were compared to those held on the Ridom SpaServer, and DiversiLab rep-PCR dendrograms were generated using the DiversiLab analysis software version 3.3.

    Results: Overall 35 spa types and 34 rep-PCR types were characterised.
    Several spa types corresponded to more than one ST including: t002 (CC5 - ST5, ST73, ST575, ST641, ST835); t044 (CC80 - ST583, ST728, ST80); t064 (CC8 - ST609, ST612, ST1173), t127 (CC1 - ST1, ST872, ST1005); t437 (CC59 - ST59, ST922); t3029 (ST733 and CC9 - ST834, ST584). Only one spa type was identified in more than one CC or singleton; t3029 in CC9 and singleton ST733, which is closely related to CC9.
    Several rep-PCR types also corresponded to more than one ST including: "1a" (CC5 - ST5, ST575, ST641); "1b" (CC5 - ST5, ST73, ST835), "3" (CC1 - ST1, ST872); "4a" (CC8 - ST8, ST612); "6a" (CC8 - ST8, ST576, ST923, ST1173); "8" (CC59 - ST59, ST87, ST922, ST952); "11" (CC80 - ST583, ST728). Two rep PCR types were identified in more than one CC or singleton; "12" in CC1 and C72 and "25" in CC9 and singleton ST732. However neither of these rep PCR types were identical with both differing by one band.

    Conclusion: Although a low level of discrimination of the STs were achieved by spa
    and rep-PCR typing, both methods were able to reliably predict CCs and
    singletons previously determined with the MLST database eBURSTv3
    program