Do not ferment Mtl and this capability was restored by complementation with the entire locus in strains Liv1098 (SH1000 mtlD::tet pMJH71) and Liv1097 (SH1000 mtlABFD::tet pMJH71). Weak growth of Liv1023 was observed. doi:10.1371/journal.pone.0067698.gFigure 5. Survival on linoleic acid agar. Comparative survival of strains on BHI agar supplemented with 1 mM linoleic acid. Strains SH1000 (open circles), Liv1023 (SH1000 mtlD::tet) (filled squares), Liv1024 (SH1000 mtlABFD::tet) (open triangles) and Liv1098 (SH1000 mtlD::tet pMJH71) were diluted in PBS and equivalent volumes 10457188 were plated onto the agar. SE from triplicate experiments is shown with error bars inside symbols. doi:10.1371/journal.pone.0067698.gS. aureus Mannitol Utilisation and SurvivalFigure 7. Growth of S. aureus in the Avasimibe web presence of mannitol and linoleic acid. Bacteria were cultured on BHI agar containing either no or added mannitol (0.1 M, 0.5 M) in the presence (black bars) or absence (white bars) of 1 mM linoleic acid. Differences in viable cells recovered in the presence of mannitol were significantly reduced compared to the absence of mannitol (P = 0.03 and P = 0.001 for 0.1 M and 0.5 M, respectively). doi:10.1371/journal.pone.0067698.gbetween SH1000 and the mtl mutants, as judged by spectrophotometric analysis of methanol-extracted cells from overnight and 2 day-old cultures (data not shown).Virulence of mtlD MutantFigure 6. Growth phenotype of mtlD inactivated S. aureus. (A) Culture of strains in broth containing Mtl at 37uC. Liv1023 (SH1000 mtlD::tet) ( ) had a significantly reduced growth rate (P,0.01, Student’s t-test) compared to wild-type (SH1000) ( ), Liv1024 (SH1000 mtlABFD::tet) (m), strains Liv1098 (SH1000 mtlD::tet pMJH71) ( ) and Liv1097 (SH1000 mtlABFD::tet pMJH71) (#). Calculated doubling times between 2 h and 3 h of growth: SH1000 = 0.33, Liv1023 (SH1000 mtlD::tet) = 1.7. Error bars indicate 1 SEM (n = 3). (B) Culture of strains in BHI broth at 25uC. Liv1023 (SH1000 mtlD::tet) ( ) has a significantly reduced growth rate (P,0.001, Student’s t-test) compared to wild-type (SH1000) ( ), Liv1024 (SH1000 mtlABFD::tet) (m), strains Liv1098 (SH1000 mtlD::tet pMJH71) ( ) and Liv1097 (SH1000 mtlABFD::tet pMJH71) (#). Calculated doubling times between 5 h and 13 h of growth: SH1000 = 2.46, Liv1023 = 3.13. Representative dataset from triplicate assay. doi:10.1371/journal.pone.0067698.gNNThe decreased in vitro AFA survival and reduced H2O2 MIC of the mtlD mutant prompted testing of its virulence compared to the isogenic parent strain using a previously described model of experimental septic arthritis (Figure 8). This model was tested to determine whether inactivation of the mtlABFD locus affected virulence, since its contribution to metabolism in vivo is unknown and the model generates abscesses where AFAs accumulate [14]. This revealed that SH1000 mtlD did not have reduced virulence, at least under the conditions studied [6,32,33].Table 2. Sugar alcohols present in S. aureus strains.Relative mean concentration Metabolite Arabitol Mannitol Mannitol-P Ribitol Sorbitol-6-P SH1000 116.1 (3.9) 417.6 (29.5) ND 240 (22.3) 149.4 (18.2) Liv1023 107.5 (19.7) 1351.4 (82.5) 8161.3 (119.7 214.8 (17.6) ND Liv1024 51.4 (7.9) 5.9 (0.9) ND 272.2 (27.1) NDResistance and Cell Surface Properties of mtl MutantsA range of antimicrobial agents were tested to determine if the observed reduced resistance of Liv1023 (SH1000 mtlD::tet) Gracillin biological activity extended beyond AFAs. Growth and MICs were comparable betwe.Do not ferment Mtl and this capability was restored by complementation with the entire locus in strains Liv1098 (SH1000 mtlD::tet pMJH71) and Liv1097 (SH1000 mtlABFD::tet pMJH71). Weak growth of Liv1023 was observed. doi:10.1371/journal.pone.0067698.gFigure 5. Survival on linoleic acid agar. Comparative survival of strains on BHI agar supplemented with 1 mM linoleic acid. Strains SH1000 (open circles), Liv1023 (SH1000 mtlD::tet) (filled squares), Liv1024 (SH1000 mtlABFD::tet) (open triangles) and Liv1098 (SH1000 mtlD::tet pMJH71) were diluted in PBS and equivalent volumes 10457188 were plated onto the agar. SE from triplicate experiments is shown with error bars inside symbols. doi:10.1371/journal.pone.0067698.gS. aureus Mannitol Utilisation and SurvivalFigure 7. Growth of S. aureus in the presence of mannitol and linoleic acid. Bacteria were cultured on BHI agar containing either no or added mannitol (0.1 M, 0.5 M) in the presence (black bars) or absence (white bars) of 1 mM linoleic acid. Differences in viable cells recovered in the presence of mannitol were significantly reduced compared to the absence of mannitol (P = 0.03 and P = 0.001 for 0.1 M and 0.5 M, respectively). doi:10.1371/journal.pone.0067698.gbetween SH1000 and the mtl mutants, as judged by spectrophotometric analysis of methanol-extracted cells from overnight and 2 day-old cultures (data not shown).Virulence of mtlD MutantFigure 6. Growth phenotype of mtlD inactivated S. aureus. (A) Culture of strains in broth containing Mtl at 37uC. Liv1023 (SH1000 mtlD::tet) ( ) had a significantly reduced growth rate (P,0.01, Student’s t-test) compared to wild-type (SH1000) ( ), Liv1024 (SH1000 mtlABFD::tet) (m), strains Liv1098 (SH1000 mtlD::tet pMJH71) ( ) and Liv1097 (SH1000 mtlABFD::tet pMJH71) (#). Calculated doubling times between 2 h and 3 h of growth: SH1000 = 0.33, Liv1023 (SH1000 mtlD::tet) = 1.7. Error bars indicate 1 SEM (n = 3). (B) Culture of strains in BHI broth at 25uC. Liv1023 (SH1000 mtlD::tet) ( ) has a significantly reduced growth rate (P,0.001, Student’s t-test) compared to wild-type (SH1000) ( ), Liv1024 (SH1000 mtlABFD::tet) (m), strains Liv1098 (SH1000 mtlD::tet pMJH71) ( ) and Liv1097 (SH1000 mtlABFD::tet pMJH71) (#). Calculated doubling times between 5 h and 13 h of growth: SH1000 = 2.46, Liv1023 = 3.13. Representative dataset from triplicate assay. doi:10.1371/journal.pone.0067698.gNNThe decreased in vitro AFA survival and reduced H2O2 MIC of the mtlD mutant prompted testing of its virulence compared to the isogenic parent strain using a previously described model of experimental septic arthritis (Figure 8). This model was tested to determine whether inactivation of the mtlABFD locus affected virulence, since its contribution to metabolism in vivo is unknown and the model generates abscesses where AFAs accumulate [14]. This revealed that SH1000 mtlD did not have reduced virulence, at least under the conditions studied [6,32,33].Table 2. Sugar alcohols present in S. aureus strains.Relative mean concentration Metabolite Arabitol Mannitol Mannitol-P Ribitol Sorbitol-6-P SH1000 116.1 (3.9) 417.6 (29.5) ND 240 (22.3) 149.4 (18.2) Liv1023 107.5 (19.7) 1351.4 (82.5) 8161.3 (119.7 214.8 (17.6) ND Liv1024 51.4 (7.9) 5.9 (0.9) ND 272.2 (27.1) NDResistance and Cell Surface Properties of mtl MutantsA range of antimicrobial agents were tested to determine if the observed reduced resistance of Liv1023 (SH1000 mtlD::tet) extended beyond AFAs. Growth and MICs were comparable betwe.