ABC-F Proteins Mediate Antibiotic Resistance through Ribosomal Protection

  1. Alex J. O’Neill
  1. Antimicrobial Research Centre, Astbury Centre for Structural Molecular Biology and School of Molecular and Cellular Biology, University of Leeds, Leeds, United Kingdom
  1. Address correspondence to Alex J. O’Neill, a.j.oneill{at}leeds.ac.uk.
  1. Editor Gerard D. Wright, McMaster University

  1. FIG 1 

    Phylogenetic tree and antibiotic resistance profiles of the ARE ABC-F proteins found in representative Gram-positive pathogens. The tree was generated using the maximum likelihood method with the MEGA 6.0.6 software package (47). An overview of the antibiotic resistance phenotypes conferred by the different subgroups of determinant is given at the right of the figure, denoted by colored boxes (although variations in individual resistance phenotypes within each subgroup are not shown).

  2. FIG 2 

    Vga(A) protects an S. aureus-derived T/T assay from inhibition by virginiamycin M. (A) Column 1 shows an uninhibited T/T assay with no addition of exogenous protein, whilst column 2 shows an uninhibited assay with the addition of 4 µM Vga(A). In columns 3 and 5, 4 µM Vga(A) added to a T/T assay mixture containing ≥IC90 of virginiamycin M (VGM [column 3]) rescued protein synthesis. In columns 3, 4, and 8, addition of 4 µM heat-denatured (Denat.) Vga(A) (column 4) or 4 µM fusidic acid resistance protein FusB (column 8) failed to rescue protein synthesis from inhibition by virginiamycin M (column 3). In columns 6 and 7, addition of 4 µM Vga(A) to a T/T assay mixture containing ≥IC90 of fusidic acid (FA) did not rescue protein synthesis. (B) Dose-dependent rescue of protein synthesis by Vga(A) from inhibition with ≥IC90 of virginiamycin M. Results are means from at least three independent determinations, and error bars show standard deviations.

  3. FIG 3 

    Lsa(A) mediates dose-dependent protection of a S. aureus-derived transcription/translation assay from inhibition by virginiamycin M (A) and lincomycin (B). Results are means from at least three independent determinations, and error bars show standard deviations.

  4. FIG 4 

    Recapitulation of resistance phenotypes associated with ARE ABC-F proteins in vitro. (A) When expressed in E. coli, Vga(A) does not confer resistance to virginiamycin M (21); addition of Vga(A) to an E. coli T/T assay containing ≥IC90 of virginiamycin M (VGM) also failed to restore translational activity. (B) ATPase activity is essential for Vga(A) function (21), and abrogation of ATPase activity of the N-terminal ABC domain rendered Vga(A) inactive when expressed in S. aureus RN4220; the purified ATPase-deficient Vga(A)E105Q protein also failed to protect staphylococcal translation from inhibition by virginiamycin M in vitro. (C) A single-amino-acid substitution in the interdomain linker expands the resistance spectrum of Vga(A) to encompass lincomycin (20). Addition of the purified Vga(A)K219T to a staphylococcal T/T assay inhibited with a >IC90 of lincomycin (LNC) restored translational activity, while addition of the wild-type protein did not. Results are means from at least three independent determinations, and error bars show standard deviations.

  5. FIG 5 

    Lsa(A) prevents binding of lincomycin to staphylococcal ribosomes, and displaces ribosome-bound lincomycin. (A) Preincubation of increasing concentrations of Lsa(A) with 0.5 µM staphylococcal ribosomes caused a reduction in binding of [3H]lincomycin. (B) Preincubation of 0.5 µM ribosomes with a 50× excess of unlabeled lincomycin (LNC) decreased subsequent binding by [3H]lincomycin (column 2 versus 1). Preincubation with 4 µM BSA did not protect ribosomes from binding by [3H]lincomycin (columns 3 and 1). Addition of 4 µM Lsa(A) resulted in decreased association of [3H]lincomycin with ribosomes (column 4 versus 1). (C) Addition of a 50× excess of unlabeled lincomycin caused dissociation of [3H]lincomycin prebound to staphylococcal ribosomes (column 2 versus 1), as did addition of 4 µM Lsa(A) (column 4 versus 1); however, addition of BSA did not (column 3 versus 1). Results are means from at least three independent determinations, and error bars show standard deviations.