Full hydrodynamic reversibility of the weak dimerization of vancomycin and elucidation of its interaction with VanS monomers at clinical concentration

Abstract

The reversibility and strength of the previously established dimerization of the important glycopeptide antibiotic vancomycin in four different aqueous solvents (including a medically-used formulation) have been studied using short-column sedimentation equilibrium in the analytical ultracentrifuge and model-independent SEDFIT-MSTAR analysis across a range of loading concentrations. The change in the weight average molar mass M _w with loading concentration was consistent with a monomer-dimer equilibrium. Overlap of data sets of point weight average molar masses M _w(r) versus local concentration c(r) for different loading concentrations demonstrated a completely reversible equilibrium process. At the clinical infusion concentration of 5 mg.mL^-1 all glycopeptide is dimerized whilst at 19 µg.mL^-1 (a clinical target trough serum concentration), vancomycin was mainly monomeric (<20% dimerized). Analysis of the variation of M _w with loading concentration revealed dissociation constants in the range 25-75 μM, commensurate with a relatively weak association. The effect of two-fold vancomycin (19 µg.mL^-1) appears to have no effect on the monomeric enterococcal VanS kinase involved in glycopeptide resistance regulation. Therefore, the 30% increase in sedimentation coefficient of VanS on adding vancomycin observed previously is more likely to be due to a ligand-induced conformational change of VanS to a more compact form rather than a ligand-induced dimerization.

Figure 1

Chemical structure of vancomycin. The disaccharide composed of vancosamine and glucose (purple) is attached at the para-position of the phenyl group of (2-[α-4-L-epi-vancosaminyl]-β-1-D-glucosyl)-D-phenyl glycine (with residue 4). Also shown are residue 1 (green): N-methyl-D-leucine; residue 2 (light orange): m-chloro-β-hydroxy-D-tyrosine; residue 3 (red): asparagine; residue 5 (grey/green): p-hydroxy-D-phenylglycine; residue 6 (pink): m-chloro- β-hydroxy-D-tyrosine; and residue 7 (dark orange): m,m-dihydroxy-L-phenylglycine. Black dotted lines highlight the groups that hydrogen bond with the D-Ala-D-Ala substrate in peptidoglycan. Redrawn from ref..

Figure 2

Sedimentation equilibrium SEDFIT-MSTAR output for analysis of vancomycin. Solvent: 0.9% NaCl at 7.0 °C at a loading concentration of ~1.25 mg.mL⁻¹ in conventional (12 mm) path length cells. (a) concentration (fringe displacement units) versus radial displacement from the centre of rotation, r (b) log concentration versus the square of the radial displacement (c) extrapolation of the M* function to the cell base to yield the “whole distribution” weight average (apparent) molar mass M _w,app = (2.4 ± 0.1) kDa; (d) plot of the point average apparent molar mass (local molar mass) M _w,app(r) – obtained by taking the derivative of the data from plot (b) versus local concentration c(r) in the analytical ultracentrifuge cell. The value at the hinge point (where c(r) = the cell loading concentration) – dashed line - yields another estimate for the whole distribution M _w,app ~ (2.4 ± 0.1) kDa. Because of the low molar mass and low concentration, non-ideality effects will be negligible and M _w,app = M _w.

Figure 3

Change of weight average molar mass M_w of vancomycin with concentration. From sedimentation equilibrium analysed by SEDFIT-MSTAR for four different solvent data sets. Squares: 10 mM HEPES. Diamonds: 10 mM HEPES + 100 mM NaCl. Up triangles: 10 mM HEPES = 100 mM NaCl + 20% (v/v) glycerol. Down triangles: 0.9% NaCl in deionised, distilled water. Solid symbols – molar masses M _w,app obtained from M* analysis. Open symbols – molar masses obtained from hinge point analysis. Because of the low molar masses, non-ideality effects can be assumed to be negligible and M _w,app = M _w. Solid line is a standard French curve fit to the data.

Figure 4

Comparison of the effects of different solvent conditions on the molar mass -concentration behaviour. As Fig. 3 but datasets for each of the solvent conditions shown separately. (a) 10 mM HEPES. (b) 10 mM HEPES + 100 mM NaCl. (c) 10 mM HEPES = 100 mM NaCl + 20% glycerol. (d) 0.9% (w/v) NaCl in deionised, distilled water.

Figure 5

Diagnostic sedimentation equilibrium plots confirming a completely reversible dimerization. Weight average molar mass values M _w(r) at individual radial positions in the ultracentrifuge cell plotted against local concentration c(r) in interference fringe units for different concentrations: violet (0.625 mg.mL⁻¹), blue (1.25 mg.mL⁻¹), green (2.5 mg.mL⁻¹), orange (5.0 mg.mL⁻¹) and red (10.0 mg.mL⁻¹). (a) 10 mM HEPES. (b) 10 mM HEPES + 100 mM NaCl. (c) 10 mM HEPES = 100 mM NaCl + 20% glycerol. (d) 0.9% NaCl in deionised, distilled water. For a completely reversible self-association the plots should lie, within experimental error on the same curve shown as a standard French curve fit to the data.

Figure 6

Kegeles-Rao evaluation of the association constants k ₂ (mL/g), K ₂ (M⁻¹), dissociation constant K_d (μM) and standard free energy of association ΔG° for the dimerization of vancomycin. Least squares fitting to the Kegeles-Rao equation^, Y(c) ≡ M ₁{M _w(c) − M ₁}/{(2M ₁ − M _w(c))²} = k ₂.c. where the M _w(c) are the weight average molar masses (averaged over the whole macromolecular distributions) at different loading (a) 10 mM HEPES. (b) 10 mM HEPES + 100 mM NaCl. (c) 10 mM HEPES = 100 mM NaCl + 20% glycerol. (d) 0.9% NaCl in deionised, distilled water. Lines shown are obtained by least squares analysis. Values are given in Table 1.

Figure 7

Sedimentation Equilibrium SEDFIT-MSTAR output for analysis of VanS in the presence of vancomycin. Solvent: 10 mM HEPES = 100 mM NaCl + 20% glycerol, pH~7.9, I = 0.1 at 7.0 °C at a loading concentration of ~0.3 mg.mL⁻¹ (VanS) supplemented with 19 μg.mL⁻¹ vancomycin in long (20mm) path length cells. (a) concentration (fringe displacement units) versus radial displacement from the centre of rotation, r (b) log concentration versus the square of the radial displacement (c) extrapolation of the M* function to the cell base to yield the “whole distribution” weight average (apparent) molar mass M _w,app = (45.0 ± 1.0) kDa; (d) plot of the point average apparent molar mass (local molar mass) M _w,app(r) – obtained by taking the derivative of the data from plot (b) versus local concentration c(r) in the analytical ultracentrifuge cell. The value at the hinge point (where c(r) = the cell loading concentration) – dashed line - yields another estimate for the whole distribution M _w,app ~ (45.0 ± 1.0) kDa. Because of the low concentration, non-ideality effects will be negligible and M _w,app = M _w.