CONFORMER SELECTION AND DIFFERENTIAL RESTRICTION OF LIGAND MOBILITY BY A PLANT LECTIN - CONFORMATIONAL BEHAVIOR OF GAL-BETA-1-3GLCNAC-BETA-1-R, GAL-BETA-1-3GALNAC-BETA-1-R AND GAL-BETA-1-2GAL-BETA-1-R' IN THE FREE STATE AND COMPLEXED WITH GALACTOSIDE-SPECIFIC MISTLETOE LECTIN ASREVEALED BY RANDOM-WALK AND CONFORMATIONAL-CLUSTERING MOLECULAR-MECHANICS CALCULATIONS, MOLECULAR-DYNAMICS SIMULATIONS AND
M. Gilleron et al., CONFORMER SELECTION AND DIFFERENTIAL RESTRICTION OF LIGAND MOBILITY BY A PLANT LECTIN - CONFORMATIONAL BEHAVIOR OF GAL-BETA-1-3GLCNAC-BETA-1-R, GAL-BETA-1-3GALNAC-BETA-1-R AND GAL-BETA-1-2GAL-BETA-1-R' IN THE FREE STATE AND COMPLEXED WITH GALACTOSIDE-SPECIFIC MISTLETOE LECTIN ASREVEALED BY RANDOM-WALK AND CONFORMATIONAL-CLUSTERING MOLECULAR-MECHANICS CALCULATIONS, MOLECULAR-DYNAMICS SIMULATIONS AND, European journal of biochemistry, 252(3), 1998, pp. 416-427
To study conformational parameters of ligands before and after complex
formation with the galactoside-binding agglutinin of Viscum album L.
(VAA) in solution, combined computer-assisted random walk molecular me
chanics (RAMM) calculations extended by conformational clustering anal
ysis (CCA), molecular dynamics (MD) simulations as well as two-dimensi
onal rotating-frame nuclear Overhauser effect (ROE) and two-dimensiona
l nuclear Overhauser effect (NOE) spectroscopy NMR experiments were em
ployed. Derivatives of the naturally occurring disaccharides Gal beta
1-3GlcNAc beta 1-R and Gal beta 1-3GalNAc beta 1-R as well as of a syn
thetic high-affinity binding partner, i.e. the disaccharide Gal beta 1
-2Gal beta 1-R', were chosen as ligands in this study. The disaccharid
es displayed inherent flexibility in the valley of the global minimum
between Phi/Psi combinations of (40 degrees/60 degrees) and (40 degree
s/-60 degrees). Calculations of the de-N-acetylated sugars revealed th
at presence of this group did not markedly influence the distribution
of low-energy conformers in the Phi, Psi, E plot. Occupation of side m
inima at Phi Psi(180 degrees/0 degrees) or (0 degrees/180 degrees) is
either unlikely or low according to the results of MD simulations and
RAMM calculations extended by CCA. Notably, these side minima define c
onformations which are not stable during a MD simulation. Transitions
to other minima occur already a few picoseconds after the start of the
simulation. NMR experiments of the foe-state ligand confirmed the val
idity of the data sets obtained by the calculations. Following the des
cription of the conformational space in the free-state NMR experiments
were performed for these disaccharides complexed with VAA. They yield
ed two interresidual contacts for Gal beta 1-3GlcNAc beta 1-R and Gal
beta 1-2Gal beta 1-R'. The ligand conformations in the complex did not
deviate markedly from those of a minimum conformation in the foe stat
e. One-and two-dimensional transferred nuclear Overhauser enhancement
(TRNOE) experiments at different mixing times excluded the influence o
f spin-diffusion effects. When the NOE build-up curves in the three st
udied cases were compared, the residual mobility of the penultimate ca
rbohydrate unit of Gal beta 1-3GalNAc beta 1-R was observed to be high
er than that of the respective hexopyranose unit of the other two boun
d ligands. Due to the availability of the conformational parameters of
Gal beta 1-2Gal beta 1-R' in association with a galectin, namely the
beta-galactoside-binding protein from chicken liver, it is remarkable
to note that this ligand displays different conformations in the bindi
ng sites of either the plant or the animal lectin. They correspond to
local energy-minimum conformations in the Phi, Psi, E plot and substan
tiate differential conformer selection by these two lectins with ident
ical nominal monosaccharide specificity.