A. Almond et Jk. Sheehan, Glycosaminoglycan conformation: do aqueous molecular dynamics simulations agree with x-ray fiber diffraction?, GLYCOBIOLOG, 10(3), 2000, pp. 329-338
Glycosaminoglycan-protein interactions are biologically important and requi
re an appreciation of glycan molecular shape in solution, which is presentl
y unavailable. In previous studies we found strong similarity between aqueo
us molecular dynamics (MD) simulations and published x-ray diffraction refi
nements of hyaluronan, We have applied a similar approach here to chondroit
in and dermatan, attempting to clarify some of the issues raised by the x-r
ay diffraction literature relating to chondroitin and dermatan sulfate. We
predict that chondroitin has the same beta(1-->4) linkage conformation as h
yaluronan, and that their average beta(1-->3) conformations differ. This is
explained by changes in hydrogen-bonding across this linkage, resulting fr
om its axial hydroxyl, causing a different sampling of left-handed helices
in chondroitin (2.5- to 3.5-fold) as compared with hyaluronan (3,0- to 4.0-
fold). Few right-handed helices, which lack intramolecular hydrogen-bonds,
were sampled during our MD simulations. Thus, we propose that the 8-fold he
lix observed in chondroitin-6-sulfate, represented in the literature as an
8(3) helix (right-handed), though it has never been refined, is more likely
to be 8(5) (left-handed) helix. Molecular dynamics simulations implied tha
t C-4(1) and S-2(0), but not C-1(4), forms of iduronate could be used in re
finements of dermatan x-ray fiber diffraction patterns. Current models of 8
-fold dermatan sulfate chains containing 4C1 iduronate refine to right-hand
ed helices, which possess no intramolecular hydrogen-bonds. However, MD sim
ulations predict that models containing S-2(0) iduronate could provide bett
er (85 helix) starting structures for refinement. Thus, the 8-fold dermatan
sulfate refinement (83 helix) could be in error.