INFLUENCE OF SULFATE AND CARBOXYLATE GROUPS ON THE CONFORMATION OF CHONDROITIN SULFATE RELATED DISACCHARIDES

H-1 NMR and C-13 NMR spectral parameters of eight sulfated uronic acid containing disaccharides 1-8 were used to determine the conformational preferences that depend on the pattern of sulfation. Three sulfated derivatives of benzyl beta-D-Gal-(1 --> 4)-beta-D-GlcA (1), its 6'-sulfate 2, 4'-sulfate 3, and 4',6'-disulfate 4 were used as models for the beta-(1 --> 4) glycosidic linkage of chondroitin sulfates and three sulfated derivatives of benzyl beta-D-GlcA-(1 --> 3)-beta-D-Gal (5), its 6-sulfate 6, 4-sulfate 7, and 4,6-disulfate 8 were used as models for the beta-(1 --> 3) glycosidic linkage of chondroitin sulfates. To determine the dependence of conformational preferences on the charged groups, the sulfated disaccharides 2, 3, and 4 were compared to their unsulfated parent compound 1, and 6, 7, and 8 were compared to their parent compound 5. The 3J(H-5,H-6) coupling constants were determined by high-order analysis of the spin systems, and from these the preferred populations of the hydroxymethyl groups were calculated. Selective ID NOEs and ROEs were measured from H-1' across the glycosidic linkage to obtain the average distance of the protons adjacent to the glycosidic linkage. Derivatives of beta-D-Gal-(1 --> 4)-beta-D-GlcA carrying a sulfate group in the 6'-position (2) and in the 4'- and 6'-position (4) show a slight repulsive effect between the 6'-sulfate groups and the carboxylate group as expressed in small changes of the preferred populations of the glycosidic linkage and the sulfonyloxymethyl group. The 4-sulfate groups in 3 and 4 do not show a significant influence on the glycosidic linkage. However, the two sulfate groups in 4 exhibit a repulsive effect leading to a very high population of the gt conformation of the sulfonyloxymethyl group. In contrast hereto, the sulfate group at C-4 of beta-D-GlcA-(1 --> 3)-beta-D-Gal disaccharides 7 and 8 and the carboxylate group exert an attractive interaction that leads to a change of the conformation of the glycosidic linkage in 7 and 8 by about 30-degrees. The 6-sulfate groups of the disaccharides 6 and 8 show a slight repulsive interaction with the carboxylate and/or 4-sulfate group. Changes in C-13 NMR chemical shifts support the interpretation obtained from the NOE and ROE analysis. Relative NOE ratios were used to estimate the correlation time to tau(c) = 3.5-4.0 X 10(-10) s for 1-8, and these were found to be in agreement with calculated T1 values. The charge-charge interaction energies observed are fairly small in 1-8. The largest effect can be observed in the rotamer population of the C-6 groups of the disulfates 4 and 8, which leads to an effective interaction energy of the two sulfate groups of almost-equal-to 1.0 kcal/mol. This is equivalent to an effective dielectric constant of epsilon almost-equal-to 10, assuming that full charges are expressed on the sulfate groups. Ensemble average values for observable epsilon of 1-8 were obtained from Metropolis Monte Carlo simulations (MMC) that were run with dielectriC constants epsilon almost-equal-to 2-80 using the GEGOP program. The results of the calculations agree well with the experimental data at dielectric constants of epsilon almost-equal-to 10.