the addition of the sulfate moiety until as late in the synthetic scheme as possible. In the final synthetic step, only high-yielding deprotection reactions remain. Only the disappearance of unsulfated starting material, in a high-yielding sulfation step, or the complete deprotection of sulfated product is usually monitored. TLC on cellulose [ 3 ], or silica gel plates [4] has been useful for analyzing mono-or di-anionic species or molecules with a significant lipophilicity. In our laboratory, we have recently found it necessary to follow the chemical modification of highly sulfated carbohydrates, where structural alterations occur without a change in molecular charge. Information on reaction completion and composition of product mixtures (containing components having three or more negative charges) is needed. Based on this requirement, we anticipated that CE would provide a rapid and sensitive method to follow both chemical reactions and product purification. Extensive studies in our laboratory had failed to successfully apply TLC to the analysis of these highly charged compounds. The use of CE is described here as an analytical tool for monitoring chemical reactions of sulfated sugars. This methodology should aid in developing new methods for the chemical synthesis and modification of sulfated organic compounds.
Trisulfated disaccharide 1 (Scheme 1) was obtained by enzymatic depolymerization of heparin as previously described by our group [ 5 ]. Acylation of hydroxyl groups on heparin and heparin fragments has been shown using acid anhydrides and ion-pairing agents [ 6 ]. However, no information on the regioselectivity of acylation or the degree of desulfation of unrecovered material (if present ) is known. As part of our continuing effort to investigate the biological activity of derivatized heparins, we have begun to examine the chemical synthesis and derivatization of heparin oligosaccharides. Introduction of hydroxyl protecting groups on 1 were required for subsequent synthetic manipulations. Acetylation, pivaloylation, and benzylation reactions of I were followed by CE using absorbance detection at 232 nm. Detection at this wavelength is based on the presence of the chromophoric unsaturated uronic acid residue. These analyses result in a time-course distribution of reaction products. Samples for analysis (2-5/zL) were removed from the mixture (using a 10-/zL syringe through the rubber septa) and diluted (quenching the reaction) using 10-20/xL of deionized, distilled water at 0Β°C. These diluted aliquots were then analyzed directly on CE (Fig. 1). Purified products were co-injected with reaction samples or compared by similarity in migration time to verify peak identity.
CE analysis of the acetylation of 1 is shown in Fig. 1A. Disaccharide 1 having a migration time of 25 rain is observed only at time 0 (electropherogram a). At 1 h, a duster of peaks having a major component 3 are detected between 20 and 23 min (electropherogram b). The peaks converged into a single peak 2 (electropherogram c), while peak 1 disappeared