Using pulse field gradient NMR diffusion measurements to define molecular size distributions in glycan preparations

Glycans comprise perhaps the largest biomass in nature, and more and more glycans are used in a number of applications, including those as pharmaceutical agents in the clinic. However, defining glycan molecular weight distributions during and after their preparation is not always straightforward. Here, we use pulse field gradient (PFG) (1)H NMR self-diffusion measurements to assess molecular weight distributions in various glycan preparations. Initially, we derived diffusion coefficients, D, on a series of dextrans with reported weight-average molecular weights from about 5 kDa to 150 kDa. For each dextran sample, we analyzed 15 diffusion decay curves, one from each of the 15 major (1)H resonance envelopes, to provide diffusion coefficients. By measuring D as a function of dextran concentration, we determined D at infinite dilution, D(inf), which allowed estimation of the hydrodynamic radius, R(h), using the Stokes-Einstein relationship. A plot of log D(inf) versus log R(h) was linear and provided a standard calibration curve from which R(h) is estimated for other glycans. We then applied this methodology to investigate two other glycans, an alpha-(1-->2)-L-rhamnosyl-alpha-(1-->4)-D-galacturonosyl with quasi-randomly distributed, mostly terminal beta(1-->4)-linked galactose side-chains (GRG) and an alpha(1-->6)-D-galacto-beta(1-->4)-D-mannan (Davanat), which is presently being tested against cancer in the clinic. Using the dextran-derived calibration curve, we find that average R(h) values for GRG and Davanat are 76+/-6 x 10(-10) m and 56+/-3 x 10(-10) m, with GRG being more polydispersed than Davanat. Results from this study will be useful to investigators requiring knowledge of polysaccharide dispersity, needing to study polysaccharides under various solution conditions, or wanting to follow degradation of polysaccharides during production.