Novel approaches to the analysis of polysaccharide structures.

Recently, atomic force microscopy has been used to image a variety of polysaccharides and map their distribution on cell surfaces. The mechanical response of polysaccharides to tensile stress has been investigated in single-molecule force spectroscopy experiments. Small-angle X-ray scattering has provided a probe of polysaccharide structure operating in a size range (2-25 nm) that is intermediate between those accessible using nuclear magnetic resonance and light scattering.

The influence of sidechains on the calculated dimensions of three related bacterial polysaccharides.

The effect of van der Waals interactions between sidechain and backbone on the shape of three bacterial polysaccharides in solution has been investigated. The three polymers, namely, gellan, welan, and rhamsan, share the same four-sugar backbone repeating-unit. Gellan is unbranched, whereas welan and rhamsan display comblike branching. Consequently, the effect of chain branching on backbone conformation may be investigated. Van der Waals repulsive interactions of sidechain and backbone serve to limit, somewhat, the range of conformational freedom of the welan backbone in comparison to that of gellan. Attractive side-chain-backbone interactions, which may be as significant as 2-3 kcal/mol, predominate over much of the accessible conformational space of the welan backbone. Despite the strength of these interactions, the unperturbed shape of welan in solution is calculated to be very similar to that of the unbranched gellan. Attractive sidechain-backbone interactions in rhamsan have a modest influence on the conformational characteristics of the rhamsan backbone. The calculated, unperturbed conformation in solution is slightly more extended than that of gellan and welan, but the fundamental shape of the chain is changed only slightly. Significant differences in the physical properties of these polymers seem not to arise from differences in their random-coil conformations provoked by van der Waals interactions of sidechain and backbone. Other contributions to the sidechain-backbone interaction, e.g., hydrogen bonding, could be involved; interchain interactions are also likely to be important.

The dependence of glucan conformational dynamics on linkage position and stereochemistry.

The 13C NMR T1 relaxation times for the (1-->4)-linked maltooligomers (Mi) and the (1-->6)-linked isomaltooligomers (IMi) with i = 2, 4, 6, and 8 were measured in aqueous solution at 22 and 65 degrees C at a concentration (3%) low enough to have removed concentration-dependent effects on the measured T1 values. Separate T1 values were measured for each carbon in the residue at the reducing end of the oligosaccharide, in the residue at the non-reducing end, and in the interior, i.e., non-terminal, residue(s). Analogous data for the corresponding high polymers show that at 22 degrees C the relaxation times for the carbons of the interior residues of the oligomers have converged to their high chain length asymptotes at about i = 10. This observation suggests that at room temperature polymeric motions in the frequency domain effective for 13C NMR relaxation at a magnetic field strength of 11.7 T have a "wavelength" of the order of 10 residues. The relaxation times characterizing the two ends of the chain are different, with longer T1 values for the carbons of the reducing end than for those of the non-reducing end. Carbons of alpha-anomeric residues at the reducing end have shorter relaxation times than those of the corresponding beta-anomeric reducing sugars. Carbons of the interior residues have T1 values shorter than the carbons of either type of terminal residue. For oligomers of a given dp there is no T1 difference between oligomers of the Mi and IMi series at room temperature. This observation is seemingly at odds with the great differences in the inherent conformational freedom of the (1-->4)- and (1-->6)-linkages. At elevated temperatures the orientational relaxation behavior of the two series of oligomers measured by 13C T1 values show interesting differences, and in the case of the Mi series, structure develops in the chain length dependence of the T1 values.

The sequence statistics and solution conformation of a barley (1----3, 1----4)-beta-D-glucan.

The sequence statistics and aqueous solution conformation of the 40 degrees water-soluble (1----3,1----4)-beta-D-glucan isolated from barley (Hordeum vulgare) have been modeled realistically using the known sequence-distributions of (1----3) and (1----4) linkages, theoretical conformational analysis, and the statistical mechanical theory of polymer-chain conformation. This barley beta-glucan fraction consists of (1----4)-beta-glucooligosaccharides, predominantly of d.p. 4 or less, joined by single beta-(1----3) linkages. Approximate treatments of the sequence statistics which do not take into account the small mole fraction (approximately 2%) of (1----4)-beta-glucooligosaccharides of d.p. approximately 10 significantly underestimate the chain extension in solution. A correct prediction of the observed chain extension is achieved when these longer, highly extended (1----4)-beta-glucooligosaccharide blocks are included in a model which randomly incorporates all (1----4)-beta-glucooligosaccharide segments in the proportions observed experimentally. Chain flexibility in the 40 degrees water-soluble beta-glucan fraction is shown to arise principally from the isolated beta-(1----3) linkages; blocks of two or more contiguous beta-(1----3) linkages provide a source of additional flexibility which may influence the properties of barley beta-glucan fractions containing a significant proportion of such sequences.

Imaging of carrageenan macrocycles and amylose using noncontact atomic force microscopy.

Samples of kappa-carrageenan, iota-carrageenan, and synthetic amylose have been examined by atomic force microscopy (AFM). All samples were spray deposited from aqueous solutions onto freshly cleaved mica, air dried, and imaged in air using noncontact atomic force microscopy (NCAFM). Images of single stranded amylose and carrageenan are presented. At relatively low polymer concentrations in the presence of NaCl iota-carrageenan formed circles that appear to be predominantly head-to-tail associated unimeric duplex (double stranded) structures. At higher iota-carrageenan concentrations the polymer forms circles and aggregates that appear to involve dimeric duplex structure. Direct comparison of synthetic amylose molecular weights determined from NCAFM images with results from solution measurements showed that NCAFM provides an excellent way to measure amylose molecular weight and molecular weight distribution. It is shown that synthetic amylose is single stranded in aqueous solution and that the chain length distribution is broader than the Poisson distribution anticipated from polymerization theory.

Thermal treatment of semi-dilute aqueous xanthan solutions yields weak gels with properties resembling hyaluronic acid.

Semi-dilute (ca 2 g/dl) aqueous xanthan (mean molar mass ca 1 x 10(6) g/mol), when heated in the presence of 0.1 M NaCl to a temperature above the order<-->disorder transition temperature, forms highly viscoelastic solutions when returned to room temperature. The steady shear and dynamic rheological behaviour of these solutions discloses a weak gel structure, the viscosity of which is unusually sensitive to the rate of shear. In shear thinning behaviour these heat and salt treated xanthan solutions mimic the properties of the aqueous hyaluronic acid solutions widely used in viscosurgical techniques. The double stranded model of native xanthan is invoked to interpret the observed behaviour of heat and salt treated semi-dilute aqueous xanthan.

The reliability of wormlike polysaccharide chain dimensions estimated from electron micrographs.

Electron micrographs of alginate, xylinan, xanthan, and scleroglucan were prepared by vacuum-drying aqueous glycerol-containing solutions, and then heavy-metal, low-angle rotary replicated. Quantitative methods for excluding streamlining effects and deformation artifacts were developed and applied to the digitized polymer contours prior to analysis of stiffness. The apparent macromolecular dimensionalities were not obtainable on the basis of the change in the scaling coefficient alpha relating the rms end-to-end distance and the contour length, mean value of r2(1/2) approximately L alpha, for chains subject to the excluded volume effect in two and three dimensions. Using a two-dimensional model, the persistence length of these molecules was estimated to be (9 +/- 1) nm (alginate), (25 +/- 4) nm (xylinan), (30 +/- 4) nm (single-stranded xanthan), (68 +/- 7) nm (double-stranded xanthan), and (80 +/- 10) nm (scleroglucan). Monte Carlo calculations for wormlike chains close to an interacting surface or confined to the region between two surfaces showed that (1) strongly adsorbed molecules are essentially two-dimensional and (2) molecules restricted to the space between two surfaces separated by a distance less than 20% of the persistence length are two-dimensional in their directional correlation. The somewhat low estimates of the persistence lengths obtained from the electron micrographs compared with those reported from solution measurements can be accounted for by the adoption of a strictly two-dimensional model in the analysis, whereas the absorbed polymers are most likely intermediate between the two-and three-dimensional cases. The model calculations and the analysis of the electron micrographs suggest that stiffness parameters are obtainable from the electron micrographs when the proper theoretical description are used in the analysis.

Imaging of individual biopolymers and supramolecular assemblies using noncontact atomic force microscopy.

A variety of biopolymers is imaged using noncontact atomic force microscopy. Samples are prepared by aerosol spray deposition of aqueous solutions on freshly cleaved mica followed by air drying. The distributions of contour lengths and chain or fibril thicknesses normal to the mica substrate can be measured for individual polymer molecules or molecular assemblies. In many cases it is possible to conclude that the structures imaged and quantitatively analyzed are representative of those present in solution and not artifacts of the deposition/dessication process. Imaging of linear and cyclic triple helices of the polysaccharide scleroglucan is demonstrated. Measurements of the triple helix thickness normal to the mica surface are analyzed, and successful measurements of the molecular weight distribution and mean molar mass are described. It is demonstrated that the extent of chain association in the polysaccharide xanthan can be modulated by the addition of low molecular weight salts. The contour length and chain thickness distributions in a xanthan fraction are presented. Increases in the extent of chain association with increasing polymer concentration are documented for the gelling polysaccharide gellan, and the formation of stiff fibrillar gellan aggregates in the presence of added low molecular salt is demonstrated. Images are presented of the polysaccharide kappa-carrageenan in its disordered, and presumably single-stranded, state. Biopolymers other than polysaccharides can be imaged by the same technique; this is demonstrated with the fibrous protein collagen. In general it is shown that aerosol spray deposition of biopolymer samples can be used in conjunction with noncontact atomic force microscopy to provide a fast, reliable, and reproducible method for assessing the size and shape distributions of individual biological macromolecules and macromolecular assemblies in solution with a minimum of time and effort devoted to sample preparation.

A general treatment of the configuration statistics of polysaccharides.

A treatment of the configurational statistics of polysaccharides is given in the isomeric state approximation. All classes of linear polysaccharides of specifiec chemical sequence are treated simultaneously. Chain tortuosity arising from torsional motions about the chemical bonds of the glycosidic linkages is recognized explicitly as is the possibility for conformational isomerism of the sugar residues. Valence angles and lengths are taken to be fixed at the equilibrium values, and pyranose residues in their chair conformations are treated as inflexible constituents of the skeletal structure. Pyranose and furanose forms capable of pseudorotation may be incorporated as rigid skeletal entities as well, provided suitable attention is given to the selection and interpretation of the conformational isomeric states included. Separation of the configuration energy into independent contributions is shown to be impossible in general. Methods are described for assessing the influence of neighbor interactions on the populations of the several conformers of the sugar residues. The relative conformational free energy of the flexible and chair form conformers of pyranose sugars is discussed, and appropriate measures of polysaccharide chain flexibility and stiffness are suggested.

Comparison of the conformational dynamics of the (1----4)- and (1----6)-linked alpha-D-glucans using 13C-NMR relaxation.

The conformational dynamics of alpha-(1----4)- and alpha-(1----6)-glucan homooligomers in the nanosecond time domain have been compared by measuring the 13C-nmr longitudinal relaxation times T1 for carbons of the terminal and interior sugar residues. Measurements are reported on monomeric glucose and on oligomers containing up to ten glucose residues at room temperature in aqueous solution at concentrations of 3 and 20 g/dL. The carbons of terminal residues display longer relaxation times than do those of interior residues, presumably as a consequence of a greater degree of conformational mobility of the chain ends. The T1s of the reducing terminal residues of all oligomers are significantly longer than those of the corresponding nonreducing termini, a phenomenon that we associate tentatively with the anomeric equilibrium at the reducing end. Carbons of the reducing terminal residues in the beta-anomeric form relax more slowly than their alpha-anomeric counterparts. At 20 g/dL the mean T1s for carbons of the terminal and interior residues attain asymptotic behavior with increasing chain length at a chain length of about six residues, and carbons of the alpha-(1----4)-linked maltooligomers relax significantly more slowly than those of the corresponding alpha-(1----6)-linked isomaltooligomers. The T1s of both glucan series increase with decreasing concentration. This concentration dependence disappears below 3 g/dL, where the T1s of the two series of homoligomers are no longer distinguishable. This suggests that in dilute aqueous solution at room temperature viscous damping effects predominate over contributions to the T1-sensitive conformational dynamics from structural differences in the glycosidic linkage region. At 3 g/dL the approach to long chain-length asymptotic behavior is more protracted than at 20 g/dL, and the T1s of carbons of interior oligomeric residues appear to match the corresponding high-polymer behavior at a chain length of eight and greater.