Submitted to: Cellulose
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: August 27, 2003
Publication Date: January 2, 2004
Citation: French, A.D., Johnson, G.P. 2004. What crystals of small analogs are trying to tell us about cellulose structure. Cellulose. 11:5-22. Interpretive Summary: Cellulose, the major component of cotton fiber, is a chain molecule composed of hundreds or even thousands of glucose units. Understanding the detailed structures of cellulose is fundamental to understanding the biology of almost all members of the plant kingdom, including cotton, and to industries based on the use of cotton, among many others. Advances in our knowledge of cellulose through experiment continue, but other useful information can come from both theoretical and experimental studies on small molecules that are related to cellulose, such as cellobiose which has only two glucose units, connected in the same manner as in cellulose. This work extrapolates from the details of molecular structure determined through crystallography to give cellulose models. All but one of these models was a helix with two to three glucose units per helix turn. Also, this survey of small molecule structures resulted in lists of expected values for the shapes of the glucose units and the linkage bond angle between them. This work is primarily of interest to scientists who are doing structural studies of cellulose. It is also of broad interest to scientists dealing in many ways with various cellulosic materials.
Technical Abstract: The molecular geometries from crystal structures of 23 small molecules such as cellobiose were reviewed and extrapolated to give model cellulose chains. Within a given model, all monosaccharide units and their linkages are identical so the models are regular helices. Despite fairly large ranges for the glycosidic linkage torsion angles phi And psi, 29 degrees and 57 degrees, respectively, there is little variation in the n and h parameters of the model helices. They are extended, with h values (the advance per residue along the helix axis) of 5.04-5.27 A. Some models were slightly right-handed, with n values up to 2.12 residues per helix turn. Left-handed models were in the majority, and their n values were as large as -2.91. These results are consistent with known structures of cellulose and its derivatives. An exception comes from a heavily derivatized cellobiose molecule. It yields right-handed helices with n equals 4.5 and h equals 3 A. Because one half turn of this helix reverses the direction of the chain in a compact region, the linkage geometry is a model for chain-folding. Other derivatives that are unable to form the 03...05' hydrogen bond gave left-handed helices. The puckering of the glucose rings was also surveyed. A number of rings in small molecule structures are puckered to a degree that is similar to the puckering determined for methyl cellotrioside, cellotetraose, cellulose I beta and cellulose II.