Carbohydrate Research 337 (2002) 1861 – 1871 www.elsevier.com/locate/carres Disaccharide conformational maps: 3D contours or 2D plots?  Carlos A. Stortz,* Alberto S. Cerezo Departamento de Quı ´mica Orga ´nica -CIHIDECAR, Facultad de Ciencias Exactas y Naturales, Uniersidad de Buenos Aires, Ciudad Uniersitaria, 1428 Buenos Aires, Argentina Received 21 June 2002; accepted 29 July 2002 Abstract The potential energy surfaces of several -(1 3)- and -(1 4)-linked disaccharides were obtained and plotted in terms of energy versus  glycosidic angle. These plots were compared to those obtained previously in the way of the usual 3D contour maps, which relate the energy with the two glycosidic angles ( and  ). Given the usually small variations of the  angle in the low-energy regions (at least using MM3), both kinds of graphs lead to similar conclusions concerning flexibility measurements by two different methods and assessment of the effects of sulfation and/or hydroxyl group orientation. Only second-order effects were found with some sulfated disaccharides, not changing the general conclusions. The computational efforts required to produce those plots are smaller, and the plots are easier to interpret. Besides, the conversion of a 3D map into a 2D plot leaves the possibility of constructing 3D maps of carbohydrates including a second variable different to , e.g., the second  angle of a trisaccharide or the  angle of a 6-linked disaccharide. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Disaccharides; Conformational analysis; Molecular mechanics; MM3; Ramachandran map 1. Introduction Conformational analysis of disaccharides is usually accompanied by the generation of a Ramachandran- like 3D contour map as a tool in understanding their conformational features 1–3 and predicting the likeli- hood of different molecular conformations. 4 In these maps the energy is determined for all mutual orienta- tions of the monosaccharide residues expressed by the glycosidic angles  and . First studies were carried out by rigid residue analysis, 3 but, by 1979, flexible residue analysis was initiated by allowing all variables to re- lax. 5–7 From the late 1980s on, many disaccharides were modeled using different force fields (and even QM methods), and the results were expressed with 3D con- tour maps. 4,8–17 The map is actually an adiabatic repre- sentation of the true conformational hypersurface that depends not only on the glycosidic angles but also on many other variables, among them the orientation of secondary hydroxyl groups and hydroxymethyl groups, the different degrees of puckering of the rings, etc., which might complicate reaching true adiabaticity. 18 However, it is usually considered that the most impor- tant energy variations are those related with the glyco- sidic angles  and , and thus, the 3D contour adiabatic map against these two variables is the usual output of disaccharide conformational analysis. When looking at the maps, particularly those pro- duced with MM3 (one of the force fields considered to be more reliable for carbohydrates), 14–23 the conclu- sions drawn from these maps are that (especially for -linked disaccharides) ‘a trough centered at a more or less fixed  angle is observed, and it contains the main minima, each of which exhibits a clearly different  angle’. 14,20 This  angle matches that expected as an expression of the exo -anomeric effect, although it is doubtful if it occurs solely on grounds of this effect, as it has been shown that other force fields (as MM2) with a weaker parameterization of this effect, also give rise to similar maps, 19,24 and C -disaccharides, for which exo -anomeric effect cannot occur, also exhibit similar angles. 25,26 A similar behavior for the  angle was found in X-ray crystallographic analysis of different disaccharides and derivatives. 27  Throughout this paper, ‘3D contours’ indicate traditional Ramachandran contour maps, while ‘2D plots’ indicate x – y graphs representing energy versus a single angle. * Corresponding author. Tel./fax: +54-11-4576-3346 E -mail address: stortz@qo.fcen.uba.ar (C.A. Stortz). 0008-6215/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII:S0008-6215(02)00257-4