The topology of multidimensional potential energy surfaces: Theory and application to peptide structure and kinetics

Abstract

Topological characteristics of multidimensional potential energy surfaces are explored and the full conformation space is mapped on the set of local minima. This map partitions conformation space into energy-dependent or temperature-dependent “attraction basins’’ and generates a “disconnectivity’’ graph that reflects the basin connectivity and characterizes the shape of the multidimensional surface. The partitioning of the conformation space is used to express the temporal behavior of the system in terms of basin-to-basin kinetics instead of the usual state-to-state transitions. For this purpose the transition matrix of the system is expressed in terms of basin-to-basin transitions and the corresponding master equation is solved. As an example, the approach is applied to the tetrapeptide, isobutyryl-(ala)3-NH-methyl (IAN), which is the shortest peptide that can form a full helical turn. A nearly complete list of minima and barriers is available for this system from the work of Czerminiski and Elber. The multidimensional potential energy surface of the peptide is shown to exhibit an overall “funnel’’ shape. The relation between connectivity and spatial proximity in dihedral angle space is examined. It is found that, although the two are similar, closeness in one does not always imply closeness in the other. The basin to basin kinetics is examined using a master equation and the results are interpreted in terms of kinetic connectivity. The conformation space of the peptide is divided up in terms of the surface topography to model its “folding’’ behavior. Even in this very simple system, the kinetics exhibit a “trapping’’ state which appears as a “kinetic intermediate,’’ as in the folding of proteins. The approach described here can be used more generally to classify multidimensional potential energy surfaces and the time development of complex systems.