Spin Label EPR-Based Characterization of Biosystem Complexity

Abstract

Following the widely spread EPR spin-label applications for biosystem characterization, a novel approach is proposed for EPR-based characterization of biosystem complexity. Hereto a computational method based on a hybrid evolutionary optimization (HEO) is introduced. The enormous volume of information obtained from multiple HEO runs is reduced with a novel so-called GHOST condensation method for automatic detection of the degree of system complexity through the construction of two-dimensional solution distributions. The GHOST method shows the ability of automatic quantitative characterization of groups of solutions, e.g. the determination of average spectral parameters and group contributions. The application of the GHOST condensation algorithm is demonstrated on four synthetic examples of different complexity and applied to two physiologically relevant examples--the determination of domains in biomembranes (lateral heterogeneity) and the study of the low-resolution structure of membrane proteins.

Structural characterization of liposomes made of diether archaeal lipids and dipalmitoyl-L-α-phosphatidylcholine

“…To obtain the best fit of a calculated EPR spectrum to the experimental one, the multi-run hybrid evolutionary optimization algorithm is used [21], together with a newly developed GHOST condensation procedure [22].…”

SDSL-ESR-based protein structure characterization

“…Another important indicator of successful spectral optimization is an equal contribution of different runs into the final set of solutions, which is measured in terms of the run flatness parameter. This value should be above 70% (Kavalenka et al 2005;Š trancar et al 2005).…”

“…Thus altogether, there are 7N c -1 spectral parameters, which have to be resolved by the spectral optimization routine. Taking into account the resolution limit of spin label ESR to be around 30 parameters, this allows the use of at most four spectral components (Stopar et al 2006;Š trancar 2007;Š trancar et al 2005). An hybrid evolutionary optimization (HEO), a combination of a genetic algorithm and a downhill-simplex local search (Filipič and Š trancar 2001;Š trancar et al 2005) is used to find the set of spectral parameters that produces the best fit to the experimental ESR spectrum.…”

“…The hybrid evolutionary optimization routine starts with a random initialization of a population of 400 solutions, and continues with the tournament selection and application of genetic operators (i.e., three-point crossover, uniform mutation, and local improvements performed with downhillsimplex) for 100 generations (Filipič and Š trancar 2001;Š trancar et al 2005). The elite set is used to keep track of the best individuals.…”

“…From all restricted rotamers of the modelled conformational space of the spin label, the distribution of nitroxide NO vectors can be characterized and compared with the related characteristics extracted from the ESR spectra. The cone model of spin label motion that is used in the analysis of experimental ESR spectra (Kavalenka et al 2005;Stopar et al 2006;Š trancar et al 2005), is parameterized with the angles # 0 and u 0 , both defined in the range (0; p/2). These angles describe the amplitude and anisotropy of the spin label rotational motion within a cone, respectively.…”