Analysis and prediction of mitochondrial targeting signals.

Publisher Summary This chapter describes the known mitochondrial-targeting signals, their analysis, and available methods to predict the presence of such signals in eukaryotic proteins. The best-characterized mitochondrial-targeting signal is the matrix-targeting signal—also called the presequence. Analysis of a large number of mitochondrial presequences suggested that most of them have the potential to form a positively charged amphiphilic α-helix in which segregation of positively charged and hydrophobic residues on opposite faces of the helix occurs. Two-dimensional nuclear magnetic resonance, fluorescence methods, and circular dichroism measurements demonstrated the ability of targeting sequences to form amphipathic α-helices in membranes or membrane-like environments, whereas in aqueous solution, they are essentially unstructured. The targeting of mitochondrial precursor proteins from the cytosol to the organelle and the subsequent intra-mitochondrial sorting depend on specific targeting and sorting information within the precursor sequence. In the case of the cleavable presequences, the characteristics of this signal are well understood and various bioinformatics and experimental tools are available to analyze them. The prediction provided by these programs is based mainly on physicochemical parameters, such as the abundance of certain amino acids and the hydrophobicity in certain regions, and/or analysis of the plain residue patterns in the amino acid sequences. Several bioinformatics tools, such as TargetP, PSORT II, MITOPRED, MitoProt II, and Predotar are also presented in the chapter.