COMPUTATIONAL CHARACTERIZATION OF POTENTIAL RNA-BINDING SITES IN ARENAVIRUS NUCLEOCAPSID PROTEINS

A nucleocapsid protein of an RNA virus was characterised using computa tional methods. Similarity searches using standard algorithms and more sensitive methods based on profiles were performed. Also, secondary s tructure prediction and statistical methods were used. The results sho w that the protein belongs to a unique well-characterised family, with three regions with potential RNA binding capacity. The amino-terminal region is found to contain a mixed-charge segment similar to proteins that bear nucleic acid-protein interaction capacity. The middle-regio n has a slight homology to the nucleolar protein Fibrillarin containin g an atypical RNP-1 conserved octamer. Finally, the carboxyl-terminal region has a putative zinc-finger.