A bipartite model of 2-5A-dependent RNase L

The 2-5A-dependent RNase (RNase L) is a tightly regulated endoribonuclease of higher vertebrates that is catalytically active only after engaging unusual effector molecules consisting of the 2',5'-linked oligoadenylates, p1-3A(2'p5'A)>/=2 (2-5A). Progressive truncations from either terminus have provided insight into the structure, function, and regulation of RNase L. We determined that deletion of the N-terminal 335 amino acids of RNase L, about 45% of the enzyme, produced a constitutively active endoribonuclease, thus effectively eliminating the requirement for 2-5A. The truncated nuclease had 6-fold lower catalytic activity against an oligo(rU) substrate than wild type RNase L. However, the two enzymes showed identical RNA cleavage site preferences with an mRNA as substrate. The repressor function required only the last three of a series of nine ankyrin-like repeats present in the N-terminal part of RNase L. In contrast, the entire ankyrin repeat region was necessary and sufficient for 2-5A binding activity. Deletion of a 10-amino acid sequence near the C terminus of RNase L, between residues 710 and 720, eliminated both the catalytic and RNA substrate binding functions of the enzyme. The ability to bind native RNase L in response to 2-5A required amino acid sequences near both termini of the protein. A bipartite model for the structure of RNase L emerged in which the regulatory functions of the molecule are located in the N-terminal half, while the catalytic domain is present in the C-terminal half.