“…Sense triplets are decoded on the ribosome by specific transfer RNAs via base pairing between codons in mRNA and anticodon triplets in tRNA+ In contrast, nonsense (termination, stop) codons are decoded in two canonical ways+ UGA can be decoded as selenocysteine by a specialized tRNA-translation factor complex (selenocysteinyl-tRNA Sec : SelB) that directs SectRNA Sec to a unique message structure (Hüttenhofer & Bock, 1998)+ More frequently UGA, UAA, and UAG are decoded as stop-translation signals+ A single-polypeptide release factor (RF) termed eRF1 performs this function in eukaryotes and two proteins, RF1 and RF2, do so in bacteria (reviewed in Tate et al+, 1996;Buckingham et al+, 1997;)+ The class-1 release factor proteins RF1, RF2, and eRF1, are distinguished by their ability to trigger peptidyl-tRNA hydrolysis, catalyzed by the ribosomal peptidyl transferase when a stop codon enters the A site+ Although the nature of this hydrolytic signal remains unknown, it is clear that class-1 RF proteins strongly compete with suppressor tRNAs in vitro and in vivo (Weiss et al+, 1984;Curran & Yarus, 1988;Eggertsson & Soll, 1988;Drugeon et al+, 1997;Le Goff et al+, 1997)+ They therefore probably overlap the ribosomal decoding site+ As has been suggested (Cantor, 1979;Nissen et al+, 1995;Ito et al+, 1996;Nakamura et al+, 1996;Kisselev et al+, 2000) and proven recently by crystallography (Song et al+, 2000), eRF1 adopts an extended, perhaps tRNA-like overall shape+ Other proteins acting at the ribosomal A site, such as elongation factor EF-G and bacterial ribosome recycling factor (RRF; Nyborg et al+, 1996;Selmer et al+, 1999), also emulate the tRNA shape+ These structural findings support a "tRNA-analog" hypothesis ; that is, that termination codons may be decoded directly by class-1 RFs, whose similar shape allows mimicry of tRNA action+ Thus RFs may contact stop codons in the decoding (A) site on the ribosome+ These facts suggest that RF action would be effectively inhibited by other strongly bound RNAs+ Selected RNA ligands for RF protein(s) therefore might act as highly specific inhibitors of stop codon translation+ This would be useful for biochemical dissection of the pathway for translation termination+ In addition, be-cause a single eRF1 protein decodes all termination codons, such RNA ligands would likely comprise a new type of omnipotent termination suppressor, increasing stop codon readthrough by interfering with the usual termination pathway+ Therefore we have isolated eRF ligands using selection-amplification (for review, see Wilson & Szostak, 1999)+ Starting with a randomized mixture of oligoribonucleotides, repetitive selection for RNA-RF binding yielded RNAs with high affinity ...…”