The first attempt to explain the role of RF3 in translation 2 Corresponding author termination proposed a dual function, first to stimulate the email: ehrenberg@xray.bmc.vv.se binding of RF1 and RF2 (RF1/2) to the ribosome, and secondly to catalyse the dissociation of the factors from Ribosomes complexed with synthetic mRNA and pepthe ribosome after hydrolysis of peptidyl-tRNA in a tidyl-tRNA, ready for peptide release, were purified by reaction dependent on guanine nucleotide, thereby stimugel filtration and used to study the function of release lating the recycling of RF1/2 between ribosomes factor RF3 and guanine nucleotides in the termination (Goldstein and Caskey, 1970). of protein synthesis. The peptide-releasing activity ofRecently, an extended version of the first action of RF3 RF1 and RF2 in limiting concentrations was stimulated proposed by Goldstein and Caskey (1970) has been by the addition of RF3 and GTP, stimulated, though strongly propagated ; Nakamura et al., to a lesser extent, by RF3 and a non-hydrolysable GTP 1996): the 'ternary complex model'. This proposal was analogue, and inhibited by RF3 and GDP or RF3 inspired by the recent discovery of structural similarity without guanine nucleotide. With short incubation between EF-G on one hand and the ternary complex times allowing only a single catalytic cycle of RF1 or between EF-Tu·GTP and aminoacyl-tRNA on the other RF2, peptide release activity was independent of RF3 (AEvarsson et al., 1994;Czworkowski et al., 1994; Nissen and guanine nucleotide. RF3 hydrolysis of GTP to et al., 1995). Domain 4 in EF-G structurally mimics the GDP ⍣ P i was dependent only on ribosomes and not anticodon stem in tRNA, and domains I (the G-domain) on RF1 or RF2. RF3 affected neither the rate of and II in EF-G have sequence similarity with EF-Tu. The association of RF1 and RF2 with the ribosome nor the model was supported further by sequence similarities catalytic rate of peptide release. A model is proposed between RF3 on one hand and EF-Tu and EF-G on the which explains how RF3 recycles RF1 and RF2 by other, as well as by sequence similarities between the displacing the factors from the ribosome after the part of EF-G that mimics the tRNA anticodon stem, and release of peptide.RF1/2. From these data, it was suggested that RF3, GTP Keywords: protein synthesis/release factors/RF3/ and RF1 or RF2 may form a ternary complex like that ribosome/termination between EF-Tu, GTP and aminoacyl-tRNA, either off or on the ribosome (Ito et al

A complete translation system has been assembled from pure initiation, elongation and termination factors as well as pure aminoacyl‐tRNA synthetases. In this system, ribosomes perform repeated rounds of translation of short synthetic mRNAs which allows the time per translational round (the recycling time) to be measured. The system has been used to study the influence of release factor RF3 and of ribosome recycling factor RRF on the rate of recycling of ribosomes. In the absence of both RF3 and RRF, the recycling time is ∼40 s. This time is reduced to ∼30 s by the addition of RF3 alone and to ∼15 s by the addition of RRF alone. When both RF3 and RRF are added to the translation system, the recycling time drops to <6 s. Release factor RF3 is seen to promote RF1 cycling between different ribosomes. The action of RRF is shown to depend on the concentration of elongation factor‐G. Even in the presence of RRF, ribosomes do not leave the mRNA after termination, but translate the same mRNA several times. This shows that RRF does not actively eject mRNA from the terminating ribosome. It is proposed that terminating ribosomes become mobile on mRNA and ready to enter the next translation round only after two distinct steps, catalysed consecutively by RF3 and RRF, which are slow in the absence of these factors.