TRENDS in Biochemical Sciences Vol.27 No.9 September 2002 448 Opinion http://tibs.trends.com 0968-0004/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0968-0004(02)02149-7 Opinion ‘Programmed ribosomal frameshifts’ (PRFs) often make translating ribosomes slip by one base in either the 5′ (-1) or the 3′ (+1) direction (reviewed in [1–3]). For viruses that use PRF, the efficiencies of frameshift events are crucial: they determine the stoichiometry of viral structural and enzymatic proteins that are available for virus particle assembly, and therefore control virus propagation (reviewed in [4]). Thus, it is important to understand how frameshifting efficiencies are controlled. The general mechanism involves specific mRNA signals that induce ribosomes to stall over special ‘slippery’ mRNA sequences, allowing the ribosome-bound tRNAs to slip and subsequently pair their non-wobble bases with the out-of-frame codons. Changing these signals alters frameshifting efficiencies. Whereas cis-acting signals are well-characterized, trans-acting factors and the biophysical parameters that contribute to the determination of PRF efficiencies are less well understood. Genetic, biochemical, molecular and pharmacological methods have been used to this end. It is believed that the basic parameters that can affect PRF efficiencies include: (1) elongation-cycle step-specific changes in the residence time of ribosomes at PRF signals; (2) changes in the stabilities of ribosome-bound tRNAs owing to alterations in codon–anticodon interactions and intrinsic ribosomal components such as ribosomal proteins and rRNAs; and (3) defects in the ability of the translational apparatus to recognize and correct errors. This article focuses on the first of these by integrating descriptions of -1 and +1 PRF into the context of elongation kinetics. Although some aspects of the resulting model are controversial, we have found them useful both as a guide to further characterize PRF at the molecular and biochemical levels and for the identification of new targets for antiviral therapeutics. Aspects of the second and third parameters are discussed within the context of this ‘integrated model’of PRF. An ‘integrated model’ of PRF As PRF occurs during translation elongation, the models describing -1 and +1 PRF must be understood within this context. The elongation cycle can be envisioned as a series of forward chemical reactions that have been partitioned into at least nine discrete steps [5]. However, to emphasize the relationship between frameshifting and the tRNA occupancy states of the ribosome, we have simplified this into four general stages: (1) selection and insertion of cognate aminoacyl-tRNA (aa-tRNA) into the A-site; (2) accommodation of the 3′-end of the aa-tRNA into the peptidyltransferase center; (3) peptidyl transfer; and (4) translocation (Fig. 1b). Frameshifting in the –1 direction is used by many RNA viruses, from complex retroviruses such as HIV-1, to the simple L-A (‘killer’) totivirus of the yeast Saccharomyces cerevisiae. The -1 PRF is directed by a bipartite cis- acting mRNA signal that comprises a heptameric slippery site (X XXY YYZ) in which the incoming reading frame is indicated by spaces, followed by a 3′ mRNA secondary structure (Fig. 1c) (reviewed in [1,4,6]). Typically, the secondary structure is an mRNA pseudoknot, which consists of two nested stems, the loop of one stem forming the base-pairs of the second. These structures cause elongating ribosomes to pause over the slippery site, such that upon a -1 shift, the non-wobble bases of both the A- and P-site tRNAs can re-pair with the new -1 frame codons. There is also strong genetic evidence supporting the notion that the +1 PRF of the Ty1 retrotransposable element of yeast is driven by ribosomal pausing [7]. In this case, the ‘hungry’AGG codon forces elongating ribosomes to pause over the CUU AGG C slippery heptamer (incoming reading frame is indicated by spaces), awaiting delivery of the rare aa-tRNA Arg to the A-site (Fig. 1a). Slippage of the ribosome–peptidyl-tRNA complex makes the new A-site codon GGC, which corresponds to a highly abundant aa-tRNA Gly . Importantly, both -1 and +1 PRF are kinetically driven; that is, they occur when ribosomes are forced to pause by cis-acting signals on the mRNAs. When –1 and +1 PRF are considered as alternatives to the forward reaction, it becomes apparent that the tRNA occupancy states of the ribosome during the two shifts are different: the -1 shift occurs when both A- and P-sites are occupied, whereas the +1 shift occurs when the A-site is empty. Thus viewed, changes in the kinetics of each of the specific sub-steps in the An ‘integrated model’ of programmed ribosomal frameshifting Jason W. Harger, Arturas Meskauskas and Jonathan D. Dinman Many viral mRNAs, including those of HIV-1, can make translating ribosomes change reading frame.Altering the efficiencies of programmed ribosomal frameshift (PRF) inhibits viral propagation.As a new target for potential antiviral agents,it is therefore important to understand how PRF is controlled. Incorporation of the current models describing PRF into the context of the translation elongation cycle leads us to propose an ‘integrated model’ of PRF both as a guide towards further characterization of PRF at the molecular and biochemical levels,and for the identification of new targets for antiviral therapeutics. Published online: 25 July 2002 Jason W. Harger Graduate School of Biomedical Sciences, Rutgers University and The University of Medicine & Dentistry of New Jersey, Robert Wood Johnson Medical School, Dept of Molecular Genetics and Microbiology, Piscataway, NJ 08854, USA; and Dept of Cell Biology and Molecular Genetics. Arturas Meskauskas Jonathan D. Dinman* Dept of Cell Biology and Molecular Genetics, University of Maryland, College Park, MD 20742, USA. *e-mail: jd280@ umail.umd.edu