Ribosomal frameshifting and transcriptional slippage: From genetic steganography and cryptography to adventitious use

Atkins, John F.; Loughran, Gary; Bhatt, Pramod R.; Firth, Andrew E.; Baranov, Pavel V.

HIV also illustrates pairing of sequence just 5 of the shift site with sequence 3 of the shift site to form a structural base 'anchoring helix' that facilitates a functionally important transition between two alternate forms of the upper part of the structure mostly specified 3 of the shift site (509) . Other recent HIV experiments are also relevant (135, 510) . Despite this case nearly all 3 recoding stimulators known are specified wholly 3 of the shift site, even though in a few cases their extreme proximity to the shift site, just 3 nts for mammalian antizyme 1 +1 frameshifting, would require at least unwinding of the 'lower' part of the structure before the shift site is in the ribosomal P-/A-sites. Whether the important feature is a 'set up' for the remaining part of the structure or an alternative considered in the nascent peptide section above, is unknown. Irrespective of whether the antizyme pseudoknot stimulates the shift directly or indirectly, the extent of conservation of its constituent stems, but not its unpaired components, illustrates the value of what can be termed 'phylogenetic probing' for structure function inferences (270) (Figure 14) . The diversity of modulators 5 of the shift site is at least matched by that of 3 stimulators. As mentioned, the majority of intra-mRNA 3 stimulatory signals, have their 5 ends 6-9 nts 3 of the shift site and are thought to be at the mRNA unwinding site within the mRNA entrance tunnel (511, 512) when the shift site is in the P-/A-sites. Expression in heterologous systems (513, 514) and in homologous systems (98, 406, 515) , have revealed differences between bacteria and eukaryotic systems in optimal spacer lengths or within eukaryotes for specific 3 stimulators, in one case with implications for frameshift directionality. For a minority of 3 stimulators their 5 ends are 12-15 nts 3 of the shift site and are likely to have their influence at the outer edge of the mRNA entrance channel. These stimulators can involve either intra-mRNA structure or 'linear' mRNA with the characterized ones being associated with trans-acting factors (23,58,59). While the effectiveness of stimulators in general is selected to act at different efficiencies in different cases, the upper 'strength' limit for 3 intra-mRNA structural stimulations is that they must not act as a complete 'roadblock' to ribosome progression (516) . The type of 3 structures range from simple stem loops, as with E. coli dnaX −1 frameshifting (392, 403, 517, 518) , likely plant sobemoviruses (463, (519) (520) (521) and explored systematically with the simian retrovirus shift site (522) , to more elaborate stem loop structures such as in bacterial IS911 (188) and HIV (509, 510) to compact pseudoknots as in poleroviruses, e.g. beet western yellows virus ( Figure 13B ), potato leaf roll virus and sugarcane yellow leaf virus (523, 524) , larger pseudoknots with all components specified nearby (see below), to pseu-doknots in which a lateral bulge-loop of an extended stemloop structure 3 -adjacent to the frameshift site interacts with an apical loop of a distant stem-loop structure (ALIL) (45, 110, 525) . A subset of related sequences can use either stem loop or pseudoknot stimulators, for instance among IS elements (187) , different alphaviruses (140) , and probably different flaviviruses (70, 120, 123, 124) . The great majority of the frameshift stimulatory pseudoknots studied are for −1 frameshifting, but some are known for +1 frameshift