ANALYSIS OF THE ROLE OF THE PSEUDOKNOT COMPONENT IN THE SRV-1 GAG-PRORIBOSOMAL FRAMESHIFT SIGNAL - LOOP LENGTHS AND STABILITY OF THE STEM REGIONS

The simian retrovirus-1 (SRV-1) gag-pro frameshift signal was identifi ed in previous work, and the overall structure of the pseudoknot invol ved was confirmed (ten Dam E, Brierley I, Inglis S. Pleij C, 1994, Nuc leic Acids Res 22:2304-2310). Here we report on the importance of spec ific elements within the pseudoknot. Some mutations in stem S1 that ma intain base pairing have reduced frameshift efficiencies. This indicat es that base pairing in itself is not sufficient. In contrast, framesh ifting correlates qualitatively with the calculated stability of mutat ions in S2. The stems thus play different roles in the frameshift even t. The nature of the base in L1 has little influence on frameshift eff iciency. It is however required to bridge S2; deleting it lowers frame shifting from 23 to 9%. In L2, frameshift efficiency was not affected in a mutant that changed 10 of 12 bases. This makes it unlikely that t he primary sequence of L2 plays a role in -1 frameshifting, in contras t to readthrough in Moloney murine leukemia virus (Wills N, Gesteland R, Atkins J, 1994, EMBO J 13:4137-4144). Deletions of 2 and 3 bases ga ve more frameshifting than the wild type, probably reflecting the incr eased stability of the pseudoknot due to a shorter loop L2. Deleting e ven more bases reduces frameshifting compared to wild-type levels. At this point, stress will build up in L2, and this will reduce overall p seudoknot stability.