Replication of positive (+)-strand RNA virus genomes is a fundamental process in a virus's life cycle. Turnip crinkle virus (TCV) and an associated satellite RNA (satC) share 151 nt of 3' terminal sequence, which are predicted to fold into four phylogenetically-inferred hairpins (from the 3' end: Pr hairpin, H5, H4b and H4a). Pr hairpin is part of the core promoter (Pr) required for satC (Song and Simon, 1995) and TCV (Sun and Simon, in press) negative (-)-strand synthesis. To identify other regulatory cis-acting elements throughout satC, individual deletions of six other predicted hairpins (H5, H4b, H4a, M1H, H6, and H2) were performed. These deletions significantly reduce accumulation of (+)-strand monomers and differentially affect accumulation of (+)-strand dimers and (-)-strands in Arabidopsis protoplasts.

Results from in vivo genetic selection and mutational analyses of satC H5 indicate that robust satC accumulation in vivo requires specific sequences in the large symmetrical internal loop (LSL) and a stable stem and specific base pairs in the lower stem. The upper stem-loop has considerable plasticity. Moreover, H5 may be involved in accumulation of both strands. Mutational analyses also suggest that the LSL and/or the 3' terminus may have other functions in addition to forming a pseudoknot (pseudoknot 1), which is required for replication.

An RNA conformational switch from a pre-active structure to an active structure appears required to regulate initiation of satC (-)-strand synthesis (Zhang et al., 2006). Results from mutational analyses suggest that H4a and H4b function as a unit and pseudoknot 2, formed between H4b and sequence flanking the 3' side of H5 and whose disruption reduces satC accumulation in vivo, stabilizes the pre-active satC structure. In addition, an upstream element (DR) may help to promote the switch.

Step-wise conversion of satC and TCV 3' terminal homologous sequences into the counterpart's sequence revealed the importance of having the cognate core promoter. The satC Pr is a substantially better promoter than the TCV Pr when assayed in vitro. These results suggest that the TCV Pr requires upstream elements for full functionality and that evolution of satC generated a Pr that functions more efficiently by itself.