A design principle for a single-stranded RNA genome that replicates with less double-strand formation

Usui, Kimihito; Ichihashi, Norikazu; Yomo, Tetsuya

We next attempted to extract a simple index that quantitatively predicts the dsRNA ratio from all of the data obtained above. According to the model in Figure 1 and the results in Figure 3 , the total numbers of G or C bases in the inserted loops were important factors for determining the dsRNA ratio. Based on this notion, we plotted the total GC numbers in loops of the inserted sequences against total RNA synthesis ( Figure 6A ) or the dsRNA ratio (Figure 6B ). Total RNA synthesis was almost independent, but the dsRNA ratio strongly correlated with the GC number in loops of the inserted sequences (correlation efficient is 0.81), indicating that the GC number in loops provides a useful index to predict the dsRNA ratio. According to this index, we obtained a simple rule that RNA must obey to replicate as a single strand: 'fewer (ideally <4) GC bases in loops. ' We also plotted other parameters (the sum of the loop size and the decrease of Gibbs's free energy if the inserted loop regions form a partial dsRNA) against the dsRNA ratio, but we observed weaker correlations than the GC number in loops (Supplementary Figure S1 ).