Figure 3. Inter-segment interactions between ssRNAs S11 and S1-S10, identified via RNA-RNA SELEX.

SELEX-enriched sequences were aligned against S1-S10 genomic segments using probabilistic score-based filtering (Materials and methods) to construct the resulting histogram plots. Each histogram peak corresponds to a sequence with a potential to interact with the S11 RNA alone (shown as black peaks), or in the presence of NSP2 (peaks shown in red). These histograms are plotted in the order of decreasing CCF amplitudes (see Figure 1B). Each plot is scaled to the size of the longest segment S1 (3302 nts), and the 3’-end of each RNA segment is marked with a vertical dotted line. Sequences that can stably base-pair with S11 RNA in the presence of NSP2 are highlighted (magenta boxes). A number of genomic sequences identified in the RNA-RNA SELEX experiments may be sequestered by stable local secondary structures (peaks identified by cyan boxes). Insets – RNA-RNA interaction sites between RNA S11 (bottom strand shown in red), and its interacting partners S3, S6, S5 and S9 (top strand, shown in black).

Figure 3—figure supplement 1. Sequence alignments of SELEX-enriched RNAs against a non-rotavirus control RNA (Non-RV).

(A) Multiple RNAs interacting with S11 enriched in the presence of NSP2 were aligned to Non-RV RNA sequence using a Bernoulli score of 12 (blue line) or 14 (red line). Only a small number of sequences (~170) with a score of 14 or above align to the Non-RV sequence, suggesting that this region could potentially interact (albeit weakly) with S11 RNA. This result is consistent with a non-zero CCF amplitude, observed for S11:Non-RV control RNA, incubated in the presence of NSP2 (Figure 4A, CCF shown as a dashed magenta line). (B) A correlation between the CCF amplitudes of pairwise RNA-RNA interactions examined by FCCS (Figure 1B) and the number of ‘filtered’ sequence alignments (using a Bernoulli score of 14 or above over 200 hits per identified peak) from the RNA-RNA SELEX against S11 RNA. At the 0.05 level, the slope is significantly different from zero, with a Pearson’s r = 0.73.

Figure 3—figure supplement 2. RNA sequences identified in the RNA-RNA SELEX experiments, with a potential to interact with S11 RNA.

RNA strands shown in green correspond to the genomic sequences represented by the peaks shown in Figure 3, identified using the Bernoulli score of 14 or above. (A) RV genomic sequences, involved in S11:S3 inter-segment interactions, confirmed by mutagenesis studies and FCCS (see Figure 4—figure supplement 1A). (B) An additional peak identified in S3 RNA that may also pair with S11, albeit with a significantly lower stability of the resulting RNA-RNA contact (−11.1 kcal/mol, corresponding to a K_d value of ~10 nM). (C–G) RNA sequences present in S6, S4, S1 and S10 RNAs (strands shown in green) with a potential to base-pair with S11 RNA. These sequences are expected to be sequestered by secondary structures and/or tertiary contacts, which may preclude these interactions with S11 RNA (see Figure 3—figure supplement 4).

Figure 3—figure supplement 3. Accessibility of the region 272–287 of S10 RNA probed using a complementary RNA sequence, derived from S11 RNA.

(A) ATTO647-labelled S10 RNA was incubated with the Cy3-labelled 17-mer RNA Seq11_84–100 (Supplementary file 2) in the presence of NSP2, as described in Materials and methods. No significant interactions between the two RNAs were observed (CCF = 0, in blue), suggesting that the probed region 272–287 is sequestered. (B) A control heat-annealing reaction of the two labelled RNA samples, resulting in a non-zero cross-correlation (CCF ~0.2), confirming that the sequence 84–100 of S11 RNA can base-pair with S10 RNA, only after heat-annealing of a mixture of the two RNAs. (C) Heat-annealing of labelled RNAs S10 and S11 results in similar CCF amplitude as in (B), suggesting that region 272–287 of S10 RNA has low accessibility for base-pairing with S11 RNA in the presence of NSP2.

Figure 3—figure supplement 4. The minimum free energy (MFE) secondary structure calculations of the complementary sequences with a potential for stable base-pairing with S11 RNA.

Thermodynamic ensemble of secondary structures and base-pairing probabilities were calculated for the entire genomic segment sequences using RNAfold, as described in Materials and methods. These RNA sequences were extracted from the computed centroid RNA structures, that is, showing minimal base pair distances to all other secondary structures in the Boltzmann ensemble. The identified S11 RNA-interacting sequences within RNAs S3 and S5 have high predicted accessibilities (regions 1415–1427 in S3 and 307–325 in S5), in agreement with the observed high CCF amplitudes for RNA combinations S3:S11 and S5:S11. Regions 1124–1136 in S6 RNA and 495–507 in S9 have lower accessibilities, while the region 1243–1260 in S6 is sequestered in a stable intra-molecular helix, consistent with the S6 RNA model proposed by Li et al. (2010), and thus is unlikely to stably base-pair with S11 RNA. RNA sequences of S4, S1 and S10 segments, identified as peaks in the RNA-RNA SELEX, showing low accessibilities of the S11-interacting sequences. The colour map shows the calculated base-pairing probabilities of the analysed nucleotides (blue = 0, red = 1).