Fig 1.
Phylogenetic comparison of tandem riboswitches across Bacillaceae and Vibrionaceae.
A) 48 Bacillaceae and 37 Vibrionaceae tandem riboswitches were clustered based on aptamer sequence and structure across both aptamers of the riboswitch. After phylogenetic clustering, individual aptamers were colored based on the class of gene being regulated and the bacterial family of origin (Vibrionaceae TP are orange, Bacillaceae TP are purple, Bacillaceae GCV are green). Clusters have been labeled with the bacterial family and gene class being regulated. Bootstrap support values are displayed for 100 replicates when > = 70. B,C) Phylogenetic clustering of 48 tandem riboswitches, separated into aptamer-1 (B) and aptamer-2 (C), taken from the Bacillaceae family and colored according to class of gene being regulated (GCV are green, TP are purple). All trees are midpoint rooted.
Fig 2.
Clustering of tandem riboswitch aptamers across Bacillaceae and Vibrionaceae.
A) Bacillaceae and Vibrionaceae tandem riboswitch aptamers were clustered using RNAmountAlign as a distance metric (threshold of 5). All represented Bacillaceae riboswitches regulate GCV (top), while Vibrionaceae riboswitches regulate TP (bottom). Aptamers are colored based on aptamer type, purple for aptamer-1 and green for aptamer-2. B) Network density was calculated for each aptamer in both networks across a range of RNAmountAlign thresholds. Dotted red line indicates the RNAmountAlign threshold (5) at which the networks in A were visualized. C) Box blots represent all pairwise edge-weights within each aptamer type. **** p-value < 2x10-16.
Fig 3.
Clustering of glycine riboswitch aptamers identified within the Bacilli class of bacteria.
A) Aptamers within the Bacilli bacterial class were identified and clustered based on RNAmountAlign pairwise similarity (visualized at threshold of 12). B) Sub-clusters (communities) were identified using the four community detection functions within R’s igraph package. Two communities were identified that contain two different aptamer types: aptamer-1 and singlet type-1, and aptamer-2 and singlet type-2 that regulate GCV and TP respectively. Network shows visualization of the community detection algorithm cluster_fast_greedy (as implemented by R). Node colors correspond to distinct clusters detected. C) The two sub-clusters containing different aptamer types were parsed from the overall network, the tandem aptamers’ partners were added to the set (as an out group within the same context), and graph clustering was visualized (RNAmountAlign threshold of 5). D) Edge density between aptamer groups was calculated for networks generated across a range of RNAmountAlign edge-weight thresholds. Dotted red line indicates the RNAmountAlign threshold (5) at which the networks in (C) were visualized.
Fig 4.
Consensus structures of Bacilli riboswitches within a given genomic context display conservation between tandem and singleton aptamers.
Consensus secondary structure of the singleton and tandem riboswitches delineated by the genomic context. Conservation and covariation of base pairing generated using R2R with the individual covariance models. Tandem (A) and singleton (B) riboswitches regulating GCV. Tandem (C) and singleton (D) riboswitches regulating TP.
Fig 5.
Clustering of random glycine riboswitch aptamers across the bacterial kingdom.
A) Network visualization of 150 randomly chosen aptamers regulating GCV and clustered based on RNAmountAlign pairwise similarity (threshold -5). B) Network visualization of 150 randomly chosen aptamers regulating TP and clustered based on RNAmountAlign pairwise similarity (threshold -5). Only singletons that could be classified as type-1 or type-2 were included in this set.
Fig 6.
Model of glycine riboswitch evolution.
Model proposed for the evolution and divergence of the glycine riboswitch. In this model a progenitor tandem riboswitch conserves one of the tandem aptamers based on the genomic context of the riboswitch, while the other slowly degrades down to the minimalistic components required for tertiary interaction to drive gene regulation. In this way, tandem glycine riboswitches may degrade into functional singleton tandem riboswitches.