Figure 1.
The flow diagram in the left describes the phylogenetic reconstruction of trees of rRNA molecules and substructures. The structures of rRNA molecules were first decomposed into substructures, including helical stem tracts and unpaired regions. Structural features of these substructures (e.g., length) were coded as phylogenetic characters and assigned character states according to an evolutionary model that polarizes character transformation towards an increase in molecular order (character argumentation). Coded characters (s) are arranged in data matrices, which can be transposed for cladistic analyses. Phylogenetic analysis using MP methods generate rooted phylogenetic trees of either molecules or substructures. Only trees of substructures are presented in this study. The flow diagram in the right shows the reconstruction of trees of proteomes and trees of protein domain structures. A census of domain structures in proteomes of hundreds of completely sequenced organisms is used to compose a data matrix and its transposed matrix, which are then used to build phylogenomic trees describing the evolution of individual protein structures and entire molecular repertoires, respectively. Elements of the matrix (g) represent genomic abundances of architectures (at FSF level of hierarchical classification of structure) in proteomes. Trees of proteomes will be described elsewhere, but are largely congruent with traditional classification. Embedded in the tree of rRNA substructures and tree of protein domains are timelines that assign age to molecular structures. These ages can be “painted” onto 2D or 3D structural models of the ribosome, generating evolutionary heat maps. Evolutionary information from RNA and protein structures is finally combined to generate a model of structural evolution.
Figure 2.
A strict consensus of 6 most-parsimonious trees (33,876 steps; CI = 0.168615, RI = 0.710934; HI = 0.831385; g1 = −1.425648) retained after a heuristic search with TBR branch swapping and simple addition sequence is colored according to relative age (nd) of extant (labeled taxa) or evolving (nodes) helical elements of structure. A total of 92 informative characters representing the structure of SSU and LSU rRNA in 93 organisms from the three superkingdoms were combined and analyzed. Bootstrap support (BS) values >50% are shown for individual nodes. Top and middle panels show evolutionary heat maps of Thermus thermophillus rRNA SSU and LSU rRNA secondary and crystal (2WDK and 2WDL) structures, respectively, with helices colored according to their age (nd). The lowest panel shows a primordial processivity core highlighted within the 70S ribosomal ensemble. Functional centers are highlighted in tree and heat maps.
Figure 3.
Timeline of development of the functional centers of the ribosome.
A, The relative age (nd) of different rRNA helices (colored circles) increases from left to right and SSU and LSU functional elements are indicated with squares and rhomboids, respectively. Pie charts below each time point show the percentage of SSU and LSU helices appearing at that time, and the two periods of evolutionary transition are shaded. B, Timeline of structures in bridges. The age of bridge interactions is assigned as the age of first acceptor element of the donor-acceptor pair forming the bridge (red lines). C, Timeline of helices that interact with the different arms of tRNA. D, Timelines of helices that form the functional centers of the ribosome. The PTC is highlighted with a red box. E, History of functions. The width of the arrows portrays the increase of elements forming the center and time taken for its development. F, Timeline of A-minor interactions in SSU and LSU rRNA. Names with capital letters indicate the donor and in small case indicate the acceptor of the A-minor interaction.
Figure 4.
Evolutionary accretion of molecular structures and establishment of A-minor interactions.
A. Cumulative plots describing ribosomal accretion of rRNA helices and r-proteins in the evolutionary timeline. Timelines at the top show the first appearance of individual structural-domains in rRNA subunits. Periods of evolutionary transition are shaded in grey. Note the rapid increase of structural complexity after the first transition, where processivity and peptide synthesis came together. B. Accumulation of A-minor interactions associated with individual rRNA subunits in ribosomal history. Plots describe the cumulative number of A-minor interactions as function of ribosomal age, as interactions accumulate in the evolutionary timeline of rRNA structure. The rapid increase in the number of A-minor interactions after the first transition, where processivity and peptide synthesis came together.
Figure 5.
Relative age of r-proteins and their interaction with rRNA helices.
A, Backbone of universal tree describing the evolution of 1,730 FSF domain structures from 749 genomes (541,383 steps; CI = 0.028, RI = 0.783; g1 = −0.111). The Venn diagram shows occurrence of FSFs in the three superkingdoms. B, rRNA helices establishing contacts with universal r-proteins. The relative age of the rRNA helices (nd) increases from left to right and r-proteins are ordered by age (from bottom to top) with corresponding ndP value. The number of nucleotides at each time point involved in RNA-protein interactions is proportional to the size of squares (SSU) and rhomboids (LSU). r-proteins contacts are colored according to the age of the helix that makes the most ancient contact or is inferred from Figure S2. C, Evolutionary heat map of SSU r-proteins. D, Evolutionary heat map of LSU r-proteins. The 3D structures show the relative age of the rRNA helices and the relative age of r-proteins interacting with them.
Figure 6.
Similarity of ancestral rRNA structures to in vitro evolved ribozymes.
A, Models of secondary and tertiary structure of L1 RNA ligase, RNA polymerase, and aminoacyl-tRNA synthetase (AARS) ribozymes. The long helix (stem A) of the 3-stemmed L1 RNA ligase molecule harbors the catalytic site and the junction of the three helical regions P1–P2, P4–P5 and P6–P7 at the center of the tripod-like RNA polymerase structure is the catalytic center. B, Alignment scores (top panels) and Z-score tests of statistical significance (bottom panels) for individual alignments of L1 ligase and rRNA helices of different age. Z-scores were derived from the alignment of 1,000 randomized sequences. Alignment scores of structures with Z-scores over 3 (horizontal dashed line) are significant at 0.01% confidence levels and are colored in red. C, Structural make up (pie charts) and frequency (bars) of rRNA helices of different age sharing structural features with the ribozymes. Only helices associated with functional centers (green pies) are labeled.
Figure 7.
A chronological representation of the evolution of the ribosome shows that very early in ribosomal evolution (nd<0.3) rRNA helices interacted with r-proteins to form a processivity core that mediated nucleotide interactions, which later (nd = 0.3) served as center for coordinated and balanced RNP accretion leading to modern ribosomal function. The purple structure indicates extant mRNA, which is used as structural reference for location of primitive functional centers. We envision the primordial ribosome had replicative functions that likely involved RNA, so the mRNA molecule from the crystallographic model should be regarded as placeholder for the ancient coding molecule. rRNA is rendered as ribbon representation, mRNA and proteins as rendered as space-filling representations.