The Rodin-Ohno hypothesis that two enzyme superfamilies descended from one ancestral gene: an unlikely scenario for the origins of translation that will not be dismissed

Abstract

Background: Because amino acid activation is rate-limiting for uncatalyzed protein synthesis, it is a key puzzle in understanding the origin of the genetic code. Two unrelated classes (I and II) of contemporary aminoacyl-tRNA synthetases (aaRS) now translate the code. Observing that codons for the most highly conserved, Class I catalytic peptides, when read in the reverse direction, are very nearly anticodons for Class II defining catalytic peptides, Rodin and Ohno proposed that the two superfamilies descended from opposite strands of the same ancestral gene. This unusual hypothesis languished for a decade, perhaps because it appeared to be unfalsifiable.

Results: The proposed sense/antisense alignment makes important predictions. Fragments that align in antiparallel orientations, and contain the respective active sites, should catalyze the same two reactions catalyzed by contemporary synthetases. Recent experiments confirmed that prediction. Invariant cores from both classes, called Urzymes after Ur = primitive, authentic, plus enzyme and representing ~20% of the contemporary structures, can be expressed and exhibit high, proportionate rate accelerations for both amino-acid activation and tRNA acylation. A major fraction (60%) of the catalytic rate acceleration by contemporary synthetases resides in segments that align sense/antisense. Bioinformatic evidence for sense/antisense ancestry extends to codons specifying the invariant secondary and tertiary structures outside the active sites of the two synthetase classes. Peptides from a designed, 46-residue gene constrained by Rosetta to encode Class I and II ATP binding sites with fully complementary sequences both accelerate amino acid activation by ATP ~400 fold.

Conclusions: Biochemical and bioinformatic results substantially enhance the posterior probability that ancestors of the two synthetase classes arose from opposite strands of the same ancestral gene. The remarkable acceleration by short peptides of the rate-limiting step in uncatalyzed protein synthesis, together with the synergy of synthetase Urzymes and their cognate tRNAs, introduce a new paradigm for the origin of protein catalysts, emphasize the potential relevance of an operational RNA code embedded in the tRNA acceptor stems, and challenge the RNA-World hypothesis.

Figure 1

Aminoacyl-tRNA synthetases have three important functions in the cell. They use ATP to activate the alpha carboxyl group, making the amino acid more reactive (Activation). This activation step adds the adenosine moiety as an “affinity tag” to the amino acid, increasing its binding affinity by ~1000-fold (Retention) This slows the release of a very reactive species and enables subsequent steps that enhance the fidelity of the final step. Finally, they catalyze transfer of the activated carboxyl group to the 3’ CCA terminus of tRNA (Acylation), completing the translation of the genetic code by the covalent linkage to the tRNA anticodon that is interpreted by the 30S ribosomal subunit in response to codons in mRNA. The approximate rate accelerations achieved by contemporary enzymes indicated are based on comparisons of kcat/K_M values to the uncatalyzed rates, estimated from experimental rates of model reactions, as described in [38,39].

Figure 2

Detailed comparison between active sites of Class I and Class II aminoacyl-tRNA synthetases. Substrate binding sites for ATP and amino acid are buried in Class I, and much closer to the surface in Class II enzymes. The color of the active-site spheres illustrates that conserved active-site residues that motivated Rodin and Ohno to advance their sense/antisense coding hypothesis are drawn exclusively from the set of substrates activated by the other class. The median transfer free energy from water to cyclohexane is favorable for Class I substrates and unfavorable for Class II substrates.

Figure 3

Superposition of four Class I and four Class II aaRS. Specific enzymes are colored differently and labeled. Full-length contemporary monomers are shown as surfaces that are 65% transparent, to reveal the invariant cores, shown as cartoons inside the surfaces.

Figure 4

Partial order structure alignment of four Class I and Class II Urzymes. Despite the strong similarities between the four Urzymes from each class, POSA identifies appropriate sub-classification.

Figure 5

The Rodin-Ohno hypothesis holds that ancestral forms of Class I and II aminoacyl-tRNA synthetases (aaRS) had fully complementary nucleic acid coding sequences and that contemporary aaRS descended from opposite strands of a single gene. The schematic in A illustrates how this hypothesis leads directly to the concept of aaRS Urzymes. Antiparallel alignment of amino acid sequences for the Class-defining motifs (HIGH and KMSKS from Class I; Motifs 1 and 2 from Class II) reveals that neither anticodon-binding domain, nor a long insertion in each Class can physically be a part of such an ancestral gene. As a result, the only portions of the two Classes consistent with the hypothesis (B) are about 120-130 residues long. These fragments coincide with invariant tertiary structural cores shared by all superfamily members. Moreover, these segments together compose a minimal active site, containing binding sites for all three substrates (C). Amino acid and ATP determinants are reflected across the gene sequence, while tRNA binding determinants are related by two-fold rotation [40].

Figure 6

Quantitative framework in which to assess the catalytic significance of Urzymes and various other putative stages of aaRS evolution. A. Rate accelerations estimated from experimental data for single substrate (red) and bi-substrate (Black, Bold) reactions adapted from [75] to include uncatalyzed and catalyzed rates of bi-substrate reactions of the ribosome [74], amino acid activation [39] and kinases [106]. Second order rate constants (black bars) were converted into comparable units by multiplying by 0.002 M, which is the ATP concentration used to assay the catalysts shown in B. B. Experimental rate accelerations estimated from steady state kinetics as kcat/K_M for a series of catalysts derived from Class I and Class II aminoacyl-tRNA synthetases ([38,39] and data of V. Weinreb, L. Li, M. Collier, and K. Gonzalez-Rivera presented here in a subsequent section). Vertical scales in A and B are the same, and the origin of the histogram in B has been set equal to the uncatalyzed rate of amino acid activation in (AAact) in A. Red bars denote Class I Tryptophanyl- and Leucyl-tRNA synthetase constructs, blue bars denote Class II Histidyl-tRNA synthetase constructs, and green denotes a ribozymal catalyst [97], included for comparison. Research presented in A, B was originally published in [37]. © The American Society for Biochemistry and Molecular Biology. C. Class I LeuRS and Class II HisRS Urzyme amino acid specificities. Amino acids with more negative ΔGk_cat/K_M values indicate higher activities. By ~1 kcal/mole (light bands) or ~ five-fold, each Urzyme prefers substrates from the class to which it belongs (dark bands). Nonetheless, both activate a range of non-cognate amino acids, and are promiscuous.

Figure 7

tRNA^Trp Acylation by TrpRS Urzyme. A. Model of the putative interaction between TrpRS Urzyme and tRNA^Trp, derived from the crystal structure of the complex between human TrpRS and tRNA^Trp[57]. Autoradiograms documenting the acylation of tRNA^Trp by Wild Type TrpRS, TrpRS Urzyme, and two intermediate modular constructs, containing either CP1 or the anticodon-binding domain. B. Model for interaction of tRNA^His with HisRS4 Urzyme and autoradiograms showing acylation by HisRS1, 2, and 4 as in A. Spheres show Trp-5’AMP, and His-5’ sulfoamyladenylate. These data were published originally in [37] © The American Society for Biochemistry and Molecular Biology.

Figure 8

Quantitative analysis of the catalytic contributions of Motif 3 and a short N-terminal extension of the Motif 1 helix to the acceleration of histidine activation by HisRS Urzyme. Graphics include the Histidyl-5‘adenylate as spheres. Details of the constructions are described in [38].

Figure 9

Factorial analysis of TrpRS inter-domain effects. A. The catalytic activity of the Urzyme facilitates a full-factorial analysis of the benefits of adding either the CP1 or anticodon-binding domain, jointly with their synergistic effect in the full-length enzyme. B. Free energy histograms for the factorial design in A, showing nearly identical patterns in which the CP1 and ABD actually diminish both specificity for tryptophan vs tyrosine and acylation of tRNA^Trp. The entire difference between Urzyme and full-length enzyme is achieved only via the synergistic participation of both domains missing in the Urzyme.

Figure 10

Evidence for sense/antisense ancestry of the secondary structures connecting catalytic peptides in Class I and Class II aaRS. Frequency distributions of codon middle base-pairing in control (A,C) and antisense alignments of a 94-residue Urgene excerpted from ~200 TrpRS and ~200 HisRS contemporary coding sequences (B). Distributions under the RO hypothesis (B) have significantly higher mean values than do those for two samples exhibiting the Null hypothesis that predicts a pairing frequency of 0.25 (one base in four). D,E. Domain and evolutionary time-dependent estimates for codon middle-base pairing between antisense alignments of TrpRS and HisRS. D. Breakdown of the three consensus domains. The nine columns arise from comparing sequences for one synthetase another when broken down by domain. E. Codon middle-base pairing between reconstructed ancestral sequences derived from phylogenetic trees of TrpRS and HisRS Urgene sequences increases as the trees approach the root. Dotted line; all sequences, Solid line bacterial sequences only. (From Chandrasekaran, et al., Mol. Biol. Evol. 2013, 30:1588-1604).

Figure 11

A bona fide sense/antisense gene with amino acid activation activity expressed from both strands. A. Sequence and structural homology of the Class I HIGH signature and the F1 ATPase Walker A sequence. Glycine residues occur in exactly the same locations. The difference between the two sequences is that the Class I signature is specific for the α-phosphate group, whereas the Walker A signature is specific for the β-phosphate group. B. A designed sense/antisense gene coding for the 46-residue Class I aaRS and Class II ATP binding sites on opposite strands. C. Time course of leucine activation by the Class I 46mer. D. Time course of histidine activation by the Class II 46mer. Both plots have data for two concentrations of amino acid (10, 32 mM) and peptide (0.44, 1.75 μM). Scatter within each set of points is very noisy, so statistical analyses of various dependences are given in Table 3.

Figure 12

Illustrative use of Bayes’s Theorem to estimate the enhancement of posterior probability under the RO hypothesis given by the catalytic activities of the 46-residue ATP binding sites expressed from a designed, sense/antisense gene. All probability distribution functions are bivariate normal probability distributions. A. Prior probabilities centered at a mean rate acceleration of 1000 (green), and for the null hypothesis (brown) that the peptides are catalytically inactive. The likelihood function derived from experimental results obtained for Class I and II ATP binding sites (blue) is centered at a rate acceleration of 400-fold. Standard deviations of all three distributions in A are 0.4 log₁₀ units, corresponding to a 40% uncertainty. X and Y axes are logarithmic in the rate enhancement, and 1000-fold is drawn from the ratio of the uncatalyzed rates of activated amino acid assembly to form peptides to that for amino acid activation by ATP. B. Posterior probabilities obtained for the null prior (purple) and for the two active peptides (turquoise). The integrated experimental posterior probability under the RO hypothesis is larger than that under the Null hypothesis by ~10¹⁴. The latter posterior probability (brown) has been multiplied by a factor of 10¹⁴ to be visible.