The relation between frequencies of amino acids and ordered trinucleotides

The long and winding road of molecular data in phylogenetic analysis

The use of molecules and reactions as evidence, markers and/or traits for evolutionary processes has a history more than a century long. Molecules have been used in studies of intra-specific variation and studies of similarity among species that do not necessarily result in the analysis of phylogenetic relations. Promoters of the use of molecular data have sustained the need for quantification as the main argument to make use of them. Moreover, quantification has allowed intensive statistical analysis, as a condition and a product of increasing automation. All of these analyses are subject to the methodological anxiety characteristic of a community in search of objectivity (Suárez-Díaz and Anaya-Muñoz, Stud Hist Philos Biol Biomed Sci 39:451-458, 2008). It is in this context that scientists compared and evaluated protein and nucleic acid sequence data with other types of molecular data - including immunological, electrophoretic and hybridization data. This paper argues that by looking at long-term historical processes, such as the use of molecular evidence in evolutionary biology, we gain valuable insights into the history of science. In that sense, it accompanies a growing concern among historians for big-pictures of science that incorporate the fruitful historical research on local cases of the last decades.

An alternative look at code evolution: using non-canonical codes to evaluate adaptive and historic models for the origin of the genetic code

The canonical code has been shown many times to be highly robust against point mutations; that is, mutations that change a single nucleotide tend to result in similar amino acids more often than expected by chance. There are two major types of models for the origin of the code, which explain how this sophisticated structure evolved. Adaptive models state that the primitive code was specifically selected for error minimization, while historic models hypothesize that the robustness of the code is an artifact or by-product of the mechanism of code evolution. In this paper, we evaluated the levels of robustness in existing non-canonical codes as well as codes that differ in only one codon assignment from the standard code. We found that the level of robustness of many of these codes is comparable or better than that of the standard code. Although these results do not preclude an adaptive origin of the genetic code, they suggest that the code was not selected for minimizing the effects of point mutations.

Collecting, comparing, and computing sequences: the making of Margaret O. Dayhoff's Atlas of Protein Sequence and Structure, 1954-1965

Collecting, comparing, and computing molecular sequences are among the most prevalent practices in contemporary biological research. They represent a specific way of producing knowledge. This paper explores the historical development of these practices, focusing on the work of Margaret O. Dayhoff, Richard V. Eck, and Robert S. Ledley, who produced the first computer-based collection of protein sequences, published in book format in 1965 as the Atlas of Protein Sequence and Structure. While these practices are generally associated with the rise of molecular evolution in the 1960s, this paper shows that they grew out of research agendas from the previous decade, including the biochemical investigation of the relations between the structures and function of proteins and the theoretical attempt to decipher the genetic code. It also shows how computers became essential for the handling and analysis of sequence data. Finally, this paper reflects on the relationships between experimenting and collecting as two distinct "ways of knowing" that were essential for the transformation of the life sciences in the twentieth century.

History, objectivity, and the construction of molecular phylogenies

Despite the promises made by molecular evolutionists since the early 1960s that phylogenies would be readily reconstructed using molecular data, the construction of molecular phylogenies has both retained many methodological problems of the past and brought up new ones of considerable epistemic relevance. The field is driven not only by changes in knowledge about the processes of molecular evolution, but also by an ever-present methodological anxiety manifested in the constant search for an increased objectivity-or in its converse, the avoidance of subjectivity. This paper offers an exhaustive account of the methodological and conceptual difficulties embedded in each of the steps required to elaborate molecular phytogenies. The authors adopt a historical perspective on the field in order to follow the development of practices that seek to increase the objectivity of their methods and representations. These include the adoption and development of explicit criteria for evaluation of evidence, and of procedures associated with methods of statistical inference, quantification and automation. All these are linked to an increasing use of computers in research since the mid 1960s. We will show that the practices of objectivity described are highly dependent on the problems and tools of molecular phylogenetics.

Origin of the genetic code: first aminoacyl-tRNA synthetases could replace isofunctional ribozymes when only the second base of codons was established

Analysis of the updated compilation of more than 8,000 tRNA gene sequences confirmed our previously reported finding that in pairs of consensus tRNAs with complementary anticodons, their second bases in the acceptor stems are also complementary. This dual complementarity points to the following: (1) the operational code embodied in the acceptor stem, and the classic genetic code embodied in the anticodon could have had the same common ancestor; (2) new tRNAs most likely entered primitive translation in pairs with complementary anticodons; and (3) this process of code expansion was directed by the primordial double-strand coding. However, we did not find the dual complementarity when testing all tRNA pairs in which anticodons were complementary only at the central position, but not complementary at least at one of the flanking two positions. This observation, together with certain additional evidence, suggests that both codes were still being shaped (with only the second base established at the time) when the first protein aminoacyl-tRNA synthetases could have already started replacing their ribozymic precursors.

The standard genetic code enhances adaptive evolution of proteins

The standard genetic code, by which most organisms translate genetic material into protein metabolism, is non-randomly organized. The Error Minimization hypothesis interprets this non-randomness as an adaptation, proposing that natural selection produced a pattern of codon assignments that buffers genomes against the impact of mutations. Indeed, on the average any given point mutation has a lesser effect on the chemical properties of the utilized amino acid than expected by chance. Might it also, however, be the case that the non-random nature of the code effects the rate of adaptive evolution? To investigate this, here we develop population genetic simulations to test the rate of adaptive gene evolution under different genetic codes. We identify two independent properties of a genetic code that profoundly influence the speed of adaptive evolution. Noting that the standard genetic code exhibits both, we offer a new insight into the effects of the "error minimizing" code: such a code enhances the efficacy of adaptive sequence evolution.

Modulation of base-specific mutation and recombination rates enables functional adaptation within the context of the genetic code

The persistence of life requires populations to adapt at a rate commensurate with the dynamics of their environment. Successful populations that inhabit highly variable environments have evolved mechanisms to increase the likelihood of successful adaptation. We introduce a 64 x 64 matrix to quantify base-specific mutation potential, analyzing four different replicative systems, error-prone PCR, mouse antibodies, a nematode, and Drosophila. Mutational tendencies are correlated with the structural evolution of proteins. In systems under strong selective pressure, mutational biases are shown to favor the adaptive search of space, either by base mutation or by recombination. Such adaptability is discussed within the context of the genetic code at the levels of replication and codon usage.

Genetic codes as evolutionary filters: subtle differences in the structure of genetic codes result in significant differences in patterns of nucleotide substitution

The codon-degeneracy model (CDM) predicts that patterns of nucleotide substitution in protein-coding genes are largely determined by the relative frequencies of four-fold (4f), two-fold, and non-degenerate sites, the attributes of which are determined by the structure of the governing genetic code. The CDM thus further predicts that genetic codes with alternative structures will "filter" molecular evolution differentially. A method, therefore, is presented by which the CDM may be applied to the unique structure of any genetic code. The mathematical relationship between the proportion of transitions at 4f degenerate nucleotide sites and the transition-to-transversion ratio is described. Predictions for five individual genetic codes, relative to the relationship between code structure and expected patterns of nucleotide substitution, are clearly defined. To test this "filter" hypothesis of genetic codes, simulated DNA sequence data sets were generated with a variety of input parameter values to estimate the relationship between patterns of nucleotide substitution and best-fit estimates of transition bias at 4f degenerate sites for both the universal genetic code and the vertebrate mitochondrial genetic code. These analyses confirm the prediction of the CDM that, all else being equal, even small differences in the structure of alternative genetic codes may result in significant shifts in the overall pattern of nucleotide substitution.

The case for an error minimizing standard genetic code

Since discovering the pattern by which amino acids are assigned to codons within the standard genetic code, investigators have explored the idea that natural selection placed biochemically similar amino acids near to one another in coding space so as to minimize the impact of mutations and/or mistranslations. The analytical evidence to support this theory has grown in sophistication and strength over the years, and counterclaims questioning its plausibility and quantitative support have yet to transcend some significant weaknesses in their approach. These weaknesses are illustrated here by means of a simple simulation model for adaptive genetic code evolution. There remain ill explored facets of the 'error minimizing' code hypothesis, however, including the mechanism and pathway by which an adaptive pattern of codon assignments emerged, the extent to which natural selection created synonym redundancy, its role in shaping the amino acid and nucleotide languages, and even the correct interpretation of the adaptive codon assignment pattern: these represent fertile areas for future research.

No accident: genetic codes freeze in error-correcting patterns of the standard genetic code

The standard genetic code poses a challenge in understanding the evolution of information processing at a fundamental level of biological organization. Genetic codes are generally coadapted with, or 'frozen' by, the protein-coding genes that they translate, and so cannot easily change by natural selection. Yet the standard code has a significantly non-random pattern that corrects common errors in the transmission of information in protein-coding genes. Because of the freezing effect and for other reasons, this pattern has been proposed not to be due to selection but rather to be incidental to other evolutionary forces or even entirely accidental. We present results from a deterministic population genetic model of code-message coevolution. We explicitly represent the freezing effect of genes on genetic codes and the perturbative effect of changes in genetic codes on genes. We incorporate characteristic patterns of mutation and translational error, namely, transition bias and positional asymmetry, respectively. Repeated selection over small successive changes produces genetic codes that are substantially, but not optimally, error correcting. In particular, our model reproduces the error-correcting patterns of the standard genetic code. Aspects of our model and results may be applicable to the general problem of adaptation to error in other natural information-processing systems.

Regularities in mutational variability in selected protein families and the Markovian model of amino acid replacement

Three families of proteinase inhibitors and the trypsin family were the subjects of the analysis of amino acid replacements at aligned positions. This approach concerned some specific types of replacement and the mechanisms that can be involved in their control. The usefulness of the Markovian model for interpretation of mutational replacement within homologous proteins was examined. The same sequences were also analyzed with the use of the non-Markovian algorithm of genetic semihomology. This study leads to the conclusion that the Markovian model is not suitable for the interpretation of protein mutational variability since: (1) The information about the history of a variable unit is included in its genetic code. (2) This information plays an important role in the probability of further possible changes of the unit.

Load minimization of the genetic code: history does not explain the pattern

The average effect of errors acting on a genetic code (the change in amino-acid meaning resulting from point mutation and mistranslation) may be quantified as its 'load'. The natural genetic code shows a clear property of minimizing this load when compared against randomly generated variant codes. Two hypotheses may be considered to explain this property. First, it is possible that the natural code is the result of selection to minimize this load. Second, it is possible that the property is an historical artefact. It has previously been reported that amino acids that have been assigned to codons starting with the same base come from the same biosynthetic pathway. This probably reflects the manner in which the code evolved from a simpler code, and says more about the physicochemical mechanisms of code assembly than about selection. The apparent load minimization of the code may therefore follow as a consequence of the fact that the code could not have evolved any other way than to allow biochemically related amino acids to have related codons. Here then, we ask whether this 'historical' force alone can explain the efficiency of the natural code in minimizing the effects of error. We therefore compare the error-minimizing ability of the natural code with that of alternative codes which, rather than being a random selection, are restricted such that amino acids from the same biochemical pathway all share the same first base. We find that although on average the restricted set of codes show a slightly higher efficiency than random ones, the real code remains extremely efficient relative to this subset P = 0.0003. This indicates that for the most part historical features do not explain the load- minimization property of the natural code. The importance of selection is further supported by the finding that the natural code's efficiency improves relative to that of historically related codes after allowance is made for realistic mutational and mistranslational biases. Once mistranslational biases have been considered, fewer than four per 100,000 alternative codes are better than the natural code.

A new algorithm for analysis of the homology in protein primary structure

A new algorithm for analysis of the homology and genetic semihomology in protein sequence is described. It assumes the close relation between the compared amino acids and their codons in related proteins. The algorithm is based on the network of the genetic relationship between amino acids and, thus differs from the commonly used statistical matrices. The results obtained by using this method are more comprehensive than used at present, and reflect the actual mechanism of protein differentiation and evolution. They concern: (1) location of homologous and semihomologous sites in compared proteins; (2) precise estimation of insertion/deletion gaps in non-homologous fragments; (3) analysis of internal homology and semihomology; (4) precise location of domains in multidomain proteins; (5) estimation of genetic code of non-homologous fragments; (6) construction of genetic probes; (7) studies on differentiation processes among related proteins; (8) estimation of the degree of relationship among related proteins; (9) studies on the evolution mechanism within homologous protein families and (10) confirmation of actual relationship of sequences showing low degree of homology.

The beta-sheets of proteins, the biosynthetic relationships between amino acids, and the origin of the genetic code

Two forces are generally hypothesised as being responsible for conditioning the origin of the organization of the genetic code: the physicochemical properties of amino acids and their biosynthetic relationships (relationships between precursor and product amino acids). If we assume that the biosynthetic relationships between amino acids were fundamental in defining the genetic code, then it is reasonable to expect that the distribution of physicochemical properties among the amino acids in precursor-product relationships cannot be random but must, rather, be affected by some selective constraints imposed by the structure of primitive proteins. Analysis shows that measurements representing the 'size' of amino acids, e.g. bulkiness, are specifically associated to the pairs of amino acids in precurso-product relationships. However, the size of amino acids cannot have been selected per se but, rather, because it reflects the beta-sheets of proteins which are, therefore, identified as the main adaptive theme promoting the origin of genetic code organization. Whereas there are no traces of the alpha-helix in the genetic code table. The above considerations make it necessary to re-examine the relationship linking the hydrophilicity of the dinucleoside monophosphates of anticodons and the polarity and bulkiness of amino acids. It can be concluded that this relationship seems to be meaningful only between the hydrophilicity of anticodons and the polarity of amino acids. The latter relationship is supposed to have been operative on hairpin structures, ancestors of the tRNA molecule. Moreover, it is on these very structures that the biosynthetic links between precursor and product amino acids might have been achieved, and the interaction between the hydrophilicity of anticodons and the polarity of amino acids might have had a role in the concession of codons (anticodons) from precursors to products.

The phylogeny of tRNAs seems to confirm the predictions of the coevolution theory of the origin of the genetic code

An extensive analysis of the evolutionary relationships existing between transfer RNAs, performed using parsimony algorithms, is presented. After building up an estimate of the tRNA ancestral sequences, these sequences are then compared using certain methods. The results seem to suggest that the coevolution hypothesis (Wong, J.T., 1975, Proc. Natl. Acad. Sci. USA 72, 1909-1912) that sees the genetic code as a map of the biosynthetic relationships between amino acids is further supported by these results, as compared to the hypotheses that see the physicochemical properties of amino acids as the main adaptative theme that led to the structuring of the genetic code.

The phylogeny of tRNA molecules and the origin of the genetic code

The evolutionary relationships between transfer RNA (tRNA) molecules are analyzed by parsimony algorithms. The position of the topologies expected on the basis of the hypotheses made to explain the origin of the genetic code, on the frequency distribution of all the possible tree topologies of the evolutionary relationships between tRNAs seems to lead to the following conclusion: The hypothesis (Wong, J. T., Proc. Natl. Acad. Sci. USA, 1975, 72: 1909-1912) that sees the genetic code as a map of the biosynthetic relationships between amino acids seems to occupy a statistically significant position on these frequency distributions, thus reflecting a significant part of the tRNA phylogeny.

On concerted origin of transfer RNAs with complementary anticodons

Pairs of antiparallely oriented consensus tRNAs with complementary anticodons show surprisingly small numbers of mispairings within the 17-bp- long anticodon stem and loop region. Even smaller such complementary distances are shown by illegitimately complementary anticodons, i.e. those with allowed pairing between G and U bases. Accordingly, we suppose that transfer RNAs have emerged concertedly as complementary strands of primordial double helix-like RNA molecules. Replication of such molecules with illegitimately complementary anticodons might generate new synonymous codons for the same pair of amino acids. Logically, the idea of tRNA concerted origin dictates very ancient establishment of direct links between anticodons and the type of amino acids with which pre-tRNAs were to be charged. More specifically, anticodons (first of all, the 2nd base) could selectively target 'their' amino acids, reaction of acylating itself being performed by another non-specific site of pre-tRNA or even by another ribozyme. In all, the above findings and speculations are consistent to the hypercyclic concept (Eigen and Schuster, 1979), and throw new light on the genetic code origin and associated problems. Also favoring this idea are data on complementary codon usage patterns in different genomes.

A method to identify protein sequences that fold into a known three-dimensional structure

The inverse protein folding problem, the problem of finding which amino acid sequences fold into a known three-dimensional (3D) structure, can be effectively attacked by finding sequences that are most compatible with the environments of the residues in the 3D structure. The environments are described by: (i) the area of the residue buried in the protein and inaccessible to solvent; (ii) the fraction of side-chain area that is covered by polar atoms (O and N); and (iii) the local secondary structure. Examples of this 3D profile method are presented for four families of proteins: the globins, cyclic AMP (adenosine 3',5'-monophosphate) receptor-like proteins, the periplasmic binding proteins, and the actins. This method is able to detect the structural similarity of the actins and 70- kilodalton heat shock proteins, even though these protein families share no detectable sequence similarity.

Optimization and the genetic code

The present paper will focus on the relation between the structure of the table of the genetic code and the evolution of primitive organisms: it will be shown that the organization of the code table according to an optimization principle based on the notion of resistance to errors can provide a criterium for selection. The ordered aspect of the genetic code table makes this result a plausible starting point for studies of the origin and evolution of the genetic code: these could include, besides a more refined optimization principle at the logical level, some effects more directly related to the physico-chemical context, and the construction of realistic models incorporating both aspects.

On the appearance and role of a spacer group in the protein amino acids

Of the 20 protein amino acids, 16 have a methylene group at the beta position, and a further three bear a methine group. No aromatic, carboxamido, carboxylic carbon, or hetero atoms are attached directly to the alpha carbon, but they are separated by this methylene or occasionally by a longer n-alkylene spacer group. Therefore, the structure of the protein amino acids should rather be formulated as H2N-CH((CH2)n-R')-COOH instead of the generally accepted H2N-CH(R)-COOH. The appearance of and the role played by the spacer group are discussed in an evolutionary context. It is suggested that the spacer group appeared as a result of prebiotic selection, based on the relative abundance, racemization rate, and suitability for thermal polymerization of the protein amino acids and their homologs with various spacer group lengths. At the biotic level of evolution the requirements for ribosomal polymerization, as well as the abilities of polypeptides to maintain a stable and flexible three-dimensional structure and to bind ligands are considered and are proposed to have been responsible for the possible exclusion of longer spacer groups. It is concluded that the general role of the spacer group is to ensure the uniformity of the constant regions H2N-CH(-)-COOH and the individuality of the R' contact groups by spatially separating them.

Noise immunity of the genetic code

Error detection and correction properties are fundamental for informative codes. Hamming's distance allows us to study this noise resistance. We present codes characterized by the resistance optimization to nonsense mutational effects. The calculation of the cumulated Hamming's distance allowing to determine the number of optimal codes and their structure can be detailed. The principle of these laws of optimization of resistance consists of choosing constituent codons connected by mutational neighbouring in such a way that random application of mutations on such a code minimize the occurrence of nonsense n-uplets or terminators. New coding symmetries are then described and screened using Galois's polynomials properties and Baudot's code. Such a study can be applied to any length of the codons. Here we present the principles of this optimization for the most simple doublet codes. Another constraint is discussed: the distribution of optimal subcodes for synonymity and the frequencies of utilization of the different codons. We compare these results to those of the present genetic code, and we observe that all coded amino acids (except the particular case of SER) are using optimal sub-codes of synonymity. This work suggests that the appearance of the genetic code was provoked by mutations while optimizing on several levels its resistance to their effects. Thus genetic coding would have been the best automata that could be produced in prebiotic conditions.

Analysis of structural similarities between brain Thy-1 antigen and immunoglobulin domains. Evidence for an evolutionary relationship and a hypothesis for its functional significance

The Thy-1 membrane glycoprotein from rat brain is shown to have structural and sequence homologies with immunoglobulin (Ig) domains on the basis of the following evidence. 1. The two disulphide bonds of Thy-1 are both consistent with the Ig-fold. 2. The molecule contains extensive beta-structure as shown by the c.d. spectrum. 3. Secondary structure prediction locates beta-strands along the sequence in a manner consistent with the Ig-fold. 4. On the basis of rules derived from known beta-sheet structures, a three-dimensional structure with the Ig-fold is predicted as favourable for Thy-1. 5. Sequences in the proposed beta-strands of Thy-1 and known beta-strands of Ig domains show significant sequence homology. This homology is statistically more significant than for the comparison of proposed beta-strand sequences of beta 2-microglobulin with Ig domains. An hypothesis is presented for the possible functional significance of an evolutionary relationship between Thy-1 and Ig. It is suggested that both Thy-1 and Ig evolved from primitive molecules, with an Ig fold, which mediated cell--cell interactions. The present-day role of Thy-1 may be similar to that of the primitive domain.

beta 2-Microglobulin: methods and clinical applications

beta 2-Microglobulin is a low molecular weight protein that is found in most biological fluids. It was originally isolated from urine of cadmium-poisoned patients. Its amino acid sequence was established and shown to be structurally related to immunoglobulin constant domains. With the aid of antibodies specific against beta 2-microglobulin, the protein was detected on the membranes of all nucleated cells, normal and neoplastic. Measuring the quantity of beta 2-microglobulin showed that high levels are present in patients with renal tubular deficiencies and several other pathological conditions including neoplastic diseases. Extremely high levels were detected in seminal fluid and colostrum. Despite the structural relationship to immunoglobulins, no immunological relationship was demonstrated with these proteins using antibodies specific for beta 2-microglobulin. However, such antibodies are cytotoxic to all cells carrying beta 2-microglobulin on their surfaces. The discovery that beta 2-microglobulin is an integral part of the histocompatibility antigens of human and murine origin stimulated further research and interest in this molecule. Several groups of investigators have shown that beta 2-microglobulin is the low molecular weight chain and is noncovalently bound to a high molecular weight chain which carries the histocompatibility antigens. The structure of the histocompatibility antigens of lymphocytes (HLA) was shown by immunochemical as well as biological methods, and it is now well accepted. The antibodies against beta 2-microglobulin are extremely useful in the isolation of the histocompatibility antigens for sequence studies. Furthermore, the antibody to beta 2-microglobulin revealed that other structures may be bound to beta 2-microglobulin such as phytohemoagglutimin (PHA) receptors, mixed lymphocyte culture (MLC) antigens, etc. Murine thymus leukemia (TL) antigen also contains beta 2-microglobulin as an integral part of its structure; other tumor antigens may have a similar structure. Through all these studies, beta 2-microglobulin emerged as the best known membrane protein that can serve as a model for study of the arrangement and the function of the cell membrane.

Linear and inverted repetitions in protein sequences

An extensive search for internal regularities in amino acid sequences has been made, using both the genetic code and the relative frequencies of amino acid alternatives in homologous proteins. The two methods give very similar results and strongly suggest the occurrence of significant linear and inverted repetitions (similar sequences of opposite polarity) in several proteins. A hypothesis is developed to explain the occurrence of such internal regularities in proteins. This hypothesis is based on a process of duplication of an ancestral loop in which a symmetrical arrangement of amino acid allows stabilization by interaction between the amino acid side chains.

Initiation of protein synthesis at an unusual position in an immunoglobulin gene?

The amino acid sequence of urinary beta(2)-microglobulin has been partially determined and found to be related to the constant region of IgG immunoglobulin heavy chain. beta(2)-Microglobulin is present in normal individuals. Its gene may have evolved from an immunoglobulin gene by the use of an unusually located start signal for initiating synthesis of the polypeptide.