Evolution of the mitochondrial genetic code I. Origin of AGR serine and stop codons in metazoan mitochondria

Summary AGA and AGG (AGR) are arginine codons in the universal genetic code. These codons are read as serine or are used as stop codons in metazoan mitochondria. The arginine residues coded by AGR in yeast orTrypanosoma are coded by arginine CGN throughout metazoan mitochondria. AGR serine sites in metazoan mitochondria are occupied mainly in corresponding sites in yeast orTrypanosoma mitochondria by UCN serine, AGY serine, or codons for amino acids other than serine or arginine. Based on these observations, we propose the following evolutionary events. AGR codons became unassigned because of deletion of tRNA Arg (UCU) and elimination of AGR codons by conversion to CGN arginine codons. Upon acquisition by serine tRNA of pairing ability with AGR codons, some codons for amino acids other than arginine mutated to AGR, and were caputed by anticodon GCU in serine tRNA. During vertebrate mitochondrial evolution, AGR stop codons presumably were created from UAG stop by deletion of the first nucleotide U and by use of R as the third nucleotide that had existed next to the ancestral UAG stop.     

On the possible origin and evolution of the genetic code

The genetic code is examined for indications of possible preceding codes that existed during early evolution. Eight of the 20 amino acids are coded by ‘quartets’ of codons with four-fold degeneracy, and 16 such quartets can exist, so that an earlier code could have provided for 15 or 16 amino acids, rather than 20. If two-fold degeneracy is postulated for the first position of the codon, there could have been 10 amino acids in the code. It is speculated that these may have been phenylalanine, valine, proline, alanine, histidine, glutamine, glutamic acid, aspartic acid, cysteine and glycine. There is a notable deficiency of arginine in proteins, despite the fact that it has six codons. Simultaneously, there is more lysine in proteins than would be expected from its two codons, if the four bases in mRNA are equiprobable and are arranged randomly. It is speculated that arginine is an ‘intruder’ into the genetic code, and that it may have displaced another amino acid such as ornithine, or may even have displaced lysine from some of its previous codon assignments. As a result, natural selection has favored lysine against the fact that it has only two codons. The introduction of tRNA into protein synthesis may have been a cataclysmic and comparatively sudden event, since duplication of tRNA takes place readily, and point mutations could rapidly differentiate members of the family of duplicates from each. Two tRNAs for different amino acids may have a common ancestor that existed more recently than the separation of the prokaryotes and eukaryotes. This is shown by homology of twoE. coli tRNAs for glycine and valine, and two yeast tRNAs for arginine and lysine.     

The "intruder" hypothesis and selection against arginine

The “intruder” hypothesis for the presence of more arginine codons in the genetic code than are needed leads to the expectation for selection against arginine in protein synthesis. This selection is therefore a consequence of the intruder hypothesis rather than a substitute for it, as claimed by Wallis (1974).     

RNA-ligand chemistry: a testable source for the genetic code

In the genetic code, triplet codons and amino acids can be shown to be related by chemical principles. Such chemical regularities could be created either during the code's origin or during later evolution. One such chemical principle can now be shown experimentally. Natural or particularly selected RNA binding sites for at least three disparate amino acids (arginine, isoleucine, and tyrosine) are enriched in codons for the cognate amino acid. Currently, in 517 total nucleotides, binding sites contain 2.4-fold more codon sequences than surrounding nucleotides. The aggregate probability of this enrichment is 10(-7) to 10(-8), had codons and binding site sequences been independent. Thus, at least some primordial coding assignments appear to have exploited triplets from amino acid binding sites as codons.     

线粒体遗传密码及基因组遗传密码的对称分析

病毒、细菌和真核生物的氨基酸编码都使用相同的遗传密码,表明它们可能有共同的来源。但人和牛的线粒体的遗传密码和基因组的遗传密码相比,出现以下不同;（1）ATA编码甲硫氮酸M而不是异亮氨酸I。（2）TGA不再是终止密码子X而编码色氨酸W。（3）AGA和AGG不再是精氨酸R的密码子而变为终止密码子X。应用高维空间拓扑分析的方法,对线粒体遗传密码和基因组遗传密码的6维编码空间进行对称性分析,得到如下结果：（1）线粒体遗传密码的起始密码子是2个而不是1个。（2）线粒体遗传密码的终止密码子是4个而不是3个。（3）线粒体遗传密码空间只有2、4、6三种偶数简并度而没1、3两种奇数简并度,表明其对称度较高。（4）线粒体遗传密码空间除丝氨酸S分成两个平行的子空间之外,终止密码子X亦分成两个平行的子空间,表明其连通度较低。（5）线粒体遗传密码一基因组遗传密码相比,共有3个简并平面出现变异,即：1001λλ（M和I）,011λ1λ（W和X）,以及1011λλ（S和X或S和R）。（6）基因组遗传密码的1、3两种奇数简并度可能来源于线粒体遗传密码的1001λλ平面和011λ1λ平面的对称性破缺。对线粒体遗传密码变异的生物学意义及遗传密码的起源进行了分析和讨论。     

遗传密码子研究进展

作为生命信息的基本遗传单位,基因组遗传密码的破译对于人们加深对生命本质的认识具有重要的理论价值和现实意义。目前,遗传密码子的研究重心已由遗传密码子的破译及反常密码子的发现转入到遗传密码子的起源与进化及扩张等研究。遗传密码子的起源与进化是当今基因组学研究的热点命题之一,相关的学说、假设层出不穷,但尚未取得实质性突破。另一方面,无义密码子的再定义及遗传密码的扩张等研究却极大的丰富和发展了遗传密码子的科学内涵,推动了生命科学研究的发展。文章综述了遗传密码子的多态性、起源与进化、无义密码子的再定义及遗传密码的扩张等方面的研究进展,并就其应用价值作了评述,期待为其在基因组学、医学等相关领域的应用研究提供参考。     

Evolution of the Genetic Code by Incorporation of Amino Acids that Improved or Changed Protein Function

Fifty years have passed since the genetic code was deciphered, but how the genetic code came into being has not been satisfactorily addressed. It is now widely accepted that the earliest genetic code did not encode all 20 amino acids found in the universal genetic code as some amino acids have complex biosynthetic pathways and likely were not available from the environment. Therefore, the genetic code evolved as pathways for synthesis of new amino acids became available. One hypothesis proposes that early in the evolution of the genetic code four amino acids—valine, alanine, aspartic acid, and glycine—were coded by GNC codons (N = any base) with the remaining codons being nonsense codons. The other sixteen amino acids were subsequently added to the genetic code by changing nonsense codons into sense codons for these amino acids. Improvement in protein function is presumed to be the driving force behind the evolution of the code, but how improved function was achieved by adding amino acids has not been examined. Based on an analysis of amino acid function in proteins, an evolutionary mechanism for expansion of the genetic code is described in which individual coded amino acids were replaced by new amino acids that used nonsense codons differing by one base change from the sense codons previously used. The improved or altered protein function afforded by the changes in amino acid function provided the selective advantage underlying the expansion of the genetic code. Analysis of amino acid properties and functions explains why amino acids are found in their respective positions in the genetic code.     

Recent emergence of the modern genetic code: a proposal

This article proposes that the genetic code was not fully formed before the divergence of life into three kingdoms. Rather, at least arginine and tryptophan evolved after the diversification of archaea, bacteria and eukaryotes, and were spread by horizontal gene transfer. Evidence for this hypothesis is based on data suggesting that enzymes for biosynthesis of arginine and tryptophan, and for arginine tRNA ligase, have shorter divergence times than the underlying lineages. Also, many of these genes display "star" phylogenies. This proposal is an extension of the idea that the genetic code was unified because of the evolutionary pressure from horizontal gene transfer. These considerations further undermine the need to postulate the existence of a "last common ancestor"; a simpler model would be that multiple lineages gave rise to life today.     

Possible evolutionary steps in the genetic code

In mammalian mitochondrial codes, fourfold codons wobble-pair with UNN anticodons so that U wobbles with U, C, A and G. Twofold pyrimidine-terminated codons pair with GNN and twofold purine-terminated codons pair with UNN. These properties enable a prediction to be made for evolution of the universal genetic code. It was postulated (1) that an archetypal code of 16 quartets coded for 15 amino acids. If this code used UNN anticodons, then duplication of tRNA genes, followed by mutations in the anticodons and aminoacylation sites, would give rise to the present universal code.     

An RNA-amino acid complex and the origin of the genetic code

The group I RNAs, of which the Tetrahymena ribosomal RNA intron is the most investigated example, catalyze their own splicing reactions. Splicing is initiated at a conserved site on the RNA that facilitates attack by exogenous guanosine (or its nucleotides) on the exon-intron junction. The guanosine site in the RNA's catalytic center also binds arginine, and is quite selective for the arginine side chain. This amino acid-RNA interaction is stereoselective, and L-arginine is preferred. Immediately at the site at which arginine binds there is one of only four RNA triplets in 92 group I RNA sequences: AGA/G and CGA/G. Thus the arginine contact site is within any of four different codons for arginine. Mutation of the conserved G in the middle of the triplet decreases affinity for the amino acid, showing that binding is sequence-specific. A pathway for the origin of the genetic code for arginine is suggested, based on the existence and properties of this sequence-specific, amino acid-specific RNA complex. The existence of a proto-ribosome related to the group I RNAs seems the most likely hypothesis. This notion is used to distinguish three periods in the development of the code. Restrained and exuberant hypotheses about the origin of the genetic code are distinguished, and some objections to these hypotheses are considered.     

Evolution of the aminoacyl-tRNA synthetases and the origin of the genetic code

The aminoacyl-tRNA synthetases exist as two enzyme families which were apparently generated by divergent evolution from two primordial synthetases. The two classes of enzymes exhibit intriguing familial relationships, in that they are distributed nonrandomly within the codon-amino acid matrix of the genetic code. For example, all XCX codons code for amino acids handled by class II synthetases, and all but one of the XUX codons code for amino acids handled by class I synthetases. One interpretation of these patterns is that the synthetases coevolved with the genetic code. The more likely explanation, however, is that the synthetases evolved in the context of an already-established genetic code—a code which developed earlier in an RNA world. The rules which governed the development of the genetic code, and led to certain patterns in the coding catalog between codons and amino acids, would also have governed the subsequent evolution of the synthetases in the context of a fixed code, leading to patterns in synthetase distribution such as those observed. These rules are (1) conservative evolution of amino acid and adapter binding sites and (2) minimization of the disruptive effects on protein structure caused by codon meaning changes.     

A hardware interpretation of the evolution of the genetic code

A quantitative rationale for the evolution of the genetic code is developed considering the principle of minimal hardware. This principle defines an optimal code as one that minimizes for a given amount of information encoded, the product of the number of physical devices used by the average complexity of each device. By identifying the number of different amino acids, number of nucleotide positions per codon and number of base types that can occupy each such position with, respectively, the amount of information, number of devices and the complexity, we show that optimal codes occur for 3, 7 and 20 amino acids with codons having a single, two and three base positions per codon, respectively. The advantage of a code of exactly 4 symbols is deduced, as well as a plausible evolutionary pathway from a code of doublets to triplets. The present day code of 20 amino acids encoded by 64 codons is shown to be the most optimal in an absolute sense. Using a tetraplet code further evolution to a code in which there would be 55 amino acids is in principle possible, but such a code would deviate slightly more than the present day code from the minimal hardware configuration. The change from a triplet code to a tetraplet code would occur at about 32 amino acids. Our conclusions are independent of, but consistent with, the observed physico-chemical properties of the amino acids and codon structures. These correlations could have evolved within the constrains imposed by the minimal hardware principle.     

Point counter point

Summary  The proposal by Schultz and Yarus is that changes in the genetic code result from ambiguous reading of codons. This is a simplistic catchall scheme. Our codon capture hypothesis was accompanied by case studies of each incident; for example, AAA changing to asparagine from lysine was preceded by all AAA lysine codons mutating to AAG under GC pressure, with disappearance of lysine anticodon UUU, followed by appearance of a new anticodon IUU for asparagine which would wobble-pair with AAU, AAC, and AAA (Ohama et al. 1990a). Ambiguous coding would not confine itself to changes in the genetic code to accommodate the proposal by Schultz and Yarus, but would extend throughout the genome if their idea is correct. Under these circumstances, much impairment of the accuracy in codon reading that is needed for maintenance of the constant sequences of amino acids in proteins would occur. Surely the net effect would be deleterious. Our conclusion is that the proposal by Schultz and Yarus is a “simple and easy answer to a complex and difficult problem,” and is not acceptable.     

Speculations on the evolution of the genetic code IV the evolution of the aminoacyl-tRNA synthetases

An evolutionary scheme is postulated in which a primitive code, involving only guanine and cytosine, would code for glycine(GG.), alanine(GC), arginine(CG.) and proline(CC). There evolves from this primitive code families of related amino acids as the code expands. The evolution of the aminiacyl-tRNA synthetases are considered to be indicators for the evolution of the genetic code. The postulated model for the evolution of the genetic code is used to give an evolutionary interpretation to the recent work on the structure and sequences of the aminoacyl-tRNA synthetases.     

Genetic Code Evolution Started with the Incorporation of Glycine,Followed by Other Small Hydrophilic Amino Acids

We propose that glycine was the first amino acid to be incorporated into the genetic code, followed by serine, aspartic and/or glutamic acid—small hydrophilic amino acids that all have codons in the bottom right-hand corner of the standard genetic code table. Because primordial ribosomal synthesis is presumed to have been rudimentary, this stage would have been characterized by the synthesis of short, water-soluble peptides, the first of which would have comprised polyglycine. Evolution of the code is proposed to have occurred by the duplication and mutation of tRNA sequences, which produced a radiation of codon assignment outwards from the bottom right-hand corner. As a result of this expansion, we propose a trend from small hydrophilic to hydrophobic amino acids, with selection for longer polypeptides requiring a hydrophobic core for folding and stability driving the incorporation of hydrophobic amino acids into the code.     

Unusual genetic codes and a novel gene structure for tRNA(AGYSer) in starfish mitochondrial DNA

The nucleotide sequence of a 3849-bp fragment of starfish mitochondrial genome was determined. The genes for NADH dehydrogenase subunits 3, 4, 5, and COIII, and three kinds of (tRNA(UCNSer), tRNA(His), and tRNA(AGYSer) were identified by comparing with the genes of other animal mitochondria so far elucidated. The gene arrangement of starfish mitochondrial genome was different from those of vertebrate and insect mitochondrial genomes. Comparison of the protein-encoding nucleotide sequences of starfish mitochondria with those of other animal mitochondria suggested a unique genetic code in starfish mitochondrial genome; both AGA and AGG (arginine in the universal code) code for serine, AUA (isoleucine in the universal code but methionine in most mitochondrial systems) for isoleucine, and AAA (lysine) for asparagine. It was also inferred that these AGA and AGG codons are decoded by serine tRNA(AGYSer) originally corresponding to AGC and AGU codons. This situation is similar to the case of Drosophila mitochondrial genome. Variations in the use of AGA and AGG codons were discussed on the basis of the evolution of animals and decoding capacity of various tRNA(AGYSer) species possessing different sizes of the dihydrouridine (D) arm.     

The Genetic Code Is One in a Million

Statistical and biochemical studies of the genetic code have found evidence of nonrandom patterns in the distribution of codon assignments. It has, for example, been shown that the code minimizes the effects of point mutation or mistranslation: erroneous codons are either synonymous or code for an amino acid with chemical properties very similar to those of the one that would have been present had the error not occurred. This work has suggested that the second base of codons is less efficient in this respect, by about three orders of magnitude, than the first and third bases. These results are based on the assumption that all forms of error at all bases are equally likely. We extend this work to investigate (1) the effect of weighting transition errors differently from transversion errors and (2) the effect of weighting each base differently, depending on reported mistranslation biases. We find that if the bias affects all codon positions equally, as might be expected were the code adapted to a mutational environment with transition/transversion bias, then any reasonable transition/transversion bias increases the relative efficiency of the second base by an order of magnitude. In addition, if we employ weightings to allow for biases in translation, then only 1 in every million random alternative codes generated is more efficient than the natural code. We thus conclude not only that the natural genetic code is extremely efficient at minimizing the effects of errors, but also that its structure reflects biases in these errors, as might be expected were the code the product of selection. Received: 25 July 1997 / Accepted: 9 January 1998     

Structuring of the genetic code took place at acidic pH

I have observed that in multiple regression the number of codons specifying amino acids in the genetic code is positively correlated with the isoelectric point of amino acids and their molecular weight. Therefore basic amino acids are, on average, codified in the genetic code by a larger number of codons, which seems to imply that the genetic code originated in an acidic 'intracellular' environment. Moreover, I compare the proteins from Picrophilus torridus and Thermoplasma volcanium, which have different intracellular pH and I define the ranks of acidophily for the amino acids. A simple index of acidophily (AI), which can be easily obtained from acidophily ranks, can be associated to any protein and, therefore, can also be associated to the genetic code if the number of synonymous codons attributed to the amino acids in the code is assumed to be the frequency with which the amino acids appeared in ancestral proteins. Finally, the sampling of the variable AI among organisms having an intracellular pH less than or equal to 6.6 and those having a non-acidic intracellular pH leads to the conclusion that the value of the genetic code's AI is not typical of proteins of the latter organisms. As the genetic code's AI value is also statistically not different from that of proteins of the organisms having an acidic intracellular pH, this supports the hypothesis that the structuring of the genetic code took place in acidic pH conditions.     

Block structure and stability of the genetic code

It is known that different codons may be unified into larger groups related to the hierarchical structure, approximate hidden symmetries, and evolutionary origin of the universal genetic code. Using a simplified evolutionary motivated two-letter version of genetic code, the general principles of the most stable coding are discussed. By the complete enumeration in such a reduced code it is strictly proved that the maximum stability with respect to point mutations and shifts in the reading frame needs the fixation of the middle letters within codons in groups with different physico-chemical properties, thus, explaining a key feature of the universal genetic code. The translational stability of the genetic code is studied by the mapping of code onto de Bruijn graph providing both the compact visual representation of mutual relationships between different codons as well as between codons and protein coding DNA sequence and a powerful tool for the investigation of stability of protein coding. Then, the results are extended to four-letter codes. As is shown, the universal genetic code obeys mainly the principles of optimal coding. These results demonstrate the hierarchical character of optimization of universal genetic code with strictly optimal coding being evolved at the earliest stages of molecular evolution. Finally, the universal genetic code is compared with the other natural variants of genetic codes.     

The evolutionary change of the genetic code as restricted by the anticodon and identity of transfer RNA

The discovery of non-universal genetic codes in several mitochondria and nuclear systems during the past ten years has necessitated a reconsideration of the concept that the genetic code is universal and frozen, as was once believed. Here, the flexibility of the relationship between codons and amino acids is discussed on the basis of the distribution of non-universal genetic codes in various organisms insofar as has been observed to date. Judging from the result of recent investigations into tRNA identity, it would appear that the non-participation of the anticodon in recognition by aminoacyl-tRNA synthetase has significantly influenced the variability of codons.