Problems and paradigns. Evolution of mitochondrial genomes and the genetic code

An Expanded Ribosomal Phylogeny of Cyanobacteria Supports a Deep Placement of Plastids

The phylum Cyanobacteria includes free-living bacteria and plastids, the descendants of cyanobacteria that were engulfed by the ancestral lineage of the major photosynthetic eukaryotic group Archaeplastida. Endosymbiotic events that followed this primary endosymbiosis spread plastids across diverse eukaryotic groups. The remnants of the ancestral cyanobacterial genome present in all modern plastids, enable the placement of plastids within Cyanobacteria using sequence-based phylogenetic analyses. To date, such phylogenetic studies have produced conflicting results and two competing hypotheses: (1) plastids diverge relatively recently in cyanobacterial evolution and are most closely related to nitrogen-fixing cyanobacteria, or (2) plastids diverge early in the evolutionary history of cyanobacteria, before the divergence of most cyanobacterial lineages. Here, we use phylogenetic analysis of ribosomal proteins from an expanded data set of cyanobacterial and representative plastid genomes to infer a deep placement for the divergence of the plastid ancestor lineage. We recover plastids as sister to Gloeomargarita and show that the group diverges from other cyanobacterial groups before Pseudanabaena, a previously unreported placement. The tree topologies and phylogenetic distances in our study have implications for future molecular clock studies that aim to model accurate divergence times, especially with respect to groups containing fossil calibrations. The newly sequenced cyanobacterial groups included here will also enable the use of novel cyanobacterial microfossil calibrations.

Hidden in plain view: degeneracy in complex systems

Degeneracy is a word with two meanings. The popular usage of the word denotes deviance and decay. In scientific discourse, degeneracy refers to the idea that different pathways can lead to the same output. In the biological sciences, the concept of degeneracy has been ignored for a few key reasons. Firstly, the word "degenerate" in popular culture has negative, emotionally powerful associations that do not inspire scientists to consider its technical meaning. Secondly, the tendency of searching for single causes of natural and social phenomena means that scientists can overlook the multi-stranded relationships between cause and effect. Thirdly, degeneracy and redundancy are often confused with each other. Degeneracy refers to dissimilar structures that are functionally similar while redundancy refers to identical structures. Degeneracy can give rise to novelty in ways that redundancy cannot. From genetic codes to immunology, vaccinology and brain development, degeneracy is a crucial part of how complex systems maintain their functional integrity. This review article discusses how the scientific concept of degeneracy was imported into genetics from physics and was later introduced to immunology and neuroscience. Using examples of degeneracy in immunology, neuroscience and linguistics, we demonstrate that degeneracy is a useful way of understanding how complex systems function. Reviewing the history and theoretical scope of degeneracy allows its usefulness to be better appreciated, its coherency to be further developed, and its application to be more quickly realized.

From end to end: tRNA editing at 5'- and 3'-terminal positions

During maturation, tRNA molecules undergo a series of individual processing steps, ranging from exo- and endonucleolytic trimming reactions at their 5'- and 3'-ends, specific base modifications and intron removal to the addition of the conserved 3'-terminal CCA sequence. Especially in mitochondria, this plethora of processing steps is completed by various editing events, where base identities at internal positions are changed and/or nucleotides at 5'- and 3'-ends are replaced or incorporated. In this review, we will focus predominantly on the latter reactions, where a growing number of cases indicate that these editing events represent a rather frequent and widespread phenomenon. While the mechanistic basis for 5'- and 3'-end editing differs dramatically, both reactions represent an absolute requirement for generating a functional tRNA. Current in vivo and in vitro model systems support a scenario in which these highly specific maturation reactions might have evolved out of ancient promiscuous RNA polymerization or quality control systems.

Biological evidence for the world's smallest tRNAs

Due to their function as adapters in translation, tRNA molecules share a common structural organization in all kingdoms and organelles with ribosomal protein biosynthesis. A typical tRNA has a cloverleaf-like secondary structure, consisting of acceptor stem, D-arm, anticodon arm, a variable region, and T-arm, with an average length of 73 nucleotides. In several mitochondrial genomes, however, tRNA genes encode transcripts that show a considerable deviation of this standard, having reduced D- or T-arms or even completely lack one of these elements, resulting in tRNAs as small as 66 nts. An extreme case of such truncations is found in the mitochondria of Enoplea. Here, several tRNA genes are annotated that lack both the D- and the T-arm, suggesting even shorter transcripts with a length of only 42 nts. However, direct evidence for these exceptional tRNAs, which were predicted by purely computational means, has been lacking so far. Here, we demonstrate that several of these miniaturized armless tRNAs consisting only of acceptor- and anticodon-arms are indeed transcribed and correctly processed by non-encoded CCA addition in the mermithid Romanomermis culicivorax. This is the first direct evidence for the existence and functionality of the smallest tRNAs ever identified so far. It opens new possibilities towards exploration/assessment of minimal structural motifs defining a functional tRNA and their evolution.

[Genetic diversity of microorganisms]

Microorganisms are important components of the biosphere in maintaining the ecological balance. With the development of molecular biology techniques, researches on the microbial genetic diversity have been developed from morphological and/or protein levels to molecular level. The development of high-throughout sequencing and metagenomics technology not only provide more abundant information and powerful evidence for understanding microbial diversities, but also have great significance for rational utilization and protection of biological resources. The advances in research on genetic diversity of microorganisms, such as separation and identification, population genetic structure, speciation, phylogeny, and evolution of microorganisms, were discussed in this paper.

The contributions of wobbling and superwobbling to the reading of the genetic code

Reduced bacterial genomes and most genomes of cell organelles (chloroplasts and mitochondria) do not encode the full set of 32 tRNA species required to read all triplets of the genetic code according to the conventional wobble rules. Superwobbling, in which a single tRNA species that contains a uridine in the wobble position of the anticodon reads an entire four-fold degenerate codon box, has been suggested as a possible mechanism for how tRNA sets can be reduced. However, the general feasibility of superwobbling and its efficiency in the various codon boxes have remained unknown. Here we report a complete experimental assessment of the decoding rules in a typical prokaryotic genetic system, the plastid genome. By constructing a large set of transplastomic knock-out mutants for pairs of isoaccepting tRNA species, we show that superwobbling occurs in all codon boxes where it is theoretically possible. Phenotypic characterization of the transplastomic mutant plants revealed that the efficiency of superwobbling varies in a codon box-dependent manner, but--contrary to previous suggestions--it is independent of the number of hydrogen bonds engaged in codon-anticodon interaction. Finally, our data provide experimental evidence of the minimum tRNA set comprising 25 tRNA species, a number lower than previously suggested. Our results demonstrate that all triplets with pyrimidines in third codon position are dually decoded: by a tRNA species utilizing standard base pairing or wobbling and by a second tRNA species employing superwobbling. This has important implications for the interpretation of the genetic code and will aid the construction of synthetic genomes with a minimum-size translational apparatus.

Purification of human mitochondrial ribosomal L7/L12 stalk proteins and reconstitution of functional hybrid ribosomes in Escherichia coli

Bacterial ribosomal L7/L12 stalk is formed by L10, L11, and multiple copies of L7/L12, which plays an essential role in recruiting initiation and elongation factors during translation. The homologs of these proteins, MRPL10, MRPL11, and MRPL12, are present in human mitochondrial ribosomes. To evaluate the role of MRPL10, MRPL11, and MRPL12 in translation, we over-expressed and purified components of the human mitochondrial L7/L12 stalk proteins in Escherichia coli. Here, we designed a construct to co-express MRPL10 and MRPL12 using a duet expression system to form a functional MRPL10-MRPL12 complex. The goal is to demonstrate the homology between the mitochondrial and bacterial L7/L12 stalk proteins and to reconstitute a hybrid ribosome to be used in structural and functional studies of the mitochondrial stalk.

Regulation of mitochondrial ribosomal protein S29 (MRPS29) expression by a 5'-upstream open reading frame

Mitochondrial ribosomal protein S29 (MRPS29) is a mitochondrial pro-apoptotic protein also known as death associated protein 3 (DAP3). Over-expression of MRPS29 has been reported to induce apoptosis in several different human cell lines while conferring resistance in glioma and Ataxia telangiectasia cells. These two contradictory reports led us to investigate the MRPS29-induced apoptosis further. Cyber searches of the EST databases revealed the presence of a splice variant of MRPS29 mRNA containing an upstream open reading frame (uORF) at the 5' untranslated region (UTR). In this study, we confirmed the presence of this uORF using real-time RT-PCR and investigated its role in MRPS29 expression.

Maintenance and expression of the S. cerevisiae mitochondrial genome--from genetics to evolution and systems biology

As a legacy of their endosymbiotic eubacterial origin, mitochondria possess a residual genome, encoding only a few proteins and dependent on a variety of factors encoded by the nuclear genome for its maintenance and expression. As a facultative anaerobe with well understood genetics and molecular biology, Saccharomyces cerevisiae is the model system of choice for studying nucleo-mitochondrial genetic interactions. Maintenance of the mitochondrial genome is controlled by a set of nuclear-coded factors forming intricately interconnected circuits responsible for replication, recombination, repair and transmission to buds. Expression of the yeast mitochondrial genome is regulated mostly at the post-transcriptional level, and involves many general and gene-specific factors regulating splicing, RNA processing and stability and translation. A very interesting aspect of the yeast mitochondrial system is the relationship between genome maintenance and gene expression. Deletions of genes involved in many different aspects of mitochondrial gene expression, notably translation, result in an irreversible loss of functional mtDNA. The mitochondrial genetic system viewed from the systems biology perspective is therefore very fragile and lacks robustness compared to the remaining systems of the cell. This lack of robustness could be a legacy of the reductive evolution of the mitochondrial genome, but explanations involving selective advantages of increased evolvability have also been postulated.

Superwobbling facilitates translation with reduced tRNA sets

Some bacterial and most organelle genomes do not encode the full set of 32 tRNA species required to read all codons according to Crick's wobble rules. 'Superwobble', in which a tRNA species with an unmodified U in the wobble position reads all four nucleotides in the third codon position, represents one possible mechanism for how a reduced tRNA set could still suffice. We have tested the superwobble hypothesis by producing knockout mutants for the pair of plastid glycine tRNA genes. Here we show that, whereas the tRNA gene with U in the wobble position is essential, the gene with G in this position is nonessential, demonstrating that the U-containing anticodon can indeed read all four glycine triplets. We also show that the price for superwobbling is a reduced translational efficiency, which explains why most organisms prefer pairs of isoaccepting tRNAs over the superwobbling mechanism.

A comparative genomics analysis of codon reassignments reveals a link with mitochondrial proteome size and a mechanism of genetic code change via suppressor tRNAs

Using a comparative genomics approach we demonstrate a negative correlation between the number of codon reassignments undergone by 222 mitochondrial genomes and the mitochondrial genome size, the number of mitochondrial ORFs, and the sizes of the large and small subunit mitochondrial rRNAs. In addition, we show that the TGA-to-tryptophan codon reassignment, which has occurred 11 times in mitochondrial genomes, is found in mitochondrial genomes smaller than those which have not undergone the reassignment. We therefore propose that mitochondrial codon reassignments occur in a wide range of phyla, particularly in Metazoa, due to a reduced "proteomic constraint" on the mitochondrial genetic code, compared to the nuclear genetic code. The reduced proteomic constraint reflects the small size of the mitochondrial-encoded proteome and allows codon reassignments to occur with less likelihood of lethality. In addition, we demonstrate a striking link between nonsense codon reassignments and the decoding properties of naturally occurring nonsense suppressor tRNAs. This suggests that natural preexisting nonsense suppression facilitated nonsense codon reassignments and constitutes a novel mechanism of genetic code change. These findings explain for the first time the identity of the stop codons and amino acids reassigned in mitochondrial and nuclear genomes. Nonsense suppressor tRNAs provided the raw material for nonsense codon reassignments, implying that the properties of the tRNA anticodon have dictated the identity of nonsense codon reassignments.

A structural model for the large subunit of the mammalian mitochondrial ribosome

Protein translation is essential for all forms of life and is conducted by a macromolecular complex, the ribosome. Evolutionary changes in protein and RNA sequences can affect the 3D organization of structural features in ribosomes in different species. The most dramatic changes occur in animal mitochondria, whose genomes have been reduced and altered significantly. The RNA component of the mitochondrial ribosome (mitoribosome) is reduced in size, with a compensatory increase in protein content. Until recently, it was unclear how these changes affect the 3D structure of the mitoribosome. Here, we present a structural model of the large subunit of the mammalian mitoribosome developed by combining molecular modeling techniques with cryo-electron microscopic data at 12.1A resolution. The model contains 93% of the mitochondrial rRNA sequence and 16 mitochondrial ribosomal proteins in the large subunit of the mitoribosome. Despite the smaller mitochondrial rRNA, the spatial positions of RNA domains known to be involved directly in protein synthesis are essentially the same as in bacterial and archaeal ribosomes. However, the dramatic reduction in rRNA content necessitates evolution of unique structural features to maintain connectivity between RNA domains. The smaller rRNA sequence also limits the likelihood of tRNA binding at the E-site of the mitoribosome, and correlates with the reduced size of D-loops and T-loops in some animal mitochondrial tRNAs, suggesting co-evolution of mitochondrial rRNA and tRNA structures.

The evolution of tRNA-Leu genes in animal mitochondrial genomes

Animal mitochondrial genomes usually have two transfer RNAs for leucine: one, with anticodon UAG, translates the four-codon family CUN, while the other, with anticodon UAA, translates the two-codon family UUR. These two genes must differ at the third anticodon position, but in some species the genes differ at many additional sites, indicating that these genes have been independent for a long time. Duplication and deletion of genes in mitochondrial genomes occur frequently during the evolution of the Metazoa. If a tRNA-Leu gene were duplicated and a substitution occurred in the anticodon, this would effectively turn one type of tRNA into the other. The original copy of the second tRNA type might then be lost by a deletion elsewhere in the genome. There are several groups of species in which the two tRNA-Leu genes occur next to one another (or very close) on the genome, which suggests that tandem duplication has occurred. Here we use RNA-specific phylogenetic methods to determine evolutionary trees for both genes. We present evidence that the process of duplication, anticodon mutation, and deletion of tRNA-Leu genes has occurred at least five times during the evolution of the metazoa-once in the common ancestor of all protostomes, once in the common ancestor of echinoderms and hemichordates, once in the hermit crab, and twice independently in mollusks.

Characterisation of a non-canonical genetic code in the oxymonad Streblomastix strix

The genetic code is one of the most highly conserved characters in living organisms. Only a small number of genomes have evolved slight variations on the code, and these non-canonical codes are instrumental in understanding the selective pressures maintaining the code. Here, we describe a new case of a non-canonical genetic code from the oxymonad flagellate Streblomastix strix. We have sequenced four protein-coding genes from S.strix and found that the canonical stop codons TAA and TAG encode the amino acid glutamine. These codons are retained in S.strix mRNAs, and the legitimate termination codons of all genes examined were found to be TGA, supporting the prediction that this should be the only true stop codon in this genome. Only four other lineages of eukaryotes are known to have evolved non-canonical nuclear genetic codes, and our phylogenetic analyses of alpha-tubulin, beta-tubulin, elongation factor-1 alpha (EF-1 alpha), heat-shock protein 90 (HSP90), and small subunit rRNA all confirm that the variant code in S.strix evolved independently of any other known variant. The independent origin of each of these codes is particularly interesting because the code found in S.strix, where TAA and TAG encode glutamine, has evolved in three of the four other nuclear lineages with variant codes, but this code has never evolved in a prokaryote or a prokaryote-derived organelle. The distribution of non-canonical codes is probably the result of a combination of differences in translation termination, tRNAs, and tRNA synthetases, such that the eukaryotic machinery preferentially allows changes involving TAA and TAG.

The variability of the mitochondrial genome in human aging: a key for life and death?

The impressive performance of the research in mitochondrial genetics and human aging in the last decade outlines a new scenery in which the inherited variation of the mitochondrial genome (mtDNA) may play a role in rate and quality of aging. This variation in humans was initially looked at as nearly neutral, and useful just for the reconstruction of human population history. However, recent data suggest that different mtDNA molecules are qualitatively different from each other. The aim of this paper is to discuss current ideas on the relationships among mitochondrial function, mtDNA inherited variation, and aging. The main processes where the mitochondrion is involved and the importance these processes have on aging and death of individuals will be described. A possible connection between programmed death phenomena (mitoptosis, apoptosis, phenoptosis) and rate and quality of aging will be discussed. Finally, the possible role played in these processes by the mtDNA germline variation will be explored.

A mitochondrial-like aconitase in the bacterium Bacteroides fragilis: implications for the evolution of the mitochondrial Krebs cycle

Aconitase and isocitrate dehydrogenase (IDH) enzyme activities were detected in anaerobically prepared cell extracts of the obligate anaerobe Bacteroides fragilis. The aconitase gene was located upstream of the genes encoding the other two components of the oxidative branch of the Krebs cycle, IDH and citrate synthase. Mutational analysis indicates that these genes are cotranscribed. A nonpolar in-frame deletion of the acnA gene that encodes the aconitase prevented growth in glucose minimal medium unless heme or succinate was added to the medium. These results imply that B. fragilis has two pathways for alpha-ketoglutarate biosynthesis-one from isocitrate and the other from succinate. Homology searches indicated that the B. fragilis aconitase is most closely related to aconitases of two other Cytophaga-Flavobacterium-Bacteroides (CFB) group bacteria, Cytophaga hutchinsonii and Fibrobacter succinogenes. Phylogenetic analysis indicates that the CFB group aconitases are most closely related to mitochondrial aconitases. In addition, the IDH of C. hutchinsonii was found to be most closely related to the mitochondrial/cytosolic IDH-2 group of eukaryotic organisms. These data suggest a common origin for these Krebs cycle enzymes in mitochondria and CFB group bacteria.

Origin and evolution of the mitochondrial proteome

The endosymbiotic theory for the origin of mitochondria requires substantial modification. The three identifiable ancestral sources to the proteome of mitochondria are proteins descended from the ancestral alpha-proteobacteria symbiont, proteins with no homology to bacterial orthologs, and diverse proteins with bacterial affinities not derived from alpha-proteobacteria. Random mutations in the form of deletions large and small seem to have eliminated nonessential genes from the endosymbiont-mitochondrial genome lineages. This process, together with the transfer of genes from the endosymbiont-mitochondrial genome to nuclei, has led to a marked reduction in the size of mitochondrial genomes. All proteins of bacterial descent that are encoded by nuclear genes were probably transferred by the same mechanism, involving the disintegration of mitochondria or bacteria by the intracellular membranous vacuoles of cells to release nucleic acid fragments that transform the nuclear genome. This ongoing process has intermittently introduced bacterial genes to nuclear genomes. The genomes of the last common ancestor of all organisms, in particular of mitochondria, encoded cytochrome oxidase homologues. There are no phylogenetic indications either in the mitochondrial proteome or in the nuclear genomes that the initial or subsequent function of the ancestor to the mitochondria was anaerobic. In contrast, there are indications that relatively advanced eukaryotes adapted to anaerobiosis by dismantling their mitochondria and refitting them as hydrogenosomes. Accordingly, a continuous history of aerobic respiration seems to have been the fate of most mitochondrial lineages. The initial phases of this history may have involved aerobic respiration by the symbiont functioning as a scavenger of toxic oxygen. The transition to mitochondria capable of active ATP export to the host cell seems to have required recruitment of eukaryotic ATP transport proteins from the nucleus. The identity of the ancestral host of the alpha-proteobacterial endosymbiont is unclear, but there is no indication that it was an autotroph. There are no indications of a specific alpha-proteobacterial origin to genes for glycolysis. In the absence of data to the contrary, it is assumed that the ancestral host cell was a heterotroph.

Organellar genes: why do they end up in the nucleus?

Many mitochondrial and plastid proteins are derived from their bacterial endosymbiotic ancestors, but their genes now reside on nuclear chromosomes instead of remaining within the organelle. To become an active nuclear gene and return to the organelle as a functional protein, an organellar gene must first be assimilated into the nuclear genome. The gene must then be transcribed and acquire a transit sequence for targeting the protein back to the organelle. On reaching the organelle, the protein must be properly folded and modified, and in many cases assembled in an orderly manner into a larger protein complex. Finally, the nuclear copy must be properly regulated to achieve a fitness level comparable with the organellar gene. Given the complexity in establishing a nuclear copy, why do organellar genes end up in the nucleus? Recent data suggest that these genes are worse off than their nuclear and free-living counterparts because of a reduction in the efficiency of natural selection, but do these population-genetic processes drive the movement of genes to the nucleus? We are now at a stage where we can begin to discriminate between competing hypotheses using a combination of experimental, natural population, bioinformatic and theoretical approaches.

Why mitochondrial genes are most often found in nuclei

A very small fraction of the proteins required for the propagation and function of mitochondria are coded by their genomes, while nuclear genes code the vast majority. We studied the migration of genes between the two genomes when transfer mechanisms mediate this exchange. We could calculate the influence of differential mutation rates, as well as that of biased transfer rates, on the partitioning of genes between the two genomes. We observe no significant difference in partitioning for haploid and diploid cell populations, but the effective size of cell populations is important. For infinitely large effective populations, higher mutation rates in mitochondria than in nuclear genomes are required to drive mitochondrial genes to the nuclear genome. In the more realistic case of finite populations, gene transfer favoring the nucleus and/or higher mutation rates in the mitochondrion will drive mitochondrial genes to the nucleus. We summarize experimental data that identify a gene transfer process mediated by vacuoles that favors the accumulation of mitochondrial genes in the nuclei of modern cells. Finally, we compare the behavior of mitochondrial genes for which transfer to the nucleus is neutral or influenced by purifying selection.

Mechanisms of human mitochondrial DNA maintenance: the determining role of primary sequence and length over function

Although the regulation of mitochondrial DNA (mtDNA) copy number is performed by nuclear-coded factors, very little is known about the mechanisms controlling this process. We attempted to introduce nonhuman ape mtDNA into human cells harboring either no mtDNA or mutated mtDNAs (partial deletion and tRNA gene point mutation). Unexpectedly, only cells containing no mtDNA could be repopulated with nonhuman ape mtDNA. Cells containing a defective human mtDNA did not incorporate or maintain ape mtDNA and therefore died under selection for oxidative phosphorylation function. On the other hand, foreign human mtDNA was readily incorporated and maintained in these cells. The suicidal preference for self-mtDNA showed that functional parameters associated with oxidative phosphorylation are less relevant to mtDNA maintenance and copy number control than recognition of mtDNA self-determinants. Non-self-mtDNA could not be maintained into cells with mtDNA even if no selection for oxidative phosphorylation was applied. The repopulation kinetics of several mtDNA forms after severe depletion by ethidium bromide treatment showed that replication and maintenance of mtDNA in human cells are highly dependent on molecular features, because partially deleted mtDNA molecules repopulated cells significantly faster than full-length mtDNA. Taken together, our results suggest that mtDNA copy number may be controlled by competition for limiting levels of trans-acting factors that recognize primarily mtDNA molecular features. In agreement with this hypothesis, marked variations in mtDNA levels did not affect the transcription of nuclear-coded factors involved in mtDNA replication.

Why have organelles retained genomes?

The observation that chloroplasts and mitochondria have retained relics of eubacterial genomes and a protein-synthesizing machinery has long puzzled biologists. If most genes have been transferred from organelles to the nucleus during evolution, why not all? What selective pressure maintains genomes in organelles? Electron transport through the photosynthetic and respiratory membranes is a powerful - but dangerous - source of energy. Recent evidence suggests that organelle genomes have persisted because structural proteins that maintain redox balance within bioenergetic membranes must be synthesized when and where they are needed, to counteract the potentially deadly side effects of ATP-generating electron transport.

The genome of Rickettsia prowazekii and some thoughts on the origin of mitochondria and hydrogenosomes

The sequence of an alpha-proteobacterial genome, that of Rickettsia prowazekii, is a substantial advance in microbial and evolutionary biology. The genome of this obligately aerobic intracellular parasite is small and is apparently still undergoing reduction, reflecting gene losses attributable to its intracellular parasitic lifestyle. Evolutionary analyses of proteins encoded in the genome contain the strongest phylogenetic evidence to date for the view that mitochondria descend from alpha-proteobacteria. Although both Rickettsia and mitochondrial genomes are highly reduced, it appears that genome reduction in these lineages has occurred independently. Rickettsia's genome encodes an ATP-generating machinery that is strikingly similar to that of aerobic mitochondria. But it does not encode homologues for the ATP-producing pathways of anaerobic mitochondria or hydrogenosomes, leaving an important issue regarding the origin and nature of the ancestral mitochondrial symbiont unresolved.

The genome sequence of Rickettsia prowazekii and the origin of mitochondria

We describe here the complete genome sequence (1,111,523 base pairs) of the obligate intracellular parasite Rickettsia prowazekii, the causative agent of epidemic typhus. This genome contains 834 protein-coding genes. The functional profiles of these genes show similarities to those of mitochondrial genes: no genes required for anaerobic glycolysis are found in either R. prowazekii or mitochondrial genomes, but a complete set of genes encoding components of the tricarboxylic acid cycle and the respiratory-chain complex is found in R. prowazekii. In effect, ATP production in Rickettsia is the same as that in mitochondria. Many genes involved in the biosynthesis and regulation of biosynthesis of amino acids and nucleosides in free-living bacteria are absent from R. prowazekii and mitochondria. Such genes seem to have been replaced by homologues in the nuclear (host) genome. The R. prowazekii genome contains the highest proportion of non-coding DNA (24%) detected so far in a microbial genome. Such non-coding sequences may be degraded remnants of 'neutralized' genes that await elimination from the genome. Phylogenetic analyses indicate that R. prowazekii is more closely related to mitochondria than is any other microbe studied so far.

Reductive evolution of resident genomes

Small, asexual populations are expected to accumulate deleterious substitutions and deletions in an irreversible manner, which in the long-term will lead to mutational meltdown and genome decay. Here, we discuss the influence of such reductive processes on the evolution of genomes that replicate within the domain of a host genome.

A chimeric disposition of the elongation factor genes in Rickettsia prowazekii

An exceptional disposition of the elongation factor genes is observed in Rickettsia prowazekii, in which there is only one tuf gene, which is distant from the lone fus gene. In contrast, the closely related bacterium Agrobacterium tumefaciens has the normal bacterial arrangement of two tuf genes, of which one is tightly linked to the fus gene. Analysis of the flanking sequences of the single tuf gene in R. prowazekii shows that it is preceded by two of the four tRNA genes located in the 5' region of the Escherichia coli tufB gene and that it is followed by rpsJ as well as associated ribosomal protein genes, which in E. coli are located downstream of the tufA gene. The fus gene is located within the str operon and is followed by one tRNA gene as well as by the genes secE and nusG, which are located in the 3' region of tufB in E. coli. This atypical disposition of genes suggests that intrachromosomal recombination between duplicated tuf genes has contributed to the evolution of the unique genomic architecture of R. prowazekii.

Genomic evolution drives the evolution of the translation system

Our thesis is that the characteristics of the translational machinery and its organization are selected in part by evolutionary pressure on genomic traits have nothing to do with translation per se. These genomic traits include size, composition, and architecture. To illustrate this point, we draw parallels between the structure of different genomes that have adapted to intracellular niches independently of each other. Our starting point is the general observation that the evolutionary history of organellar and parasitic bacteria have favored bantam genomes. Furthermore, we suggest that the constraints of the reductive mode of genomic evolution account for the divergence of the genetic code in mitochondria and the genetic organization of the translational system observed in parasitic bacteria. In particular, we associate codon reassignments in animal mitochondria with greatly simplified tRNA populations. Likewise, we relate the organization of translational genes in the obligate intracellular parasite Rickettsia prowazekii to the processes supporting the reductive mode of genomic evolution. Such findings provide strong support for the hypothesis that genomes of organelles and of parasitic bacteria have arisen from the much larger genomes of ancestral bacteria that have been reduced by intrachromosomal recombination and deletion events. A consequence of the reductive mode of genomic evolution is that the resulting translation systems may deviate markedly from conventional systems.

Evolutionary changes in the genetic code

1. The genetic code was thought to be identical ("universal") in all biological systems until 1981, when it was discovered that the coding system in mammalian mitochondria differed from the universal code in the use of codons AUA, UGA, AGA and AGG. 2. Many other differences have since been discovered, some in mitochondria of various phyla, others in bacteria, ciliated protozoa, algae and yeasts. 3. The original thesis that the code was universal and "frozen" depended on the precept that any mutational change in the code would be lethal, because it would produce widespread alterations in the amino acid sequences of proteins. Such changes would destroy protein function, and hence would be intolerable. 4. The objection was "by-passed" by nature. It is possible for a codon to disappear from mRNA molecules, often as a result of directional mutation pressure in DNA: thus all UGA stop codons can be replaced by UAA. 5. The missing UGA codon can then reappear when some UGG tryptophan codons mutate to UGA. The new UGA codons will be translated as tryptophan, as is the case in non-plant mitochondria and Mycoplasma. Therefore, no changes have taken place in the amino acid sequences of proteins. 6. Variations of this procedure have occurred, affecting various codons, and discoveries are still being made. The findings illustrate the evolutionary interplay between tRNA, release factors and codon-anticodon pairing.