Skew of Mononucleotide Frequencies, Relative Abundance of Dinucleotides, and DNA Strand Asymmetry

Based on 152 mitochondrial genomes and 36 bacterial chromosomes that have been completely sequenced, as well as three long contigs for human chromosomes 6, 21, and 22, we examined skews of mononucleotide frequencies and the relative abundance of dinucleotides in one DNA strand. Each group of these genomes has its own characteristics. Regarding mitochondrial genomes, both C_pG and G_pT are underrepresented, while either G_pG or C_pC or both are overrepresented. The relative frequency of nucleotide T vs A and of nucleotide G vs C is strongly skewed, due presumably to strand asymmetry in replication errors and unidirectional DNA replication from single origins. Exceptions are found in the plant and yeast mitochondrial genomes, each of which may replicate from multiple origins. Regarding bacterial genomes, the ``universal' rule of C_pG deficiency is restricted to archaebacteria and some eubacteria. In other eubacteria, the most underrepresented dinucleotide is either T_pA or G_pT. In general, there are significant T vs A and G vs C skews in each half of the bacterial genome, although these are almost exactly canceled out over the whole genome. Regarding human chromosomes 6, 21, and 22, dinucleotide C_pG tends to be avoided. The relative frequency of mononucleotides exhibits conspicuous local skews, suggesting that each of these chromosomal segments contains more than one DNA replication origin. It is concluded that, when there are several replicons in a genomic region, not only the number of DNA replication origins but also the directionality is important and that the observed patterns of nucleotide frequencies in the genome strongly support the hypothesis of strand asymmetry in replication errors. Received: 1 November 2000 / Accepted: 12 March 2001     

A New Aspect to the Origin and Evolution of Eukaryotes

One of the most important omissions in recent evolutionary theory concerns how eukaryotes could emerge and evolve. According to the currently accepted views, the first eukaryotic cell possessed a nucleus, an endomembrane system, and a cytoskeleton but had an inefficient prokaryotic-like metabolism. In contrast, one of the most ancient eukaryotes, the metamonada Giardia lamblia, was found to have formerly possessed mitochondria. In sharp contrast with the traditional views, this paper suggests, based on the energetic aspect of genome organization, that the emergence of eukaryotes was promoted by the establishment of an efficient energy-converting organelle, such as the mitochondrion. Mitochondria were acquired by the endosymbiosis of ancient α-purple photosynthetic Gram-negative eubacteria that reorganized the prokaryotic metabolism of the archaebacterial-like ancestral host cells. The presence of an ATP pool in the cytoplasm provided by this cell organelle allowed a major increase in genome size. This evolutionary change, the remarkable increase both in genome size and complexity, explains the origin of the eukaryotic cell itself. The loss of cell wall and the appearance of multicellularity can also be explained by the acquisition of mitochondria. All bacteria use chemiosmotic mechanisms to harness energy; therefore the periplasm bounded by the cell wall is an essential part of prokaryotic cells. Following the establishment of mitochondria, the original plasma membrane-bound metabolism of prokaryotes, as well as the funcion of the periplasm providing a compartment for the formation of different ion gradients, has been transferred into the inner mitochondrial membrane and intermembrane space. After the loss of the essential function of periplasm, the bacterial cell wall could also be lost, which enabled the naked cells to establish direct connections among themselves. The relatively late emergence of mitochondria may be the reason why multicellularity evolved so slowly. Received: 29 May 1997 / Accepted: 9 October 1997     

Evolutionary Comparisons of RecA-Like Proteins Across All Major Kingdoms of Living Organisms

Protein sequences with similarities to Escherichia coli RecA were compared across the major kingdoms of eubacteria, archaebacteria, and eukaryotes. The archaeal sequences branch monophyletically and are most closely related to the eukaryotic paralogous Rad51 and Dmc1 groups. A multiple alignment of the sequences suggests a modular structure of RecA-like proteins consisting of distinct segments, some of which are conserved only within subgroups of sequences. The eukaryotic and archaeal sequences share an N-terminal domain which may play a role in interactions with other factors and nucleic acids. Several positions in the alignment blocks are highly conserved within the eubacteria as one group and within the eukaryotes and archaebacteria as a second group, but compared between the groups these positions display nonconservative amino acid substitutions. Conservation within the RecA-like core domain identifies possible key residues involved in ATP-induced conformational changes. We propose that RecA-like proteins derive evolutionarily from an assortment of independent domains and that the functional homologs of RecA in noneubacteria comprise an array of RecA-like proteins acting in series or cooperatively. Received: 25 October 1996 / Accepted: 31 December 1996     

Periodicities of 10-11bp as indicators of the supercoiled state of genomic DNA

DNA sequences contain information about the bendability and native conformation of DNA. For example, a repetition of certain dinucleotides at distances of 10-11bp supports wrapping around nucleosomes and supercoiled structures of bacterial DNA. We analyzed 86 eubacterial genomes, 16 archaea, and six genomes of higher eukaryotes. First, we discuss whether or not the observed periodicities represent indeed bendability signals. This claim is confirmed since: (1) dinucleotide signals are of comparable size to mononucleotide signals, (2) the signals are present in non-coding DNA as well, and (3) repeat masking has only a minor effect on 10-11bp periodicities. Moreover, the periodicities persist up to 150bp, comparable to the nucleosome size. We show that doublet peaks in Caenorhabditis elegans and some prokaryotes can be traced back to long-ranging modulations. In mammalian genomes, we find consistently spectral peaks as observed earlier in human chromosomes 20, 21 and 22. It has been shown in previous studies that archaea have periods of 10bp, whereas eubacteria exhibit 11bp periodicities. These differences reflect different supercoiled states of microbial DNA. Is the period of 10bp an archaeal or a thermophilic feature? This question is addressed by relating periodicities to optimal growth temperatures. It turns out that the archaea Methanopyrus kandleri (t(opt)=80 degrees C) and a Halobacterium strain (t(opt)=42 degrees C) both have longer periods of about 11bp. Eubacterial genomes have consistently periods around 11bp indicative of negative supercoiling.     

Evidence for the early divergence of tryptophanyl- and tyrosyl-tRNA synthetases

Each amino acid is attached to its cognate tRNA by a distinct aminoacyl-tRNA synthetase (aaRS). The conventional evolutionary view is that the modern complement of synthetases existed prior to the divergence of eubacteria and eukaryotes. Thus comparisons of prokaryotic and eukaryotic aminoacyl-tRNA synthetases of the same type (charging specificity) should show greater sequence similarities than comparisons between synthetases of different types—and this is almost always so. However, a recent study [Ribas de Pouplana L, Furgier M, Quinn CL, Schimmel P (1996) Proc Natl Acad Sci USA 93:166–170] suggested that tryptophanyl- (TrpRS) and tyrosyl-tRNA (TyrRS) synthetases of the Eucarya (eukaryotes) are more similar to each other than either is to counterparts in the Bacteria (eubacteria). Here, we reexamine the evolutionary relationships of TyrRS and TrpRS using a broader range of taxa, including new sequence data from the Archaea (archaebacteria) as well as species of Eucarya and Bacteria. Our results differ from those of Ribas de Pouplana et al.: All phylogenetic methods support the separate monophyly of TrpRS and TyrRS. We attribute this result to the inclusion of the archaeal data which might serve to reduce long branch effects possibly associated with eukaryotic TrpRS and TyrRS sequences. Furthermore, reciprocally rooted phylogenies of TrpRS and TyrRS sequences confirm the closer evolutionary relationship of Archaea to eukaryotes by placing the root of the universal tree in the Bacteria. Received: 7 December 1996 / Accepted: 11 February 1997     

Primary Structure and Phylogenetic Relationships of a Malate Dehydrogenase Gene from Giardia lamblia

The lactate and malate dehydrogenases comprise a complex protein superfamily with multiple enzyme homologues found in eubacteria, archaebacteria, and eukaryotes. In this study we describe the sequence and phylogenetic relationships of a malate dehydrogenase (MDH) gene from the amitochondriate diplomonad protist, Giardia lamblia. Parsimony, distance, and maximum-likelihood analyses of the MDH protein family solidly position G. lamblia MDH within a eukaryote cytosolic MDH clade, to the exclusion of chloroplast, mitochondrial, and peroxisomal homologues. Furthermore, G. lamblia MDH is specifically related to a homologue from Trichomonas vaginalis. This MDH topology, together with published phylogenetic analyses of β-tubulin, chaperonin 60, valyl-tRNA synthetase, and EF-1α, suggests a sister-group relationship between diplomonads and parabasalids. Since these amitochondriate lineages contain genes encoding proteins which are characteristic of mitochondria and α-proteobacteria, their shared ancestry suggests that mitochondrial properties were lost in the common ancestor of both groups. Received: 14 September 1998 / Accepted: 29 December 1998     

The Evolution of the Conserved ATPase Domain (CAD): Reconstructing the History of an Ancient Protein Module

The AAA proteins (ATPases Associated with a variety of cellular Activities) are found in eubacterial, archaebacterial, and eukaryotic species and participate in a large number of cellular processes, including protein degradation, vesicle fusion, cell cycle control, and cellular secretory processes. The AAA proteins are characterized by the presence of a 230 to 250-amino acid ATPase domain referred to as the Conserved ATPase Domain or CAD. Phylogenetic analysis of 133 CAD sequences from 38 species reveal that AAA CADs are organized into discrete groups that are related not only in structure but in cellular function. Evolutionary analyses also indicate that the CAD was present in the last common ancestor of eubacteria, archaebacteria, and eukaryotes. The eubacterial CADs are found in metalloproteases, while CAD-containing proteins in the archaebacterial and eukaryotic lineages appear to have diversified by a series of gene duplication events that lead to the establishment of different functional AAA proteins, including proteasomal regulatory, NSF/Sec, and Pas proteins. The phylogeny of the CADs provides the basis for establishing the patterns of evolutionary change that characterize the AAA proteins. Received: 28 January 1997 / Accepted: 8 May 1997     

Periodicity in prokaryotic and eukaryotic genomes identified by power spectrum analysis

We used a power spectrum method to identify periodic patterns in nucleotide sequence, and characterized nucleotide sequences that confer periodicities to prokaryotic and eukaryotic genomes and genomes. A 10-bp periodicity was prevalent in hyperthermophilic bacteria and archaebacteria, and an 11-bp periodicity was prevalent in eubacteria. The 10-bp periodicity was also prevalent in the eukaryotes such as the worm Caenorhabditis elegans. Additionally, in the worm genome, a 68-bp periodicity in chromosome I, a 59-bp periodicity in chromosome II, and a 94-bp periodicity in chromosome III were found. In human chromosomes 21 and 22, approximately 167- or 84-bp periodicity was detected along the entire length of these chromosomes. Because the 167-bp is identical to the length of DNA that forms two complete helical turns in nucleosome organization, we speculated that the respective sequences may correspond to arrays of a special compact form of nucleosomes clustered in specific regions of the human chromosomes. This periodic element contained a high frequency of TGG. TGG-rich sequences are known to form a specific subset of folded DNA structures, and therefore, the sequences might have potential to form specific higher order structures related to the clustered occurrence of a specific form of the speculated nucleosomes.     

Evolutionary Relationship Between Translation Initiation Factor eIF-2γ and Selenocysteine-Specific Elongation Factor SELB: Change of Function in Translation Factors

Eubacterial and eukaryotic translation initiation systems have very little in common, and therefore the evolutionary events that gave rise to these two disparate systems are difficult to ascertain. One common feature is the presence of initiation, elongation, and release factors belonging to a large GTPase superfamily. One of these initiation factors, the γ subunit of initiation factor 2 (eIF-2γ), is found only in eukaryotes and archaebacteria. We have sequenced eIF-2γ gene fragments from representative diplomonads, parabasalia, and microsporidia and used these new sequences together with new archaebacterial homologues to examine the phylogenetic position of eIF-2γ within the GTPase superfamily. The archaebacterial and eukaryotic eIF-2γ proteins are found to be very closely related, and are in turn related to SELB, the selenocysteine-specific elongation factor from eubacteria. The overall topology of the GTPase tree further suggests that the eIF-2γ/SELB group may represent an ancient subfamily of GTPases that diverged prior to the last common ancestor of extant life. Received: 2 January 1998 / Accepted: 1 June 1998     

Gene Descent, Duplication, and Horizontal Transfer in the Evolution of Glutamyl- and Glutaminyl-tRNA Synthetases

In translation, separate aminoacyl-tRNA synthetases attach the 20 different amino acids to their cognate tRNAs, with the exception of glutamine. Eukaryotes and some bacteria employ a specific glutaminyl-tRNA synthetase (GlnRS) which other Bacteria, the Archaea (archaebacteria), and organelles apparently lack. Instead, tRNA^Gln is initially acylated with glutamate by glutamyl-tRNA synthetase (GluRS), then the glutamate moiety is transamidated to glutamine. Lamour et al. [(1994) Proc Natl Acad Sci USA 91:8670–8674] suggested that an early duplication of the GluRS gene in eukaryotes gave rise to the gene for GlnRS—a copy of which was subsequently transferred to proteobacteria. However, questions remain about the occurrence of GlnRS genes among the Eucarya (eukaryotes) outside of the ``crown' taxa (animals, fungi, and plants), the distribution of GlnRS genes in the Bacteria, and their evolutionary relationships to genes from the Archaea. Here, we show that GlnRS occurs in the most deeply branching eukaryotes and that putative GluRS genes from the Archaea are more closely related to GlnRS and GluRS genes of the Eucarya than to those of Bacteria. There is still no evidence for the existence of GlnRS in the Archaea. We propose that the last common ancestor to contemporary cells, or cenancestor, used transamidation to synthesize Gln-tRNA^Gln and that both the Bacteria and the Archaea retained this pathway, while eukaryotes developed a specific GlnRS gene through the duplication of an existing GluRS gene. In the Bacteria, GlnRS genes have been identified in a total of 10 species from three highly diverse taxonomic groups: Thermus/Deinococcus, Proteobacteria γ/β subdivision, and Bacteroides/Cytophaga/Flexibacter. Although all bacterial GlnRS form a monophyletic group, the broad phyletic distribution of this tRNA synthetase suggests that multiple gene transfers from eukaryotes to bacteria occurred shortly after the Archaea–eukaryote divergence.     

Heat Shock Protein 70 Family: Multiple Sequence Comparisons, Function, and Evolution

The heat shock protein 70 kDa sequences (HSP70) are of great importance as molecular chaperones in protein folding and transport. They are abundant under conditions of cellular stress. They are highly conserved in all domains of life: Archaea, eubacteria, eukaryotes, and organelles (mitochondria, chloroplasts). A multiple alignment of a large collection of these sequences was obtained employing our symmetric-iterative ITERALIGN program (Brocchieri and Karlin 1998). Assessments of conservation are interpreted in evolutionary terms and with respect to functional implications. Many archaeal sequences (methanogens and halophiles) tend to align best with the Gram-positive sequences. These two groups also miss a signature segment [about 25 amino acids (aa) long] present in all other HSP70 species (Gupta and Golding 1993). We observed a second signature sequence of about 4 aa absent from all eukaryotic homologues, significantly aligned in all prokaryotic sequences. Consensus sequences were developed for eight groups [Archaea, Gram-positive, proteobacterial Gram-negative, singular bacteria, mitochondria, plastids, eukaryotic endoplasmic reticulum (ER) isoforms, eukaryotic cytoplasmic isoforms]. All group consensus comparisons tend to summarize better the alignments than do the individual sequence comparisons. The global individual consensus ``matches' 87% with the consensus of consensuses sequence. A functional analysis of the global consensus identifies a (new) highly significant mixed charge cluster proximal to the carboxyl terminus of the sequence highlighting the hypercharge run EEDKKRRER (one-letter aa code used). The individual Archaea and Gram-positive sequences contain a corresponding significant mixed charge cluster in the location of the charge cluster of the consensus sequence. In contrast, the four Gram-negative proteobacterial sequences of the alignment do not have a charge cluster (even at the 5% significance level). All eukaryotic HSP70 sequences have the analogous charge cluster. Strikingly, several of the eukaryotic isoforms show multiple mixed charged clusters. These clusters were interpreted with supporting data related to HSP70 activity in facilitating chaperone, transport, and secretion function. We observed that the consensus contains only a single tryptophan residue and a single conserved cysteine. This is interpreted with respect to the target rule for disaggregating misfolded proteins. The mitochondrial HSP70 connections to bacterial HSP70 are analyzed, suggesting a polyphyletic split of Trypanosoma and Leishmania protist mitochondrial (Mt) homologues separated from Mt-animal/fungal/plant homologues. Moreover, the HSP70 sequences from the amitochondrial Entamoeba histolytica and Trichomonas vaginalis species were analyzed. The E. histolytica HSP70 is most similar to the higher eukaryotic cytoplasmic sequences, with significantly weaker alignments to ER sequences and much diminished matching to all eubacterial, mitochondrial, and chloroplast sequences. This appears to be at variance with the hypothesis that E. histolytica rather recently lost its mitochondrial organelle. T. vaginalis contains two HSP70 sequences, one Mt-like and the second similar to eukaryotic cytoplasmic sequences suggesting two diverse origins. Received: 29 January 1998 / Accepted: 14 May 1998     

Detecting Changes in the Functional Constraints of Paralogous Genes

We describe a new procedure to determine whether regional alterations in the evolutionary constraints imposed on paralogous proteins have occurred. We used as models the A and B (alternatively called α and β) subunits of V/F/A-ATPases, originated by a gene duplication more than 3 billion years ago. Changes associated to three major splits (eubacteria versus Archaea-eukaryotes; Archaea versus eukaryotes; and among free-living bacteria and symbiotic mitochondria) were studied. Only in the first case, when we compared eubacterial or mitochondrial F-ATPases versus eukaryotic vacuolar V-ATPases or archaeal A-ATPases, constraint changes were observed. Modifications in the degree of regional constraining were not detected for the other two types of comparisons (V-ATPases versus A-ATPases and within F-ATPases, respectively). When the rates of evolution of the two subunits were compared, it was found that F-ATPases regulatory subunits evolved faster than catalytic subunits, but the opposite was true for A- and V-ATPases. Our results suggest that, even for universal and essential proteins, selective constraints may be occasionally altered. On the other hand, in some cases no changes were detected after periods of more than 2.2 billion years. Received: 24 February 2000 / Accepted: 12 August 2000     

Phylogenetic Differences in Content and Intensity of Periodic Proteins

Many proteins exhibit sequence periodicity, often correlated with a visible structural periodicity. The statistical significance of such periodicity can be assessed by means of a chi-squared-based test, with significance thresholds being calculated from shuffled sequences. Comparison of the complete proteomes of 45 species reveals striking differences in the proportion of periodic proteins and the intensity of the most significant periodicities. Eukaryotes tend to have a higher proportion of periodic proteins than eubacteria, which in turn tend to have more than archaea. The intensity of periodicity in the most periodic proteins is also greatest in eukaryotes. By contrast, the relatively small group of periodic proteins in archaea also tend to be weakly periodic compared to those of eukaryotes and eubacteria. Exceptions to this general rule are found in those prokaryotes with multicellular life-cycle phases, e.g., Methanosarcina sp., or Anabaena sp., which have more periodicities than prokaryotes in general, and in unicellular eukaryotes, which have fewer than multicellular eukaryotes. The distribution of significantly periodic proteins in eukaryotes is over a wide range of period lengths, whereas prokaryotic proteins typically have a more limited set of period lengths. This is further investigated by repeating the analysis on the NRL-3D database of proteins of solved structure. Some short-range periodicities are explicable in terms of basic secondary structure, e.g., alpha helices, while middle-range periodicities are frequently found to consist of known short Pfam domains, e.g., leucine-rich repeats, tetratricopeptides or armadillo domains. However, not all can be explained in this way.Reviewing Editor: Dr. John Oakeshott     

Nucleosomes Shape DNA Polymorphism and Divergence

An estimated 80% of genomic DNA in eukaryotes is packaged as nucleosomes, which, together with the remaining interstitial linker regions, generate higher order chromatin structures [1]. Nucleosome sequences isolated from diverse organisms exhibit ∼10 bp periodic variations in AA, TT and GC dinucleotide frequencies. These sequence elements generate intrinsically curved DNA and help establish the histone-DNA interface. We investigated an important unanswered question concerning the interplay between chromatin organization and genome evolution: do the DNA sequence preferences inherent to the highly conserved histone core exert detectable natural selection on genomic divergence and polymorphism? To address this hypothesis, we isolated nucleosomal DNA sequences from Drosophila melanogaster embryos and examined the underlying genomic variation within and between species. We found that divergence along the D. melanogaster lineage is periodic across nucleosome regions with base changes following preferred nucleotides, providing new evidence for systematic evolutionary forces in the generation and maintenance of nucleosome-associated dinucleotide periodicities. Further, Single Nucleotide Polymorphism (SNP) frequency spectra show striking periodicities across nucleosomal regions, paralleling divergence patterns. Preferred alleles occur at higher frequencies in natural populations, consistent with a central role for natural selection. These patterns are stronger for nucleosomes in introns than in intergenic regions, suggesting selection is stronger in transcribed regions where nucleosomes undergo more displacement, remodeling and functional modification. In addition, we observe a large-scale (∼180 bp) periodic enrichment of AA/TT dinucleotides associated with nucleosome occupancy, while GC dinucleotide frequency peaks in linker regions. Divergence and polymorphism data also support a role for natural selection in the generation and maintenance of these super-nucleosomal patterns. Our results demonstrate that nucleosome-associated sequence periodicities are under selective pressure, implying that structural interactions between nucleosomes and DNA sequence shape sequence evolution, particularly in introns.     

Dynamics of Tandem Repetitive Afa-Family Sequences in Triticeae, Wheat-Related Species

The Afa-family sequences in wheat-related species, Triticeae, are tandem repetitive sequences of 340 bp. All the analyzed Triticeae species carried the sequences in their genomes, though the copy numbers varied about 100-fold among the species. The nucleotide fragments amplified by PCR were cloned and sequenced, and their behavior in the evolution of Triticeae was analyzed by the neighbor-joining (NJ) method. The sequences in genomes with many copies of this family clustered at independent branches of the phylogenic tree, whereas the sequences in genomes with a few copies did not. This may suggest that Afa-family sequences had amplified several times in the evolution of Triticeae, each using a limited number of different master copies. In addition, the sequences of the A and B genomes of hexaploid common wheat indicated that the Afa-family sequences had not evolved in a concerted manner between the genomes. Furthermore, the sequences of each chromosome of the D genome of this species indicated that the sequences had amplified on all over the D-genome chromosomes in a short period. Received: 1 September 1997 / Accepted: 19 January 1998     

Gene Conversion Drives Within Genic Sequences: Concerted Evolution of Ribosomal RNA Genes in Bacteria and Archaea

Multiple copies of a given ribosomal RNA gene family undergo concerted evolution such that sequences of all gene copies are virtually identical within a species although they diverge normally between species. In eukaryotes, gene conversion and unequal crossing over are the proposed mechanisms for concerted evolution of tandemly repeated sequences, whereas dispersed genes are homogenized by gene conversion. However, the homogenization mechanisms for multiple-copy, normally dispersed, prokaryotic rRNA genes are not well understood. Here we compared the sequences of multiple paralogous rRNA genes within a genome in 12 prokaryotic organisms that have multiple copies of the rRNA genes. Within a genome, putative sequence conversion tracts were found throughout the entire length of each individual rRNA genes and their immediate flanks. Individual conversion events convert only a short sequence tract, and the conversion partners can be any paralogous genes within the genome. Interestingly, the genic sequences undergo much slower divergence than their flanking sequences. Moreover, genomic context and operon organization do not affect rRNA gene homogenization. Thus, gene conversion underlies concerted evolution of bacterial rRNA genes, which normally occurs within genic sequences, and homogenization of flanking regions may result from co-conversion with the genic sequence. Received: 31 March 2000 / Accepted: 15 June 2000     

Molecular phylogenies based on ribosomal protein L11, L1, L10, and L12 sequences

Summary Available sequences that correspond to the E. coli ribosomal proteins L11, L1, L10, and L12 from eubacteria, archaebacteria, and eukaryotes have been aligned. The alignments were analyzed qualitatively for shared structural features and for conservation of deletions or insertions. The alignments were further subjected to quantitative phylogenetic analysis, and the amino acid identity between selected pairs of sequences was calculated. In general, eubacteria, archaebacteria, and eukaryotes each form coherent and well-resolved nonoverlapping phylogenetic domains. The degree of diversity of the four proteins between the three groups is not uniform. For L11, the eubacterial and archaebacterial proteins are very similar whereas the eukaryotic L11 is clearly less similar. In contrast, in the case of the L12 proteins and to a lesser extent the L10 proteins, the archaebacterial and eukaryotic proteins are similar whereas the eubacterial proteins are different. The eukaryotic L1 equivalent protein has yet to be identified. If the root of the universal tree is near or within the eubacterial domain, our ribosomal protein-based phylogenies indicate that archaebacteria are monophyletic. The eukaryotic lineage appears to originate either near or within the archaebacterial domain. Correspondence to: P. Dennis     

Base Composition Skews, Replication Orientation, and Gene Orientation in 12 Prokaryote Genomes

Variation in GC content, GC skew and AT skew along genomic regions was examined at third codon positions in completely sequenced prokaryotes. Eight out of nine eubacteria studied show GC and AT skews that change sign at the origin of replication. The leading strand in DNA replication is G-T rich at codon position 3 in six eubacteria, but C-T rich in two Mycoplasma species. In M. genitalium the AT and GC skews are symmetrical around the origin and terminus of replication, whereas its GC content variation has been shown to have a centre of symmetry elsewhere in the genome. Borrelia burgdorferi and Treponema pallidum show extraordinary extents of base composition skew correlated with direction of DNA replication. Base composition skews measured at third codon positions probably reflect mutational biases, whereas those measured over all bases in a sequence (or at codon positions 1 and 2) can be strongly affected by protein considerations due to the tendency in some bacteria for genes to be transcribed in the same direction that they are replicated. Consequently in some species the direction of skew for total genomic DNA is opposite to that for codon position 3. Received: 2 February 1998 / Accepted: 15 June 1998