MBE Advance Access originally published online on December 5, 2006
Molecular Biology and Evolution 2007 24(2):482-497; doi:10.1093/molbev/msl184
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Articles |
Origin of the Bacterial SET Domain Genes: Vertical or Horizontal?
* School of Biological Sciences, University of Nebraska–Lincoln
Department of Genetic Engineering, Centro de Investigación y de Estudios Avanzados—Unidad Irapuato, Guanajuato, CP 36821, México
Faculty of Agriculture, University of Jordan, Amman, Jordan; and
Protein Array Center, Invitrogen Corporation, Branford, Connecticut
E-mail: zavramova2{at}unl.edu.
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResults and DiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
The presence of Supressor of variegation-Enhanser of zeste-Trithorax (SET) domain genes in bacteria is a current paradigm for lateral genetic exchange between eukaryotes and prokaryotes. Because a major function of SET domain proteins is the chemical modification of chromatin and bacteria do not have chromatin, there is no apparent functional requirement for the existence of bacterial SET domain genes. Consequently, their finding in only a small fraction of pathogenic and symbiotic bacteria was taken as evidence that bacteria have obtained the SET domain genes from their hosts. Furthermore, it was proposed that the products of the genes would, most likely, be involved in bacteria–host interactions. The broadened scope of sequenced bacterial genomes to include also free-living and environmental species provided a larger sample to analyze the bacterial SET domain genes. By phylogenetic analysis, examination of individual chromosomal regions for signs of insertion, and evaluating the chromosomal versus SET domain genes' GC contents, we provide evidence that SET domain genes have existed in the bacterial domain of life independently of eukaryotes. The bacterial genes have undergone an evolution of their own unconnected to the evolution of the eukaryotic SET domain genes. Initial finding of SET domain genes in predominantly pathogenic and symbiotic bacteria resulted, most probably, from a biased sample. However, a lateral transfer of SET domain genes may have occurred between some bacteria and a family of Archaea. A model for the evolution and distribution of SET domain genes in bacteria is proposed.
Key Words: horizontal gene transfer • SET domain genes • chromatin proteins in bacteria • bacterial domain of life • lateral gene transfer
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
In eukaryotes, chromatin structure provides an additional level of gene regulation by modulating genes' accessibility to the transcriptional machinery. Factors that alter chromatin structure are defined as epigenetic. They provide "memory" of transcriptional states that are faithfully reproduced after each round of cell division and throughout the development of an organism. Two antagonistically acting groups of genes, the Polycomb repressors and the Trithorax activators, are responsible for maintaining the activity of homeotic genes in higher eukaryotes. Ever since the discovery of a highly conserved (
130 amino acids) peptide Supressor of variegation-Enhanser of zeste-Trithorax (SET) in proteins belonging to both the repressor Su(var)3-9, E(z) and the activator (Trithorax and Ash1) groups, it has been expected that the SET domain peptide plays some important role. However, this role has remained a mystery until the recent discovery that SET domain peptides can methylate lysines at specific locations on the histone tails (Rea et al. 2000
). Modified amino acids on the histone tails provide tags that are "read" by other complexes creating or destroying affinities for chromatin regulators. The combinatorial nature of these modifications commands transitions between active and inactive states, extending the informational potential of the genetic code (Felsenfeld and Groudine 2003). These amino-terminal modifications constitute a "histone code" and a molecular basis of epigenetic regulation (Jenuwein and Allis 2001).
Because SET domain–containing proteins may modify chromatin structure, not surprisingly, SET domain–encoding genes have been found in all eukaryotic genomes sequenced so far, from the unicellular primitive eukaryotes to the multicellular animals and plants. Absence of SET domain genes from a large number of bacterial genomes has provided support for an assumption that they have appeared with the occurrence of the eukaryotes (Stephens et al. 1998; Alvarez-Venegas and Avramova 2002; Aravind and Iyer 2003). Analyzing the genome of the obligate intracellular pathogen Chlamydia trachomatis, Stephens et al. (1998) identified a SET domain gene and suggested that it has originated via horizontal gene transfer (HGT) from a eukaryotic host. As additional bacterial genomes were sequenced, SET domain genes were identified in more pathogenic and symbiotic bacterial species. A logical assumption was made that the presence of SET domain genes in bacterial species is a consequence of their contacts with eukaryotic cells and that bacteria have acquired the gene through HGT (Alvarez-Venegas and Avramova 2002; Aravind and Iyer 2003). Several facts supported this assumption: First, bacteria lack chromatin structure (no need for epigenetic regulation); second, among more than 390 sequenced bacterial genomes (at the time of this study), only 83 carry SET domain sequences, and the majority of these bacteria are pathogens or symbionts; third, closely related species differ in whether they carry a SET domain–encoding gene; fourth, with the exception of 3 Methanosarcina species, sequenced Archaebacteria lack SET-related genes.
Transfer of eukaryotic genes is considered common in parasitic and symbiotic bacteria. For instance, the intimate association between Chlamidiaceae and host cells might favor horizontal gene flow (Koonin et al. 2001). However, analysis of multiple Chlamidiae genomes has indicated little genomic exchange with other genera (Read et al. 2000; Brinkman et al. 2002). The idea of a widely spread HGT, especially between species from the different domains of life, has become a hotly debated issue. Opinions range from HGT being overwhelming and rampant, obscuring possible phylogenetic relationship between the species, to being overemphasized, limited, and insufficient to "unroot" the tree of life (reviewed in Glandsdorff 2000; Ochman et al. 2000; Woese 2000; Brown 2003).
Here, we revisited the paradigm for a gene transfer across the eukaryotic and bacterial domains of life and analyzed the phylogeny of the SET domain–encoding genes found in bacteria. The broadened scope of sequenced bacterial genomes, to include also free-living and environmental species, provided an impetus to reexamine the distribution of SET domain sequences in a larger sample of bacterial genomes. Our goals were to determine, first, whether the presence of a SET domain in closely related species would be connected to their lifestyles, free versus parasitic; second, whether the earlier conclusion that SET domain genes have been horizontally acquired by pathogenic and symbiotic bacteria were biased by the available sample of sequenced genomes selectively representing pathogenic and agronomicaly significant species; and third, whether phylogenetic relationships between the bacterial SET domain genes could suggest occurrence and evolution unrelated to the eukaryotic SET domain genes.
Our analyses indicate that SET domain genes have existed in the bacterial domain of life and that their initial finding in pathogens and symbionts resulted from a biased sample. Importantly, bacterial SET domain genes have undergone an evolution of their own, unconnected to the evolution of the eukaryotic SET domain genes. Absence of SET domain sequences in the majority of currently available bacterial genomes, apparently, reflects gene loss. Phylogenetic and chromosome analyses of the SET domain gene–containing regions of Chlorobium, Bacillus, and Methanosarcinal genomes, however, suggest a possible HGT between some bacteria and Archaea.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Phylogenetic Analyses of Bacterial SET Domain–Containing Proteins
From approximately 400 completely and partially sequenced bacterial genomes in the National Center for Biotechnology Information (NCBI) database, we have retrieved 83 bacterial species encoding putative SET domain proteins (table 1). Excluding identical sequences from very closely related genomes, we analyzed 45 SET domain proteins found in 39 bacterial species; duplicate genes representing paralogous functions within the same species' genomes are included. Eukaryotic entries were selected to represent a broad cross section of proteins found in unicellular, filamentous, and multicellular organisms. Members from different SET domain families, as recognized in plant and animal systems (Baumbusch et al. 2001
; Alvarez-Venegas and Avramova 2002; Marmostein 2003), are included. SET domain peptides have a tripartite structure of conserved N- and C-boxes separated by an inserted middle module of variable length and composition (Aravind and Iyer 2003; Marmostein 2003; see also fig. 3). To achieve optimal alignment, eukaryotic proteins with insertions comparable to the lengths of insertions in the bacterial proteins (20–30 amino acids) were selected. A full list of aligned sequences is summarized in the supplementary figure SF1, Supplementary Material online. Database searches were performed with Blast and PSI-Blast programs on the NCBI nonredundant database. SET domain–containing proteins were collected by TBlastX and PSI-Blast searches (E value 0.001). Pairwise alignments were compiled using the ClustalW program (Chenna et al. 2003). Phylogenetic and bootstrap analyses using the Protpars method from the PHYLIP package and the Seqboot program (500 pseudoreplicates) were employed. Unrooted majority rule consensus trees were built with the CONSENSE and plotted using the TREEVIEW programs (Page 1996). The fitch function was used to make minimum evolution trees using Phylip software (Felsenstein 1989). Minimal evolution (ME) trees with 10 global rearrangements and 10 random jumbles are shown in supplementary figures SF2 and SF3, Supplementary Material online.
|
|
Phylogenetic analyses for supportive bacterial 23S ribosomal RNAs and bacterial 50S ribosomal protein L3 were performed as specified in the legends of the supplementary figures SF4 and SF5, Supplementary Material online.
Genome analyses and localization of bacterial SET domain genes were carried individually for each of the bacterial genomes carrying SET domain genes. To outline syntenic regions flanking the SET gene, bacterial genomes were compared by pairwise alignment. Comparisons were carried out at both DNA sequence levels and at amino acid level using published genome data.
|
Results and Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Distribution of SET Domain–Encoding Genes among Bacteria
Analysis of overall distribution of SET domain–encoding genes among sequenced bacterial genomes (NCBI) revealed the following facts (summarized in table 1): First, the retrieved bacterial SET domain genes are present in most of the known bacterial domains (Cyanobacteria, photosynthetic green sulfur, Flexobacter–Bacteroides, Spirochaetaes, Chlamidiae, Planctomycetaes, and low G + C Gram-positive and
, ß, 
, and 
Gram-negative bacteria). Clearly, SET domain genes have existed before the separation of these branches. Second, although most species represent obligatory pathogens and symbionts, SET domain genes are found also in environmental species (opportunistic pathogens or symbionts) and in free-living organisms for which no symbiotic relationships have been found. Examples of the first group are Leptospira interrogans, Burkholderia fungorum, and Bradyrhisobium japonicum. The second include Ralstonia matallidurans, Rhodopseudomonas palustris, Chlorobium tepidum, Rubrivivax gelatinosus, and Verrucomicrobium spinosum. SET domain genes exist in organisms living at arctic temperatures (Polaromonas) and in hot springs (Chlorobium). Thus, it sounds unlikely that bacteria living on its own and under extreme conditions have acquired the SET domain gene through a eukaryote. Third, among the SET domain–containing bacteria, we identified 6 species that contain more than 1 copy of a SET gene. This finding is particularly important because it illustrates duplication events and evolution of bacteria-specific SET domain paralogs. We note also the highly variable presence of SET domain genes in closely related species. Of the available sequenced genomes of the - and 
-subdivisions, only 1 representative of the -subdivision (Myxoccocus xanthus) carries a SET domain gene, whereas none of the -subdivision has been found yet. In contrast, all reported species from the -subdivision (the Xylella and Xanthomonas species) have SET domain genes. In the ß-subdivision, all members of the Burkholderiacea and the Bordetella families have SET genes, whereas no member from the Neisseriaceae family carries any. Burkholderia cepacia and Ru. gelatinosus (from an undefined ß-subdivision group) have 2 SET domain genes in each genome.
Among 29 genomes from the -proteobacterial subdivision, available at the time of this study, only 3 species (Br. japonicum, Rh. palustris, and Mesorhizobium loti) carried SET domain genes. It is remarkable that both Br. japonicum and Me. loti have 2 SET domain genes each, whereas closely related species (symbiotic Rhizobiaceae and pathogenic Rickettsiales) may have none.
Outside the proteobacterial domain, the distribution of the SET domain–containing genomes is both broad and sporadic. Chlamydial species (with the exception of Parachlamydia, UWE25) carry SET domain genes, whereas none of the 23 Actinobacteria genomes has it. Species from Cyanobacteria (all reported Nostoc species), Planctomycetaes (Pirellula), photosynthetic high-temperature inhabiting (Chlorobium), Flexobacter–Bacteroides (Cytophaga), Verrucomicrobia (Verrucomicrobium), and Spirochaetales (Leptospira) have SET domain genes. Some have more than 1 copy, indicating paralogy. Interestingly, among more than 60 sequenced low G + C Gram-positive bacteria, only 3 Bacilli (Bacillus cereus, Bacillus anthracis, and Bacillus thuringiensis) have SET domain genes. Lastly, only 3 mesophylic Archaea species carry SET domain genes, intriguingly similar to the SETs of the Bacilli.
Collectively, these observations argue that SET genes have existed in the bacterial domain of life before the occurrence of eukaryotes. Absence of SET domain genes from many bacterial genomes could reflect gene loss.
The Eukaryotic–Bacterial SET Domain Tree
Currently, the 2 major arguments supporting a lateral gene transfer between the eukaryotic and bacterial kingdoms are: 1) the presence of SET domain genes in the obligatory intracellular parasites from the orders Chlamydophila/Chlamydia and 2) the high levels of similarity of the SET domain proteins from the pathogenic -Proteobacteria (Xanthomonas and Xylella) and eukaryotes. The first argument is based on the assumption that intracellular Chlamydiae, having diverged from eubacteria some 2 billion years ago (Horn et al. 2004 and references therein), constitute an isolated niche sheltered from exchanges with other bacteria. However, living within eukaryotic hosts is a plausible condition enabling a horizontal acquisition of a SET domain gene (Stephens et al. 1998; Aravind and Iyer 2003). The second argument, based on the high similarity (2 x 10–20, 38% identical) of the SET domain protein of the plant pathogen Xylella fastidiosa to a rice SET protein, is taken as evidence for a recent transition from a host genome to the genome of the invading bacteria (Aravind and Iyer 2003).
To examine the relationships between eukaryotic and bacterial SET domain genes, we reconstructed phylogenetic trees by several different approaches. Phylogenetic analysis is an objective approach for determining the occurrence and the directionality of HGT, despite some recognized limitations (Stanhope et al. 2001). Thereby, we carried out additional analyses to provide independent support employing different (genome-based) approaches.
The reconstructed maximum parsimony (MP) tree (fig. 1) shows all eukaryotic entries clustered on 2 related branches (69% bootstrap). The support is not very strong, but 2 observations relevant for our further discussion are that eukaryotic proteins do not intermix with proteins of bacterial origin and that a similar distribution pattern is consistently reproduced by trees built by different techniques (ME tree; supplementary fig. SF2, Supplementary Material online) and with different combinations of eukaryotic entries (data not shown). According to the criteria of Stanhope et al. (2001), these results did not support an HGT from eukaryotes to prokaryotes. Interestingly, however, they illustrated complex relationships among the bacterial proteins including a possible HGT among Bacilli and the archaeal species (see further below).
|
The Bacterial SET Domain Tree
To explore SET domain protein relationships among bacteria, we reconstructed trees using only the bacterial SET domain sequences (fig. 2; supplementary fig. SF3, Supplementary Material online). As controls, we built trees for bacterial genes less likely to be subjected to HGT. rRNA-based trees are considered "immune" to HGT and are largely supported by genome trees (van Berkum et al. 2003
). Other genes, postulated to be less prone to horizontal shuffling, code for highly integrated elements tightly coupled with a functioning integral system, like individual ribosomal proteins (Woese 1998). MP trees for the 23S rRNA genes and for the conserved 50S ribosomal subunit protein, L3, of the bacterial species carrying SET domain genes (supplementary figs. SF4 and SF5, Supplementary Material online) are in general agreement with the SET domain protein tree, although some relationships are not well supported.
|
Closer examination of the bacterial SET domain tree revealed that the bacterial proteins could be separated into 2 distinct domains, arbitrarily called here Domain 1 and Domain 2 (fig. 2). The positioning of each protein within a Domain reflects a characteristic structural feature of its SET domain. Structural studies of eukaryotic SET domain proteins have led to the discovery of an unusual fold, the "knot" (Jacobs et al. 2002
). Two conserved peptides, known as the N-terminal and the C-terminal boxes, flank a nonconserved insertion module of variable length and composition (fig. 3). The N- and the C-terminal boxes form the knot and carry the 2 most conserved amino acid motifs involved in the formation of the "loop" and the "thread." Overlapping with these motifs are the NHXC and the GEELXXXY consensus sequences involved in the cofactor-binding and the enzyme active sites, respectively (reviewed in Marmostein 2003; fig. 3). Relevant for our analysis is the Cys residue in the NHXC box. While Asn and his amino acids are conserved in all known SET domain proteins, the Cys is conserved only in a subset (Marmostein 2003). As a rule, proteins with a Cys in the box carry also a conserved motif (CXCXXXXC) downstream of the SET domain, known as the post-SET domain. Structural studies have shown that the C from the NHXC box and the CXCXXXXC motif may coordinate a zinc atom to form a Zn finger. It plays a role in the substrate specificity of the SET domain methylases (Xiao et al. 2003). A Cys in the NHXC box predicts presence of the post-SET motif, whereas absent C is always accompanied by absent post-SET domain. No violation of this structural rule has been found so far and the bacterial SET domain peptides make no exception (fig. 3). From hereon, presence or absence of post-SET domain motifs in the bacterial proteins is denoted as (+)pSET and (–)pSET, respectively. All proteins in Domain 1 carry the (+)pSET version, whereas those on Domain 2 lack the post-SET motif. This structural feature reflects their distribution into distinct phylogenetic domains. We suggest that the absence of a post-SET domain represents a secondary event in the evolution of the bacterial SET domain function resulting from a single mutation of the C in the NHXC box. In the absence of evolutionary pressure to keep the Zn finger, a loss of the post-SET domain subsequent to this mutation is a likely outcome.
Domain 1: The (+)pSET Domain Bacterial Proteins
Grouped in Domain 1 are proteins of the (+)pSET type (fig. 3). The SET domain proteins of the ß-, -, and -Proteobacteria display common origins, in agreement with the respective 23S RNA and S50-L3 protein trees (supplementary figs. SF4 and SF5, Supplementary Material online). Incongruent is the clustering of V. spinosum with Proteobacteria in all SET domain trees (figs. 1 and 2; supplementary figs. SF2 and SF3, Supplementary Material online). Verrucomicrobium spinosum is a free-living environmental species from the Chlamidiae/Verrucomicrobia group (Schlesner 2004) and the clustering of its SET domain protein with Proteobacteria, but not Chlamydia, might indicate HGT for the V. spinosum gene. However, it is possible also that the SET gene has an ancestral origin in V. spinosum, a possibility discussed later in more detail.
Representatives of the - and ß-subdivisions segregate into the most populous and best-supported clades. Two species, B. cepacia and Ru. gelatinosus, have 2 (+)pSET domain copies in their genomes each (ZP_00425417 and YP_771959 and ZP_00241588 and ZP_00243950, respectively). These genes form sister groups with proteins from other species and, thus, represent paralogs. The -group is represented by 2 highly related species, Br. japonicum and Rh. palustris, with 80% identical SET domain proteins. We note that the (+)pSET version of these species is slightly diverged (fig. 3). Two free-living species, Pirellula sp. and Nostoc punctiforme, have related (+)pSET proteins. Pirellula, a marine bacterium from the order Planctomycetales, is considered to be of an independent monophyletic origin with no clear ancestral relationships to the other bacteria (Glockner et al. 2003). However, the SET domain protein of Pirellula sp. always clusters with the SET domain protein of the Cyanobacterium (Nostoc), pointing to a common, albeit distant, ancestry.
Domain 2: The (–)pSET Domain Bacterial Proteins
A major distinction of Domain 2 members is that all proteins are of the (–)pSET type, with the exception of one L. interrogans copy. Species of broadly diverse origins, including 3 members of the archaebacterial, Methanosarcinae family are grouped in Domain 2. The SET domain genes from 2 rhizobial species are the only representatives of Proteobacteria in this phylogenetic domain.
The Origin and the Evolution of the Bacterial SET Domain Genes May Be Traced in the Spirochaetae L. interrogans
Leptospira interrogans carries 3 SET domain copies: 1 (+)pSET and 2 (–)pSET types only weakly related to each other. The (+)pSET (NP_71078) is weakly (6 x 10–6) similar to one of the (–)pSET (NP_710330
[GenBank]
), and there is no significant similarity to the other (–)pSET (NP_714158
[GenBank]
). The 2 (–)pSET copies are distantly related (8 x 10–3), suggesting that each of the 3 L. interrogans genes represents an ancient paralog descending from a distinct ancestral gene line (fig. 5).
|
Both a facultative parasite and a saprophyte that can strive on its own, L. interrogans is related to the strictly parasitic spirochaetae, Borrelia burgdorferei and Treponema pallidum, but the latter do not have SET domain genes. Only 315 genes are shared between the 3 species and it is thought that species-specific genes provide L. interrogans with opportunities not required for the obligatory parasitic spirochaetae (Ren et al. 2003
). Clearly, the 3 SET domain genes belong to the category distinguishing the free-living organism from the strictly parasitic relatives. Among eubacteria, spirochaete are evolutionarily primitive and their origins are not clear. On the 23S RNA tree, L. interrogans was remotely related to the Proteobacteria and Chlamydiae species, whereas its 50S L3 protein related it to the Bacteroides–Flexobacter group (Cytophaga hutchinsonii) and to the green-sulfur cyanobacterium, Ch. tepidum (supplementary figs. SF4 and SF5, Supplementary Material online). Through its 3 SET domain genes, L. interrogans related phylogenetically to all these bacterial groups and, thus, may be defined as a bearer of gene copies descending from ancient paralogs (see also fig. 5).
The (+)pSET L. interrogans protein, NP_710786
[GenBank]
, lacks a significant portion of the N-box sequences that accounts for its positioning in Domain 2, despite the presence of a post-SET region (fig. 3). The other 2 L. interrogans SET domain copies represent 2 different versions of the (–)pSET. Each copy is related to SET domain proteins present in unrelated bacterial lineages: one (NP_710330
[GenBank]
) is most similar to the SET proteins found in 2 -proteobacterial rhizobial species, Me. loti and Br. japonicum; the other (–)pSET protein (NP_714158
[GenBank]
) is related to the Chlamydiacea proteins, to a member of the large Cytophaga–Flavobacterium–Bacteroides subphylum, Cy. hutchinsonii, to Chlorobium, and to 3 proteins from Bacillales. Thereby, the 3 genes of L. interrogans represent bacterial SET domain gene paralogs. The relationships between the SET domain copies of L. interrogans and the other bacteria may provide important clues on the relationship and the evolution of these genes in bacteria and are analyzed in more detail.
The L. interrogans (–)pSET NP_710330 Protein and the Rhizobial Symbionts
The spirochaetal gene (NP_710330) is much closer to the 2 genes found in Me. loti (7 x 10–17 and 2 x 10–12) and to 1 of the 2 genes found in Br. japonicum (7 x 10–15) than to any of the copies in its own genome.
The 2 rhizobial symbionts Me. loti and Br. japonicum are closely related and share common ancestry (Kaneko et al. 2000, 2002). Their (–)pSET domain proteins are related between themselves (5 x 10–18) and with the L. interrogans (NP_710330
[GenBank]
, 7 x 10–17 and 7 x 10–15, respectively), and all 4 proteins form a common clade. We may propose that the gene has been present in an ancestor before the separation of the L. interrogans from the Me. loti–Br. japonicum branch but later, after the separation of Me. loti and Br. japonicum, the gene has been duplicated in Me. loti. The fact that the 2 Me. loti proteins share higher homology between themselves (2 x 10–28) than to any other SET domain protein supports this model.
The relationships between the SET domain proteins of the -proteobacterial rhizobial species, however, are more complex due to facts that none of the other available sequenced genomes of rhizobial symbionts carry SET domain genes and that in Br. japonicum, there is a second copy of a SET domain gene less similar to the other Br. japonicum and Me. loti genes. This Br. japonicum NP_772427
[GenBank]
protein is of the (+)pSET type and is phylogenetically related and highly similar to the protein found in the free-living -proteobacterium, Rh. palustris ZP_0000868 (8 x 10–67; 80% identical). The post-SET domain motif (CXCXXC) in the proteins of these species slightly differs from the (CXCXXXXC) motif found in all other known SET domain proteins (fig. 3). Compared with other bacterial and eukaryotic SET domain proteins, the Br. japonicum and Rh. palustris proteins showed similarity only to one of the L. interrogans proteins, NP_714158
[GenBank]
(7 x 10–11 and 2 x 10–11, respectively). The 2 copies of Br. japonicum were less similar to each other (1 x 10–5, 27% identical) than to proteins found in other species, indicating gene paralogy.
Mesorhizobium loti, Br. japonicum, and Rh. palustris are of a common ancestral origin (Larimer et al. 2004) and, thus, their SET domain genes may reflect the evolutionary history of the genes. We may suggest that their common ancestor has carried 2 paralogous genes, a (+)pSET and a(–)pSET: Br. japonicum has inherited both, Rh. palustris has inherited the (+)pSET, whereas Me. loti has inherited the (–)pSET. Subsequently, in Me. loti, this gene has undergone species-specific duplication (Model, fig. 5).
The L. interrogans Protein NP_714158 and the Cytophaga Subphylum
On the phylogenetic trees, the L. interrogans protein, NP_714158
[GenBank]
, clusters with the 2 copies of Cy. hutchinsonii, a member of the Cytophaga–Flavobacterium–Bacteroides subphylum (Gherna and Woese 1992) (figs. 1 and 2; supplementary figs. SF2 and SF3, Supplementary Material online). The 2 (–)pSET Cytophaga proteins (most similar to each other, 3e –09, than to any other SET domain protein) apparently represent a gene duplication event that had occurred after the separation of the species from its closest SET domain relatives. Thereby, the L. interrogans protein NP_714158
[GenBank]
represents a version of the (–)pSET domain protein existing in bacteria before the separation of Spirochatales, Chlamydiae, Cytophaga, Chlorobia, and Bacillae (fig. 5).
The SET Domain Genes of Chlamydia
A distinctive Gram-negative family of uncertain evolutionary origin deeply separated from other eubacteria, Chlamydiae is thought to represent one of the kingdom-level branches on the phylogentic tree (Pace 1997). A distant relationship to Planctomycetales (Amann et al. 1997) and Cyanobacteria (Brinkman et al. 2002) was suggested. On the 23S RNA and the S50-L3 protein trees, Chlamydiacea did not show supported relationship to any particular bacterial group (supplementary figs. SF4 and SF5, Supplementary Material online).
Phylogenetic analysis of the SET domain proteins positioned Chlamydiacea in a clade of its own, among the (–)pSET domain–type bacteria, with no significant relationship to the eukaryotic cluster (figs. 1 and 2). Compared with eukaryotic and bacterial SET domain proteins available in the database, the chlamydial SET domain proteins showed significant similarity only to each other (1 x 10–86–1 x 10–65) and to one of the L. interrogans proteins, NP_714158 [GenBank] (6 x 10–11). This gene version is related also to the proteins found in Br. japonicum, Cytophaga, Chlorobium, and Bacillus (see also fig. 5), suggesting ancestral bacterial origins for the chlamydial genes. Interestingly, the SET domain protein of the environmental V. spinosum (Chlamydiae/Verrucomicrobia group) belongs to the (+)pSET domain type. A plausible scenario is that a common ancestor has carried both the (+)pSET and the (–)pSET ancestral forms and that Chlamydiae have inherited one of the ancestral copies, the Verrucomicrobia lineage has kept the other, whereas the parachlamydial environmental relative, UWE25, has lost both.
Thereby, the phylogenetic relationships of the SET domain proteins of the Chlamydiae–Verrucomicrobia group do not support a horizontal gene transfer from a mammalian genome, as earlier suggested (Stephens et al. 1998; Aravind and Iyer 2003). The SET domain proteins of V. spinosum and Chlamydiae cluster into separate clades representing 2 paralogs bared by a common ancestor.
The SET Domain Proteins of Species Unrelated by Origin Cluster Together
Unrelated by the species origin, the SET domain proteins of the free-living green-sulfur photosynthetic bacterium Ch. tepidum, of 3 archaeal (Methanosarcina) species, Methanosarcina acetivorans, Methanosarcina barkeii, and Methanosarcina mazei, and of 3 Bacillus species, share a well-supported clade. It is remarkable that the proteins from these 3 groups always segregate together, independent of the method of tree construction (figs. 1 and 2; supplementary figs. SF2 and SF3, Supplementary Material online). The SET domain genes of the Bacilli, Chlorobium, and Methanosarcina are closer than relationship between their genomes, in general. Chlorobium tepidum has many genes for metabolic processes that are more similar to archaeal species than to other bacteria, and it was suggested that extensive HGT between Archaea and hot-spring bacteria had occurred (Nelson et al. 1999). Archaea has been defined as a separate domain of life (Woese and Fox 1977), although many of its metabolic pathways resemble more bacterial than eukaryotic counterparts, suggesting exchange between bacteria and Archaea (Doolittle and Logsdon 1998; Koonin et al. 2001). Methanosarcinae have large genomes and numerous COGs that no other Archaea members have, but, notably, most of these functions are present in various bacteria. Methanosarcinae are considered a "sink" for horizontally acquired bacterial genes. About one-third of M. mazei ORFs have significant hits in the bacterial genomes referred to as "bacteria-like" (Deppenmeier et al. 2002). It is thought that HGT from bacteria, Gram-positive in particular, has played an important evolutionary role in shaping its physiology (Brown 2003).
Within the large group of sequenced Gram-positive bacteria, only the 3 closely related Bacillus species, Ba. anthracis, Ba. cereus, and Ba. thuringiensis, have SET domain genes. They are 100% identical among themselves and surprisingly similar to the SET domain proteins from the 3 Methanosarcinal species (2 x 10–25, 51% identical). Because genome-tree analysis has unequivocally supported the monophyly of Archaea, the bacteria-like metabolic genes found in Archaea members might have been transferred from bacteria (Galagan et al. 2002). The segregation of the Methanosarcinae and the Bacillus SET domain proteins in the same clade is also consistent with an HGT (Stanhope et al. 2001).
The relationships are more complex due to the clustering in the same clade of the SET domain protein of Chlorobium. Whole-genome analysis has suggested that HGT might have occurred between thermophylic eubacteria and archaea (Nelson et al. 1999; Eisen et al. 2002). However, Methanosarcinae do not live under extreme conditions and do not share habitats with Ch. tepidum, but they may with Ba. anthracis, Ba. cereus, and Ba. thuringiensis (Bintrim et al. 1997; Radnedge et al. 2003). To explore possibilities for a lateral exchange of genetic material between bacteria, in general, and between the Bacillae and Methanosarcinae, in particular, we examined the genomic regions surrounding the SET domains in all bacterial genomes for molecular signs of HGT at the chromosomal level.
Analysis of the SET Domain–Containing Genomic Regions for Signs of Insertion
A major argument in support of a horizontal transfer of SET domain genes to bacteria has been the fact that the majority of sequenced bacterial genomes do not carry the gene. The patchy distribution of genes and gene clusters on prokaryote chromosomes reveals an underlying process of recurrent loss and gain (Doolittle 1999; Ochman et al. 2000; Makarova and Koonin 2003).
Although our phylogenetic data did not support eukaryotic origins of the bacterial SET domain genes, they left open a possibility that SET domain genes might have been exchanged between genomes of bacteria and/or Archaebacteria. To explore this possibility, we analyzed the chromosomal regions around the SET domain genes for recognizable signs. Hallmarks of HGT are presence of specific elements affecting integration, such as remnants of mobile elements, rearrangements bordered by identical copies of insertion sequences (IS), clusters of transposases, and interrupted synteny. Extra chromosomes, plasmid vectors, or pathogenicity- and symbiosis islands on the main chromosomes of some pathogenic and symbiont species are considered potential vehicles of foreign genes (Salanoubat et al. 2002; Parkhill et al. 2003). Indirect evidence for ancestral versus horizontal origin of bacterial genes is their GC content. The average GC ratio is considered characteristic for a particular microbial genome, and regions in the DNA with changed ratios are thought to reflect recent horizontal transfer. Consequently, we analyzed the genome arrangements and the GC contents of the SET domain genes in the genomes of parasitic and symbiont species as well as in the Chlorobium, Bacilli, and Methanosarcinae clade.
The SET Domain Genes in All Pathogenic and Symbiont Species Are Chromosomally Located
- and ß-Proteobacterial Genomes.
In - and ß-proteobacterial genomes, genes involved in pathogenicity or other kinds of host–bacterial interaction systems are located on specific "islands" bordered by phage integrases. Whole-genome differences between the Xylella and Xanthomonas families are limited to phage-associated chromosomal rearrangements and deletions (Bhattacharyya et al. 2002; Van Sluys et al. 2003). The SET domain genes in Xylella/Xanthomonas genomes are not associated with the pathogenicity islands (Alfano and Collmer 1997; Galan and Collmer 1999) and, according to consensus criteria, the SET domain genes are not involved in interactions with the host. The GC content of the SET domain genes of Xanthomonas and Xylella are 49–51%, and their overall genomic contents are 51–53%, respectively.
Phylogenetically, X. fastidiosa is placed at the base of the -Proteobacteria (Bhattacharyya et al. 2002), implying that the remaining members of the subfilum have inherited the SET domain gene. The chromosomal location of the SET domain gene is consistent with a role in bacterial functions not necessarily related to interactions with the host.
ß-Proteobacterial Genomes.
Among ß-proteobacterial genomes, SET domain genes are found in virulent species from the Bordetella family and in the large Burkholderia family including bacteria with broad ecological habitats: environmental biodegradative bacteria (B. fungorum), plant pathogens (Ralstonia solanacearum), and free-living soil bacteria (Ralstonia matallidurans). The SET domain genes are among the conserved core genes shared by the 3 Bordetella (Bordetella bronchiseptica, Bordetella pertusis, and Bordetella parapertusis) genomes. The 3 Bordetella genomes are highly rearranged, each rearrangement bordered by identical copies of IS 1001 or IS 1002 elements (Parkhill et al. 2003). Despite an overall lack of colinearity of the 3 chromosomes, the SET genes are found in remarkably conserved syntenic regions away from the virulence systems. This implies that the SET domain genes, together with linked sequences come from a common ancestor. The SET domain genes of the 3 Bordetella species have GC content (64–66%) comparable to the reported respective genomic contents (67–69%).
The SET domain genes in R. solanacearum, B. fungorum, and R. matallidurans are on the main chromosomes, away from regions associated with effectors of host interactions or the ability to colonize (Coenye and Vandamme 2003). The GC content for the SET genes of B. fungorum, R. matallidurans, and R. solanacearum (61–65%) are comparable to the GC ratios of overall chromosomal contents (62–66%) supporting ancestral origins for the SET domain genes. Furthermore, in B. fungorum and R. matallidurans, the SET domain genes are in largely syntenic regions, whereas in R. solanacearum, the synteny is "broken" due to the abundant presence of integrases and transposases in this genome (Salanoubat et al. 2002). This fact is relevant because the conserved synteny in the SET domain regions of B. fungorum and R. matallidurans indicates that the gene has been present before the separation of these species.
-Proteobacteria.
The –Proteobacteria, Me. loti and Br. japonicum, provide examples of species involved in symbiotic relationships with eukaryotes. It is thought that the rhizobial lineages have diverged well before the evolution of the legumes and that the genes for the symbiosis were subsequently acquired by lateral transfer (Mergaert et al. 1997; Broughton and Perret 1999). Mesorhizobium loti carries a mobile symbiotic island that can convert a soil saprophyte into a symbiont (Sullivan et al. 2002). The pair of the SET genes, however, is on the chromosome flanking the island, making it unlikely that they belong to the mobile category. In the closely related Br. japonicum, the putative island is not known to be mobile providing no basis for possible mobility of its SET domain genes (Goettfert et al. 2001; Kaneko et al. 2002). The GC contents are 60.4% and 61.2% for the Me. loti and 62% for the Br. japonicum (–)pSET domain genes, whereas the genomic GC content is 64% for Me. loti and 66% for Br. japonicum.
The second, (+)pSET domain, copy of Br. japonicum is in a 90-kb region colinear with Rh. palustris. This remarkable synteny is interrupted by a block of genes, with best hits to genes found in other ß- (B. fungorum, R. matallidurans, and R. solanacearum) and - (Xanthomonas axonopodis and Xanthomonas campestris) bacterial chromosomes. These findings are relevant because they reflect the close relationships between the -, ß-, and -proteobacterial species, supporting a common origin for the (+)pSET proteobacterial genes.
The Genomes of Chlamydiae.
The genomes of Chlamydiae are small, preserving only the minimum of genes required for their exclusive lifestyle as obligatory intracellular parasites. It was suggested that parasitic Chlamidiae have recruited SET domain genes from eukaryote hosts and are using them as an invading tool (Stephens et al. 1998; Aravind and Iyer 2003). However, our phylogenetic analysis did not support a relationship between the chlamydial and eukaryotic SET domain genes (fig. 1; supplementary fig. SF2, Supplementary Material online). Whole-genome analysis of Chlamydiacea revealed absence of genes typical for chromosomal rearrangements, indicating lack of mechanisms for entry of foreign DNA (Read et al. 2000, 2003; Horn et al. 2004). It was suggested that Chlamydia and Chlamydophila could not exchange genes with their hosts, or other bacteria, and that the eukaryotic-similar genes found in their genomes reflected not lateral, but ancient evolutionary relationships (Brinkman et al. 2002). The GC content for all genes (or ORFs) of chlamydial species infecting only humans is 41%. This is much lower than the GC contents of their mammalian hosts (52%) and may be taken as evidence of a lack of exchange with eukaryotic DNA. The GC content of the Chlamydia/Chlamydophila SET domain genes is 41.7–43.1%, in good agreement with the other ORFs (Brinkman et al. 2002). In addition, the significant synteny around the SET domain genes on all chlamydial chromosomes (data not shown) is inconsistent with a recent transfer from a host genome.
A Possible HGT of SET Domain Genes among Bacterial and Archaebacterial Species
Our phylogenetic analysis did not exclude possible horizontal transfers involving the free-living hot-spring bacterium Ch. tepidum, the Bacilli, and the Methanosarcinae species (figs. 1 and 2; supplementary figs. SF2 and SF3, Supplementary Material online). It is thought that Methanosarcinae owe much of its ecological success to ancient HGT. Nearly 30% of its genes are considered to be of bacterial origin (Deppenmeier et al. 2002). Among Archaea, only the 3 mesophylic Methanosarcinae species have a SET domain gene. Examination of the DNA regions flanking the SET domain genes in the genomes of Chlorobium, Bacilli, and the Methanosarcinae support an exchange that might have involved members of the 3 groups.
The Chlorobium SET domain protein (NP_661845 [GenBank] ) is most similar to the Methanosarcinal NP_634869 [GenBank] (5 x 10–11) and to the Bacillus NP_834732 [GenBank] (1 x 10–07) SET domain proteins. The genes flanking the Chlorobium SET domain gene (shaded area in fig. 4) have best hits in the genomes of the 3 Methanosarcinae and the 3 Bacilli, although these highly similar genes are not in the immediate vicinity of the Methanosarcinal and Bacillal SET domain genes (fig. 4). Chlorobium genes outside the shaded box are found in a broad spectrum of bacteria (marked as "bacterial genes"). An exception is NP_661849 [GenBank] , having its most similar counterpart in Met. acetivorans (1 x 10–0, 62% identical).
|
In the 3 Bacilli, the SET domain genes are in syntenic regions on the main chromosomes. Immediately upstream of the SET domain gene is a gene encoding a methyl-accepting chemotaxis protein that has most similar counterparts in the genomes of Methanosarcina barkerii and Met. acetivorans (1 x 10–23). This gene is absent from M. mazei altogether. At the other flank of the SET domain in Bacillus are genes with best hits in the genome of Chlorobium.
The genes around the SET domain genes in Met. barkerii and in M. mazei (shaded regions) are closest to genes from the 3 Bacilli chromosomes forming short "Bacillus-like islands" (fig. 4). We note that the colinearity around the SET domain genes is preserved in M. mazei and Met. barkerii, but not in Met. acetivorans, despite the fact that M. mazei and Met. acetivorans are closer (Galagan et al. 2002). In agreement, the SET domain proteins of M. mazei and Met. acetivorans are 92% identical, whereas the M. mazei and Met. barkerii are only 76% identical. Because the genomes of M. mazei and Met. barkerii are more distantly related, the preserved synteny around the SET domain genes indicates that this arrangement has existed in the common ancestor and that in Met. acetivorans the gene has been internally relocated after the separation of Met. acetivorans from M. mazei.
The GC contents of the SET domain genes of all species are similar to the overall genome contents and, thus, do not offer a clue as to the origin of the genes. In Chlorobium, the established contents for the gene is 57.1% versus the genomic 57%, in the 3 Methanosarcinae, 40.6–43.0% versus 41.5–42.7%, and in the 3 Bacilli, 33.8–34.3% versus 35.1–35.3% for the SET domain genes and reported genomes, respectively. The high similarity between the SET domain proteins of M. mazei and Ba. cereus (3 x 10–25, 51% identical), together with their phylogenetic relationship incongruent with the origin of the species, support a lateral exchange.
Evolution of the Bacterial SET Domain Genes: A Model
Based on our data, we propose a model for an ancestral origin of the bacterial SET domain genes (fig. 5). A central postulate is that an ancient (+)pSET gene version has been duplicated and diverged into 2 ((–)pSET and (–)*pSET) copies and that all 3 gene versions were present in a common bacterial ancestor (CBA). The SET domain genes of extant bacterial genomes, accordingly, represent individual copies, or combination of copies, descending from the 3 ancestral genes. The spirochaetae (L. interrogans) has inherited (and subsequently diverged) all 3 gene copies; other bacteria carry 1, or combinations of 2, of the primordial lineages. This assumption may account for the relationship between extant bacterial SET domain genes and the genes found in L. interrogans. For example, the predecessors of Nostoc and Pirellula lines have inherited the (+)pSET gene version and have modified it in a species-specific mode; the putative ancestor of the Chlamydia/Verrucomicrobia group has had 2 genes: 1 (+)pSET and 1 (–)pSET version. The Verrucomicrobia lineage has preserved the (+)pSET gene, whereas extant Chlamydiae have kept the (–)pSET copy. The same (–)pSET version has been inherited also by the Cytophaga, Chlorobium, and Bacillus lineages where it has undergone an evolution of its own and was, perhaps, transferred to the Methanosarcinae.
According to our model, the common ancestor of Proteobacteria has retained 2 ancestral copies: the (+)pSET and a (–)*pSET different from the version retained by the ancestors of Cytophaga, Chlorobium, and Bacillus described above. Furthermore, during the proteobacterial evolution, the ancestor of the -lineage has retained both copies, whereas the progenitor of the ß-, -, and -groups has kept only the ancestral (+)pSET copy. Subsequently, in the -subdivision, Br. japonicum has inherited both ancestral (+)pSET and (–)*pSET copies, and Rh. palustris has kept only the (+)pSET copy, whereas Me. loti has retained only the (–)*pSET version. In Me. loti, the gene has been subsequently duplicated for species-specific functions, absent from Br. japonicum and Rh. palustris.
The (+)pSET gene version has been largely conserved in Proteobacteria, has been duplicated, and has evolved for bacteria-specific roles. The high similarity between the (+)pSET domain proteins of Proteobacteria and the Verrucomicrobial member V. spinosum might reflect the striking conservation of a gene inherited from the CBA. The monophyletic origin of the ß- and -subdivisions is congruent with ancestral relationships among their SET domain genes. Furthermore, the similarity between the eukaryotic SET domain proteins and the proteins of Xanthomonas/Xylella may have another interesting implication vis a vis a recent hypothesis that the eukaryotic genome has arisen from a fusion of a proteobacterium (P) and an archaeal eocyte (Rivera and Lake 2004). This links, at the deepest levels, prokaryotes with eukaryotes and suggests possible common origins for the eukaryotic and the protobacterial SET domain genes.
Why Do Bacteria Have SET Domain Genes?
Unambiguous answers to this question would be provided by "bench studies," none of which, surprisingly, have been carried out to this date. Based on our analyses, we can make a few predictions. The facts that bacteria do not carry histones, the established substrate for the SET domain activity, and that SET domain genes were found predominantly in pathogenic and symbiotic bacteria have suggested that bacterial SET domain proteins might be involved in interactions with the host. Extra chromosomes, plasmid vectors, secretion regions, pathogenicity and symbiosis islands in many pathogenic and symbiont species, are potential vehicles for "interacting factors" with the host cells. Our genomic analysis, however, showed that in none of the examined genomes were the SET domain genes located at these mobile structures. Because all bacteria have their SET domain genes on the main chromosomes, it is unlikely that the gene products are involved in direct host–bacterial interactions.
Most likely, the bacterial SET domain genes are involved in bacterial cell-specific functions. For example, the spirochaetae L. interrogans, related to the strictly parasitic B. burgdorferei and T. pallidum, carries 3 SET domain genes paralogs, whereas the other 2 genomes carry none. Parasitic bacteria thriving in a more homeostatic niche tend to delete expendable sequences from their genomes (Brinkman et al. 2002). We suggest that the SET domain genes provide L. interrogans with survival opportunities not needed by its parasitic relatives.
An interesting observation is that bacterial species that undergo unique types of developmental cycles carry SET domain genes: M. xanthus undergoes developmental regulation to produce multicellular fruiting bodies (Julien et al. 2000) and Chlamydia has biphasic developmental cycle controlled by a set of specific genes (Belland et al. 2003). In C. trachomatis, the SET domain gene is among the late expressing genes. Should proximity of genes be a reflection of their coordinated function, it might be informative that on all Chlamydia chromosomes, the SET gene is immediately preceded by a gene encoding a protein involved in cell division (FtsK). On the other flank is a histone-like gene, possibly involved in the compaction of the bacterial chromosome during the formation of the metabolically inactive compact elementary body. It is tempting to suggest that the SET domain genes may be involved in bacterial processes that are distant predecessors of eukaryotic developmental mechanisms. In agreement, the only archaebacterial group carrying SET domain genes, Methanosarcinae, are unique among Archaea with their ability to form complex multicellular structures suggestive of differentiation (Galagan et al. 2002).
An often-made reference for a relationship between archaeal and eukaryotic systems is the archaeal chromatin. Many euryarcheota carry homologs of eukaryotic histones that can compact DNA (Reeve et al. 1997). Some carry 1 copy (M. mazei, Met. acetivorans, and Met. barkerii) and some have more than one that form dimers and tetramers (White and Bell 2002). However, because none of the Archaea (except Methanosarcinae) carries SET domain genes, it is clear that archaeal histones are not targets of the SET domain protein activity. Moreover, the archaeal histones have the characteristic histone fold, but no tails, indicating that histone-tail modifications do not take place even in species that have the modifying activity. These facts argue that unlike eukaryotes, Methanosarcinae, apparently do not use a "histone code" to regulate chromosome activity and gene expression (Strahl and Allis 2000). However, the M. mazei SET domain protein carries specific methyltransferase activity for a protein, MC1-, associating with DNA (Manzur and Zhu 2005). It provided the first example of a SET domain function outside the eukaryotic domain of life.
A major result of our analyses, thereby, is the conclusion that the SET domain genes found in extant bacteria are, most likely, of bacterial origin. The presence of SET domain genes in members of all main clades of the bacterial domain of life indicated that ancestral versions of the gene have existed before the separation of the known bacterial divisions. The increased sample of sequenced bacterial genomes allowed us to establish presence of SET domain genes in free-living and environmental species that are unlikely to have acquired their SET domain genes from a eukaryotic donor. Apparently, the initial finding of SET domain genes in pathogenic and symbiont species was a result of a biased sample.
The apparently monophyletic origin of the SET domain proteins of the ß- and -bacteria, congruent with the monophyletic origin of the species, supports ancestral relationships among the bacterial SET domain genes. This conclusion is important because it implies that the high similarity of the Xanthomonas/Xylella and the eukaryotic SET domain proteins might reflect not horizontal gene transger from eukaryotes but a common origin from distant proteobacterial genomes (Rivera and Lake 2004). The absence of SET domain genes in the majority of currently sequenced bacterial genomes may be attributed to gene loss. Gene loss has played a significant role in bacterial genome evolution; unusual bacteria–eukaryotic gene similarity are thought to reflect gene loss in a related lineage (Mira et al. 2001; Salzber and Eisen 2001). The small genomes of obligatory parasites suggested that host adaptations are a consequence of gene loss, not gain of function (Brinkman et al. 2002). This argues against a "gain" of a SET gene from the host in parasitic bacteria. Dynamic genome reorganizations, considered typical for bacteria, may account for the overall lack of synteny and the loss of SET domain genes. The mosaic of short patchy synteny in genomes of closely related species is evidence of internal genome activity. In genomes where SET domain genes were within regions flanked by transposable elements and clusters of transposases, they were associated with internal genome rearrangements rather than accommodation of foreign DNAs. In most cases, the GC contents established for individual bacterial SET domain genes did not differ significantly from the GC contents of the host genomes but in the case of Me. loti and Br. japonicum, the SET domain genes had lower GC (60.4% and 61.2%) than their respective genomes (64% for Me. loti and 66%). Additional evidence did not support a lateral acquisition of these genes and, thus, it is unclear how these differences in the GC contents of the SET domain genes and the genomes carrying them might correlate, or reflect, the history of their origin.
A most compelling argument for a bacterial versus eukaryotic origin of the SET domain genes found in bacteria, however, is the evidence that the bacterial SET domain genes have undergone an evolution of their own. The segregation of the bacterial SET domain proteins into 2 phylogenetically related Domains (1 and 2) and the presence of multiple SET domain gene copies within a single genome are the 2 main lines of support. We consider the loss of the post-SET sequence (appearance of Domain 2 proteins) as a later event in the evolution of the bacterial SET domain function. A few eukaryotic families (E(z), SET7/9, SET8, and RuBisCo) also lack the post-SET domain. However, the evolution of the bacterial (–)pSET sequences is apparently independent of the evolution of eukaryotic (–)pSET: phylogenetic analysis did not reveal a relationship between the bacterial (–)pSET and the eukaryotic (–)pSET types (not shown).
Presence of multiple SET domain genes within a genome reflects duplication events that have taken place at different times. The duplication of the Me. loti gene has occurred after its speciation within the -proteobacterial subdivision, and duplication of the SET domain genes in B. cepacia and Ru. gelatinosus has occurred after the separation of the ß- from the other proteobacterial groups, whereas the 2 Cy. hutchinsonii genes are species-specific duplication of an ancient (–)pSET version after the separation of the spirochaetal, the Bacteroides–Flexobacter, and the chlamydial lineages. Existence of more than 1 copy of SET domain genes in the genomes of several species and their segregation into different clades indicates evolution of species-specific and paralogous functions in bacteria.
The phylogenetic and chromosome analyses of Chlorobium, Bacillus, and Methanosarcinal SET domain–containing species supported an HGT between bacteria and Archeae. Genomic analyses in the vicinity of the SET domain genes and phylogenetic trees are consistent with a lateral exchange between bacterial and Methanosarcinal genomes. The SET domain genes and their immediate neighbors in Chlorobium and in Ba. thuringiensis, Ba. cereus, and Ba. anthracis have best hits in the genomes of the Methanosarcinal species and vice versa. It is plausible that a Methanosarcinal ancestor has received the SET domain gene (together with a few nearby genes) from bacterial donors although the donor (Chlorobium or Bacillus) cannot be unambiguously defined. Methanosarcinae may occupy similar habitats with Ba. cereus (Bintrim et al. 1997), suggesting that an ancestor of Methanosarcinae might have acquired it from a Bacillus ancestor. The GC contents of the Methanosarcinal SET domain genes, however, indicate that if a transfer has taken place, it has not been recent as the gene has been successfully "blended" with the genomic sequences.
|
Supplementary Material |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Supplementary figures SF1–SF5 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
The authors are indebted to F. Doolittle and Camilla Nesbo for critically reading this manuscript and for providing helpful suggestions and encouragement. This work was partially supported by a University of Nebraska Lincoln-Research Cluster Grant Progect grant award and by Molecular and Cellular Biology-0343934 National Science Foundation grant to Z.A.
|
Footnotes |
|---|
Jonathan Eisen, Associate Editor
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Alfano JR and Collmer A. (1997) The type III secretion pathway of plant pathogenic bacteria: trafficking harpins avr proteins, and death. J Bacteriol 179:5655–5662.
Alvarez-Venegas R and Avramova Z. (2002) SET-domain proteins of the Su(var), E(z) and Trithorax families. Gene 285:25–37.[CrossRef][Web of Science][Medline]
Amann R, Springer N, Schonhuber W, Ludwig W, Schmid EN, Muller KD, Michel R. (1997) Obligate intracellular bacterial parasites of acanthamoebae related to Chlamydia spp. Appl Environ Microbiol 63:115–121.[Abstract]
Aravind L and Iyer LM. (2003) Provenance of SET-domain histone methyltransferases through duplication of a simple structural unit. Cell Cycle 2:369–376.[Medline]
Baumbusch LO, Thorstensen T, Krauss V, Fischer A, Naumann K, Assalkhou R, Schiltz I, Reuter G, Aalen RB. (2001) The Arabidopsis thaliana genome contains at least 29 active genes encoding SET-domain proteins that can be assigned to four evolutionary conserved classes. Nucleic Acids Res 29:4319–4333.
Belland RJ, Zhong GM, Crane DD, Hogan D, Sturdevant D, Sharma J, Beatty WL, Caldwell HD. (2003) Genomic transcriptional profiling of the developmental cycle of Chlamydia trachomatis. Proc Natl Acad Sci USA 100:8478–8483.
Bhattacharyya A, Stilwagen S, Ivanova N, et al. (22 co-authors). (2002) Whole-genome comparative analysis of 3 phytopathogenic Xylella fastidiosa strains. Proc Natl Acad Sci USA 99:12403–12408.
Bintrim SB, Donohue T, Handelsman J, Roberts GP, Goodman RM. (1997) Molecular phylogeny of Archaea from soil. Proc Natl Acad Sci USA 94:277–282.
Brinkman FSL, Blanchard JL, Cherkasov A, et al. (13 co-authors). (2002) Evidence that plant-like genes in Chlamydia species reflect an ancestral relationship between Chlamydiaceae, cyanobacteria, and the chloroplast. Genome Res 12:1159–1167.
Broughton WJ and Perret X. (1999) Genealogy of legume-Rhizobium symbioses. Curr Opin Plant Biol 2:305–311.[CrossRef][Web of Science][Medline]
Brown JR. (2003) Ancient horizontal gene transfer. Nat Rev Genet 4:121–132.[CrossRef][Web of Science][Medline]
Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, Higgins DG, Thompson JD. (2003) Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res 31:3497–3500.
Coenye T and Vandamme P. (2003) Diversity and significance of Burkholderia. Environ Microbiol 5:719–729.[CrossRef][Medline]
Deppenmeier U, Johan A, Hartsh T, et al. (23 co-authors). (2002) The genome of Methanosarcina mazei: evidence for lateral gene transfer between bacteria and archaea. J Mol Microbiol Biotech 4:453–461.
Doolittle WF. (1999) Phylogenetic classification and the universal tree. Science 284:2124–2128.[CrossRef][Web of Science][Medline]
Doolittle WF and Logsdon JM. (1998) Archaeal genomics: do archaea have a mixed heritage? Curr Biol 8:R209–R211.[CrossRef][Web of Science][Medline]
Dorman CJ and Deighan P. (2003) Regulation of gene expression by histone-like proteins in bacteria. Curr Opin Genet Dev 13:179–184.[CrossRef][Web of Science][Medline]
Eisen JA, Nelson KE, Paulsen IT, et al. (35 co-authors). (2002) The complete genome sequence of Chlorobium tepidum TLS, a photosynthetic, anaerobic, green-sulfur bacterium. Proc Natl Acad Sci USA 99:9509–9514.
Felsenfeld G and Groudine M. (2003) Controlling the double helix. Nature 421:448–453.[CrossRef][Medline]
Felsenstein J. (1989) PHYLIP—phylogeny inference package (Version 3.2). Cladistics 5:164–166.
Galagan JE, Nusbaum C, Roy A, et al. (55 co-authors). (2002) The genome of M-acetivorans reveals extensive metabolic and physiological diversity. Genome Res 12:532–542.
Galan JE and Collmer A. (1999) Type III secretion machines: bacterial devices for protein delivery into host cells. Science 284:1322–1328.
Gherna R and Woese CR. (1992) A partial phylogenetic analysis of the flavobacter-bacteroides phylum: basis for taxonomic restructuring. Syst Appl Microbiol 15:513–521.[Web of Science][Medline]
Glansdorff N. (2000) About the last common ancestor, the universal life-tree and lateral gene transfer: a reappraisal. Mol Microbiol 38:177–185.[CrossRef][Web of Science][Medline]
Glockner FO, Kube M, Bauer M, et al. (14 co-authors). (2003) Complete genome sequence of the marine planctomycete Pirellula sp. strain 1. Proc Natl Acad Sci USA 100:8298–8303.
Gottfert M, Rothlisberger S, Kundig C, Beck C, Marty R, Hennecke H. (2001) Potential symbiosis-specific genes uncovered by sequencing a 410-kilobase DNA region of the Bradyrhizobium japonicum chromosome. J Bacteriol 183:1405–1412.
Horn M, Collingro A, Schmitz-Esser S, et al. (13 co-authors). (2004) Illuminating the evolutionary history of Chlamydiae. Science 304:728–730.
Ivanova N, Sorokin A, Anderson I, et al. (23 co-authors). (2003) Genome sequence of Bacillus cereus and comparative analysis with Bacillus anthracis. Nature 423:87–91.[CrossRef][Medline]
Jacobs SA, Harp JM, Devarkonada S, Kim Y, Rastinejad F, Khorasanizadeh S. (2002) The active site of the SET domain is constructed on a knot. Nat Struct Biol 9:833–838.[Web of Science][Medline]
Jenuwein T and Allis CD. (2001) Translating the histone code. Science 293:1074–1080.
Julien B, Kaiser DA, Garza A. (2000) Spatial control of cell differentiation in Myxococcus xantus. Proc Natl Acad Sci USA 97:9098–9103.
Kaneko T, Nakamura Y, Sato S, et al. (24 co-authors). (2000) Complete genome structure of the nitrogen-fixing symbiotic bacterium Mesorhizobium loti. DNA Res 7:331–338.[Abstract]
Kaneko T, Nakamura Y, Sato S, et al. (17 co-authors). (2002) Complete genomic sequence of nitrogen-fixing symbiotic bacterium Bradyrhizobium japonicum USDA110. DNA Res 9:189–197.[Abstract]
Koonin EV, Makarova KS, Aravind L. (2001) Horizontal gene transfer in prokaryotes: quantification and classification. Annu Rev Microb 55:709–742.[CrossRef][Web of Science][Medline]
Larimer FW, Chain P, Hauser L, et al. (19 co-authors). (2004) Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nature Biotech 22:55–61.[CrossRef][Web of Science][Medline]
Makarova KS and Koonin EV. (2003) Comparative genomics of archaea: how much we learned in six years and what's next? Genome Biol 4:115.[CrossRef][Medline]
Manzur KL and Zhu M-M. (2005) An archaeal SET domain protein exhibits distinct methyltransferase activity towards DNA-associated protein MC1-a. FEBS Lett 579:3859–3865.[CrossRef][Web of Science][Medline]
Marmostein R. (2003) Structure of SET domain proteins: a new twist on histone methylation. Trends Biochem Sci 28:59–62.[CrossRef][Web of Science][Medline]
Mergaert P, Van Montagu M, Holsters M. (1997) Molecular mechanisms of Nod factor diversity. Mol Microbiol 25:811–817.[CrossRef][Web of Science][Medline]
Mira A, Ochman H, Moran NA. (2001) Deletional bias and the evolution of bacterial genomes. Trends Genet 17:580–589.[CrossRef][Web of Science][Medline]
Nelson KE, Clayton RA, Gill SR, et al. (29 co-authors). (1999) Evidence for lateral gene transfer between Archaea and Bacteria from genome sequence of Thermotoga maritime. Nature 399:323–329.[CrossRef][Medline]
Ochman HM, Lawrence JG, Groisman EA. (2000) Lateral gene transfer and the nature of bacterial innovation. Nature 405:299–304.[CrossRef][Medline]
Pace NR. (1997) A molecular view of microbial diversity and the biosphere. Science 276:734–740.
Page RDM. (1996) TREEVIEW: an application to display phylogenetic trees on personal computers. Comput Appl Biosci 12:357–358.
Parkhill J, Sebaihia M, Preston A, et al. (53 co-authors). (2003) Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet 35:32–40.[CrossRef][Web of Science][Medline]
Radnedge L, Agron PG, Hill KK, Jackson PJ, Ticknor LO, Keim P, Andersen GL. (2003) Genome differences that distinguish Bacillus anthracis from Bacillus cereus and Bacillus thuringiensis. Appl Environ Microbiol 69:2755–2764.
Rea S, Eisenhaber F, O'Carroll D, et al. (11 co-authors). (2000) Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature 406:593–599.[CrossRef][Medline]
Read TD, Brunham RC, Shen C, et al. (25 co-authors). (2000) Genome sequences of Chlamydia trachomatis MoPn and Chlamydia pneumoniae AR39. Nucleic Aids Res 28:1397–1406.
Read TD, Myers GSA, Brunham RC, et al. (21 co-authors). (2003) Genome sequence of Chlamydophila caviae (Chlamydia psittaci GPIC): examining the role of niche-specific genes in the evolution of the Chlamydiaceae. Nucleic Aids Res 31:2134–2147.
Reeve JN, Sandman K, Daniels CJ. (1997) Archaeal histones, nucleosomes, and transcription initiation. Cell 89:999–1002.[CrossRef][Web of Science][Medline]
Ren S-X, Gu G, Jiang X-G, et al. (39 co-authors). (2003) Unique physiological and pathogenic features of Leptospira interrogans revealed by whole genome sequencing. Nature 422:888–892.[CrossRef][Medline]
Rivera MC and Lake JA. (2004) The ring of life provides evidence for a genome fusion origin of eukaryotes. Nature 431:152–155.[CrossRef][Medline]
Salanoubat M, Genin S, Artiguenave F, et al. (28 co-authors). (2002) Genome sequence of the plant pathogen Ralstonia solanacearum. Nature 415:497–502.[CrossRef][Medline]
Salzber SL and Eisen JA. (2001) Lateral gene transfer or viral colonization? Science 293:1048.
Schlesner H. (2004) The Prokaryotes: an evolving electronic resource for the microbiological community: the genus Verrucomicrobium. (Springer-Verlag, New York).
Stanhope MJ, Lupas A, Italia MJ, Koretke KK, Volker C, Brown JR. (2001) Phylogenetic analysis do not support horizontal gene transfer from bacteria to vertebrates. Nature 411:940–943.[CrossRef][Medline]
Stephens RS, Kalman S, Lammel C, et al. (12 co-authors). (1998) Genome sequence of an obligate intracellular pathogen of humans: Chlamydia trachomatis. Science 282:754–759.
Strahl BD and Allis CD. (2000) The language of covalent histone modifications. Nature 403:41–45.[CrossRef][Medline]
Sullivan JT, Trzebiatowski JR, Cruickshank RW, et al. (14 co-authors). (2002) Comparative sequence analysis of the symbiosis island of Mesorhizobium loti strain R7A. J Bacteriol 184:3086–3095.
van Berkum P, Terefework Z, Paulin L, Suomalainen S, Lindstrom K, Eardly BD. (2003) Discordant phylogenies within the rrn loci of rhizobia. J Bacteriol 185:2988–2998.
Van Sluys MA, de Oliveira MC, Monteiro-Vitorello CB, et al. (57 co-authors). (2003) Comparative analysis of the complete genome sequences of Pierse's disease and citrus variegated chlorosis strains of Xylella fastidiosa. J Bacteriol 185:1018–1026.
White MF and Bell SD. (2002) Holding it together: chromatin in archaea. Trends Genet 18:621–626.[CrossRef][Web of Science][Medline]
Woese CR. (1998) The universal ancestor. Proc Natl Acad Sci USA 95:6854–6859.
Woese CR. (2000) Interpreting the universal phylogenetic tree. Proc Natl Acad Sci USA 97:8392–8396.
Woese CR and Fox GE. (1977) Phylogenetic structure of the prokaryotic domain: the primary kingdoms. Proc Natl Acad Sci USA 74:5088–5090.
Xiao B, Wilson JR, Gamblin SJ. (2003) SET domains and histone methylation. Curr Opin Struct Biol 13:699–705.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Saleh, R. Alvarez-Venegas, M. Yilmaz, O. Le, G. Hou, M. Sadder, A. Al-Abdallat, Y. Xia, G. Lu, I. Ladunga, et al. The Highly Similar Arabidopsis Homologs of Trithorax ATX1 and ATX2 Encode Proteins with Divergent Biochemical Functions PLANT CELL, March 1, 2008; 20(3): 568 - 579. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||