Previous Article | Next Article ![]()
Applied and Environmental Microbiology, September 2003, p. 5530-5535, Vol. 69, No. 9
0099-2240/03/$08.00+0 DOI: 10.1128/AEM.69.9.5530-5535.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Unité des Rickettsies, Faculté de Médecine, Université de la Méditerranée, 13385 Marseille Cedex 05, France
Received 6 June 2003/ Accepted 26 June 2003
|
|
|---|
|
|
|---|
Parachlamydia is a new genus within Chlamydiales (2) that presents Chlamydia-like developmental stages (15) and shares 80 to 90% similarity of 16S rRNA genes with Chlamydiaceae (12, 14). In contrast to the Chlamydiaceae, which naturally infect multicellular organisms such as mammals and birds, Parachlamydia acanthamoebae naturally infects free-living amoebae and has probably never been a parasite of multicellular organisms. Moreover, since Parachlamydiaceae-Chlamydiaceae divergence was contemporary with the eukaryotic radiation about 1 billion years ago (26), the presence of the tlc gene within the genome of P. acanthamoebae would preclude the hypothesis of a transfer of the tlc gene from plants and might, on the contrary, suggest its transfer from Chlamydiales to plants. Therefore, in the present study, we investigated whether the tlc gene was present within the genome of P. acanthamoebae and evaluated the genetic and phylogenetic relationships of the nonmitochondrial ADP/ATP translocase coding sequences from an evolutionary perspective.
|
|
|---|
PCR amplification and sequencing of P. acanthamoebae tlc gene.
The strategy for determining the sequence of the gene encoding the ADP/ATP translocase (i.e., tlc) of P. acanthamoebae consisted of (i) amplifying a segment of the targeted gene with degenerate primers and (ii) completing the full-length sequence by the genome-walking approach. adpF1 5'-GAAGC(TA)AAACGTTT(CT)TACGCTCT, adpF11 5'-TGTTCTGGGGGTTAGCCAA, and adpR4 5'-C(AG)TCAATAGC(AG)GCTTTICCTTT-, all designed by alignment of the tlc genes of Chlamydia muridarum, Chlamydia trachomatis, Chlamydophila pneumoniae J138, and Rickettsia conorii, succeeded in amplifying 910-bp (adpF11-R4) and 980-bp (adpF1-R4) nucleotides. These PCR products were purified by using the QIAquick PCR purification kit (Qiagen, Courtaboeuf, France) and sequenced by using the dRhodamine terminator cycle sequencing ready reaction with AmpliTaq DNA (Perkin-Elmer Biosystems, Warrington, United Kingdom). Sequences were determined on 3100 ABI Prism automated sequencer (Applied Biosystems, Courtaboeuf, France). Sequences derived from each primer were aligned, compared, and combined in a single sequence by using Autoassembler software version 2.1 (Applied Biosystems). The unknown 5' and 3' ends of this partial sequence were amplified by using the Universal Genome Walker kit (Clontech Laboratories, Palo Alto, Calif.). The validity of the sequence obtained was assessed by comparison with two additional sequences obtained by PCR amplification with primers designed from the sequence immediately flanking the open reading frame: Adp65F (5'-GATCCACGAAAGCACTCTTATT) Adp62R (5'-GGCAATCTATCACGTAATTGAAAAT).
Presence of the tlc gene in additional clades.
The nucleotide sequence of Parachlamydia tlc (GenBank accession number AF490592) and the corresponding amino acid sequence were compared to sequences available in the GenBank database by using the BLASTN and BLASTP 2.2.6 programs available on the National Center for Biotechnology Information website (www.ncbi.nlm.nih.gov) (1). An iterated profile search was also performed by using position-specific iterated BLAST (PSI-BLAST) (1).
Genetic and phylogenetic analysis.
After alignment with CLUSTAL W (28), genetic distances of nucleotides and amino acids sequences were calculated by using MEGA 2.1 software (18). With the same software, we inferred neighbor-joining (p-distance), minimum evolution (p-distance), and parsimony trees (standard parsimony) by using the amino acid and nucleotide sequences.
Time scale.
The molecular clock of the tlc gene was calibrated with seven points obtained from fossil estimates and from the estimated time of divergence of bacterial species on the basis of the 16S rRNA sequence divergence, assuming a rate of evolution of 1 to 2% per 50 million years (20). The 16S rRNA sequences were edited by removal of the longer 5' and 3' ends so that their lengths matched that of the shortest sequence. The percentage of 16S rRNA sequence divergence was calculated by using CLUSTAL W program (28), supported by the PBIL website (http://npsa-pbil.ibcp.fr/cgi-bin/align_clustalw.pl). The time of divergence of green plants and red algae, estimated by Sogin and Silberman (26), the time of divergence of monocotyledons (Oryza sativa) and dicotyledons (Citrus spp., Solanum tuberosum, and Arabidopsis thaliana) estimated by Gale et al. (13), and the estimated times of divergence of bacterial species (assuming a rate of evolution of 1.5% of 16S rRNA sequence per 50 million years) were plotted as a function of the proportion (p) of amino acid sites at which the sequence is different from that of the ancestral ADP/ATP translocase sequence, i.e., the p-distance between each node and the node that separated the more divergent sequences. The equation of the regression line, its standard error, and its r2 coefficient were calculated with Stata 7.0 (Stata Corp., College Station, Tex.) and Microsoft Excel 97 (Microsoft Corp., Redmond, Wash.). To estimate the time of divergence of red algae and green plants and of monocotyledons and dicotyledons, regression was performed similarly, using only the time scale inferred from 16S rRNA genes and without using the times estimated by Sogin and Silberman (26) and by Gale et al. (13). Similar analyses were performed by using a minimum evolution tree inferred from amino acid sequences. To further test the reliability of the time scale, we performed an omit test, which assessed how the time estimates were modified by the omission of each of the calibration point.
|
|
|---|
Presence of tlc gene in Galdieria sulfuraria (a red algae), Citrus hybrid cultivar (citrus), Oryza sativa (rice), Holospora obtusa and Caedibacter caryophilus (endosymbionts of Paramecium), Encephalitozoon cuniculi (microsporidia), and Medicago sativa (alfalfa).
By using the basic local alignment search tool (BLAST) with the amino acid sequence of the Parachlamydia ADP/ATP translocase as input, we found tlc gene sequences in additional clades, including G. sulfuraria (a red algae; GenBank accession number AJ251356 [236 of 468 identities]), Citrus hybrid cultivar (GenBank accession number AY098893 [231 of 480 identities]), O. sativa (rice; GenBank accession number AP003234 [236 of 496 identities]), H. obtusa (GenBank accession number AY120885 [183 of 465 identities]), Caedibacter caryophilus (GenBank accession number AJ441310 [45 of 53 identities]), and M. sativa (GenBank accession number AF416339 [52 of 81 identities]). However, when gap and extension penalties of 11 and 1, respectively, were used, BLAST analysis did not detect the tlc gene sequences in Wolbachia sp. (an endosymbiont of Drosophila melanogaster), cyanobacteria, protozoa, or animals. A BLAST search also identified four E. cuniculi proteins of unknown functions (GenBank accession numbers NP_586157 to NP_597260 [93 of 443 to 111 of 473 identities]). With the exception of the short tlc sequence of M. sativa (317 bp), we used all of these tlc sequences in the genetic and phylogenetic analyses.
Proteins of unknown functions related to the ADP/ATP translocase in Chlamydiales, cyanobacteria, and plant pathogen (Xylella fastidiosa).
By PSI-BLAST, we identified proteins of unknown function with significant alignments within the proteomes of two cyanobacteria (Nostoc sp. and Trichodesmium erythraeum), within that of Chlamydiales, and within different
-proteobacteria (including X. fastidiosa, a plant pathogen). We performed phylogenetic analysis with the amino acid sequences of all of these proteins of unknown function and with the four microsporidium proteins (which exhibit sequence similarity with ADP/ATP translocase but whose functions are also unknown) and with the Chlamydiales, Rickettsiales, and plant plastid ADP/ATP translocase. The tree topology was similar to that shown in Fig. 1, except that the E. cuniculi proteins and the other proteins of unknown function rooted deeply, being phylogenetically far from the ADP/ATP translocase (data not shown).
![]() View larger version (50K): [in a new window] |
FIG. 1. p-Distance neighbor-joining tree inferred from amino acid sequences of the ADP/ATP translocase of P. acanthamoebae, Chlamydiaceae, Rickettsiales, and plant and alga plastids. Bootstrap values resulting of 100 replications are present at branch points. The tree was rooted with four proteins present in E. cuniculi, whose functions are unknown and were identified by BLAST of nonmitochondrial ADP/ATP translocase. The time scale was derived from estimates obtained from the literature (13, 26) and on the basis of 16S rRNA gene sequence divergence (20). Note the congruence between the node of a given divergence (see neighbor-joining tree) and its estimated time (see time scale). , Estimated from 16S rRNA divergence (20); $, estimated by Sogin et al. (26); #, estimated by Gale et al. (13).
|
|
View this table: [in a new window] |
TABLE 1. Comparison of similarities of complete amino acid sequences (excluding alignment gaps) of ADP/ATP-translocase-encoding genes (tlc) of Chlamydiales, Rickettsiales, and plant and alga plastids
|
Time scale.
The divergences of the 16S rRNA sequence between Chlamydiales and Rickettsiales, P. acanthamoebae and Chlamydiaceae, H. obtusa and Rickettsia spp., Chlamydophila pneumoniae and Chlamydia spp., and R. conorii and Rickettsia prowazekii were of 55, 14, 16, 6, and 1.59%, respectively. Thus, if we assume a rate of evolution of 1 to 2% per 50 million years (20), the respective divergences of these organisms may have occurred more than 2.75 billions years ago, 0.7 to 1.4 billion years ago, 0.8 to 1.6 billion years ago, 300 to 600 million years ago, and 80 to 160 million years ago (see Fig. 1).
Congruence of p-distance and time scale.
The dependence of the p-distance of the neighbor-joining tree inferred from the amino acid sequences of the tlc genes on the time estimated for bacterial species on the basis of 16S rRNA sequence divergence (20) and for plastids from the time of divergence of monocotyledons and dicotyledons (13) and of G. sulfuraria (a red algae) and Viridiplantae (green plants) was nearly linear, as shown by the low extent of deviation from the regression line (r2 = 0.84) (Fig. 2). The equation of the regression line is: t = -4.11p + 1.28, where t is the time from now in billion years and "p" is the proportion of amino acid sites at which the sequence is different from that of the ancestral tlc sequence at the time of its duplication in Chlamydiales, i.e., the p-distance from a given node to the middle of the neighbor-joining tree (see Fig. 1 and 2). If p = 0, the estimated time of duplication of the tlc gene in Chlamydiales will be ca. 1.28 billion years. The reliability of this estimated time was supported by the omit test. By omitting any of each of the calibration point, the time of duplication of the tlc gene ranged from 1.20 to 1.33 billion years, with a mean of 1.28 ± 0.04 billion years. The estimated times of transfer of the tlc gene from Chlamydiales to Rickettsiales and to plant and alga plastids would be ca. 1.16 and 0.98 billion years, respectively. Similarly, the estimated times of duplication of the tlc gene in Rickettsia spp. and in A. thaliana would be about 1,017 to 1,095 and 223 million years, respectively. The extent of deviation of the regression line was lower (r2 = 0.94) when only the time of divergence of bacterial species estimated on the basis of 16S rRNA sequence divergence was used as a time scale (20). The equation of the regression line was then: t = -4.13p + 1.34. The p-distance was 0.089 between the node separating the G. sulfuraria from Viridiplantae and the midpoint of the neighbor-joining tree (Fig. 1); thus, their time of divergence is about 0.97 billion years. Similarly, the time of divergence for monocotyledons and dicotyledons was estimated to be about 507 million years. Using the minimum evolution tree inferred from the amino acid sequence, the date of duplication of the tlc gene, the date of transfer of the gene from Chlamydiales to plants, the date of divergence of algae from green plants, and the date of divergence of monocotyledons and dicotyledons are close to those estimated with the neighbor-joining tree (Table 2). By omitting any of the calibration points and using the minimum-evolution tree inferred from the amino acid sequence, the time of duplication of the tlc gene ranged from 1.19 to 1.33 billion years, with a mean of 1.27 ± 0.05 billion years, further supporting the reliability of the estimated times.
![]() View larger version (16K): [in a new window] |
FIG. 2. Congruence (r2 = 0.88) of the time scale and the p-distance of the tlc gene. The time scale was derived from independent estimates obtained from the literature (13, 26) and on the basis of 16S rRNA gene sequence divergence (20). The p-distance between each node and the midpoint of the neighbor-joining tree inferred from the amino acid sequences of the ADP/ATP translocase represent the proportion (p) of amino acid sites at which the sequence is different from that of the ancestral sequence, at the time of duplication of the Chlamydiales ADP/ATP-translocase-encoding gene (tlc). Symbols: , divergence of H. obtusa and Rickettsia spp.; , divergence of Parachlamydia spp. and Chlamydiaceae; , divergence of red algae and green plants; , divergence of C. pneumoniae and Chlamydia spp.; , divergence of monocotyledons and dicotyledons; , divergence of C. trachomatis and C. muridarum; , Rickettsia spp. Lines: solid line, regression line (t = -4.11p + 1.28); dotted lines, confidence intervals (t = -4.11p + 1.13 and t = -4.11p + 1.43).
|
|
View this table: [in a new window] |
TABLE 2. Date of duplication of Chlamydiales tlc gene and transfers of the tlc gene from Chlamydiales to Rickettsiales and from Chlamydiales to plant and alga plastidsa
|
|
|
|---|
Our study provides a scale of time derived from the divergence of 16S rRNA gene sequences (20), with or without additional time estimates obtained from the literature (13, 26). The congruence of the p-distance derived from amino acid sequences of the ADP/ATP translocase with these time scales (r2 coefficients of 0.84 and 0.94, respectively) confirms the value of the tree presented in Fig. 1. This congruence confirms the value of the calibration of the molecular clock, performed based on both a fossil estimate and another molecular clock. The congruence of the p-distance derived only from tlc nucleotide sequences with the scale of time derived from the divergence of 16S rRNA sequences is further confirmed by the estimated time of divergence of red algae from green plants of 0.97 billion years that is similar to the 1 billion years estimate of Sogin and Silberman (26). The reliability of these results is further confirmed by the fact that dates estimated by using the neighbor-joining tree inferred from the ADP/ATP translocase amino acid sequences were really close to those estimated by using the minimum-evolution tree (Table 2) and by the fact that omission of any of the calibration point of the time scale only slightly modified the estimated time of duplication of the tlc genes in the Chlamydiales (standard deviations of 0.04 and 0.05 billions years, respectively).
The time scale analysis suggests that the ancestral sequence duplicated in Chlamydiales 1.27 to 1.28 billion years ago (see Table 2). Since the estimated time of transfer of the tlc gene to Rickettsiales was about 1.15 to 1.16 billion years, the presence of tlc genes is probably due to horizontal transfer and not to the speciation of Chlamydiales from Rickettsiales, which is estimated to have occurred more than 2.75 billion years ago. Although horizontal transfer of genes between Rickettsiales and Chlamydiaceae is unlikely to occur today, since these clades do not share common host cells, such a transfer may have occurred more than one billion years ago, at a time when these bacteria may have been facultatively intracellular and may have shared a common ancestral cell host, such as a free-living amoeba. The latter hypothesis is supported by the fact that a branch of evolution of both clades still parasitizes Acanthamoeba spp. (2, 11). The absence of the tlc gene in Wolbachia sp. (a clade that diverged from Rickettsia spp. 100 to 200 million years ago) is probably due to subsequent gene loss (22). The tlc1 gene was apparently transferred to eucarya 0.97 to 0.98 billion years ago, i.e., around the time when Chlamydiaceae and Parachlamydiaceae diverged (0.7 to 1.4 billion years ago). The presence of tlc1 in red algae and in higher plant plastids and its absence in the sequenced genomes of protozoan and animals show that transfer occurred after the divergence of red algae and plants from other eucaryotes but before that of Rhodophyta (red algae) from Viridiplantae (green plants). The latter divergence is contemporary with the eukaryotic radiation, which has also been estimated to have occurred about 1 billion years ago (9, 26). Since these time evaluations are congruent with our tree representations, we believe that they reflect the true time of transfer of tlc1 to eukaryotes.
Everett et al. (10) suggested that the common Chlamydiales ancestor might be related to the cyanobacterium-chloroplast lineage; the absence of the tlc gene in the genome of Synechocystis sp., Nostoc sp., and T. erythraeum (three cyanobacteria) (16), being then explained by a subsequent gene loss (22) or by the appearance of that gene in the Chlamydiales genome after their speciation from cyanobacteria. Brinkman et al. (6) showed that the vast majority of plant-like genes in Chlamydiales correspond to plant genes that are derived from and function in the chloroplast, suggesting that the ancestral Chlamydiales might have been involved in the endosymbiotic origin of chloroplasts or have at least played a role in the chimeric events that led to the formation of plant lineages.
The existence of the tlc gene and its duplication, long before the radiation of eukaryotes, including plants and animals, demonstrates the long history of parasitism of Chlamydiales with unicellular eukaryotes and is the oldest evidence of bacterial parasitism. One branch of Chlamydiales still parasitizes protozoa, whereas the other is associated with multicellular animals.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»