| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Biological Sciences, Kent State University, Kent, OH 44242-0001, USA
2 Laboratory of Molecular Population Genetics and Evolution, M. G. Kholodny Institute of Botany, National Academy of Sciences of Ukraine, vul. Tereshchenkivska 2, Kiev, Ukraine
Correspondence
Volodymyr Dvornyk
vdvornyk{at}kent.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
).
Cyanobacteria are the only prokaryotes known so far to exhibit circadian rhythmicity (Ishiura et al., 1998; Kondo & Ishiura, 2000). Their circadian system has been comprehensively studied in the unicellular Synechococcus elongatus PCC 7942 (Ditty et al., 2003; Johnson & Golden, 1999). A core element of the system in this species is a cluster of three genes: kaiA, kaiB and kaiC (Ishiura et al., 1998). In addition to the kai genes, several other genes have been shown to participate in the regulation of circadian rhythmicity, mainly by mediating input/output signals (Iwasaki et al., 2000; Kutsuna et al., 1998; Schmitz et al., 2000; Tsinoremas et al., 1996).
Unlike S. elongatus PCC 7942, which possesses a single kaiABC cluster, other cyanobacterial species may have multiple copies of the kaiB and kaiC genes (Dvornyk et al., 2002, 2003; Kaneko et al., 2001). The kaiA gene may either occur in a single copy or be absent in some cyanobacteria (Dvornyk et al., 2003; Nakamura et al., 2003). Along with the data on the phylogenetic analysis of the kaiB and kaiC genes, this suggests that cyanobacteria probably have two types of the circadian system, kaiABC- and kaiBC-based, respectively (Dvornyk et al., 2003). The kaiBC-based system perhaps also has components involved in input/output pathways, similar to those of the kaiABC-based system of S. elongatus PCC 7942. Based on the results of evolutionary analyses of the key components (kaiB, kaiC and sasA genes), it has been hypothesized that each type of cyanobacterial circadian system has specific functional and selective constraints, which shape its evolution and preclude lateral transfers of the components between the systems (Dvornyk et al., 2003, 2004). A large proportion of these constraints are thought to be due to the emergence of the kaiA gene (Dvornyk et al., 2004, 2005).
The cpmA gene (circadian phase modifier) was reported to be involved in an output pathway of the circadian clock (Katayama et al., 1999). Inactivation of cpmA dramatically affects activity of the kaiA promoter, but has no effect on the kaiBC promoter. These data suggest that, like sasA, cpmA may also have specific evolutionary constraints within each of the two types of the cyanobacterial circadian system, and that at least some of the constraints resulted from the appearance of kaiA in the course of the system's evolution. To test this hypothesis, I performed a comprehensive evolutionary and structural analysis of the cpmA genes from cyanobacteria.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
) with the respective amino acid sequences of S. elongatus PCC 7942 (GenBank accession AAD29318.1) as a query. All sequences with bit score cutoff above 120 were considered homologues. The retrieved protein and nucleotide sequences were aligned using CLUSTALW (Thompson et al., 1994) and manually adjusted based on structural considerations (Katayama et al., 1999). Alignments of the nucleotide sequences were modified manually according to the respective amino acid alignments.
For the comparative phylogenetic analysis, I used sequences of the 16S rRNA gene and the RecA protein from the respective or closely related strains. The 16S rRNA gene is a common marker for evolutionary studies of cyanobacteria (Garcia-Pichel et al., 2001; Honda et al., 1999). RecA is encoded by a housekeeping gene, which was previously shown to be useful for phylogenetic reconstructions of bacteria (Lloyd & Sharp, 1993). The respective sequences were obtained from the public databases, aligned with CLUSTALW (Thompson et al., 1994) and adjusted by visual inspection. A list of the sequences used is given in Table 1.
|
) with the Jukes–Cantor correction for multiple substitutions at the same site and a transitions/transversions ratio of 1·2. The MEGA 3.1 software (Kumar et al., 2004) was used for the computations of dN.
To determine the DNA substitution model best fitting the data, the ModelTest 3.0 software was used (Posada & Crandall, 1998). Based on the results of this test, the Tamura–Nei model of substitutions (Tamura & Nei, 1993) with gamma distribution was employed for further phylogenetic analysis of the 16S rRNA genes.
The alternative topologies of the phylogenetic trees of the CpmA protein sequences and 16S rRNA nucleotide sequences were initially constructed using the neighbour-joining (NJ) algorithm (Saitou & Nei, 1987) and then compared using the Shimodaira–Hasegawa test (Shimodaira & Hasegawa, 1999) to find the best topology. The RecA tree was obtained using the NJ method and the Poisson correction distance assuming equal substitution rate among amino acid sites. The statistical support for the nodes of the trees was obtained by the interior branch test (Sitnikova et al., 1995). The phylogenetic trees were built using the respective options of the MEGA 3.1 software (Kumar et al., 2004). Branch lengths in the CpmA tree were estimated using maximum-likelihood (ML) approach and the WAG matrix of amino acid substitutions (Whelan & Goldman, 2001) with a gamma parameter. Because the ML test rejected the global clock hypothesis for the CpmA proteins, the local clock model (Yoder & Yang, 2000) was applied to date their evolution. The analyses were performed using the PAML software (Yang, 1997).
To detect lateral transfers, the congruency of the obtained phylogenetic trees of species (16S rRNA tree) and the CpmA proteins was estimated using the likelihood-ratio test (LRT, Kishino & Hasegawa, 1989).
Analysis of selective constraints.
The nonsynonymous/synonymous rate ratio, 
=dN/dS, is commonly used as a measure of selective pressure at the protein level. Depending on its value (=1, <1 and >1) it may indicate neutral evolution, purifying selection and positive selection, respectively. The dN/dS ratio may either vary among amino acid sites but be averaged over all sequences (site-specific model) or differ between branches of a phylogenetic tree (branch-specific model). Comprehensive description of these models has been provided elsewhere (Nielsen & Yang, 1998; Yang et al., 2000; Yang & Nielsen, 2002). All computations were performed with the PAML software (Yang, 1997). The nested models were compared pairwise by the LRT. According to the LRT, twice the log-likelihood difference, 2
=2(1–0), follows a 
2 distribution with df=p1–p0, where p is the number of free parameters in the model. To compare non-nested models, the Akaike information criterion was used: AIC=–2+2p, where p is the number of independent parameters in the model (Akaike, 1974). The model with the lowest AIC is considered the most appropriate.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). However, no CpmA homologues have been found either in the photosynthetic 
-proteobacteria (i.e. Rhodobacter, Rhodospirillum and Rhodopseudomonas) or in Chloroflexus, which all are reported to have kai genes and/or a kaiBC operon laterally transferred from cyanobacteria (Dvornyk et al., 2003). In the screened genomes, the cpmA gene usually occurred in a single copy. The only exception was Trichodesmium erythraeum, which contained two copies. Those were identical by their nucleotide sequence, but different in length: one gene is shorter by 90 nucleotides from the 3'-end (Table 1
).
The 16S rRNA tree obtained featured three major clades, which corresponded to Cyanobacteria, Proteobacteria and Archaea (Fig. 1). Importantly, the cyanobacterial clade had 100 % statistical support. The topology of this clade was essentially in accordance with the conventional taxonomy of cyanobacteria (Castenholz, 1989), as well as with previously reported molecular data (Lyra et al., 2001; Turner, 1997). It indicated that apparently no horizontal transfers of 16S rRNA genes have occurred between the cyanobacterial species studied.
|
). Clade C1 contained the genes only from cyanobacteria. Clade C2 comprised the genes from proteobacteria, while clade C4 included those mainly from methanogenic archaea. The phylogenetic position of clade C3 deserves particular consideration for two important reasons. First, it definitely does not match the taxonomic position of the respective cyanobacterial species (Turner, 1997) and their positions in the 16S rRNA tree (Fig. 1). Nor does it correspond to the position of these species in the RecA phylogeny (Fig. 2). The most obvious explanation of the incongruence is lateral transfer from some early branching non-photosynthetic bacteria, as the tree topology suggests. This assumption is supported by the results of the Kishino–Hasegawa test (Kishino & Hasegawa, 1989), which strongly rejected the null hypothesis of 16S rRNA and CpmA tree compatibility (P<0·001) using either the 16S rRNA or the CpmA sequence data and with all alternative trees.
|
). Clade C1 includes the cpmA genes from the species with all three kai genes. The only exception is Gloeobacter violaceus, which has no kai genes (Nakamura et al., 2003). This incongruence, when analysed against the topologies of the CpmA tree and the 16S rRNA and RecA subtrees of cyanobacteria (Figs 1 and 2), suggests that the cpmA gene of G. violaceus was laterally transferred from one of the evolutionarily younger species. Clade C3 contains cpmA from cyanobacteria, which possess either all three clock genes (Synechococcus sp. WH 8102) or only two of them, kaiB and kaiC (three Prochlorococcus strains). The topology of this clade is also characteristic of the other components of the circadian system (i.e. kaiBC, sasA and ldpA; Dvornyk et al., 2004). Along with data from the phylogenetic analyses of 16S rRNA (Fig. 1) and RecA (Fig. 2), this suggests that the kaiA gene was laterally transferred to Synechococcus sp. WH 8102 from some other cyanobacteria. However, the taxa from which the transfer occurred can not be determined due to the limited data available.
Structure, polymorphism and mutation rates of the cpmA genes
The length of the CpmA protein homologues varied from 216 (Synechococcus sp. WH 8102) to 333 (Geobacter metallireducens) amino acid residues. According to the NCBI Conserved Domain Database (Marchler-Bauer et al., 2003), CpmA belongs to COG1691, the superfamily of NCAIR mutase (PurE)-related proteins, which catalyse the carboxylation of 5'-phosphoribosyl-5-amino-4-imidazole (AIR) to 5'-phosphoribosyl-5-aminoimidazole-4-carboxylic acid in purine biosynthesis (Meyer et al., 1992; Tiedeman et al., 1989; Watanabe et al., 1989). However, this classification is based on the general function prediction only. An analysis of the cpmA nucleotide polymorphism suggested that the gene probably consists of two functional domains. Each of them occupies about half of the gene. The N-terminal domain of the encoded protein is 85–203 amino acid residues long (corresponding to residues 1–120 of the S. elongatus PCC 7942 CpmA sequence). The C-terminal domain is less variable in length and spans 131–158 residues. It contains two hydrophobic cores (Katayama et al., 1999). This domain is homologous to the N-terminal segments of the other related proteins, AIR carboxylase (pfam00731) and NCAIR mutase (COG0041), though this homology is weak.
The two domains of the cpmA gene showed striking differences in the level of nucleotide polymorphism (Table 2). The C-terminal domain was much more highly conserved than the N-terminal one in all four major clades. The genes of C3 had a lower mean dN than those of C2. To test whether this difference is due to the more stringent evolutionary constraints in C3, the age of the respective nodes and the nonsynonymous substitution rate were estimated. The time of S. elongatus PCC 7942 speciation, about 1000 MYA (Dvornyk et al., 2003), was used as a calibration point (node 1, Fig. 1). Based on this, the time of nodes 2, 3 and 4 was estimated to be 1207·75±1·16, 1040·35±2·54 and 2018·04±7·74 MYA, respectively. Importantly, the time estimates for the nodes 1 and 3 were similar; this was in general agreement with the 16S rRNA tree, which suggested a close time of origin for the genera Synechococcus and Prochlorococcus. The respective mutation rates for nonsynonymous sites are given in Table 2. As the data suggest, the cpmA homologues from subfamilies C2 and C3 have the highest mutation rate, while the genes from two other subfamilies have a much lower one. The mutation rates in C2 and C3 are similar, so the differences in their dN estimates are likely to be due to the different evolutionary ages of the subfamilies.
|
), as compared to C2 and C3, the lower mutation rate may be due to the evolutionary constraints related to this function.
Selective constraints in evolution of cpmA genes
The site-specific likelihood models assume variable selective pressures among sites but no variation among branches of a phylogenetic tree (Nielsen & Yang, 1998; Yang et al., 2000). The parameter estimates and log-likelihood values for these models are given in Table 3. Among the discrete models, M3 fits the data significantly better than M0, M1 or M2, according to the LRT (for nested models) or AIC criteria (for nonnested comparisons: M1/M3, M2/M3). This suggests heterogeneous selective pressures along whole cpmA sequences. However, values smaller than 1 indicate no positive selection on the whole set of the cpmA genes (Table 3). Among the continuous distribution models, model M8 was significantly better than M7 (P<0·0001). However, the mean value estimated under the best model (M8) is 0·066, and the proportion of sites under positive selection p1=0, again implying strong purifying selection. These data favour the notion that, for the whole dataset, the main force acting is negative selection, with evidence for heterogeneous selection rates on different codons.
|
ratio among lineages, I proceeded from the hypothesis that selective constraints of the cpmA gene in the kaiABC-based circadian system might be influenced by the appearance of the kaiA gene and subsequent acquisition by cpmA of a new function, namely regulation of the kaiA promoter activity (Katayama et al., 1999). This might result in positive selection along branch a (Fig. 1). On the other hand, the lateral transfer of the cpmA gene from non-photosynthetic proteobacteria to photosynthetic cyanobacteria, if resulting in an acquisition of a new function by the gene, might also be followed by the period of positive selection (along branch b). I compared the log likelihood values under two models: the one-ratio model (M0 in the notation of Goldman & Yang, 1994), which assumes the same parameter for the entire tree, and the branch-specific-ratio model, which assumes the same dN/dS ratio (0) for the background lineages, while specifying either the same (1=a=b, model 1) or different (a
b, model 2) ratios for lineages a and b (Fig. 1). The likelihoods of both models (1=–22132·94 and 2=–22132·56) were significantly better (P<0·0001) than that of M0 (0=–22167·19), suggesting rejection of the one-ratio model and, consequently, different dN/dS ratios along the lineages a and b as compared to the background ratio. The LRT statistic for the comparison of models 1 and 2 is 2=2x0·38=0·76, with P=0·38 and df=1. It indicates that model 2 is not significantly better than model 1 and thus the ratios for lineages a and b are similar.
To test whether 1 is significantly higher than 1, the log likelihood value was calculated under the two-ratios model but with 1=1 fixed, giving the log likelihood value –22133·73. The two-ratios model that does not place the constraint on 1 is not significantly better (1=–22132·94); the test statistic is 2=2x0·79=1·58, with P=0·21 and df=1. So 1 is not significantly greater than 1 at the 5 % significance level (see Yang, 1998). However, this model averages the value along the sequence. Therefore, I further applied Yang and Nielsen's branch-site model B (Yang & Nielsen, 2002), which is an extension of the M3 model and potentially allows for positive selection at some sites in both background and foreground lineages. This model suggested no sites under positive selection in the background branches (
=0·085) and about 8 % positively selected sites in the foreground branches (
=35·182). However, model B yielded the log likelihood value –21333·31, which does not fit the data better than the site-specific model M3 (Table 3). This suggests that, although dN/dS ratios vary along the different branches as well as among codons, they are probably not above 1 and thus do not indicate any positive selection.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Evolution of the cpmA genes of cyanobacteria has involved at least one influential horizontal transfer from proteobacteria, most likely from the early branching ones (Fig. 1). The fact that clade C3 is monophyletic suggests that transfer of this subfamily probably occurred before the speciation of Prochlorococcus. Based on the results of the molecular clock analysis, the time of the transfer may be estimated as about 1000 MYA. This is close to the time of origin of two other circadian-related genes, kaiA and ldpA (Dvornyk et al., 2003; Dvornyk, 2005). This correspondence is probably not accidental and may be in favour of the assumption that the most important events resulting in the ‘evolutionary upgrade’ of the kaiBC-based system to the kaiABC-based one (e.g. appearance of kaiA, ldpA and maybe some other as yet unknown genes) occurred about 1000 MYA (Dvornyk et al., 2003).
Horizontal (lateral) gene transfer is important in the evolution and adaptation of prokaryotes (Gogarten et al., 2002; Koonin et al., 2001). Recent data showed that, in the evolution of the cyanobacterial circadian system, lateral transfers of its various elements have been a common event (Dvornyk et al., 2003, 2004; Dvornyk & Nevo, 2004). However, except for a few documented transfers of the kaiB and kaiC genes from cyanobacteria to proteobacteria (Dvornyk et al., 2003), the other transfers have occurred only between cyanobacteria with the same type of system, either the kaiABC- or the kaiBC-based one (Dvornyk et al., 2004; Dvornyk, 2005).
The phylogenetic trees of all the components of the cyanobacterial circadian system studied so far (e.g. the kaiBC operon, sasA, ldpA and cpmA) have featured two clades (corresponding to C1 and C3 in Fig. 1). While the topology of each clade may vary, a set of the species within the clade is always the same. That is, none of the species of clade C3 ever appears in C1, and vice versa. The observed major congruence between the phylogenies of the various components of the circadian system (e.g. the kaiBC operon, sasA, ldpA and cpmA) has several important implications. First, it indicates that evolution of the genes involved in circadian control is probably system-specific, i.e. the genes within each of the two types of the cyanobacterial circadian system evolve concordantly. Second, each type of circadian system has specific functional and selective constraints, which preclude horizontal transfers of the components between the systems. The only exception to this rule is Synechococcus sp. WH 8102, which has all three kai genes. This might suggest that its circadian system is the kaiABC-based one. However, phylogenetic analyses of all currently known components of the system unambiguously place this species together with Prochlorococcus, which has the kaiBC-based system. Based on the analyses of the 16S rRNA (Fig. 1) and RecA (Fig. 2) sequences, Synechococcus sp. WH 8102 is phylogenetically positioned between S. elongatus PCC 7942 and Prochlorococcus. The tree topologies of the different circadian components may correspond merely because the 16S rRNA and RecA trees are similar and the rate of lateral transfer is low. However, this is unlikely in the case being considered. As mentioned above, the phylogenetic trees of the elements of the cyanobacterial circadian system studied so far are not fully congruent within the two main clades (Dvornyk et al., 2003, 2004). The incongruence may be due to a number of analytical and biological factors (Rokas et al., 2003), e.g. the choice of optimality criterion (Huelsenbeck, 1995), limited data availability (Cummings et al., 1995), taxon sampling (Graybeal, 1998), specific assumptions in the modelling of sequence evolution (Yang et al., 1994), natural selection or genetic drift (Maddison, 1997; Martin & Burg, 2002), and others. Among the possible reasons, horizontal transfers of the cpmA homologues seem to be the most feasible explanation of the observed incongruence. Apparently, horizontal transfers have been rather common between the species of the same clade. For example, multiple transfers of the kaiBC operon between closely related filamentous cyanobacteria in the Nostocaceae were recently reported (Dvornyk & Nevo, 2004). As follows from the results of the molecular phylogenetic analyses using small ribosomal subunit RNA (Turner, 1997), 16S rRNA (Fig. 1) and RecA (Fig. 2), the unicellular cyanobacteria S. elongatus PCC 7942 and Prochlorococcus are close relatives and thus clustered together. This might suggest that they should either have the same type of circadian system or experience horizontal transfers of some its components. However, in the phylogenies of all known circadian-related genes, S. elongatus PCC 7942 and Prochlorococcus sp. always appear in clearly separated clades (Dvornyk et al., 2004; Dvornyk, 2005; Dvornyk & Knudsen, 2005). This means that the cyanobacterial system of S. elogatus PCC 7942 is different from that of Prochlorococcus sp. and is closer to the system of more phylogenetically distant filamentous cyanobacteria (Nostoc, Anabaena). So far, no apparent horizontal transfers between the different systems have been detected, except for the kaiA gene. Furthermore, the significant differences in substitution rates between the components of the two systems (Dvornyk et al., 2004) are in favour of the assumption that the functional constraints are indeed system-specific.
In S. elongatus PCC 7942, cpmA was shown to modify the phasing and amplitude of class 1 rhythms (Katayama et al., 1999), which have their circadian peak near subjective dusk and their trough at subjective dawn (Liu et al., 1995). In particular, the disruption of cpmA significantly affects expression of psbAI and psbAII (Katayama et al., 1999), which encode two forms of the photosystem II reaction centre D1 protein (Golden et al., 1986; Schaefer & Golden, 1989). Given that among the prokaryotes with the cpmA homologues (Table 1) only cyanobacteria are photosynthetic, the genes of subfamilies C1 and C3 (Fig. 1) might be expected to perform similar functions related to photosynthesis. On the other hand, an important role of cpmA in the kaiABC-based circadian system is regulation of the kaiA promoter activity (Katayama et al., 1999). In this system, kaiA is critical for circadian oscillation, because it functions as an activator of the kaiBC promoter (Ishiura et al., 1998). The exact mechanism of the effect of cpmA on circadian-like expression of the photosynthesis-related genes is still unknown. It may involve kaiA as an essential element. In such a case, in the kaiBC-based system lacking kaiA, the relevant function of cpmA is abolished and thus the genes of C3 seem not to be associated with the circadian system. However, if control of circadian-dependent expression of the photosynthesis-related genes by cpmA employs a different mechanism, without kaiA, and this mechanism is available in all cyanobacteria, another scenario is possible. It may imply that the lower mutation rate in the C1 subfamily is due to the constraints conferred by the interaction with the kaiA promoter, but in the rest, the genes of C1 and C3 perform essentially the same circadian-associated functions. While the former is probable, the latter seems more unlikely than the previous scenario. Indeed, as mentioned above, the lateral transfer of the C3 subfamily to cyanobacteria occurred about 1000 MYA, i.e. after the kaiB and kaiC genes formed an operon and joined into the circadian protosystem that happened between 3500 and 2320 MYA (Dvornyk et al., 2003), and after photosynthesis originated (Xiong et al., 2000). Moreover, the cpmA genes were probably transferred from evolutionarily old non-photosynthetic proteobacteria, i.e. a priori were not involved in regulation of photosynthesis. If the transferred cpmA genes were to acquire a new photosynthesis-related function, this process should hypothetically involve either positive Darwinian selection for functional divergence (Ohno, 1970) or relaxation of selective constraints (Zhang et al., 1998). However, as the results of the present study suggest, the cpmA genes of the C3 subfamily are under strong purifying selection. Although the genes of subfamily C1 are under strong purifying selection too, they, in contrast to C3, have evolved together with photosynthesis-related genes and the other components of the circadian system. Hence their evolution should be concordant to ensure proper interactions between the genes.
The results of the present study suggest that the functional and selective constraints of the kaiABC-based system have slowed down the rate of sequence evolution of the cpmA genes as compared to the homologues of the kaiBC-based system. This supports the hypothesis that the emergence of kaiA was one of the most influential events in the evolution of the system and resulted in changes of selective and functional constraints of the other system components (Dvornyk et al., 2004; Dvornyk, 2005). A vector of these changes probably depends on the role of a particular element in the system and may be in favour of either relaxing the constraints (e.g. for the sasA and ldpA genes: Dvornyk et al., 2004; Dvornyk, 2005) or making them more stringent (as for the cpmA gene). However, purifying selection probably remains one of the primary selective forces of their evolution.
The present study of the cpmA genes provides further evidence for the complex nature of evolution of the prokaryotic circadian system and the different time of origin of its components. It also raises new questions to be answered. For example, the system with kaiA remains functional even after disruption of some other elements, e.g. sasA (Iwasaki et al., 2000), ldpA (Katayama et al., 2003) or cpmA (Katayama et al., 1999), despite the lower fitness of the mutants. Does the kaiBC-based system do similarly? How does the N-terminal domain of cpmA contribute to the circadian function of the gene? The evolutionary data obtained provide a basis for further studies of molecular mechanisms underlying the control of circadian rhythmicity in cyanobacteria.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 3389–3402.
Castenholz, R. W. (1989). Subsection IV, order Nostocales. In Bergey's Manual of Systematic Bacteriology, pp. 1780–1793. Edited by J. T. Staley, M. P. Bryant, N. Pfennig & J. G. Holt. Baltimore: Williams & Wilkins.
Cummings, M. P., Otto, S. P. & Wakeley, J. (1995). Sampling properties of DNA sequence data in phylogenetic analysis. Mol Biol Evol 12, 814–822.[Abstract]
Ditty, J. L., Williams, S. B. & Golden, S. S. (2003). A cyanobacterial circadian timing mechanism. Annu Rev Genet 37, 513–543.[CrossRef][Medline]
Dvornyk, V. (2005). Molecular evolution of ldpA, a gene mediating circadian input signal in cyanobacteria. J Mol Evol 60, 105–112.[CrossRef][Medline]
Dvornyk, V. & Knudsen, B. (2005). Functional divergence of circadian clock proteins in prokaryotes. Genetica 124, 247–254.[CrossRef][Medline]
Dvornyk, V. & Nevo, E. (2004). Evidence for multiple lateral transfers of the circadian clock cluster in filamentous heterocystic cyanobacteria Nostocaceae. J Mol Evol 58, 341–347.[CrossRef][Medline]
Dvornyk, V., Vinogradova, O. N. & Nevo, E. (2002). Long-term microclimatic stress causes rapid adaptive radiation of kaiABC clock gene family in a cyanobacterium, Nostoc linckia, from the "Evolution Canyons" I and II, Israel. Proc Natl Acad Sci U S A 99, 2082–2087.
Dvornyk, V., Deng, H. W. & Nevo, E. (2004). Structure and molecular phylogeny of sasA genes in cyanobacteria: insights into evolution of the prokaryotic circadian system. Mol Biol Evol 21, 1468–1476.
Dvornyk, V., Vinogradova, O. N. & Nevo, E. (2003). Origin and evolution of circadian clock genes in prokaryotes. Proc Natl Acad Sci U S A 100, 2495–2500.
Garcia-Pichel, F., Lopez-Cortes, A. & Nubel, U. (2001). Phylogenetic and morphological diversity of cyanobacteria in soil desert crusts from the Colorado plateau. Appl Environ Microbiol 67, 1902–1910.
Gogarten, J. P., Doolittle, W. F. & Lawrence, J. G. (2002). Prokaryotic evolution in light of gene transfer. Mol Biol Evol 19, 2226–2238.
Golden, S. S., Brusslan, J. & Haselkorn, R. (1986). Expression of a family of psbA genes encoding a photosystem II polypeptide in the cyanobacterium Anacystis nidulans R2. EMBO J 5, 2789–2798.[Medline]
Goldman, N. & Yang, Z. (1994). A codon-based model of nucleotide substitution for protein-coding DNA sequences. Mol Biol Evol 11, 725–736.[Abstract]
Graybeal, A. (1998). Is it better to add taxa or characters to a difficult phylogenetic problem? Syst Biol 47, 9–17.[CrossRef][Medline]
Honda, D., Yokota, A. & Sugiyama, J. (1999). Detection of seven major evolutionary lineages in cyanobacteria based on the 16S rRNA gene sequence analysis with new sequences of five marine Synechococcus strains. J Mol Evol 48, 723–739.[CrossRef][Medline]
Huelsenbeck, J. P. (1995). Performance of phylogenetic methods in simulation. Syst Biol 44, 17–48.[CrossRef]
Ishiura, M., Kutsuna, S., Aoki, S., Iwasaki, H., Andersson, C. R., Tanabe, A., Golden, S. S., Johnson, C. H. & Kondo, T. (1998). Expression of a gene cluster kaiABC as a circadian feedback process in cyanobacteria. Science 281, 1519–1523.
Iwasaki, H., Williams, S. B., Kitayama, Y., Ishiura, M., Golden, S. S. & Kondo, T. (2000). A kaiC-interacting sensory histidine kinase, SasA, necessary to sustain robust circadian oscillation in cyanobacteria. Cell 101, 223–233.[CrossRef][Medline]
Johnson, C. H. & Golden, S. S. (1999). Circadian programs in cyanobacteria: adaptiveness and mechanism. Annu Rev Microbiol 53, 389–409.[CrossRef][Medline]
Kaneko, T., Nakamura, Y., Wolk, C. P. & 18 other authors (2001). Complete genomic sequence of the filamentous nitrogen-fixing cyanobacterium Anabaena sp. strain PCC 7120. DNA Res 8, 205–213.[Abstract]
Katayama, M., Tsinoremas, N. F., Kondo, T. & Golden, S. S. (1999). cpmA, a gene involved in an output pathway of the cyanobacterial circadian system. J Bacteriol 181, 3516–3524.
Katayama, M., Kondo, T., Xiong, J. & Golden, S. S. (2003). ldpA encodes an iron-sulfur protein involved in light-dependent modulation of the circadian period in the cyanobacterium Synechococcus elongatus PCC 7942. J Bacteriol 185, 1415–1422.
Kishino, H. & Hasegawa, M. (1989). Evaluation of the maximum likelihood estimate of the evolutionary tree topologies from DNA sequence data, and the branching order in Hominoidea. J Mol Evol 29, 170–179.[CrossRef][Medline]
Kondo, T. & Ishiura, M. (2000). The circadian clock of cyanobacteria. Bioessays 22, 10–15.[CrossRef][Medline]
Koonin, E. V., Makarova, K. S. & Aravind, L. (2001). Horizontal gene transfer in prokaryotes: quantification and classification. Annu Rev Microbiol 55, 709–742.[CrossRef][Medline]
Kumar, S., Tamura, K. & Nei, M. (2004). MEGA3: Integrated software for Molecular Evolutionary Genetics Analysis and sequence alignment. Brief Bioinform 5, 150–163.
Kutsuna, S., Kondo, T., Aoki, S. & Ishiura, M. (1998). A period-extender gene, pex, that extends the period of the circadian clock in the cyanobacterium Synechococcus sp. strain PCC 7942. J Bacteriol 180, 2167–2174.
Liu, Y., Tsinoremas, N. F., Johnson, C. H., Lebedeva, N. V., Golden, S. S., Ishiura, M. & Kondo, T. (1995). Circadian orchestration of gene expression in cyanobacteria. Genes Dev 9, 1469–1478.
Lloyd, A. T. & Sharp, P. M. (1993). Evolution of the recA gene and the molecular phylogeny of bacteria. J Mol Evol 37, 399–407.[Medline]
Lyra, C., Suomalainen, S., Gugger, M., Vezie, C., Sundman, P., Paulin, L. & Sivonen, K. (2001). Molecular characterization of planktic cyanobacteria of Anabaena, Aphanizomenon, Microcystis and Planktothrix genera. Int J Syst Evol Microbiol 51, 513–526.[Abstract]
Maddison, W. P. (1997). Gene trees in species trees. Syst Biol 46, 523–536.[CrossRef]
Marchler-Bauer, A., Anderson, J. B., DeWeese-Scott, C. & 24 other authors (2003). CDD: a curated Entrez database of conserved domain alignments. Nucleic Acids Res 31, 383–387.
Martin, A. P. & Burg, T. M. (2002). Perils of paralogy: using HSP70 genes for inferring organismal phylogenies. Syst Biol 51, 570–587.[CrossRef][Medline]
Meyer, E., Leonard, N. J., Bhat, B., Stubbe, J. & Smith, J. M. (1992). Purification and characterization of the purE, purK, and purC gene products: identification of a previously unrecognized energy requirement in the purine biosynthetic pathway. Biochemistry 31, 5022–5032.[CrossRef][Medline]
Nakamura, Y., Kaneko, T., Sato, S. & 16 other authors (2003). Complete genome structure of Gloeobacter violaceus PCC 7421, a cyanobacterium that lacks thylakoids. DNA Res 10, 137–145.[Abstract]
Nei, M. & Gojobori, T. (1986). Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol 3, 418–426.[Abstract]
Nielsen, R. & Yang, Z. (1998). Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene. Genetics 148, 929–936.
Ohno, S. (1970). Evolution by Gene Duplication. Berlin: Springer.
Posada, D. & Crandall, K. A. (1998). MODELTEST: testing the model of DNA substitution. Bioinformatics 14, 817–818.
Rokas, A., King, N., Finnerty, J. & Carroll, S. B. (2003). Conflicting phylogenetic signals at the base of the metazoan tree. Evol Dev 5, 346–359.[CrossRef][Medline]
Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4, 406–425.[Abstract]
Schaefer, M. R. & Golden, S. S. (1989). Light availability influences the ratio of two forms of D1 in cyanobacterial thylakoids. J Biol Chem 264, 7412–7417.
Schmitz, O., Katayama, M., Williams, S. B., Kondo, T. & Golden, S. S. (2000). CikA, a bacteriophytochrome that resets the cyanobacterial circadian clock. Science 289, 765–768.
Shimodaira, H. & Hasegawa, M. (1999). Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Mol Biol Evol 16, 1114–1116.
Sitnikova, T., Rzhetsky, A. & Nei, M. (1995). Interior-branch and bootstrap tests of phylogenetic trees. Mol Biol Evol 12, 319–333.[Abstract]
Tamura, K. & Nei, M. (1993). Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol Biol Evol 10, 512–526.[Abstract]
Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673–4680.
Tiedeman, A. A., Keyhani, J., Kamholz, J., Daum, H. A., III, Gots, J. S. & Smith, J. M. (1989). Nucleotide sequence analysis of the purEK operon encoding 5'-phosphoribosyl-5-aminoimidazole carboxylase of Escherichia coli K-12. J Bacteriol 171, 205–212.
Tsinoremas, N. F., Ishiura, M., Kondo, T., Tanaka, K., Takahashi, H., Johnson, C. H. & Golden, S. S. (1996). A sigma factor that modifies the circadian expression of a subset of genes in cyanobacteria. EMBO J 15, 2488–2495.[Medline]
Turner, S. (1997). Molecular systematics of oxygenic photosynthetic bacteria. Plant Syst Evol [Suppl] 11, 13–52.
Watanabe, W., Sampei, G., Aiba, A. & Mizobuchi, K. (1989). Identification and sequence analysis of Escherichia coli purE and purK genes encoding 5'-phosphoribosyl-5-amino-4-imidazole carboxylase for de novo purine biosynthesis. J Bacteriol 171, 198–204.
Whelan, S. & Goldman, N. (2001). A general empirical model of protein evolution derived from multiple protein families using a maximum-likelihood approach. Mol Biol Evol 18, 691–699.
Xiong, J., Fischer, W. M., Inoue, K., Nakahara, M. & Bauer, C. E. (2000). Molecular evidence for the early evolution of photosynthesis. Science 289, 1724–1730.[CrossRef][Medline]
Yang, Z. (1997). PAML: a program package for phylogenetic analysis by maximum likelihood. CABIOS 15, 555–556.
Yang, Z. (1998). Likelihood ratio tests for detecting positive selection and application to primate lysozyme evolution. Mol Biol Evol 15, 568–573.[Abstract]
Yang, Z. & Nielsen, R. (2002). Codon-substitution models for detecting molecular adaptation at individual sites along specific lineages. Mol Biol Evol 19, 908–917.
Yang, Z., Goldman, N. & Friday, A. (1994). Comparison of models for nucleotide substitution used in maximum-likelihood phylogenetic estimation. Mol Biol Evol 11, 316–324.[Abstract]
Yang, Z., Nielsen, R., Goldman, N. & Pedersen, A.-M. K. (2000). Codon-substitution models for heterogeneous selection pressure at amino acid sites. Genetics 155, 431–449.
Yoder, A. D. & Yang, Z. (2000). Estimation of primate speciation dates using local molecular clocks. Mol Biol Evol 17, 1081–1090.
Zhang, J., Rosenberg, H. F. & Nei, M. (1998). Positive Darwinian selection after gene duplication in primate ribonuclease genes. Proc Natl Acad Sci U S A 95, 3708–3713.
Received 1 August 2005;
revised 30 September 2005;
accepted 4 October 2005.
This article has been cited by other articles:
|
|
 |
|
  B. M. Tyler, S. Tripathy, X. Zhang, P. Dehal, R. H. Y. Jiang, A. Aerts, F. D. Arredondo, L. Baxter, D. Bensasson, J. L. Beynon, et al. Phytophthora genome sequences uncover evolutionary origins and mechanisms of pathogenesis. Science, September 1, 2006; 313(5791): 1261 - 1266. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
