Bot. Bull. Acad. Sin. (2001) 42: 181-186

Turner et al. Molecular phylogeny of cyanobacteria

Molecular phylogeny of nitrogen-fixing unicellular cyanobacteria

Sen Turner1,3, Tan-Chi Huang2, and Shu-Miaw Chaw2,*

1Department of Biology, Indiana University, Bloomington, Indiana 47405-3700, USA

2Institute of Botany, Academia Sinica, Taipei 115, Taiwan

(Received August 25, 2000; Accepted December 11, 2000)

Abstract. Molecular phylogenetic study was conducted using maximum likelihood tree inference methods with small subunit ribosomal RNA sequence data to ascertain the evolutionary relationships among sheathless, single-cell cyanobacteria capable of nitrogen fixation. Cyanobacterial strains of the genus Cyanothece (circumscribed by Waterbury and Rippka, 1989) fall into at least three independent lines of descent within a larger assemblage previously designated the SPM sequence group. No strong correlation between aerobic versus anaerobic nitrogen-fixing activity and phylogenetic relationships was observed. The results support a hypothesis of multiple gains and/or losses of nitrogen-fixation abilities among the sheathless, unicellular cyanobacteria.

Keywords: Aerobic; Anaerobic; Cyanobacteria; Nitrogen fixation; Phylogeny; Ribosomal RNA; Sheathless; Small subunit.

Introduction

The unicellular cyanobacteria exhibit a great diversity of physiological properties. Few of them are able to fix nitrogen, either aerobically or anaerobically. Those capable of fixing nitrogen can be grouped into two types: sheathed and sheathless, the former being enclosed in a glycoprotein sheath (glycocalyx, capsule) external to the cell wall (Castenholz and Waterbury, 1989). The first sheathed species was classified in the genus Gloeocapsa (Wyatt and Silvey, 1969); however, it was later assigned to the genus Gloeothece, based on patterns of cell division (Rippka et al., 1979). On the other hand, the sheathless isolates were primarily attributed to the genus Synechococcus (Huang and Chow, 1986; Len et al., 1986) or Aphanothece (Singh, 1973; Ni et al., 1988). More recently, Reddy et al. (1993) treated their sheathless isolates as members of Cyanothece, in agreement with the newer taxonomic criteria proposed by Waterbury and Rippka (1989).

Due to the fact that both botanical and bacteriological criteria have been used by various authorities to classify cyanobacteria, the systematics of these organisms has undergone numerous changes. (For a review of the problematic nature of cyanobacterial systematics, see Turner,

1997). The genus Cyanothece was first proposed by Komrek (1976) to accommodate some species previously placed in Synechococcus, the major feature distinguishing the former from the latter being that Cyanothece are present as single cells or in pairs, but never grouped into chains as are some Synechococcus. Subsequently, Waterbury and Rippka (1989) reserved the genus Cyanothece for unicellular cyanobacteria that lacked sheaths, divided in a single plane, and were larger than 3 m in diameter. They also noted that, with one possible exception, the seven strains they originally placed in this genus were able to express nitrogenase activity either aerobically or anaerobically. The physiologies of these two kinds of nitrogen-fixing cyanobacteria are quite distinct. The aerobic type has developed a strategy to protect nitrogenase from oxygen generated by photosynthesis, but the anaerobic type either has not or has subsequently lost the ability to do so.

Because of the unique capability of fixing nitrogen and carbon dioxide within the same cell under aerobic conditions, the sheathless aerobic type has attracted more attention and has been better characterized than the anaerobic type (Mitsui et al., 1986; Huang et al., 1990; Coln-Lpez et al., 1997). Therefore, our knowledge of the anaerobic type is very limited, and more studies are needed to justify whether the accommodation of the above two types of unicellular, nitrogen-fixing cyanobacteria in the same genus are reasonable. In this report, we present a phylogenetic analysis of small ribosomal subunit RNA (SSU rRNA) sequences to evaluate the phylogenetic relationships among sheathless, unicellular cyanobacteria with nitrogen-fixation abilities.

3 Present address: NCBI (GenBank), National Library of Medicine, NIH, Building 38A, 8600 Rockville Pike, Bethesda, Maryland 20894, USA.

*Corresponding author. Fax: 886-2-27827954; E-mail: bochaw@sinica.edu.tw


Botanical Bulletin of Academia Sinica, Vol. 42, 2001

Materials and Methods

Bacterial Growth and Gene Sequencing

Isolates of Synechococcus RF-1 (syn. Cyanothece PCC 8801) and Cyanothece PCC 7418 were obtained from the laboratory of T.-C. Huang and Institut Pasteur (Paris, France), respectively. They were grown in nitrate-free BG-11 medium (Stanier et al., 1971) supplemented with 0.01M HEPES buffer (pH 8.0). The cultures were incubated in continuous light at 28C, and the cells were harvested during log phase.

In order to avoid the amplification of pseudogenes or otherwise nonfunctional genes, or any post-transcriptionally processed intervening sequences potentially present, total RNA was extracted from fresh cells using the modified method of Raha et al. (1990) in which genomic DNA was removed by DNase I (Boehringer Mannheim). The method of Goodman and MacDonald (1979) was then used to synthesize the first strand cDNA with the AMV reverse transcriptase (Promega) and the primer 16SR (AGAAAGGAGGTGATCCAGCC) to prime the 3' end. The reverse transcription reaction contained (per l) 20 ng RNA template, 0.5 U AMV reverse transcriptase, 1 pmol primer, 5 nmole MgCl2 and 1 nmole dNTPs. The reaction was incubated at 42C for 30 mins. Three independent cDNAs were synthesized and used as templates for PCR amplification following the method of Chaw et al. (1995) except that primer 16SF (AGAGTTTGATCCTGGCTCAG) and 16SR were used. A negative control, in which template was replaced by double distilled water, and an additional control, in which total RNA served as template in the absence of reverse transcriptase, were also run during all PCR amplifications. The former was used to detect any contamination in the PCR components, and the latter to determine if there was any trace of undigested DNA present in the RNA preparation.

PCR products were purified using Promega WizardTM PCR Preps DNA Purification System, and cloned into a pGEM T-Easy vector (Promega, WI). Plasmid DNAs were purified using Qiaprep Spin Miniprep Kit (Qiagen, Hilden, Germany), and sequenced with additional internal primers on an ABI 373 automated sequencer (Applied Biosystem, CA). Independent PCR clones were sequenced using universal (M13, M13 reverse, and/or T7) primers plus additional sequencing primers (530F, TGCCAGCAGCCGC GGTA; 1170F, CGTGTCGTGAGATGTTG; 630R, TTCCGG ATAACGCTTGC; 1100R, GGGTTGCGCTCGTTGC).

The sequences generated in this study have been deposited in the GenBank/ EMBL/DDBJ databases under the accession numbers AF296872 and AF296873.

Phylogenetic Analysis

The SSU rRNA sequences of Synechococcus RF-1 and Cyanothece PCC 7418 were aligned with others in a cyanobacterial SSU rRNA database maintained by one of us (S. Turner). Initial alignment was done by eye and adjusted as necessary following secondary structure analysis of the new sequences using the structure of

Synechococcus PCC 6301 as a template (Gutell, 1993). A total of 1,377 unambiguously aligned sequence positions were used in phylogenetic analyses (Turner et al., 1999). A preliminary maximum likelihood analysis (not shown) indicated that both sequences fell within the Synechocystis/Pleurocapsa/Microcystis (SPM) SSU rRNA sequence group as previously defined (Turner, 1997; Turner et al., 1999). This sequence group corresponds to Group 5 of Honda et al. (1999) and Branch G of Wilmotte (1994).

A more refined analysis limited to the SPM sequence group was carried out using all available, approximately full-length, unique SSU rRNA sequences for strains of cyanobacteria known to fall within this group, except that representation of the genus Microcystis, for which over 70 full-length sequences are available, was limited to the single suggested type strain Microcystis aeruginosa PCC 7941. The taxa and accession numbers of the sequences in this analysis are given in Table 1. The transition/transversion (ti/tv) ratio was estimated using the program TREE-PUZZLE (Version 4.0.1, formerly named PUZZLE) under

Table 1. Taxa and corresponding sequence accession numbers.

Climacodium frauenfeldianum symbiont AF193247

Cyanobacterium stanieri PCC 7202a AF132782

(syn. Synechococcus cedrorum SAG 88.79)

Cyanothece ATCC 51142 AF132771

Cyanothece PCC 7418b AF296872

Cyanothece PCC 7424b AF132932

Dactylococcopsis salina PCC 8305a AJ000711

Euhalothece MPI 95AH10 AJ000709

Euhalothece MPI 95AH13 AJ000710

Euhalothece MPI 96N304 AJ000713

Gloeocapsa PCC 73106 b AF132784

Gloeothece membranacea PCC 6501a X78680

Halospirulina tapeticola CCC Baja-95 C1.2a Y18791

Halospirulina CCC Baja-95 C1.3 Y18790

Halospirulina MPI S3 Y18789

Halothece MPI 96P605 AJ000724

Merismopedia "glauca" B1448-1 X94705

Microcystis aeruginosa PCC 7941a AJ133171

"Oscillatoria rosea" IAM M-220 AB003164

Pleurocapsa PCC 7516 b X78681

Prochloron didemni X63141

Spirulina major PCC 6313a X75045

Spirulina "subsalsa" IAM M-223 AB003166

Spirulina P7 AF091109

Spirulina CCC Snake P. Y-85 Y18793

Spirulina MPI S4 Y18792

Stanieria cyanosphaera PCC 7437a AF132931

Synechocystis PCC 6803 D64000

Synechococcus PCC 7002 AJ000716

Synechococcus PCC 7003 b AB015059

Synechococcus PCC 7117 AB015060

Synechococcus PCC 73109 AB015061

"Synechococcus" RF-1 AF296872 (syn. Cyanothece PCC 8801)

Xenococcus PCC 7305 b AF132783

Epithets in quotes are provisional.

aSuggested type strain (Rippka and Herdman, 1992, Nbel et al., 2000).

bSuggested reference strain (Rippka and Herdman, 1992).


Turner et al. Molecular phylogeny of cyanobacteria

two models of evolution, the first being the Tamura-Nei model under the assumption of a uniform rate of change among sequence sites, and the second under the assumption that site-to-site variability in evolutionary rate follows a Gamma distribution approximated by 16 rate categories (Tamura and Nei, 1993; Strimmer and von Haeseler, 1996; Strimmer et al., 1997). Settings for the TREE-PUZZLE program were as follows: approximate quartet likelihood, exact parameter estimates, parameters estimated from data set, and parameter estimation uses set to "quartet sampling + NJ tree." Initial maximum likelihood trees were inferred with the program fastDNAml (Versions 1.0 and 1.1.1a) using the ti/tv ratio estimated by TREE-PUZZLE under a uniform rate of change (ti/tv = 1.39) (Olsen et al., 1994). Trees were inferred 40 times with random sequence addition, localized branch swapping across a maximum of five branches during tree building, and global branch swapping after addition of the last taxon. A total of nine unique topologies were found with log likelihood scores ranging from -9068.16948 (best) to -9071.49465 (worst). Each of these trees was subsequently input into the program DNArates (Version 1.0.3). This program uses the tree topology and branch lengths in conjunction with the aligned sequences to make a maximum likelihood estimate of the evolutionary rate of change for each site in the sequence alignment. The program then parses the sites into a user-defined number of rate categories, in this case 16, the rate of each category being an average of the rates of the individual sites within it (Pracht, S., Overbeek, R., and Olsen, G.J., personal communication). The ti/tv ratio used in this step and subsequent ones was that estimated by TREE-PUZZLE under a model of Gamma-distributed rates (ti/tv = 2.00).

For each set of rate categories estimated from the nine initial trees, phylogenetic trees were inferred de novo using the "categories" option of fastDNAml, the same branch-swapping scheme as before, and multiple random additions of sequences until the best tree had been found at least five times. This step resulted in a total of three tree topologies, two of which had not been found in the first tree-inference step, with log likelihood scores ranging from -7231.59110 to -7237.83215. Rates were calculated de novo using these three trees as input to the DNArates program and trees were subsequently inferred de novo as before, resulting in two new unique topologies with log likelihood scores of -7219.3025 and -7226.46717. A third round of rates determination and tree inference resulted in no change in tree topologies, but slight improvement in log likelihood scores due to further optimization of branch lengths (log likelihood = -7215.47021 and -7221.79242, respectively).

Shortly after this analysis was completed, five new sequences that also fall into the SPM sequence group became available from the public DNA databases (accession numbers Y18789 to Y18793) (Nbel et al., 2000). These were added to the best current phylogenetic tree using the "restart" option of fastDNAml and the previous ti/tv value, rate categories, and branch-swapping scheme. The combined data set of 33 taxa was also reanalyzed with

TREE-PUZZLE as before again assuming 16 Gamma-distributed rate categories. The newly estimated ti/tv ratio (2.28) was then used with the new tree as input to DNArates, and these latest rate categories along with the new ti/tv ratio were used to infer a 33 taxon tree de novo as before.

Confidence values for groupings of the final, best tree were determined by using it in conjunction with fastDNAml with the "global," "restart," and "keep" options. Based on log likelihood scores, the best 1000 trees out of 3540 topologies examined were retained and used as input to the program TreeCons (version 1.0) (Jermiin et al., 1997). Of these, 480 were found not to be significantly different from the best tree at the 1% significance level using the Kishino-Hasegawa test (Kishino and Hasegawa, 1989). These 481 trees were used to compute a majority-rule consensus tree with associated relative likelihood support (RLS) scores under a standardized, exponential weighting scheme (Class V in the TreeCons program). For presentation purposes, trees were graphically modified using the programs TreeView (Version 1.5) (Page, 1996) and MacClade (Version 3.05) (Maddison and Maddison, 1992).

Results and Discussion

Sequences that were generated by the RT-PCR method were of the same length as those of other cyanobacteria generated by DNA-based PCR in other studies. In the case of Cyanothece strain PCC 7418, the sequence is essentially identical to that generated by PCR amplification of genomic DNA (accession AJ000708) (Garcia-Pichel et al., 1998). From these results, we conclude that neither post-transcriptional modification, such as the excision of transcribed intragenic intervening sequences, nor the fidelity of RT-PCR is an issue in this study.

The optimal unrooted maximum likelihood tree is shown as a phylogram in Figure 1 with RLS scores indicated for the internal branches. The same tree is shown as a cladogram in Figure 2. Branches with RLS <90% are considered to be poorly supported and are truncated in Figure 2 (Turner et al., 1999). For purposes of discussion, RLS scores ranging from 90% to 95% will be considered indicative of weak support, from 95% to 98% indicative of moderate support, and 98% to 100% indicative of strong support.

Of the two taxa examined in this study, Cyanothece PCC 7418 falls within a clade of halophilic cyanobacteria containing strains assigned to the genera Dactylococcopsis and Euhalothece, a grouping that receives strong support (RLS = 100%). In particular, Cyanothece PCC 7418 appears most closely related to Dactylococcopsis salina PCC 8305 with moderate support (RLS = 97.6%). Halothece MPI 96P605 is the sister taxon to this clade, also with strong support (RLS = 98.5%). These results are in agreement with those of an earlier study (Garcia-Pichel et al., 1998).

Synechococcus RF-1 groups with weak support with two other cyanobacteria, Cyanothece ATCC 51142 and an unidentified unicellular endosymbiont of the marine diatom


Botanical Bulletin of Academia Sinica, Vol. 42, 2001

Climacodium frauenfeldianum (RLS = 94.9%). The association of strains RF-1 and ATCC 51142 is in agreement with their highly similar ultrastructures, particularly the presence of polymorphic bodies (inclusion granules) throughout the cytoplasm under aerobic nitrogen-fixing conditions (Chou and Huang, 1991; Reddy et al., 1993).

Strain RF-1 has been deposited in the Pasteur Culture Collection of cyanobacteria, where it has been given the designation Cyanothece PCC 8801 on the basis of phenotypic criteria. However, the results shown in Figures 1 and 2 indicate that those strains assigned to the genus Cyanothece are polyphyletic with four strains (including RF-1) falling into three distinct sequence clusters containing strains assigned to other genera. Besides Cyanothece PCC 7418 falling within the Euhalothece/Dactylococcopsis cluster and strain RF-1 grouping with Cyanothece ATCC 51142 and the Climacodium symbiont, Cyanothece PCC 7424 is related to Gloeothece membranacea PCC 6501 with strong support (RLS = 98.1%). This result suggests that strain PCC 7424 may be a sheathless

Figure 2. Unrooted phylogenetic tree as in Figure 1 but represented as a cladogram with truncation of internal branches having less than 90% RLS scores to yield polytomies. Support for internal branches is indicated by a thick horizontal line for strong support (98% RLS 100%), by a thin horizontal line for moderate support (95% RLS < 98%), and by a broken line for weak support (90% RLS < 95%). Arrows point to the strains with prominent sheaths. Nitrogenase activity is indicated as follows: +, aerobic activity; (+) anaerobic activity; -, no activity; ?, no information. Data on nitrogenase activity are from Reddy et al. (1993), Rippka et al. (1979), and Waterbury and Rippka (1989). For further details, see Figure 1 legend.

variant of Gloeothece, a possibility raised previously (Rippka and Cohen-Bazire, 1983). This pair groups in turn with Microcystis aeruginosa PCC 7941 with moderate support (RLS = 95.3%).

Irrespective of the polyphyly of the genus Cyanothece, strain RF-1 is clearly not a close relative of other strains in the SPM sequence group that have been assigned to the genus Synechococcus, specifically PCC strains 7002, 7003, 7117, and 73109. These four strains appear to be closely related to one another and with the filamentous cyanobacterium Oscillatoria IAM M-220 (RLS = 100%). In the PCC Catalogue of Strains, they have been assigned on the basis of phenotypic properties to Synechococcus Cluster 3, composed mostly of marine Synechococcus, and which generally corresponds to Marine-cluster C of Waterbury and Rippka (1989). Moreover, as indicated in Figure 2, these strains of Synechococcus are negative for nitrogenase activity.

Figure 1. Unrooted phylogenetic tree of the SPM sequence group of cyanobacteria. The tree was inferred from SSU rRNA sequences by maximum likelihood analysis with correction for site-to-site variation in evolutionary rates. RLS scores are given above, below, or alongside their corresponding internal branches. Species epithets are limited to type strains; complete names are given in Table 1. The scale for horizontal branch lengths corresponds to the number of fixed point mutations per sequence position. For further details, see text.


Turner et al. Molecular phylogeny of cyanobacteria

Cyanobacterium stanieri PCC 7202, another strain originally assigned to the genus Synechococcus (Cluster 4), also appears to be unrelated to RF-1 or to the Cluster 3 strains of Synechococcus. This contradicts an earlier proposal that strains RF-1 and ATCC 51142 may be related to strain PCC 7202 (cf. Synechococcus cedrorum SAG 88.79) based on ultrastructural similarities, specifically the distribution pattern of thylakoids within the cytoplasm (Komrek and Cepk, 1998).

In a recent microscopic and ultrastructural study of cyanobacteria assigned to the genus Cyanothece sensu Komrek, C. aeruginosa SAG 87.79 was selected as an exemplar for this species and, by extension, the genus (Komrek and Cepk, 1998). For this strain, the cellular distribution of both thylakoids and DNA, as well as other features, are distinctly different from the other strains assigned to Cyanothece discussed in the work presented here. At present, there are no DNA sequence data for strain SAG 87.79, so the extent of its molecular phylogenetic relationship with these other strains is unknown. However, given the generally good correlation between ultrastructural and phylogenetic differences, it is unlikely to be closely related to any of them. In conclusion, no strong correlation between aerobic versus anaerobic nitrogenase activity and phylogenetic relationships was observed. The results support a hypothesis of multiple gains and/or losses of nitrogen-fixation abilities among the unicellular cyanobacteria. Future investigations are necessary to further illuminate the similarities and differences among the unicellular, nitrogen-fixing cyanobacteria in order to better understand the relationship between their phenotypes and genotypes.

Acknowledgements. We are grateful to the administrator of the Bureau of Animal and Plant Health Inspection and Quarantine, Council of Agriculture, Executive Yuan for kindly issuing an import permit, No.88-P-1; and to Y. Cheng and B.H. Wang for their help with sequencing work. This study was supported by a grant to SMC and TCH from a five-year (1997-2002) Research Plan of the Institute of Botany, Academia Sinica, Taipei.

Literature Cited

Castenholz, R.W. and J.B. Waterbury. 1989. Group I. Cyanobacteria. Preface. In J.T. Staley, M.P. Bryant, N. Pfenning, and J.G. Holt (eds.), Bergey's Manual of Systematic Bacteriology, vol. 3. Williams & Wilkins, Baltimore, pp. 1710-1727.

Chaw, S.M., H.M. Sung, H. Long, A. Zharkikh, and W.H. Li. 1995. The phylogenetic positions of the conifer genera Amentotaxus, Phyllocladus, and Nageia inferred from 18S rRNA sequences. J. Mol. Evol. 41: 224-230.

Chou, H.M. and T.C. Huang. 1991. Ultrastructure of the aerobic, nitrogen-fixing unicellular cyanobacterium Synechococcus sp. RF-1.Arch. Hydrobiol. Suppl. 92, Algalogical Studies. 64: 53-59.

Coln-Lpez, M.S., D.M.Sherman, and L.A. Sherman. 1997. Transcriptional and translational regulation of nitrogenase in light-dark and continuous-light-grown cultures of the uni

cellular cyanobacterium Cyanothece sp. strain ATCC 51142. J. Bacteriol. 179: 4319-4327.

Garcia-Pichel, F., U. Nbel, and G. Muyzer. 1998. The phylogeny of unicellular, extremely halotolerant cyanobacteria. Arch. Microbiol. 169: 469-482.

Goodman, H.M. and R.J. MacDonald. 1979. Cloning of hormone genes from a mixture of cDNA molecules. Methods Enzymol. 68: 75-90.

Gutell, R.R. 1993. Collection of small subunit (16S- and 16S-like) ribosomal RNA structures. Nucl. Acids Res. 21: 181-189.

Honda, D., A. Yokota, and J. Sugiyama. 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.

Huang, T.C. and T.J. Chow. 1986. New type of nitrogen-fixing unicellular cyanobacterium (blue-green alga). FEMS Microbiol. Lett. 36: 109-110.

Huang, T.C., J. Tu, , T.J. Chow, and T.H. Chen. 1990. Circadian rhythm of the prokaryote Synechococcus sp. RF-1. Plant Physiol. 92: 531-533.

Jermiin, L.S., G.J. Olsen, K.L. Mengersen, and S. Easteal. 1997. Majority-rule consensus of phylogenetic trees obtained by maximum-likelihood analysis. Mol. Biol. Evol. 14: 1296-1302.

Kishino, H. and M. Hasegawa. 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.

Komrek, J. 1976. Taxonomic review of the genera Synechocystis SAUV. 1892, Synechococcus NG. 1849, and Cyanothece gen. nov. (Cyanophyceae). Arch. Protistenk. 118: 119-179.

Komrek, J. and V. Cepk. 1998.Cytomorphological characters supporting the taxonomic validity of Cyanothece (Cyanoprokaryota). Pl. Syst. Evol. 210: 25-39.

Len, C., S. Kumazawa, and A. Mitsui. 1986. Cyclic appearance of aerobic nitrogenase activity during synchronous growth of unicellular cyanobacteria. Curr. Microbiol. 13: 149-153.

Maddison, W.P. and D.R. Maddison. 1992. MacClade: analysis of phylogeny and character evolution. Version 3.0. Sinauer, Sunderland, Massachusetts.

Mitsui, A., S. Kumazawa, A. Takahashi, H. Ikemoto, S. Cao, and T. Arai. 1986. Strategy by which nitrogen-fixing unicellular cyanobacteria grow photoautotrophically. Nature 323: 720-722.

Ni, C.V., D.Z. Khong, Z.D. Tien, and I.N. Gogotov. 1988. Nitrogen fixation by the cyanobacteria, Aphanothece palida, isolated from a rice field soil. Microbiologia 57: 458-463.

Nbel, U., F. Garcia-Pichel, and G. Muyzers. 2000. The halotolerance and phylogeny of cyanobacteria with tightly coiled trichomes (Spirulina Turpin) and the description of Halospirulina tapeticola gen. nov., sp. nov. Int. J. Syst. Evol. Microbiol. 50: 1265-1277.

Olsen, G.J., H. Matsuka, R. Hagstrom, and R. Overbeek. 1994. fastDNAml: a tool for construction of phylogenetic trees of DNA sequences using maximum likelihood. Computer Applic. Biosci. 10: 41-48.

Page, R.D.M. 1996. TREEVIEW: An application to display phylogenetic trees on personal computers. Computer


Botanical Bulletin of Academia Sinica, Vol. 42, 2001

Applic. Biosci. 12: 357-358.

Raha, S., F. Merante, G. Proteau, and J.K. Reed. 1990. Simultaneous isolation of total cellular RNA and DNA from tissue culture cells using phenol and lithium chloride. Genet. Anal. Tech. Appl. 7: 173-177.

Reddy, K.J., J.B. Haskell, D.M. Sherman, and L.A. Sherman. 1993. Unicellular, aerobic nitrogen-fixing cyanobacteria of the genus Cyanothece. J. Bacteriol. 175: 1284-1292.

Rippka, R., and G. Cohen-Bazire. 1983. The Cyanobacteriales: a legitimate order based on the type strain Cyanobacterium stanieri ? Ann. Microbiol. (Inst. Pasteur) 134B: 21-36.

Rippka, R., and H. Herdman. 1992. Pasteur Culture Collection of Cyanobacteria Catalogue & Taxonomic Handbook. 1. Catalogue of Strains. Institut Pasteur, Paris.

Rippka, R., J. Deruelles, J.B. Waterbury, M. Herdman, and R.Y. Stanier. 1979. Generic assignments, strain histories, and properties of pure cultures of cyanobacteria. J. Gen. Microbiol. 111: 1-61.

Singh, P.K. 1973. Nitrogen fixation by the unicellular blue-green alga Aphanothece. Arch. Microbiol. 92: 59-62.

Stanier, R.Y., R. Kunisawa, M. Mandel, and G. Cohen-Bazire. 1971. Purification and properties of unicellular blue-green algae (Order Chroococcales). Bact. Rev. 35: 171-205.

Strimmer, K. and A. von Haeseler. 1996. Quartet puzzling: a quartet maximum-likelihood method for reconstructing tree

topologies. Mol. Biol. Evol. 13: 964-969.

Strimmer, K., N. Goldman, and A. von Haeseler. 1997. Bayesian probabilities and quartet puzzling. Mol. Biol. Evol. 14: 210-211.

Tamura, K. and M. Nei. 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.

Turner, S. 1997. Molecular systematics of oxygenic photosynthetic bacteria. Pl. Syst. Evol. [Suppl.] 11: 13-52.

Turner, S., K.M. Pryer, V.P.W. Miao, and J.D. Palmer. 1999. Investigating deep phylogenetic relationships among cyanobacteria and plastids by small subunit rRNA sequence analysis. J. Eukaryot. Microbiol. 46: 327-338.

Waterbury, J.B. and R. Rippka. 1989. Order Chroococcales Wettstein 1924, emend. Rippka et al. 1979. In J.T. Staley, M.P. Bryant, N. Pfenning, and J.G. Holt (eds.), Bergey's Manual of Systematic Bacteriology, vol. 3. Williams & Wilkins, Baltimore, pp. 1728-1746.

Wilmotte, A.P. 1994. Molecular evolution and taxonomy of the cyanobacteria. In D.A. Bryant (ed.), The Molecular Biology of Cyanobacteria. Kluwer, The Netherlands, pp. 1-25.

Wyatt, J.T. and J.K.G. Silvey. 1969. Nitrogen fixation by Gloeocapsa. Science 115: 908-909.

TӭMźĦl˽t

Sen Turner1 ȷ2 Q2

1 LĦwjǥͪt
2 s|Ӫs

ڭ̧Qγ̤jik˽ttökRźĦp RNA ǦCAHT{LTBT ᪺ӭMźĦtYCo{ Cyanothece ݡi̾ Waterbury and Rippka (1989) ɩwjź Ħ~tiܤ֤TؿWߺtƪtuAL̳kb@ӸjE YHeŲw X SPM ǦCsCݮﹽ񪺩TᬡʩM˽tYõLjPCHWGӭMźĦTO hoP]Ρ^C

G źĦFLTFTFݮFF˽tYFpF RNA C