rDNA analyses of planktonic heterocystous cyanobacteria, including members of the genera Anabaenopsis and Cyanospira

Isabelle Iteman1, Rosmarie Rippka1, Nicole Tandeau de Marsac1 and Michael Herdman1

Unité des Cyanobactéries (CNRS URA 2172), Département de Biochimie et Génétique Moléculaire, Institut Pasteur, 28 rue du Dr Roux, 75724 Paris Cedex 15, France1

Author for correspondence: Isabelle Iteman. Tel: +33 1 4568 8416. Fax: +33 1 4061 3042. e-mail: iiteman{at}pasteur.fr


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The taxonomic coherence and phylogenetic relationships of 11 planktonic heterocystous cyanobacterial isolates were examined by investigating two areas of the rRNA operon, the 16S rRNA gene (rrnS) and the internal transcribed spacer (ITS) located between the 16S rRNA and 23S rRNA genes. The rrnS sequences were determined for five strains, including representatives of Anabaena flos-aquae, Aphanizomenon flos-aquae, Nodularia sp. and two alkaliphilic planktonic members of the genera Anabaenopsis and Cyanospira, whose phylogenetic position was previously unknown. Comparison of the data with those previously published for individual groups of planktonic heterocystous cyanobacteria showed that, with the exception of members assigned to the genus Cylindrospermopsis, all the planktonic strains form a distinct subclade within the monophyletic clade of heterocystous cyanobacteria. Within this subclade five different phylogenetic clusters were distinguished. The phylogenetic groupings of Anabaena and Aphanizomenon strains within three of these clusters were not always consistent with their generic or specific assignments based on classical morphological definitions, and the high degree of sequence similarity between strains of Anabaenopsis and Cyanospira suggests that they may be assignable to a single genus. Ribotyping and additional studies performed on PCR amplicons of the 16S rDNA or the ITS for the 11 planktonic heterocystous strains demonstrated that they all contain multiple rrn operons and ITS regions of variable size. Finally, evidence is provided for intra-genomic sequence heterogeneity of the 16S rRNA genes within most of the individual isolates.

Keywords: 16S rRNA, internal transcribed spacer, phylogeny, sequence heterogeneity, restriction fragment length polymorphism

Abbreviations: IGS, intergenic spacer region; ITS, internal transcribed spacer of the rRNA operon; PCC, Pasteur Culture Collection of Cyanobacteria; PPFD, photosynthetic photon flux density; RB, relative binding

The GenBank accession numbers of the 16S rDNA gene sequences reported in this paper are AY038032AY038037.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In contrast to unicellular and filamentous non-heterocystous cyanobacterial strains, whose 16S rDNA sequences are dispersed in the phylogenetic tree and do not form clusters that are consistent with their classification, the heterocystous cyanobacteria form a monophyletic group (see Giovannoni et al., 1988 ; Turner, 1997 ; Wilmotte & Herdman, 2001 ). This group shows a remarkable degree of 16S rRNA gene sequence conservation, as first described by Giovannoni et al. (1988) , and includes members of both Sections IV and V defined by Rippka et al. (1979) , which correspond to the orders Nostocales and Stigonematales of the Botanical Code of Nomenclature. This separation at the ordinal level is thus not supported by phylogenetic analysis.

Classically assigned to the Nostocales, many planktonic heterocystous cyanobacteria are characterized by the synthesis of gas vesicles, bouyancy-providing structures that permit the filaments to move vertically in the water column (Walsby, 1981 ). Many also produce potent neuro- or hepatotoxins, or both (reviewed by Sivonen & Jones, 1999 ) and, as a consequence of their bouyancy, often accumulate in dense water-blooms that are dangerous for the health of fish, birds and mammals, including man. Detailed studies of phylogenetic relationships amongst these organisms have been mostly based on 16S rRNA gene sequences (Lyra et al., 1997 ; Barker et al., 1999 ; Beltran & Neilan, 2000 ; Lehtimäki et al., 2000 ; Li et al., 2000 ; Lyra et al., 2001 ; Moffitt et al., 2001 ; Saker & Neilan, 2001 ). In some cases other sequences have been examined, for example those of the DNA-dependent RNA polymerase gene (rpoC1; Fergusson & Saint, 2000 ; Wilson et al., 2000 ), the non-coding intergenic spacer of the phycocyanin operon (PC-IGS: Neilan et al., 1995 ; Barker et al., 1999 , 2000 ; Bolch et al., 1999 ) and the IGS between two adjacent copies of the gvpA gene encoding the major structural gas vesicle protein (Barker et al., 1999 ).

The above studies emphasized individual genera or monophyletic groups of heterocystous cyanobacteria such as Nodularia spp. (Lehtimäki et al., 2000 ; Moffitt et al., 2001 ), Anabaena circinalis (Beltran & Neilan, 2000 ), Anabaena/Aphanizomenon strains (Lyra et al., 2001 ) or Cylindrospermopsis (Saker & Neilan, 2001 ), and none defined the relationships between these groups. A comparison of these studies suggests the existence of five apparently monophyletic clusters of planktonic heterocystous cyanobacteria. One of these contains organisms assigned to two different genera, Anabaena and Aphanizomenon. A second cluster regroups only species of Anabaena. The third lineage is mainly composed of strains identified as Anabaena circinalis, but also contains a strain named as Anabaena affinis (Beltran & Neilan, 2000 ). However, for members of these three clusters, the strain descriptions were often insufficient to exclude the possibility that representatives of a single species may carry different specific epithets, or that strains of the same morphotype may have different names, as a consequence of the subtle definitions that discriminate between these botanical ‘species’. The two remaining clusters are respectively represented by planktonic and non-planktonic members of the genus Nodularia and isolates of the tropical planktonic species Cylindrospermopsis raciborskii. Phylogenetic studies of these latter organisms were accompanied by more detailed descriptions (Barker et al., 1999 ; Bolch et al., 1999 ; Lehtimäki et al., 2000 ; Saker & Neilan, 2001 ). The characteristic cell morphology or heterocyst differentiation patterns of members of these two clusters are more easily recognizable than those of strains of the first three clusters and thus facilitate identification, at least at the generic level. Within each of the five clusters, the 16S rDNA sequence similarity is so high as to suggest either that their members may be assignable to a single species, or that the resolving power of the analysis is insufficient to distinguish different species, or even different genera.

In this work, to futher study the taxonomic and phylogenetic relationships of the planktonic heterocystous strains, 11 well-characterized isolates were examined by investigating the 16S rRNA gene and the internal transcribed spacer (ITS) located between the 16S rRNA and 23S rRNA genes. The sequences of the 16S rRNA genes of five isolates were determined. Three strains (Aphanizomenon flos-aquae PCC 7905, Anabaena flos-aquae PCC 9302 and Nodularia sp. PCC 9350) represent three of the five clusters described above. The alkaliphilic planktonic members of the genera Anabaenopsis (PCC 9215) and Cyanospira (PCC 9501), which share a coiled trichome morphology and whose phylogenetic position was previously unknown, were examined for the first time. We also studied the relationships of these five strains to those of six additional isolates by RFLP analyses of their 16S rDNA amplicons. The number and size differences of the ITS region in all 11 strains were examined by PCR amplification. A preliminary account of some aspects of this study was presented by Iteman et al. (1999) .


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Organisms.
All cyanobacterial strains (Table 1) were axenic and were obtained from the Pasteur Culture Collection of Cyanobacteria (PCC), Unité des Cyanobactéries, Institut Pasteur (Rippka & Herdman, 1992 ; http://www.pasteur.fr/recherche/banques/PCC/). Nostoc PCC 7120 and the unicellular Synechocystis PCC 6803 were used as reference strains in certain experiments. Escherichia coli strain JM109, provided with the cloning kit (Promega), was employed for transformation with the plasmid containing the polymerase chain reaction (PCR) products (16S rDNA or ITS domain).


View this table:
[in this window]
[in a new window]
 
Table 1. Properties and origin of the strains examined in this study

 
Media and growth.
The following liquid culture media, developed from BG-110 (Rippka et al., 1979 ), were initially employed for maintenance. Medium a: BG-110+NaHCO3 (10 mM), for Anabaena flos-aquae PCC 9332, Anabaenopsis spp. PCC 9215, PCC 9216 and PCC 9608, Aphanizomenon flos-aquae PCC 7905 and Nostoc sp. PCC 7120. Medium b: medium a+NaNO3 (2 mM), for Anabaena flos-aquae PCC 9302 and PCC 9349. Medium c: BG-110+NaHCO3 (65 mM)+Na2CO3 (15 mM), for Anabaenopsis elenkinii PCC 9420 and Cyanospira spp. PCC 9501 and PCC 9502. Medium d: medium a+artificial seawater (Turks Island Salts, 20%, v/v, Merck Index 9954), for Nodularia sp. PCC 9350. In later experiments, the concentration of NaHCO3 in medium a was reduced to 5 mM (=medium e) for Anabaena flos-aquae PCC 9332, Aphanizomenon flos-aquae PCC 7905 and Nostoc PCC 7120, since cell morphology seemed more uniform as judged by microscopic inspection. Synechocystis PCC 6803 was cultivated in medium BG-11 (Rippka et al., 1979 ). Stock cultures were grown without gassing in 100 ml flasks containing 40 ml of the appropriate medium. Incubation was at 23 °C under white light (Osram Universal White) at a photosynthetic photon flux density of 10 µmol quanta m-2 s-1 (measured with a LICOR LI-185B quantum/radiometer/photometer equipped with a LI-190SB quantum sensor) provided over a light/dark cycle of 14 h/10 h. Sampling for PCR amplifications was performed on cultures with a filament density equivalent to an OD750 0·8–1·0 as judged by eye. E. coli JM109 was grown at 37 °C in Luria–Bertani medium with ampicillin (50 µg ml-1).

Salinity and temperature maxima for growth.
The growth responses to different salinities or temperatures of the 11 planktonic strains were examined in test tubes (18 mm diameter) containing 6 ml medium. For the salinity tests, medium a was used for all strains (except for strains PCC 9302 and PCC 9349, for which medium b was employed), with or without a supplement of Turks Island Salts to give the equivalent of 0·25, 0·5, 1 and 2 times the salinity of seawater. This allowed verification of whether the high NaHCO3/Na2CO3 concentration in the maintenance medium c for Anabaenopsis PCC 9420 and the two strains of Cyanospira was necessary in the presence of elevated salt concentrations. Incubation conditions were as described for the stock cultures. Testing for temperature tolerance was performed in medium a (PCC 9215, PCC 9216 and PCC 9608), medium b (PCC 9302 and PCC 9349), medium e (PCC 7905 and PCC 9332) and medium a supplemented with 0·2x Turks Island Salts, found to be optimal for the remaining strains. For lack of incubators providing both different temperatures and a light/dark regime, the cultures were incubated at 23, 30 and 37 °C, receiving an irradiance of 7 µmol quanta m-2 s-1 of continuous white light.

Photomicroscopy.
Filament suspensions were placed onto dry agarose-coated slides (2%, w/v, in sterile H2O) and photographed using Plan-Neofluar objectives (40x/0·75 Ph or 100x/130 Oil Ph3) and an Axioskop 2 microscope (Carl Zeiss Jena). Images were taken with a CCD camera (DEI-750; Optronics Engineering) fitted to the camera port of the microscope, transferred to a PC and further processed with Paint Shop Pro 4 (JASC) and Photoshop 5.0 (Adobe Europe).

DNA extraction and amplification of the 16S rRNA gene and ITS.
PCR amplification was performed on purified DNA or directly on cell lysates as described previously (Iteman et al., 2000 ). Amplification of the 16S rRNA gene was carried out by PCR using primers A2 and S17 (Table 2). The size of the amplicon was 1451 bp versus 1489 bp for the whole gene. The ITS region was amplified by using primers 322 and 340 (Table 2). The PCR products were migrated on 2·5% (w/v) agarose gels and visualized on a UV table after staining with ethidium bromide. To study the heterogeneity of the 16S rRNA genes within the genome of Anabaena flos-aquae PCC 9302, an amplification with primers 318 and S11 (Table 2) was made directly on cells following three washings in sterile H2O. The PCR products (569 bp) of four independent reactions were mixed and cloned into the pGEM-T vector (Promega) after purification with PCR purification kit (Qiagen), as described by Iteman et al. (2000) . For a random selection of the positive clones, an amplification reaction (four tubes per clone) was performed using oligonucleotides complementary to the flanking M13 primer sites of the vector, and the PCR products of each were pooled prior to sequencing.


View this table:
[in this window]
[in a new window]
 
Table 2. Primers used for PCR amplification of 16S rDNA and ITS, and for sequence determination

 
Sequencing of the 16S rRNA gene.
After purification by Wizard PCR kit (Promega), three or more independent PCR products of the 16S rDNA were mixed and sequenced directly on both strands with the T7 sequencing kit (Pharmacia) using the primers shown in Table 2, or cloned in the pGEM-T vector (Iteman et al., 2000 ); three clones of each were mixed and sequenced on both strands using the two M13 primer sites on the vector and the primers listed in Table 2. The shorter sequences of the cloned 16S rDNA of Anabaena flos-aquae PCC 9302 were sequenced from the M13 primer sites of the vector. The sequences determined have been deposited in GenBank with the following accession numbers: Anabaena flos-aquae PCC 9302, AY038032 and AY038037, representing the two 16S rRNA genes found in the majority and minority rrn operons, respectively; Anabaenopsis sp. PCC 9215, AY038033; Aphanizomenon flos-aquae PCC 7905, AY038035; Cyanospira rippkae PCC 9501, AY038036; Nodularia sp. PCC 9350, AY038034.

Ribotyping.
The genomic DNA of the 11 planktonic heterocystous strains, together with that of Synechocystis PCC 6803 as control, was prepared from large-volume cultures (400 ml). After centrifugation, the pellet was resuspended in 20 ml NET buffer (1 M NaCl, 100 mM EDTA and 6 mM Tris/HCl, pH 8·0) containing lysozyme (2 mg ml-1). After at least 3 h at 37 °C, during which lysis was verified several times by microscopic observation, the cells were centrifuged (10000 g for 15 min). The DNA released into the supernatant was recovered by precipitation with an equal volume of 2-propanol and centrifugation (10000 g for 15 min). The crude DNA pellet was resuspended in 10 ml 1xTE buffer (100 mM Tris/HCl and EDTA 1 mM, pH 7·4) and deproteinized with an equal volume of phenol/chloroform/isoamyl alcohol (25:24:1, by vol.), followed by extraction with chloroform/isoamyl alcohol (24:1, v/v). For strains PCC 7905, PCC 9332, PCC 9350 and PCC 9608, whose lysis was less efficient, the cell pellet fractions were also resuspended in 10 ml 1xTE buffer and deproteinized separately. Finally, the purified DNA of all fractions was precipitated by addition of 0·1 vol. sodium acetate solution (3 M; pH 5·2) and 1 vol. 2-propanol. After overnight incubation at 4 °C and centrifugation at 10000 g for 30 min, all DNA precipitates were dried under vacuum and resuspended in 300 µl to 1 ml 0·1xTE buffer. DNA from supernatant and pellet fractions was pooled. Five microlitres of each DNA were restricted using EcoRV or HindIII according to the manufacturer’s instructions and the restriction products were migrated on 0·8% (w/v) agarose gel (Litex FMC) in 1xTris/borate/EDTA buffer (Sambrook et al., 1989 ). After visualization of the bands by staining with ethidium bromide, the digested DNA was transferred by Southern blotting (Sambrook et al., 1989 ) from the agarose gel to a nylon membrane (Amersham Pharmacia Biotech) after alkaline denaturation (0·5 M NaOH and 1·5 M NaCl) and a step of neutralization (0·02 M NaOH and 1 M ammonium acetate). The membranes were washed with 2xSSC (300 mM NaCl and 30 mM sodium citrate, pH 7·0), air-dried and exposed under UV light to fix the DNA fragments. The non-radioactive probe labelling, the step of prehybridization and hybridization of the membrane and the detection reaction were performed using the ECL random prime labelling and detection kit according to the manufacturer’s instructions (Amersham Pharmacia Biotech). The probe was the PCR product corresponding to the 16S rDNA of Nostoc PCC 7120.

RFLP of PCR products corresponding to the 16S rRNA gene.
The PCR products (10 µl) corresponding to the 16S rRNA genes of the 11 planktonic strains and Nostoc PCC 7120 were digested with each of 12 restriction enzymes (AluI, BanII, BspMI, DdeI, HaeII, NheI, NruI, PstI, RsaI, SacI, SphI and XcmI) according to the manufacturer’s instructions. The digestions were migrated in 2·5% (w/v) agarose gel (Litex FMC) in 1x Tris/borate/EDTA buffer at 100 V. The gels were stained with ethidium bromide, visualized on a UV table and photographed. The PCR product of the 16S rRNA gene of Nostoc PCC 7120 was used as reference restriction profile. The presence or absence of each restriction band was recorded in a matrix, and the dataset was analysed by distance methods in the software package TREECON (Van de Peer & De Wachter, 1994 ).

Phylogenetic analysis.
The 16S rRNA sequences were aligned in the ARB editor (http://www.mikro.biologie.tu-muenchen.de/pub/ARB/) with a representative dataset of sequences of other heterocystous cyanobacteria. The sequences employed (with GenBank accession numbers) are shown in Table 3. Relationships between the strains were inferred by distance methods in ARB, in TREECON (Van de Peer & De Wachter, 1994 ) and by the maximum-likelihood method (fastDNAml v 1.1; Olsen et al., 1994 ). Highly variable regions of uncertain alignment and those that had gaps in more than 60% of the sequences were excluded from the analysis, resulting in a total of 1412 comparable positions. The phylogenetic tree (Fig. 2) was midpoint rooted using the non-heterocystous Leptolyngbya boryana PCC 73110 (accession number X84810) as outgroup. A second tree (Fig. 7) was inferred with the same methods but employing all positions of two short (569 bases) sequences of individual clones of the Anabaena flos-aquae PCC 9302 16S rRNA genes and the corresponding region from several close relatives.


View this table:
[in this window]
[in a new window]
 
Table 3. Published sequences employed for phylogenetic tree inference, toxicity (where known) and geographical origin of the strains

 

   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Strain properties
The characteristics of the cyanobacterial genera examined in this study have been described by Rippka et al. (2001) and representatives for which new 16S rDNA sequences were determined are illustrated in Fig. 1. Since the strain Anabaena flos-aquae PCC 9302 (Fig. 1a) is a mutant incapable of forming mature heterocysts, a more typical member of this species (PCC 9332) is also shown (Fig. 1b). The known properties of all 11 strains studied are summarized in Table 1. Organisms assigned to Anabaena flos-aquae are typically characterized by barrel-shaped or cylindrical cells, with relatively deep constrictions (Fig. 1b). However, variation in trichome morphology within this ‘species’ is evidenced by the absence of pronounced constrictions between the cells of strain PCC 9302 (Fig. 1a). Members of the genus Aphanizomenon (Fig. 1c) differ by classical definition from Anabaena flos-aquae primarily in possessing cylindrical cells, significantly longer than wide, constrictions between the cells being shallow. Trichome ends are slightly attenuated, unlike those of Anabaena flos-aquae. The distinction between Aphanizomenon flos-aquae, Anabaena flos-aquae and certain other Anabaena spp. being often difficult (compare Fig. 1a with Fig. 1c), these organisms were included in this study in an attempt to identify potential relatives for which sequence data are available. Members of the genus Nodularia are easily recognized by the discoid cells of their trichomes (Fig. 1d). Members of the genus Anabaenopsis (Jeeji-Bai et al., 1977 ; Rippka et al., 2001 ) possess 8–16 cells in short loosely coiled trichomes with terminal heterocysts (Fig. 1e). De novo heterocyst differentiation always occurs in pairs after asymmetric cell division in the centre of the trichome (Fig. 1e). Strains of Cyanospira (Fig. 1f; Table 1) differ from Anabaenopsis in possessing longer trichomes that consist of several coils in which heterocysts do not occur in pairs (Florenzano et al., 1985 ); however, filament breakage at positions adjacent to the heterocysts often produces short trichomes (Fig. 1f) that superficially resemble those of Anabaenopsis.



View larger version (169K):
[in this window]
[in a new window]
 
Fig. 1. Photomicrographs illustrating the morphological appearance of (a) Anabaena flos-aquae PCC 9302, (b) Anabaena flos-aquae PCC 9332, (c) Aphanizomenon flos-aquae PCC 7905, (d) Nodularia sp. PCC 9350, (e) Anabaenopsis sp. PCC 9215, (f) Cyanospira rippkae PCC 9501. The small cells in the centre of the Anabaenopsis trichome (e) represent an early stage of heterocyst differentiation. All phase-contrast. Bars, 10 µm (a–c), 5 µm (d–f).

 
Phylogenetic relationships between the planktonic heterocystous cyanobacteria
The sequence of the 16S rRNA gene was determined for five gas-vacuolate planktonic cyanobacterial strains (Anabaena flos-aquae PCC 9302, Aphanizomenon flos-aquae PCC 7905, Nodularia sp. PCC 9350, Anabaenopsis sp. PCC 9215 and Cyanospira rippkae PCC 9501). The sequences were compared with those of representative heterocystous cyanobacteria available in GenBank. Phylogenetic trees were inferred by both maximum-likelihood and distance methods, which gave similar results. In contrast to those published previously, the tree presented (Fig. 2) is representative of all sequences presently available for planktonic heterocystous cyanobacterial genera or species. With the exception of Cylindrospermopsis raciborskii, whose position varied in trees inferred with the different methods but was never found to group with the other planktonic organisms, the strains can be assigned to five clusters (Fig. 2) that each internally show a high degree of sequence similarity. The clusters are well supported by bootstrap analysis and partly reflect the morphological similarity of the organisms (Fig. 1). Cluster I contains several strains of Aphanizomenon, including PCC 7905, together with two strains previously identified as Anabaena flos-aquae and Anabaena sp. (Carmichael & Gorham, 1978 ; Lyra et al., 1997 ). The sequences show a minimal similarity of 99·7% (between Aphanizomenon flos-aquae PCC 7905 and Anabaena flos-aquae NRC44-1). Cluster II is composed of four strains, selected from 21 identified as Anabaena circinalis by Beltran & Neilan (2000) , that exhibit a minimal similarity of 98·8%, and a more distantly related strain identified as Anabaena flos-aquae AWQC112D (Beltran & Neilan, 2000 ) that shows a mean similarity of 98·3% to the other members. Cluster III, containing Anabaena flos-aquae PCC 9302, also comprises several other strains identified (Lyra et al., 1997 ) as planktonic species of Anabaena that exhibit a minimum similarity of 98·9%. Although strain PCC 9302 was rendered axenic from a culture of Anabaena flos-aquae NRC525-17 obtained directly from W. W. Carmichael, its 16S rDNA sequence shows only 97·2% similarity with that determined for the non-axenic strain NRC525-17 by Beltran & Neilan (2000) . Since the latter sequence clearly falls outside cluster III (Fig. 2), it is unlikely that strain PCC 9302 is coidentic with the isolate studied by Beltran & Neilan (2000) . In addition to high internal similarity, the members of clusters I to III, which contain strains named as Aphanizomenon flos-aquae or Aphanizomenon sp., Anabaena circinalis, Anabaena flos-aquae, Anabaena lemmermannii or Anabaena sp., are also closely related to each other. For example, the sequence of Aphanizomenon flos-aquae PCC 7905 (cluster I) shares 97·2% similarity with that of Anabaena flos-aquae PCC 9302 (cluster III) and 96·7% similarity with Anabaena circinalis strain 306A (cluster II), and the latter sequence has 96·6% similarity with that of strain PCC 9302.



View larger version (64K):
[in this window]
[in a new window]
 
Fig. 2. Phylogenetic tree based on 16S rDNA sequences showing the relationships between planktonic heterocystous cyanobacteria and representative non-planktonic members. Sequences determined in the present study are indicated in bold type. The scale marker represents 0·01 nucleotide substitutions per sequence position. Roman numerals I to VI indicate the clusters of planktonic strains referred to in the text. The tree, based on 1412 sequence positions (excluding highly variable regions and positions that had gaps in more than 60% of the sequences) was inferred using the Jukes & Cantor algorithm and the neighbour-joining option in the software package ARB and subjected to 1000 bootstrap cycles, the results being presented as percentage values. Only bootstrap values greater than 50% are shown. With the exception of the position of Cylindrospermopsis raciborskii AWT205, which appeared to be unstable, identical trees were obtained using the Jin & Nei algorithm with the default shape parameter ({alpha}=1·0) and the neighbour-joining option in the software package TREECON (Van de Peer & De Wachter, 1994 ) and with the default transition/transversion ratio (2·00) giving a transition/transversion parameter of 1·45585 in fastDNAml (Olsen et al., 1994 ).

 
Five Nodularia strains, including both gas-vesicle-forming (PCC 9350, BCNOD9427 and NSPI-05) and non-planktonic members (PCC 73104 and PCC 7804), form a closely related group (cluster IV) with a minimum similarity of 98·2%. Cluster V, composed of Anabaenopsis sp. PCC 9215 and Cyanospira rippkae PCC 9501, which share 98·1% sequence similarity, is new, since data were not previously available for these genera. The similarities between members of clusters IV and V are again relatively high, the 16S rDNA sequence of Cyanospira rippkae PCC 9501 showing 96·6% similarity to that of Nodularia PCC 9350. However, the subclade containing Nodularia, Anabaenopsis and Cyanospira is well separated from those containing Aphanizomenon flos-aquae PCC 7905 and Anabaena flos-aquae PCC 9302, since Nodularia PCC 9350 (cluster IV) shows only 94·4% and 94·2% sequence similarity to members of clusters I and III, respectively.

Cylindrospermopsis raciborskii (Seenayya & Subba-Raju, 1972 ) is morphologically distinct from all other planktonic cyanobacteria, being distinguished by the thin trichomes (less than 4 µm diameter) that may be loosely coiled and contain cylindrical cells. Heterocysts are conical to spear-shaped and occur exclusively at the ends of the trichome. Although represented by only a single strain in Fig. 2, organisms assigned to this species are closely related (Saker & Neilan, 2001 ), showing at least 99·8% sequence similarity, and are united into cluster VI. The present analysis reveals that this cluster is genetically distant from all other planktonic heterocystous cyanobacteria.

The closest relatives to members of clusters I to V are two strains of Anabaena cylindrica (PCC 7122 and NIES 19) that do not form gas vesicles, are motile and correspond to a species that has never been associated with bloom formation. The former strain shows 95·6% sequence similarity to, for example, Nodularia sp. PCC 7804 in cluster IV. However, it is evident that strains presently assigned to the genus Anabaena are very diverse, since strains carrying the latter generic epithet are dispersed throughout the tree, with Anabaena PCC 7108 being only distantly related both to the planktonic Anabaena spp. and to the A. cylindrica strains (Fig. 2). At the root of the heterocystous cyanobacterial tree lie three Calothrix strains, characterized by filaments containing heterocysts principally in a basal position, with the trichome tapering from base to apex (see Rippka et al., 1979 , 2001 ), and the heavily ensheathed Scytonema hofmanni PCC 7110, which differentiates heterocysts generally in intercalary positions. The available sequences of members of the genus Nostoc fall into two different branches of the tree. Three organisms cluster closely with Nostoc punctiforme PCC 73102, which has been proposed as type strain of this species (Rippka & Herdman, 1992 ; Rippka et al., 2001 ). However, the atypical Nostoc sp. strains PCC 7118 and PCC 7120 (see Rippka et al., 2001 ), together with a strain named as Anabaena variabilis (NIES 23), are clearly separated from the N. punctiforme group, being more closely related (96·9%) to Cylindrospermum stagnale PCC 7417. Finally, three strains assigned to the order Stigonematales (Chlorogloeopsis fritschii PCC 6718, Fischerella muscicola PCC 7414 and ‘Mastigocladus HTF’ PCC 7518) cluster within the heterocystous members of the order Nostocales, as previously observed (Giovannoni et al., 1988 ; Turner, 1997 ; Wilmotte & Herdman, 2001 ).

Multiple rRNA operons
A typical ribotype pattern, obtained following digestion of total DNA with HindIII, is shown in Fig. 3. This enzyme has no recognition site in any of the 16S rDNA sequences available for planktonic heterocystous strains, and resulted in four bands (Anabaenopsis strains PCC 9215, PCC 9216, PCC 9420 and PCC 9608; Cyanospira strains PCC 9501 and PCC 9502; Nodularia PCC 9350) or five bands (Anabaena flos-aquae strains PCC 9302, PCC 9332 and PCC 9349; Aphanizomenon flos-aquae PCC 7905) after Southern hybridization with the 16S rDNA probe. In contrast, restriction of genomic DNA of the unicellular strain Synechocystis PCC 6803 (whose 16S rDNA also lacks HindIII sites) indicated that this isolate contained only two copies of the operon (Fig. 3), consistent with the published genome sequence (Kaneko et al., 1996 ). Since the bands of higher molecular mass in the heterocystous strains are of size sufficient to contain two or more copies of the rrn operon, the total DNA was also restricted with EcoRV, which has a single recognition site within the 16S rDNA of each organism for which this sequence is available. With this enzyme, 8 or 10 hybridizing bands were revealed (data not shown), as expected if only a single copy of the operon was present in each HindIII band. Seven strains were thus shown to possess four copies of the rrn operon, and four strains contained five copies. Within genera for which several strains were examined, each showed a distinct ribotype pattern.



View larger version (88K):
[in this window]
[in a new window]
 
Fig. 3. Ribotyping of 11 planktonic heterocystous cyanobacteria, with the unicellular strain Synechocystis PCC 6803 included as a control, following digestion of genomic DNA with HindIII and hybridization with a Nostoc PCC 7120 16S rDNA probe. The fragment sizes indicated on the left correspond to a HindIII digest of phage {lambda} DNA.

 
The number and size of the internal transcribed spacer (ITS)
The set of primers (322–340) used to amplify the ITS gave three bands of different intensities for members of the genera Anabaenopsis, Aphanizomenon, Cyanospira and Nodularia, as previously described for Nostoc PCC 7120 (Iteman et al., 2000 ). The amplicons ranged in size from 470 to 880 bp (Fig. 4). Since the first primer recognition site is located around 150 bp before the 3' end of the 16S rRNA gene and the second is near the 5' end of the 23S rRNA gene, the amplified fragments correspond to the length of the ITS plus a total of 200 bp from the adjacent genes and therefore represent ITS regions of length 270 to 680 bp. As in Nostoc PCC 7120 (Iteman et al., 2000 ), the intermediate bands observed for each organism are most likely heteroduplexes, formed during the PCR amplification by reannealing of single strands of the different ITS species (Jensen et al., 1993 ; Iteman et al., 2000 ). Aphanizomenon PCC 7905 and Nostoc PCC 7120 shared an almost identical ITS banding pattern and thus cannot be easily distinguished by this analysis. Amongst the four strains of Anabaenopsis, three (PCC 9215, PCC 9216 and PCC 9608) had identical patterns, whereas the African isolate Anabaenopsis elenkinii PCC 9420 was markedly different. The two strains of Cyanospira (PCC 9501 and PCC 9502), despite their different specific designations (Florenzano et al., 1985 ), had identical profiles (Fig. 4) and were therefore indistinguishable in the ITS region. In contrast, the three strains of Anabaena flos-aquae each gave rise to four bands, again with possible heteroduplexes, and all showed individual ITS patterns (Fig. 4). Unlike the heterocystous strains, the unicellular Synechocystis PCC 6803 gave only a single ITS amplicon (Fig. 4), as discussed by Iteman et al. (2000) .



View larger version (76K):
[in this window]
[in a new window]
 
Fig. 4. ITS amplicons of strains of Aphanizomenon flos-aquae (Aph), Anabaenopsis (Abp), Cyanospira (Cys), Anabaena flos-aquae (Ana) and Nodularia (Nod), which are longer by 200 bp than the ITS region itself. L, 100 bp ladder. Synechocystis PCC 6803 (unicellular) and Nostoc PCC 7120 (heterocystous) are both included as controls. Reproduced with permission from Iteman et al. (1999) .

 
RFLP of the 16S rRNA gene
The 11 strains (Table 1) were examined by fingerprinting (RFLP) of the 1450 bp PCR product corresponding to the 16S rRNA gene, Nostoc PCC 7120 being used as a control of restriction profiles. Depending on the strain and the restriction enzyme employed, the patterns were composed of 1 to 7 bands between 95 and 1450 bp (Table 4). The dendrogram derived from the analysis with 12 restriction enzymes (Fig. 5) is consistent with the phylogenetic tree obtained with the 16S rRNA gene sequences (Fig. 2). The four strains of Anabaenopsis and the two strains of Cyanospira are resolved as two relatively closely related genetic entities, and cluster with Nodularia PCC 9350. Anabaena flos-aquae and Aphanizomenon strains are separated on different branches. Restriction sites that would generate the RFLP patterns observed for Aphanizomenon PCC 7905 with the 12 enzymes employed were also found by in silico analysis of the 16S rRNA gene sequences of Aphanizomenon strains TR183, BC9601 and 202, and Anabaena strain 86, in agreement with their grouping in the 16S rDNA tree (Fig. 2). Strains within the genus Cyanospira and those of Anabaena flos-aquae were not further resolved in this analysis, while the four strains of Anabaenopsis were clearly distinguished from each other, the African strain PCC 9420 being more divergent from the European isolates (PCC 9215 and PCC 9216 from Spain, PCC 9608 from Sweden).


View this table:
[in this window]
[in a new window]
 
Table 4. Restriction profiles of 16S rDNA amplicons of 11 planktonic heterocystous cyanobacteria, with Nostoc PCC 7120 as a control

 


View larger version (18K):
[in this window]
[in a new window]
 
Fig. 5. Dendrogram constructed from RFLP analysis of 16S rDNA amplicons of 12 strains of heterocystous cyanobacteria following restriction with 12 enzymes (Table 4), Nostoc PCC 7120 being employed as outgroup. All restriction bands were taken into consideration, using the Nei & Li algorithm and the neighbour-joining option in the software package TREECON (Van de Peer & De Wachter, 1994 ), with 1000 bootstrap cycles, the results being presented as percentage values. Only bootstrap values greater than 50% are shown.

 
Sequence heterogeneity within 16S rRNA genes of individual strains
In many cases, the sum of the band sizes obtained by restriction enzyme analysis considerably exceeded the length of the amplicon, 1450 bp (Table 4). This phenomenon could result from sequence variations in the multiple 16S rRNA genes within individual strains. To investigate this hypothesis, the RsaI restriction pattern deduced by in silico analysis from the 16S rRNA gene sequence of strain Anabaena flos-aquae PCC 9302, containing five fragments of 420, 406, 340, 160 and 122 bp sequentially from the 5' end of the amplicon, was compared to those of two other strains (Anabaena flos-aquae NRC525-17 and Anabaena sp. strain 90). An additional site was found within the 406 bp fragment of the latter two strains that would yield bands of 369 and 37 bp, the larger one having a size remarkably close to the 370 bp RsaI band of the strain PCC 9302.

The region covering this hypothetical RsaI site and corresponding to positions 462–1030 of Synechocystis PCC 6803 (E. coli positions 516–1083) was sequenced in 14 clones of the 16S rDNA of strain PCC 9302. Six clones contained the RsaI site at the anticipated position, the remainder did not. The two sequences of PCC 9302 differ in a total of 6 out of 569 positions. The less frequent sequence (represented by clone 2) contains the base A at position 341 that creates the GTAC site for RsaI. This change is compensated by base T at position 316 (Fig. 6) that maintains the base pairing. Since strain PCC 9302 was shown by ribotyping to possess five rrn operons, the ratio (8:6) of the different sequences among the clones suggests that three operons contain the majority 16S rDNA sequence (clone 15, Fig. 6) and that two contain the less frequent one (clone 2). The sequences were compared with that of Anabaena flos-aquae NRC525-17, which differs in 21 positions from both sequences of strain PCC 9302. The difference (1·3%) between the two 16S rDNA sequences of Anabaena flos-aquae PCC 9302 is the same as the maximal difference observed within cluster III (Fig. 2) and is sufficient to make them appear as two different taxonomic units, more distant from each other than is the major sequence from that of Anabaena strain 90 (Fig. 7).



View larger version (41K):
[in this window]
[in a new window]
 
Fig. 6. Sequence alignment of a short region of 16S rDNA from Anabaena flos-aquae PCC 9302, representing the majority (clone PCC 9302-15) and minority (clone PCC 9302-2) sequences recovered from this organism. Only the differences from the first sequence are shown. The position of the additional RsaI cutting site in clone PCC 9302-2 is shaded. The third sequence (Anfl525) is that published (Beltran & Neilan, 2000 ) for Anabaena flos-aquae strain NRC525-17, which is clearly different from both sequences of the supposedly coidentic and axenic subisolate PCC 9302.

 


View larger version (20K):
[in this window]
[in a new window]
 
Fig. 7. Phylogenetic tree showing the relationships between the two 16S rRNA genes (PCC 9302 clone 15 and PCC 9302 clone 2, indicated in bold) of Anabaena flos-aquae PCC 9302 and the equivalent region from close relatives (see Fig. 2). The tree, based on 569 sequence positions, was inferred using the Jukes & Cantor algorithm and the neighbour-joining option in the software package ARB, and was subjected to 1000 bootstrap replicates. Identical tree topologies, well supported by bootstrap analysis, were obtained using the Jin & Nei algorithm, and fastDNAml, where each branch showed a significantly positive length at P<0·01.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
It is generally accepted for bacteria that a 16S rDNA sequence similarity of 96–97% and DNA–DNA hybridization values of 70% relative binding (RB) with a {Delta}Tm of 5 °C represent the lower boundary of a species (Wayne et al., 1987 ), and that a genus may be defined by species with 95% or greater sequence similarity (Ludwig et al., 1998 ). However, Rosselló-Mora & Amann (2001) suggested that values as low as 50% RB and a {Delta}Tm of 7 °C may be employed for the delineation of species, and also showed that sequence similarity is often greater than 98–99% when total DNA similarity is as low as 10–40% RB. Similar conclusions were presented by Keswani & Whitman (2001) , who also showed that the relationship between RB and sequence similarity varies even within the same subphylum. It therefore seems that 16S rDNA sequence similarity does not always adequately reflect total genomic relationships. A comparison of cyanobacterial sequence similarity (Fig. 2) with DNA hybridization values determined by Lachance (1981) is consistent with this proposition. Nodularia PCC 7804 and PCC 73104 show 98·3% 16S rDNA similarity but only 65% RB ({Delta}Tm of 6 °C), whereas Calothrix PCC 7102 and PCC 7709 share lower (96·4%) rDNA similarity but higher DNA homology (74% RB, {Delta}Tm of 4 °C). Anabaena PCC 7122 shows 95·1% and 95·6% rDNA similarity with Cylindrospermum PCC 7417 and Nodularia PCC 7804, whereas the total DNA homology is only 23% and 19%, respectively.

It is therefore difficult at present to define the generic or specific status of the five clusters of planktonic heterocystous cyanobacteria on the basis of 16S rDNA sequence homology, and it appears that the classical morphological definitions of the corresponding taxa may in some cases be inadequate. Turner (1997) suggested that, although cyanobacterial 16S rRNA sequence comparisons are not suitable for resolution below the level of genus, generic assignments appear to correlate with sequence data. However, Lyra et al. (2001) showed a close phylogenetic relationship between non-toxic Aphanizomenon and neurotoxic Anabaena isolates originating from Finland, Norway, the Baltic Sea, Holland and Japan. The present results additionally show that strains named as Aphanizomenon flos-aquae, Anabaena flos-aquae and Anabaena circinalis, although falling into three clusters, are not entirely resolved according to their specific or generic assignments. The subclade containing the Nodularia, Anabaenopsis and Cyanospira clusters is clearly separated from the three clusters containing Anabaena and Aphanizomenon strains, although the low degree of sequence divergence (particularly for Anabaenopsis and Cyanospira) may again not necessarily be in agreement with their assignment to three different genera.

Cylindrospermopsis raciborskii is the only planktonic, potentially toxic, heterocystous cyanobacterium that does not group within the major clade containing the five clusters described above. Members of this species were originally assigned to the genus Anabaenopsis (see Rippka et al., 2001 ), but the new genus Cylindrospermopsis was created by Seenayya & Subba-Raju (1972) on the basis of the fundamentally different pattern of heterocyst differentiation in members of the former genus. As indicated by the generic epithet, members of Cylindrospermopsis have also been thought to be close to those of Cylindrospermum, since the differentiation of exclusively terminal heterocysts is common to both (Horecká & Komárek, 1979 ). The low degree of sequence similarity between Cylindrospermopsis raciborskii AWT205 and Anabaenopsis PCC 9215 (93·3%) is consistent with their generic separation. In contrast, Cylindrospermum PCC 7417 clusters with various strains of Nostoc and Anabaena (Fig. 2) and has only 92·8% sequence similarity with Cylindrospermopsis raciborskii AWT205, disputing the postulated close relationship between Cylindrospermum and Cylindrospermopsis.

The toxicity (if known) of the strains compared in Fig. 2 is summarized in Table 3. Cluster I contains strains that produce anatoxin-a (Anabaena sp. strain 86 and Anabaena flos-aquae NRC44-1) and non-toxic strains (Aphanizomenon sp. strains 202 and TR183). The toxicity of Aphanizomenon sp. strains PCC 7905 and BC9601 has not yet been examined. Strain PCC 7905 originated from a lake in the Netherlands, and Anabaena flos-aquae NRC44-1 was obtained from a lake in Canada, whereas all other members of this cluster were isolated from Scandinavia (Anabaena sp. strain 86 and Aphanizomenon sp. strain 202 from freshwater lakes in Finland; Aphanizomenon sp. strains BC9601 and TR183 from the Baltic Sea). The very close phylogenetic relationship between these six strains demonstrates that neither toxicity nor geographical origin are useful taxonomic markers. The same conclusions can be made for members of clusters II, III and IV. Cluster II regroups four isolates of Anabaena circinalis, three originating from Australia and one from Japan (strain NIES 41), but also includes one strain of Anabaena flos-aquae (AWQC112D) from Australia. Two of these produce saxitoxins (A. circinalis AWQ279B and AWT001), two are saxitoxin-negative (A. circinalis AWQC306A and A. flos-aquae AWQC112D) and the toxicity of strain NIES 41 is unknown. Cluster III contains isolates that produce the hepatotoxin microcystin (Anabaena sp. strains 66A and 90, and A. lemmermannii NIVA-CYA 83/1) and originating from Scandinavia, but also includes Anabaena flos-aquae PCC 9302, which, based on its strain history, was isolated from a lake in Canada and produces the neurotoxin anatoxin-a(s) (Carmichael & Gorham, 1978 ). In cluster IV, three strains of Nodularia (NSPI-05, PCC 9350 and PCC 7804) produce the hepatotoxin nodularin, strain PCC 73104 is non-toxic and the toxicity of strain BCNOD9427 is unknown. Again, this cluster regroups isolates from Australia (NSPI-05), Canada (PCC 73104), France (PCC 7804) and the Baltic Sea (BCNOD9427 and PCC 9350). As shown in Fig. 2, the Australian strain Nodularia sp. NSPI-05 has a very high degree of sequence similarity with one of the Baltic Sea isolates (BCNOD9427). These analyses clearly demonstrate that it is impossible on the basis of 16S rDNA sequences to distinguish toxic or non-toxic cyanobacteria or organisms from different geographical regions, and they thus contradict previous conclusions concerning the value of these markers (Beltran & Neilan, 2000 ; Lehtimäki et al., 2000 ; Moffitt et al., 2001 ).

Members of both clusters IV and V, however, share certain features that may be of significance. Most strains of Nodularia were found in environments of high salinity or high pH (Bolch et al., 1999 ; Rippka et al., 2001 ), as were the two strains of Cyanospira and three of the four Anabaenopsis isolates (Rippka et al., 2001 ) examined in the present study. Strains of these three genera cluster together in the phylogenetic tree (Figs 2 and 5), and the PCC strains studied, with the exception of Anabaenopsis PCC 9608, all show a relatively high salt tolerance (Table 1), distinguishing them from their freshwater relatives and suggesting a relationship between habitat and phylogenetic position. However, this again does not appear to be a general rule, since Aphanizomenon strains BC9601 and TR183, isolated from the Baltic Sea (Barker et al., 2000 ; Lyra et al., 2001 ) are not separated phylogenetically from freshwater isolates of Aphanizomenon (PCC 7905) or Anabaena (strains NRC44-1 and 86) (Fig. 2). A second feature of potential importance is the ability of all PCC strains of Anabaenopsis, Cyanospira and Nodularia to support more elevated temperatures than the Anabaena and Aphanizomenon strains examined (Table 1).

Using a larger set of strains than that studied by 16S rDNA sequence analysis, we have shown by RFLP that cyanobacteria assigned in the PCC to Aphanizomenon flos-aquae, Anabaena flos-aquae, Anabaenopsis, Cyanospira and Nodularia appear to each form a distinct cluster (Fig. 5), consistent with their positions in trees inferred from the 16S rDNA sequences. However, in view of morphological similarities, their close phylogenetic relatedness revealed by sequence analysis and their clustering in the 16S rDNA RFLP analysis, it is not possible at present to justify the generic separation of strains assigned to Anabaena flos-aquae and Aphanizomenon flos-aquae. Lyra et al. (1997 , 2001 ) and Lehtimäki et al. (2000) also found close relationships between strains of Anabaena and Aphanizomenon in a 16S rDNA RFLP study. However, Nodularia PCC 73104 was found to group with strains of the genus Nostoc, including the atypical Nostoc sp. PCC 7120, although sequence analyses (Lehtimäki et al., 2000 ; Lyra et al., 2001 ) demonstrated that Nodularia strains clustered more closely with members of the genera Anabaena and Aphanizomenon, which is in better agreement with the RFLP studies presented here. It therefore seems essential to employ a suitable set of restriction enzymes, which can be chosen by examination of the large number of 16S rDNA sequences now available, in order to permit a higher level of discrimination.

All planktonic heterocystous cyanobacteria examined contain several ITS regions of different length (Fig. 4), confirming the genetic heterogeneity of their rrn operons. In contrast, the two rrn operons of Synechocystis PCC 6803 contain ITS regions of identical size, consistent with the genome sequence (Kaneko et al., 1996 ). Variability in size of the ITS has been examined in detail in Nostoc PCC 7120, where the longer and shorter ITS regions were shown to be similar in sequence except for the presence and absence, respectively, of two tRNA genes (Iteman et al., 2000 ). Such differences in length of the multiple ITS regions within individual heterocystous cyanobacterial strains appear to be very common (Lu et al., 1997 ; Neilan et al., 1997 ; West & Adams, 1997 ; Barker et al., 1999 ; Boyer et al., 2001 ), and have also been observed in other bacteria such as Corynebacterium and Streptomyces (Aubel et al., 1997 ; Hain et al., 1997 ). The ITS patterns of the planktonic strains examined (Fig. 4) were useful in permitting discrimination between strains of Anabaena flos-aquae that were identical in RFLP analysis of the 16S rDNA amplicons, and separated Anabaenopsis PCC 9420 from its three relatives (strains PCC 9215, PCC 9216 and PCC 9608). Although the ITS patterns of the two strains of Cyanospira were identical, they were clearly different from those of the four Anabaenopsis strains, their closest relatives revealed by RFLP analysis of the 16S rDNA. The size of the different ITS regions is thus a useful additional taxonomic character but, as in the case of Aphanizomenon PCC 7905 and Nostoc PCC 7120, which shared an almost identical ITS banding pattern, it may in some cases not permit the distinction between unrelated organisms.

We have also demonstrated by ribotyping that all the heterocystous strains studied contain up to five rrn operons (Fig. 3), consistent with the multiple operons observed in Anabaena strain CA and Anabaena cylindrica CCAP 1403/2a (PCC 7122) by Nichols et al. (1982) and in Nostoc PCC 7120 (Ligon et al., 1991 ). The two members of the genus Cyanospira, which were indistinguishable by RFLP analysis of the 16S rDNA and whose ITS patterns were identical, were clearly separated by ribotyping, as were the Anabaena flos-aquae strains, whose ITS patterns differed but were identical in RFLP analyses. The four members of Anabaenopsis, only partially separated by ITS and RFLP studies, were again distinguished by their ribotypes. The ribotype patterns of individual strains of the same genus, revealing sequence differences in the regions outside of the 16S rDNA, are therefore highly discriminative markers.

We have demonstrated sequence heterogeneity within the 16S rRNA genes of the multiple ribosomal operons of Anabaena flos-aquae PCC 9302. This was first suggested by RFLP analysis, where restriction with RsaI produced bands whose total size exceeded the length of the amplicon, and was confirmed by sequencing a short region of 14 individual 16S rDNA clones. Two distinct sequences were determined, the differences between them corresponding to five transitions (A{leftrightarrow}G or C{leftrightarrow}T) and one transversion (A{leftrightarrow}T) (Fig. 6). This high frequency of transitional substitution suggests that the origin of the intragenomic heterogeneity in the 16S rRNA genes is due to mutation rather than to horizontal gene transfer (Gojobori et al., 1982 ; Ueda et al., 1999 ). Sequence heterogeneity within the 16S rRNA genes of individual planktonic heterocystous cyanobacteria appears to be widespread, since the RFLP data (Table 4) indicate its probable occurrence in 10 of the 11 strains examined. Interestingly, the RFLP patterns (Lyra et al., 1997 ) of Anabaena sp. strain 90, the nearest relative of strain PCC 9302 (Fig. 2), did not reveal such heterogeneity following digestion with RsaI, whereas digestion of Anabaena sp. strain 66A with MspI produced bands (approximate total of 1750 bp) in excess of the expected 1500 bp (see Fig. 2 in Lyra et al., 1997 ), although the authors did not comment on this discrepancy. Such heterogeneity has been observed only rarely in other prokaryotes, such as Haloarcula marismortui (Mylvaganam & Dennis, 1992 ; Dennis et al., 1998 ), E. coli (Cilia et al., 1996 ; Martinez-Murcia et al., 1999 ), Paenibacillus polymyxa (Nübel et al., 1996 ), Thermobispora bispora (Wang et al., 1997 ) and Phormium yellow leaf phytoplasma (Liefting et al., 1996 ) but in fact is probably more common, as reported by Clayton et al. (1995) . As previously shown for other organisms (Clayton et al., 1995 ; Yap et al., 1999 ), intra-genomic 16S rDNA sequence heterogeneity has implications for the inference of phylogenetic relationships of planktonic heterocystous cyanobacteria (Fig. 7).


   ACKNOWLEDGEMENTS
 
This work was supported by contract BIO4-CT96-0256 (BASIC) of the European programme BIOTECH (Life Sciences and Technologies, Biotechnology Programme, 1994–1998), the Institut Pasteur and the Centre National de la Recherche Scientifique (CNRS URA 2172).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Aubel, D., Renaud, F. N. & Freney, J. (1997). Genomic diversity of several Corynebacterium species identified by amplification of the 16S–23S gene spacer regions. Int J Syst Bacteriol 47, 767-772.

Barker, G. L. A., Hayes, P. K., O’Mahony, S. L., Vacharapiyasophon, P. & Walsby, A. E. (1999). A molecular and phenotypic analysis of Nodularia (cyanobacteria) from the Baltic Sea. J Phycol 35, 931-937.

Barker, G. L. A., Konopka, A., Handley, B. A. & Hayes, P. K. (2000). Genetic variation in Aphanizomenon (cyanobacteria) colonies from the Baltic Sea and North America. J Phycol 36, 947-950.

Beltran, E. C. & Neilan, B. A. (2000). Geographical segregation of the neurotoxin-producing cyanobacterium Anabaena circinalis. Appl Environ Microbiol 66, 4468-4474.[Abstract/Free Full Text]

Bolch, C. J. S., Orr, T. P., Jones, G. J. & Blackburn, S. I. (1999). Genetic, morphological, and toxicological variation among globally distributed strains of Nodularia (Cyanobacteria). J Phycol 35, 339-355.

Booker, M. J. & Walsby, A. E. (1979). The relative form resistance of straight and helical blue-green algal filaments. Br Phycol J 14, 141-150.

Boyer, S. L., Flechtner, V. R. & Johansen, J. R. (2001). Is the 16S–23S internal transcribed spacer region a good tool for use in molecular systematics and population genetics? A case study in cyanobacteria. Mol Biol Evol 18, 1057-1069.[Abstract/Free Full Text]

Carmichael, W. W. & Gorham, P. R. (1978). Anatoxins from clones of Anabaena flos-aquae isolated from lakes of Western Canada. Mitt Int Ver Limnol 21, 285-295.

Carmichael, W. W. & Gorham, P. R. (1980). Freshwater cyanophyte toxins: types and their effects on the use of microalgal biomass. In Algal Biomass: Production and Use , pp. 437-448. Edited by G. Shelef & C. J. Soeder. Amsterdam:Elsevier.

Cilia, V., Lafay, B. & Christen, R. (1996). Sequence heterogeneity among 16S rRNA sequences, and their effect on phylogenetic analyses at the species level. Mol Biol Evol 13, 451-461.[Abstract]

Clayton, R., Sutton, G., Hinkle, P. J., Bult, C. & Fields, C. (1995). Intraspecific variation in small-subunit rRNA sequences in GenBank: why single sequences may not adequately represent prokaryotic taxa. Int J Syst Bacteriol 45, 595-599.[Abstract]

Dennis, P. P., Ziesche, S. & Mylvaganam, S. (1998). Transcription analysis of two disparate rRNA operons in the halophilic archaeon Haloarcula marismortui. J Bacteriol 180, 4804-4813.[Abstract/Free Full Text]

Fergusson, K. M. & Saint, C. P. (2000). Molecular phylogeny of Anabaena circinalis and its identification in environmental samples by PCR. Appl Environ Microbiol 66, 4145-4148.[Abstract/Free Full Text]

Florenzano, G., Sili, C., Pelosi, E. & Vincenzini, M. (1985). Cyanospira rippkae and Cyanospira capsulata (gen. nov. and spp. nov): new filamentous heterocystous cyanobacteria from Magadi lake (Kenya). Arch Microbiol 140, 301-306.

Giovannoni, S. J., Turner, S., Olsen, G. J., Barns, S., Lane, D. J. & Pace, N. R. (1988). Evolutionary relationships among cyanobacteria and green chloroplasts. J Bacteriol 170, 3584-3592.[Medline]

Gojobori, T., Ishii, K. & Nei, M. (1982). Estimation of average number of nucleotide substitutions when the rate of substitution varies with nucleotide. J Mol Evol 18, 414-423.[Medline]

Hain, T., Ward-Rainey, N., Kroppenstedt, R. M., Stackebrandt, E. & Rainey, F. A. (1997). Discrimination of Streptomyces albidoflavus strains based on the size and number of 16S–23S rDNA intergenic spacers. Int J Syst Bacteriol 47, 202-206.[Abstract/Free Full Text]

Horecká, M. & Komárek, J. (1979). Taxonomic position of three planktonic blue-green algae from the genera Aphanizomenon and Cylindrospermopsis. Preslia 51, 289-312.

Iteman, I., Rippka, R., Tandeau de Marsac, N. & Herdman, M. (1999). Use of molecular tools for the study of genetic relationships of heterocystous cyanobacteria. In Marine Cyanobacteria (Bulletin de l’Institut Océanographique, Monaco, special issue 19) , pp. 13-20. Edited by L. Charpy & A. Larkum. Monaco:Institut Océanographique.

Iteman, I., Rippka, R., Tandeau de Marsac, N. & Herdman, M. (2000). Conserved structural and regulatory domains within divergent 16S rRNA–23S rRNA spacer sequences of cyanobacteria. Microbiology 146, 1275-1286.[Abstract/Free Full Text]

Jeeji-Bai, N., Hegewald, E. & Soeder, C. J. (1977). Revision and taxonomic analysis of the genus Anabaenopsis. Arch Hydrobiol Suppl 51, 3-24.

Jensen, M. A., Webster, J. A. & Straus, N. (1993). Effect of PCR conditions on the formation of the heteroduplex and single-stranded DNA products in the amplification of bacterial rDNA spacers regions. PCR Methods Appl 3, 186-194.[Medline]

Kaneko, T., Sato, S., Kotani, H. & 21 other authors (1996). Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC 6803. II. Sequence determination of the entire genome and assignment of potential protein-coding regions. DNA Res 3, 109–136.[Medline]

Keswani, J. & Whitman, W. B. (2001). Relationship of 16S rRNA sequence similarity to DNA hybridization in prokaryotes. Int J Syst Evol Microbiol 51, 667-678.[Abstract]

Lachance, M. A. (1981). Genetic relatedness of heterocystous cyanobacteria by deoxyribonucleic acid–deoxyribonucleic acid reassociation. Int J Syst Bacteriol 31, 139-147.

Lehtimäki, J., Lyra, C., Suomalainen, S., Sundman, P., Rouhiainen, L., Paulin, L., Salkinoja-Salonen, M. & Sivonen, K. (2000). Characterization of Nodularia strains, cyanobacteria from brackish waters, by genotypic and phenotypic methods. Int J Syst Evol Microbiol 50, 1043-1053.[Abstract]

Li, R., Carmichael, W. W., Liu, Y. & Watanabe, M. M. (2000). Taxonomic re-evaluation of Aphanizomenon flos-aquae NH-5 based on morphology and 16S rRNA gene sequences. Hydrobiologia 438, 99-105.

Liefting, L., Andersen, M., Beever, R., Gardner, R. & Forster, R. (1996). Sequence heterogeneity in the two 16S rRNA genes of Phormium yellow leaf phytoplasma. Appl Environ Microbiol 62, 3133-3139.[Abstract]

Ligon, P. J. B., Meyer, K. G., Martin, J. A. & Curtis, S. E. (1991). Nucleotide sequence of a 16S rRNA gene from Anabaena sp. strain PCC 7120. Nucleic Acids Res 19, 4553.[Medline]

Lu, W., Evans, H. E., McColl, M. & Saunders, V. A. (1997). Identification of cyanobacteria by polymorphisms of PCR-amplified rDNA spacer region. FEMS Microbiol Lett 153, 141-149.

Ludwig, W., Strunk, O., Klugbauer, S., Klugbauer, N., Weizenegger, M., Neumaier, J., Bachleitner, M. & Schleifer, K.-H. (1998). Bacterial phylogeny based on comparative sequence analysis. Electrophoresis 19, 554-568.[Medline]

Lyra, C., Hantula, J., Vainio, E., Rapala, J., Rouhiainen, L. & Sivonen, K. (1997). Characterization of cyanobacteria by SDS-PAGE of whole-cell proteins and PCR/RFLP of the 16S rRNA gene. Arch Microbiol 168, 176-184.[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]

Mahmood, N. A. & Carmichael, W. W. (1986). The pharmacology of anatoxin-a(s), a neurotoxin produced by the freshwater cyanobacterium Anabaena flos-aquae NRC 525-17. Toxicon 24, 425-434.[Medline]

Martin, C., Sivonen, K., Matern, U., Dierstein, R. & Weckesser, J. (1990). Rapid purification of the peptide toxins microcystin-LR and nodularin. FEMS Microbiol Lett 68, 1-6.

Martinez-Murcia, A. J., Anton, A. I. & Rodriguez-Valera, F. (1999). Patterns of sequence variation in two regions of the 16S rRNA multigene family of Escherichia coli. Int J Syst Bacteriol 49, 601-610.[Abstract]

Miao, V. P. W., Rabenau, A. & Lee, A. (1997). Cultural and molecular characterization of photobionts of Peltigera membranacea. Lichenologist 29, 571-586.

Moffitt, M. C., Blackburn, S. I. & Neilan, B. A. (2001). rRNA sequences reflect the ecophysiology and define the toxic cyanobacteria of the genus Nodularia. Int J Syst Evol Microbiol 51, 505-512.[Abstract]

Mylvaganam, S. & Dennis, P. P. (1992). Sequence heterogeneity between the two genes encoding 16S rRNA from the halophilic archaebacterium Haloarcula marismortui. Genetics 130, 399-410.[Abstract/Free Full Text]

Neilan, B. A., Jacobs, D. & Goodman, A. E. (1995). Genetic diversity and phylogeny of toxic cyanobacteria determined by DNA polymorphisms within the phycocyanin locus. Appl Environ Microbiol 61, 3875-3883.[Abstract]

Neilan, B. A., Stuart, J. L., Goodman, A. E., Cox, P. T. & Hawkins, P. R. (1997). Specific amplification and restriction polymorphisms of the cyanobacterial rRNA operon spacer region. Syst Appl Microbiol 20, 612-621.

Nichols, J. M., Foulds, I. J., Crouch, D. H. & Carr, N. G. (1982). The diversity of cyanobacterial genomes with respect to rRNA cistrons. J Gen Microbiol 128, 2739-2746.

Nübel, U., Engelen, B., Felske, A., Snaidr, J., Wieshuber, A., Amann, R. I., Ludwig, W. & Backhaus, H. (1996). Sequence heterogeneities of genes encoding 16S rRNAs in Paenibacillus polymyxa detected by temperature gradient gel electrophoresis. J Bacteriol 178, 5636-5643.[Abstract/Free Full Text]

Olsen, G. J., Matsuda, H., Hagstrom, R. & Overbeek, R. (1994). fastDNAml: a tool for construction of phylogenetic trees of DNA sequences using maximum-likelihood. Comput Appl Biosci 10, 41-48.[Abstract]

Rippka, R. & Herdman, M. (1992). Catalogue of Strains. Pasteur Culture Collection of Cyanobacterial Strains in Axenic Culture. Paris: Institut Pasteur.

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

Rippka, R., Castenholz, R. W. & Herdman, M. (2001). Subsection IV. (Formerly Nostocales Castenholz 1989b sensu Rippka, Deruelles, Waterbury, Herdman and Stanier 1979). In Bergey’s Manual of Systematic Bacteriology, 2nd edn, vol. 1, The Archaea and the Deeply Branching and Phototrophic Bacteria, pp. 562–589. Edited by D. R. Boone, R. W. Castenholz & G. M. Garrity. New York: Springler-Verlag.

Rosselló-Mora, R. & Amann, R. (2001). The species concept for prokaryotes. FEMS Microbiol Rev 25, 39-67.[Medline]

Saker, M. L. & Neilan, B. A. (2001). Varied diazotrophies, morphologies, and toxicities of genetically similar isolates of Cylindrospermopsis raciborskii (Nostocales, Cyanophyceae) from Northern Australia. Appl Environ Microbiol 67, 1839-1845.[Abstract/Free Full Text]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Schlösser, U. G. (1994). Sammlung von Algenkulturen at the University of Göttingen. Catalogue of strains. Bot Acta 107, 111-186.

Seenayya, G. & Subba-Raju, N. (1972). On the ecology and systematic position of the alga known as Anabaenopsis raciborskii (Wolosz.) Elenk. and a critical evaluation of the forms described under the genus Anabaenopsis. In Taxonomy and Biology of Blue-green Algae , pp. 52-57. Edited by T. V. Desikachary. Madras:University of Madras.

Sivonen, K. & Jones, G. (1999). Cyanobacterial toxins. In Toxic Cyanobacteria in Water. A Guide to Their Public Health Consequences, Monitoring and Management , pp. 41-111. Edited by I. Chorus & J. Bartram. London:E. & F. N. Spon on behalf of the WHO.

Turner, S. (1997). Molecular systematics of oxygenic photosynthetic bacteria. Plant Syst Evol (Suppl) 11, 13-52.

Ueda, K., Seki, T., Kudo, T., Yoshida, T. & Kataoka, M. (1999). Two distinct mechanisms cause heterogeneity of 16S rRNA. J Bacteriol 181, 78-82.[Abstract/Free Full Text]

Van de Peer, Y. & De Wachter, R. (1994). TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment. Comput Appl Biosci 10, 569-570.[Medline]

Walsby, A. E. (1981). Cyanobacteria: planktonic gas-vacuolate forms. In The Prokaryotes , pp. 224-235. Edited by M. P. Starr, H. Stolp, H. G. Trüper, A. Balows & H. G. Schlegel. Berlin, Heidelberg, New York:Springer-Verlag.

Wang, Y., Zhang, Z. & Ramanan, N. (1997). The actinomycete Thermobispora bispora contains two distinct types of transcriptionally active 16S rRNA genes. J Bacteriol 179, 3270-3276.[Abstract]

Wayne, L. G., Brenner, D. J., Colwell, R. R. & 9 other authors (1987). Report of the Ad Hoc Committee on reconciliation of approaches to Bacterial Systematics. Int J Syst Bacteriol 37, 463–464.

West, N. J. & Adams, D. G. (1997). Phenotypic and genotypic comparison of symbiotic and free-living cyanobacteria from a single field site. Appl Environ Microbiol 63, 4479-4484.[Abstract]

Wilmotte, A. & Herdman, M. (2001). Phylogenetic relationships among the cyanobacteria based on 16S rRNA sequences. Bergey’s Manual of Systematic Bacteriology, 2nd edn, vol. 1, The Archaea and the Deeply Branching and Phototrophic Bacteria, pp. 487–493. Edited by D. R. Boone, R. W. Castenholz & G. M. Garrity. New York: Springler-Verlag.

Wilmotte, A., Van der Auwera, G. & De Wachter, R. (1993). Structure of the 16S rRNA of the thermophilic cyanobacterium Chlorogloeopsis HTF (‘Mastigocladus laminosus HTF’) strain PCC 7518, and phylogenetic analysis. FEBS Lett 317, 96-100.[Medline]

Wilson, K. M., Schembri, M. A., Baker, P. D. & Saint, C. P. (2000). Molecular characterization of the toxic cyanobacterium Cylindrospermopsis raciborskii and design of a species-specific PCR. Appl Env Microbiol 66, 332-338.[Abstract/Free Full Text]

Yap, W. H., Zhang, Z. & Wang, Y. (1999). Distinct types of rRNA operons exist in the genome of the Actinomycete Thermomonospora chromogena and evidence for horizontal transfer of an entire rRNA operon. J Bacteriol 181, 5201-5209.[Abstract/Free Full Text]

Zevenboom, W., Van Der Does, J., Bruning, K. & Mur, L. R. (1981). A non-heterocystous mutant of Aphanizomenon flos-aquae, selected by competition in light-limited continuous culture. FEMS Microbiol Lett 10, 11-16.

Received 30 July 2001; revised 12 October 2001; accepted 17 October 2001.