School of Biological Sciences, University of Bristol, Woodland Road, Bristol BS8 1UG, UK
Correspondence
Sven Becker
SvenBecker{at}gmx.ch
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The buoyant density of a cell is determined by several classes of components which make the cell either more or less dense than water; of these, only the gas vesicle provides significant buoyancy. Gas vesicles are made entirely of proteins, which form a cylindrical shell enclosing a hollow space. There are two principal protein components, GvpA, which forms the ribs of the hollow cylindrical structure, and GvpC, which binds to the outer surface of the ribs, strengthening them against collapse by external pressure (for review see Walsby, 1994). Proteins encoded by other gas vesicle genes have been detected in gas vesicles of halobacteria (Shukla & DasSarma, 2004
). GvpD and GvpE have been found to function as a repressor and activator, respectively, of the promoter for gvpA in Halobacterium salinarum (Pfeifer et al., 2002
).
In cyanobacteria, changes in gas vesicle production have been studied in species of Anabaena and Planktothrix (Oscillatoria) (Oliver & Walsby, 1984; Utkilen et al., 1985
), but what determines these changes at the molecular level is not known. However, it is assumed that a change in gas vesicle production must involve changes in the expression of relevant genes. The genes involved in gas vesicle production have been identified and sequenced in several cyanobacteria; the principal protein-encoding structural genes, gvpA and gvpC, occur in various arrangements in gene clusters containing additional ORFs. Planktothrix spp. possess gvpA genes that are nearly identical to those in Anabaena and Calothrix sp.; the gvpC genes, however, show only limited similarity to those in these two organisms (Griffiths et al., 1992
; Albouy et al., 2001
). In Anabaena there are multiple gvpA cistrons in tandem repeat followed by a single copy of gvpC (Hayes & Powell, 1995
); in Planktothrix there are two or more copies of both gvpA and gvpC in an alternating arrangement (Beard et al., 1999
).
Specific promoter regions of gas vesicle gene operons have not been investigated in cyanobacteria; however, in Calothrix it has been demonstrated by Northern hybridization that gvpA and gvpC are co-transcribed (Damerval et al., 1987), while Anabaena produces single and multicistronic gvpA transcripts and also transcripts containing both gvpA and gvpC (Hayes & Powell, 1995
). The gvp transcripts from Planktothrix sp. strain CYA18 are considerably longer than predicted for the single ORFs (J. Kromkamp & P. K. Hayes, unpublished) and may include several of the alternating gvpA and gvpC genes now known to be present in many Planktothrix strains. Some strains of Planktothrix have only one length variant of gvpC, while others have two different variants, gvpC16 and gvpC20 (e.g. strain Pla 9736), or gvpC20 and gvpC28 (e.g. strain PCC 7821). The occurrence of these three length variants of gvpC in Planktothrix strains is correlated with the production of gas vesicles of different diameters and, consequently, of different strengths (Beard et al., 1999
, 2000
). This allows a selection of different gas vesicle genotypes in lakes of different depths (Walsby & Bleything, 1988
; Bright & Walsby, 1999
).
For cyanobacterial genotypes with two gvpC length variants it is not known whether the genes are expressed differentially in response to environmental conditions. In order to perform culture-independent in situ studies on single filaments, a sensitive and specific methodology is needed to detect very low amounts of gvpC transcripts. This is possible with quantitative real-time PCR. The real-time monitoring of amplicon accumulation in PCR allows calibration by the threshold-cycle method (Heid et al., 1996), i.e. quantification is based on the number of cycles required to reach a certain concentration of amplicons rather than on the concentration reached after a fixed number of cycles in end-point analysis. The threshold cycle (CT) is defined as the number of PCR cycles at which a fluorescence signal, developed by a dyetemplate complex or by TaqMan chemistry (Becker et al., 2000
), exceeds a pre-set value. The CT value is reached in few PCR cycles if a large number of templates is initially present, but requires many cycles if the reaction starts with few templates.
In this study we describe a protocol for the reverse-transcription of mRNA from a single cyanobacterial filament to cDNA, which is then used as a template in real-time PCR for the detection of different gas vesicle gene (gvp) transcripts. We demonstrate the potential of this methodology for the measurement of transcript abundance of individual genes in single filaments grown in laboratory culture or collected from their natural environment. Such studies will extend the knowledge of buoyancy regulation in cyanobacteria and their vertical movements in lakes.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Sampling of Planktothrix filaments from Lake Zürich.
On August 12 and 13 2002, water samples from 8, 10·5 and 15 m depth were taken from Lake Zürich above the deepest point of the lake, immediately concentrated by filtration as described by Walsby et al. (1998) and stored on ice. Each single filament of Planktothrix rubescens was picked under a binocular microscope with a sterile syringe and washed successively in three drops of a sterile mineral medium (Bright & Walsby, 2000
) containing all ingredients except FeNaEDTA and NaHCO3. Finally, a single filament was washed in one drop of sterile nuclease-free water (Sigma), transferred to 30 µl nuclease-free water and kept on ice before it was frozen at 20 °C. The samples were shipped to Bristol on dry ice; through delay they arrived at room temperature, but were refrozen and then kept at 70 °C until further processing.
Nucleic acid extraction and synthesis of first strand cDNA from a single filament.
Genomic DNA from Planktothrix batch cultures was extracted as described by Beard et al. (1999). The concentration of DNA was calculated from the A260 measured with a diode-array S2000 UV/Vis spectrophotometer (WPA). For RNA extraction, a single filament was picked from a Planktothrix laboratory culture as described above and transferred to 20 µl nuclease-free water (Sigma) with 20 units Rnasin ribonuclease inhibitor (Promega) and 2·2 mM Dithiothreitol (Sigma) before two cycles of freezing/thawing (20 °C/on ice) were applied in order to fracture the cells, followed by freezing of the lysate at 20 °C, storing at 70 °C and thawing on ice prior to further processing.
Before DNase treatment, the thawed lysate was centrifuged for 5 min at 5000 r.p.m. at 4 °C to pellet unlysed cells and cell debris. A 15 µl quantity of the supernatant was transferred to a fresh reaction tube and mixed with 2 units RQ1 RNase-free DNase (Promega), 2 µl RQ1 DNase 10x reaction buffer (Promega) and 40 units Rnasin ribonuclease inhibitor (Promega). After incubation at 37 °C (45 min), 3 µl stop solution (20 mM EGTA, Promega) was added and the reaction mixtures were incubated for another 10 min at 65 °C to inactivate the DNase.
Before thawing the Lake Zürich samples from 70 °C, 40 units Rnasin ribonuclease inhibitor (Promega), 0·85 µl dithiothreitol (final concentration 2·3 mM; Sigma) and 5·25 µl nuclease-free water (Sigma) were added to the 30 µl of water that contained a single filament. The samples then underwent two cycles of freezing/thawing (see above). For DNase treatment, a larger quantity (31·4 µl) of the lysate (supernatant after centrifugation, see above) was taken and the amounts of all other components of the reaction mixtures (see above) were adjusted accordingly.
For synthesis of first strand cDNA, 7·5 µl of DNase-treated lysate from a single filament was mixed with 7·5 µl 10 µM reverse primer (Table 1), incubated at 70 °C (4 min) and 60 °C (1 min), and then kept on ice prior to the addition of 10 µl of a master mix containing the following: 200 units M-MLV Reverse Transcriptase (Promega), 5 µl 5x M-MLV Reverse Transcriptase reaction buffer (Promega), 1·25 mM dNTPs, 24 units Rnasin ribonuclease inhibitor (Promega) and 1·25 mM dithiothreitol (Sigma). The reaction mixtures were incubated at 42 °C for 1 h. Control reaction mixtures without reverse transcriptase were treated as above and contained the mixture of lysate and reverse primer, 5x M-MLV Reverse Transcriptase reaction buffer (Promega), dNTPs and water.
|
Quantitative real-time PCR.
Oligonucleotide primers for gas vesicle genes gvpA and the three known gvpC length variants in Planktothrix rubescens (Table 1) were constructed using alignments of the gvp sequences published by Beard et al. (1999
, 2000)
and other sequences from the GenBank/EMBL nucleotide sequence database. The binding positions of all primers were chosen to produce amplicons of 185 to 200 bp and to achieve maximum specificity. All primer sequences were checked for specificity by a BLAST search in the GenBank/EMBL nucleotide sequence database. For quantitative PCR, 10 µl reaction mixtures contained 4 µl template (genomic DNA or cDNA), 900 nM of each primer (Table 1
) and the components of the Sybr Green PCR core reagents kit (Applied Biosystems): 0·25 units AmpliTaq Gold DNA polymerase; 0·1 units Amperase UNG; 200 µM dATP, dCTP, dGTP; 400 µM dUTP; 3 mM MgCl2 and 1 µl 10x SYBR Green buffer with ROX fluorescence dye as a passive reference for normalization. For real-time monitoring of the fluorescence increase during successive cycles in PCR, a Sequence Detection System 7000 (Applied Biosystems) was used. The thermal programme was as follows: 2 min at 50 °C, 10 min at 95 °C, then 40 cycles of 20 s at 94 °C and 1 min at 68 °C.
The threshold cycle (CT) of real-time PCR assays is defined as the number of PCR cycles at which the fluorescence signal exceeds a pre-set value (Heid et al., 1996). In all assays CT was determined at a normalized fluorescence of 0·5; loglinear standard curves were obtained by plotting the CT of reactions with various amounts of serially diluted DNA versus the dilution factor. From the slope (s) of the curves, the amplification efficiency (E) of the PCR assays was calculated as 101/s1 (Klein et al., 1999
). Differences in the CT values (
CT) between separate PCR assays containing the same species and same amount of genomic template DNA, but different primer pairs, were used to calculate ratios of gas vesicle genes in Planktothrix genotypes. This is possible because the number of amplicons with similar lengths is the same in different assays when their fluorescence signals reach the threshold (at which the CT value is determined). In the general case (with a PCR efficiency of 1) the ratio of the initial copy numbers between two assays is equal to
. Therefore, if
CT=1,
=2, i.e. one assay contained twice as many initial copies as the other. In the assay with half the initial copy number, one extra cycle is required to reach the same threshold amplicon number.
The ability of primers to discriminate between the three different gvpC length variants (Table 1) was assessed by using the three primer pairs in separate PCR assays with a template DNA species that carries only one of the target genes. If a non-target primer pair resulted in the generation of a product curve, i.e. a CT value could be determined, the difference between the CT values (
CT) of the target and the non-target assays was calculated. The percentage of template molecules that amplify with the non-target primers is given by the term 100/
.
Dissociation analysis.
The thermal programme for real-time PCR was terminated with a temperature gradient between 60 and 95 °C. Since Sybr Green fluoresces only upon binding to double-stranded DNA, the decrease of fluorescence with increasing temperature provides a measure of the melting behaviour of a certain amplicon. By determining dF/dT, the negative derivative of fluorescence (F) versus temperature (T), individual melting profiles were obtained for each amplicon. The identity of amplicons derived from cDNA was checked by comparing their melting profiles with those of gvp fragments amplified from genomic DNA.
Gel electrophoresis.
After completion of the thermal programme, 1 µl of the real-time PCR reaction mixtures was loaded on 10 % polyacrylamide (acrylamide-N,N'-methylene-bisacrylamide, 30 : 0·8) gels of 0·8 mm thickness. Marker lanes contained 2 µl of a 100 bp DNA ladder (Promega). PCR products were separated in a Mini Protean II gel electrophoresis system (Bio-Rad) with 1x TBE running buffer (pH 8) at 50 V (3 h). The gels were stained for 30 min with 1x Sybr Gold nucleic acid gel stain (Molecular Probes) in 1x TAE buffer and photographed under UV illumination.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
To test the reliability of our new real-time PCR assays, we used them to determine the relative gvp gene copy numbers in genomic DNA from five Planktothrix strains of different gvp genotypes that had been characterized by sequencing (Beard et al., 1999, 2000
). We used 103-fold diluted target DNA of strains Pla 9303, Pla 9316 and CYA 137 that had been used in the assays shown in Fig. 1
; thus an amplification efficiency of 1 was assumed. By the construction of loglinear standard curves, the amplification efficiency with DNA from strains Pla 97112 and PCC 7821 was checked and also found to be close to 1 (data not shown). With the CT values of the real-time PCR assays, the ratios of the gvp genes were calculated and compared with those revealed by sequencing the gene clusters (Beard et al., 1999
, 2000
). For genotypes 2a, 3b and 4a the calculated relative gene contents were similar to those found by sequence analysis, but in genotypes 1 and 6 the assays underestimated the copy number of gvpA relative to gvpC20 or gvpC28 copies (Table 2
).
|
|
Consequently, we decided to evaluate the effect of cDNA precipitation on real-time PCR performance with cDNA from single filaments of a Planktothrix Pla 9303 culture that had been grown in batch cultures for 20 to 57 days. One set of PCR primers (for gvpA or gvpC20) was used on subsamples of unprecipitated and precipitated cDNA derived from the same filament. For all the 20 to 57 day old cultures, irrespective of the cDNA treatment, we obtained a real-time PCR signal (CT value) (Table 3). No signal was observed in control assays with a portion of the cell lysates that were not subjected to reverse transcription (RT control). Precipitation not only led to exponential product curves that had a linear relationship before the plateau phase (data not shown), but also significantly reduced the CT value of most of the assays, indicating removal of PCR inhibitors (most probably reverse transcriptase). In only one assay (Table 3
, filament 7) did precipitation increase the CT value slightly, possibly because of the loss of a small amount of template. The carrier DNA that is used in the protocol usually prevents this, however. Dissociation analysis confirmed that gvpA amplicons were present only when precipitated cDNA was used as the template (Table 3
). It appears, therefore, that purification of cDNA by precipitation is essential for real-time RT-PCR detection of certain gvp transcripts in a single cyanobacterial filament.
|
|
|
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Real-time PCR can be calibrated by the threshold-cycle (CT) method (Heid et al., 1996), which is based on the number of amplification cycles required to reach a certain concentration of amplicons. To evaluate the overall performance and amplification efficiencies of the new PCR assays used in this study, we constructed loglinear standard curves with genomic Planktothrix DNA as a template (Fig. 1
). The assays targeting gvpA and three gvpC length variants showed amplification efficiencies between 1·01 and 1·04, indicating a doubling of the amplicon concentration in each PCR cycle. The detection limits of the assays were similar, between 0·18 and 0·82 pg DNA. With a mean detection limit of 0·4 pg DNA per assay, a molecular mass of 650 Da of one base pair (bp) and an assumed Planktothrix genome size of 4·5 Mbp, the analytical sensitivity of the 10 µl assays can be calculated as 82 genomes (Becker et al., 2000
).
If the amplification efficiency is known and genomic DNA is used as a template, differences in the CT values between real-time PCR assays can be used to calculate the copy number ratio of different target genes. For three of the five Planktothrix gas vesicle genotypes investigated, the ratios of gvpC genes were similar to those determined by DNA sequencing (Table 2), although slight deviations from the expected ratios were observed in the genotypes 3b (gvpC16) and 4a (gvpC28). However in genotypes 1 and 6, the relative gvpA copy number was underestimated by a factor of about 2·5. We assume that the observed deviations are due to variation in the amplification efficiencies of single assays that can lead to shifts of CT values. Additionally, calculations based on the slopes of standard curves as shown in Fig. 1
may overestimate the amplification efficiency of single reactions (Ramakers et al., 2003
). This can lead to amplification efficiencies slightly higher than 1 (compare Fig. 1
). Therefore, for quantitative transcription analysis, and because it is impossible to construct standard curves with cDNA from single cyanobacterial filaments, it is essential to monitor the amplification efficiencies of single reactions (see below).
Real-time PCR not only provides a sensitive and specific detection method, but, when performing a dissociation analysis with Sybr Green as a detector, also allows the amplicon identity to be checked in the same assay, a prerequisite for reliable transcription studies. The dissociation analysis also circumvents the time-consuming electrophoretic separation of PCR products and makes it possible to identify the contribution of non-specific PCR products, such as primer dimers, to the fluorescence signal.
In this study we describe a protocol for the extraction of mRNA from a single cyanobacterial filament. After reverse transcription, small amounts of cDNA were detected in real-time PCR. By comparing the CT values of these assays with the results of gel-based amplicon separation (Fig. 3) and fluorescent dissociation analysis (Fig. 4
), we showed that it is possible to detect transcripts of more than one type of gvpC variant in a single Planktothrix filament. Hence, transcripts of gvpC16 and gvpC20 or gvpC20 and gvpC28 were identified in laboratory cultures of Planktothrix strains Pla 9736 and PCC 7821, respectively (Table 4
). These results indicate that both gvpC length variants are transcribed under our laboratory conditions.
In single Planktothrix filaments from Lake Zürich (Table 5) transcripts of both gvpC16 and gvpC20 were found. These two length variants have been identified in Planktothrix isolates from this habitat (Beard et al., 2000
). The presence of different gvpC transcripts in a single cyanobacterial filament has not previously been described. Currently no information is available to indicate whether the presence of two different gvpC transcripts reflects the expression (i.e. the translation to protein) of the two genes. Immunoblotting of proteins from pure Planktothrix gas vesicles has shown that only one GvpC protein species was present in gas vesicles isolated from strains carrying two gvpC length variants (S. J. Beard, unpublished results). It remains possible, however, that gas vesicles of different diameters and strengths may co-occur within a single filament under appropriate environmental conditions; this might contribute to the efficiency of buoyancy provision in Planktothrix. We hypothesize that post-transcriptional regulation may influence the type of protein (GvpC) present in gas vesicles. Information on gas vesicle gene promoters in cyanobacteria as well as insights into a possible translational regulation of gas vesicle formation and the structural assembly of the protein cylinders are necessary.
Quantification of transcripts in a single cyanobacterial filament
The level of transcription of a particular gene can be estimated by reverse transcription of mRNA to cDNA and measuring the number of cDNA copies by quantitative PCR. To achieve quantitative results, the preservation of mRNA in samples is crucial at the time of sampling. Since it might be difficult to maintain the required low temperature for working with RNA while processing samples, they could be treated directly with reagents that preserve the mRNA profile in the cells and allow storage at room temperature for a sufficient period of time. To achieve efficient reverse transcription of different mRNA species, it might be necessary, however, to purify extracted mRNA from single cyanobacterial filaments after preservation of the transcript profile with reagents such as RNAprotect Bacteria Reagent (Qiagen) or RNAlater (Ambion) (S. Becker, unpublished results).
In this study we confirmed results of Liss (2002), who had shown that it is essential to remove PCR inhibitors (e.g. carry-over of reverse transcriptase) from small amounts of cDNA that will be used as a template in real-time PCR. This can be done by the precipitation of cDNA. From the difference in the CT values between the assays with precipitated and unprecipitated cDNA it is calculated that the input template amount would have been underestimated by a factor of up to 103 in the absence of the precipitation step (Table 3
). It may be essential to use precipitated cDNA to confirm the presence of certain gvp amplicons by dissociation analysis (Table 3
), a prerequisite for reliable transcription studies. Additionally, precipitation of cDNA appears to be essential for obtaining product curves that enable linear regression of the section before the plateau phase (data not shown), a prerequisite for quantitative analysis after Ramakers et al. (2003)
(see below).
In real-time PCR, two quantification methods based on CT values are available: absolute quantification by the construction of loglinear standard curves as shown in Fig. 1 (Heid et al., 1996
) and relative quantification by the comparative CT method (Livak & Schmittgen, 2001
). However, standard curves with cDNA from single cyanobacterial filaments are not possible and both methods are prone to quantification errors if possible deviations in the amplification efficiencies between sample and standard assays are not taken into consideration (Ramakers et al., 2003
). Therefore, as suggested for TaqMan chemistry (Becker et al., 2000
), when using Sybr Green as a detector, the amplification efficiency in real-time PCR assays needs to be monitored in order to achieve reliable quantitative results. According to Ramakers et al. (2003)
, the input template copy number in a real-time PCR assay is reflected by the input fluorescence of the assay, which can be calculated by a regression of the linear section of an exponential product curve in the early stage of PCR. This quantification method is independent from standard curves and can be used to monitor the amplification efficiencies between samples. The measurement of cDNA from single cyanobacterial filaments in which the mRNA has been preserved, and normalization of the PCR results to the filament biomass or number of cells (based on the filament length determined by image analysis, cf. Walsby & Avery, 1996
), seem feasible for quantitative transcription studies in cultures and natural samples.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Beard, S. J., Handley, B. A., Hayes, P. K. & Walsby, A. E. (1999). The diversity of gas vesicle genes in Planktothrix rubescens from Lake Zürich. Microbiology 145, 27572768.[Medline]
Beard, S. J., Davis, P. A., Iglesias-Rodríguez, D., Skulberg, O. M. & Walsby, A. E. (2000). Gas vesicle genes in Planktothrix spp. from Nordic lakes: strains with weak gas vesicles possess a longer variant of gvpC. Microbiology 146, 20092018.[Medline]
Becker, S., Böger, P., Oehlmann, R. & Ernst, A. (2000). PCR bias in ecological analysis: a case study for quantitative Taq nuclease assays in analyses of microbial communities. Appl Environ Microbiol 66, 49454953.
Bright, D. I. & Walsby, A. E. (1999). The relationship between critical pressure and width of gas vesicles in isolates of Planktothrix rubescens from Lake Zürich. Microbiology 145, 27692775.[Medline]
Bright, D. I. & Walsby, A. E. (2000). The daily integral of growth by Planktothrix rubescens calculated from growth rate in culture and irradiance in Lake Zürich. New Phytol 146, 301316.[CrossRef]
Damerval, T., Houmard, J., Guglielmi, G., Csiszàr, K. & Tandeau de Marsac, N. (1987). A developmentally regulated gvpABC operon is involved in the formation of gas vesicles in the cyanobacterium Calothrix 7601. Gene 54, 8392.[CrossRef][Medline]
Griffiths, A. E., Walsby, A. E. & Hayes, P. K. (1992). The homologies of gas vesicle proteins. J Gen Microbiol 138, 12431250.[Medline]
Hayes, P. K. & Powell, R. S. (1995). The gvpA/C cluster of Anabaena flos-aquae has multiple copies of a gene encoding GvpA. Arch Microbiol 164, 5057.[Medline]
Heid, C. A., Stevens, J., Livak, K. J. & Williams, P. M. (1996). Real-time quantitative PCR. Genome Res 6, 986994.[Abstract]
Klein, D., Janda, P., Steinborn, R., Müller, M., Salmons, B. & Günzburg, W. H. (1999). Proviral load determination of different feline immunodeficiency virus isolates using real-time polymerase chain reaction: influence of mismatches on quantification. Electrophoresis 20, 291299.[CrossRef][Medline]
Kromkamp, J. & Walsby, A. E. (1990). A computer model of buoyancy and vertical migration in cyanobacteria. J Plankton Res 12, 161183.
Liss, B. (2002). Improved quantitative real-time RT-PCR for expression profiling of individual cells. Nucleic Acids Res 30, e89.
Livak, K. J. & Schmittgen, T. D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the method. Methods 25, 402408.[CrossRef][Medline]
Oliver, R. L. (1994). Floating and sinking in gas-vacuolate cyanobacteria. J Phycol 30, 161173.
Oliver, R. L. & Walsby, A. E. (1984). Direct evidence for the role of light-mediated gas vesicle collapse in the buoyancy regulation of Anabaena flos-aquae (cyanobacteria). Limnol Oceanogr 29, 879886.
Pfeifer, F., Gregor, D., Hofacker, A., Plosser, P. & Zimmermann, P. (2002). Regulation of gas vesicle formation in halophilic archaea. J Mol Microbiol Biotechnol 4, 175181.[Medline]
Ramakers, C., Ruijter, J. M., Deprez, R. H. & Moorman, A. F. (2003). Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci Lett 339, 6266.[CrossRef][Medline]
Shukla, H. D. & DasSarma, S. (2004). Complexity of gas vesicle biogenesis in Halobacterium sp. strain NRC-1: identification of five new proteins. J Bacteriol 186, 31823186.
Skulberg, O. M. & Skulberg, R. (1985). Planktic species of Oscillatoria (Cyanophyceae) from Norway. Characterization and classification. Arch Hydrobiol Suppl 71, 157174.
Utkilen, H. C., Oliver, R. L. & Walsby, A. E. (1985). Buoyancy regulation in a red Oscillatoria unable to collapse gas vacuoles by turgor pressure. Arch Hydrobiol 102, 319329.
Walsby, A. E. (1994). Gas vesicles. Microbiol Rev 58, 94144.[Medline]
Walsby, A. E. & Avery, A. (1996). Measurement of filamentous cyanobacteria by image analysis. J Microbiol Methods 26, 1120.[CrossRef]
Walsby, A. E. & Bleything, A. (1988). The dimensions of cyanobacterial gas vesicles in relation to their efficiency in providing buoyancy and withstanding pressure. J Gen Microbiol 134, 26352645.
Walsby, A. E., Avery, A. & Schanz, F. (1998). The critical pressures of gas vesicles in Planktothrix rubescens in relation to the depth of winter mixing in Lake Zürich, Switzerland. J Plankton Res 20, 13571375.[Abstract]
Received 14 June 2004;
revised 14 September 2004;
accepted 14 October 2004.
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
J MED MICROBIOL | ALL SGM JOURNALS |