SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

N2 fixation is an O2-sensitive process and some filamentous diazotrophic cyanobacteria that grow performing oxygenic photosynthesis confine their N2 fixation machinery to heterocysts, specialized cells that maintain a reducing environment adequate for N2 fixation. Respiration is thought to contribute to the diazotrophic metabolism of heterocysts and the genome of the heterocyst-forming cyanobacterium Anabaena sp. PCC 7120 bears three gene clusters putatively encoding cytochrome c oxidases. Transcript analysis of these cox gene clusters through RNA/DNA hybridization identified two cox operons, cox2 and cox3, that are induced after nitrogen step-down in an NtcA- and HetR-dependent manner and appear to be expressed specifically in heterocysts. In contrast, cox1 was expressed only in vegetative cells. Expression of cox2 and cox3 occurred at an intermediate stage (about 9 h) during the process of heterocyst development following nitrogen step-down. Inactivation of genes in the two inducible cox operons, but not separately in either of them, strongly reduced nitrogenase activity and prevented diazotrophic growth in aerobic conditions. These results show that the nitrogen-regulated cytochrome c oxidase-type respiratory terminal oxidases Cox2 and Cox3 are essential for heterocyst function in Anabaena sp. PCC 7120.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Heterocysts are the sites of N2 fixation in some filamentous cyanobacteria such as those of the genera Anabaena and Nostoc. In oxic cultures subjected to nitrogen step-down, these specialized cells are formed from semi-regularly spaced vegetative cells generating a pattern of about one heterocyst to ten vegetative cells in the filament (Wolk, 2000). Heterocyst differentiation does not take place in mutants with defects in the global nitrogen regulation gene ntcA (Frías et al., 1994; Wei et al., 1994) or in the development regulatory gene hetR (Buikema and Haselkorn, 1991; Black et al., 1993). Whereas cyanobacteria perform oxygenic photosynthesis, the enzyme nitrogenase that reduces N2 to ammonia is extremely sensitive to oxygen. Heterocysts are thought to provide a reducing microenvironment adequate for the expression of the nif genes that encode nitrogenase and for the operation of this enzyme (Fay, 1992). Different factors appear to contribute to the establishment of the micro-oxic environment (reviewed in Wolk et al., 1994): heterocysts do not carry out oxygenic photosynthesis (although they retain the capacity for photosystem I-dependent photosynthetic reactions), bear additional envelope layers that diminish diffusion of gases into them and appear to have an increased respiratory activity. Cyanobacteria are known to express an aa3-type cytochrome c oxidase (Schmetterer, 1994), and this type of oxidase has been identified in heterocysts (Häfele et al., 1988).

A cluster of genes, coxBAC, encoding the three subunits of an aa3-type cytochrome c oxidase has been characterized in the unicellular cyanobacterium Synechocystis sp. PCC 6803 (Schmetterer et al., 1994) and in the heterocyst-forming cyanobacterium Anabaena variabilis ATCC 29413 (Schmetterer et al., 2001). CoxBAC is essential for chemoheterotrophic growth on both strains (Pils et al., 1997; Schmetterer et al., 2001) but is not needed for diazotrophic growth of A. variabilis (Schmetterer et al., 2001). A second cluster of genes homologous to coxBAC has been identified in Synechocystis sp. PCC 6803 (Kaneko et al., 1996) that would encode ARTO (alternative respiratory terminal oxidase, Pils et al., 1997), also known as CtaII (Howitt and Vermaas, 1998). This enzyme has been interpreted to represent a cytochrome bo-type quinol oxidase (Howitt and Vermaas, 1998), but recent results have suggested that, in vivo, it can oxidize cytochrome c553 (Pils and Schmetterer, 2001). Alternative respiratory terminal oxidase is not a typical cytochrome c oxidase because it lacks the binding site for the binuclear center CuA in subunit II (CoxB) and a Mg2+ binding site in subunit I (CoxA) (Howitt and Vermaas, 1998) and shows very low, perhaps negligible, activity in the standard in vitro assays of oxidation of reduced horse heart cytochrome c (Schmetterer et al., 1994).

Anabaena sp. PCC 7120 is a heterocyst-forming cyanobacterium whose complete genomic sequence is available ( Kaneko et al., 2001 ). This sequence bears three gene clusters that are homologous to coxBAC of Synechocystis sp. PCC 6803 and of A. variabilis . According to the degree of identity of their putative coxA gene products to A. variabilis CoxA, we have named these gene clusters cox1 (99% identity), cox2 (67% identity) and cox3 (54% identity) respectively (see Fig. 1 ). Interestingly, only the polypeptides encoded by cox1 and cox2 are predicted to bear the Cu A and Mg 2+ binding motifs and, consistently, the cox3 -encoded polypeptides are most similar in sequence to ARTO polypeptides. In this work, we sought the identification of cox gene cluster(s) whose products are required for heterocyst function and diazotrophic growth in Anabaena sp. PCC 7120. We first analysed the expression pattern of each cox gene cluster and then generated mutants of the cox2 and cox3 clusters, whose expression appears to be restricted to N 2 -fixing cells.

image

Figure 1. Schematic representation of the three gene clusters containing genes homologous to cox genes found in the genome of Anabaena sp. PCC 7120 (drawn from data in http:www.kazusa.or.jpcyanobaseAnabaena /). The ORF name, length in bp, proposed gene name and (when appropriate) percentage identity of the putative protein product to the A. variabilis CoxB, CoxA or CoxC polypeptide ( Schmetterer et al., 2001 ) are shown for each ORF. Triangles indicate the approximate location of oligonucleotide primers used for PCR amplification or for primer extension analysis (see text). The insertion sites of the C.K3 or C.S3 gene-cassettes to generate mutants in cox2 or cox3 are also shown.

Download figure to PowerPoint

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Expression of cox genes

Expression of the cox1 gene cluster was tested by RNA/DNA hybridization using RNA isolated from whole filaments of Anabaena sp. PCC 7120 grown with dinitrogen, nitrate or ammonium as the nitrogen source and, as probes, DNA fragments internal to the coxB1 and coxA1 genes. A band of ∼ 4.1 kb as well as some hybridization signals likely corresponding to degraded RNA were observed independently of the nitrogen source (Fig. 2) indicating that the cox1 gene cluster was expressed in vegetative cells. To test whether in the N2-fixing filament this gene cluster was expressed in the heterocysts as well as in the vegetative cells, an RNA preparation from isolated heterocysts was hybridized with the coxB1 probe. No indication of expression in the heterocysts was obtained, whereas a control RNA isolated from N2-grown whole filaments showed again the 4.1 kb hybridization band (Fig. 2B). Primer extension analysis carried out with two different primers (CB1-1 and CB1-2) identified a putative transcription start point 75 bp upstream from the coxB1 start codon (not shown). Because the length from this transcription start point to the end of coxC1 is 3.80 kb, the 4.1 kb transcript could cover the whole cox1 gene cluster, which would thus represent an operon.

image

Figure 2. Expression analysis of the cox1 gene cluster. RNA from Anabaena sp. PCC 7120 whole filaments grown with ammonium, nitrate or N 2 as the nitrogen source, or from isolated heterocysts (H), was probed with DNA fragments of coxB1 (A, B), coxA1 (C) or, as a loading and transfer control, rnpB . The size in kb of the largest detected transcript (marked by a triangle) is indicated.

Download figure to PowerPoint

Expression of the cox2 gene cluster was analysed as above using as a probe a DNA fragment from the coxB2 gene. A hybridization band corresponding to a transcript size of 3.7 kb was observed with RNA isolated from diazotrophically grown filaments but not with RNA from filaments grown with combined nitrogen (Fig. 3A). To investigate whether this transcript could correspond to expression in the heterocysts, an RNA preparation from isolated heterocysts was tested. The results obtained indicated a high level of expression of coxB2 in the heterocysts (Fig. 3A). Quantification of the hybridization signals (see Experimental procedures for details) indicated that the level of expression of coxB2 in the heterocysts was about 10-fold higher than that detected for the whole filaments grown on N2. This result suggests that coxB2 is expressed only in the heterocysts. Primer extension analysis with two different primers (CB2-1 and CB2-7) identified a putative transcription start point, 174 bp upstream from the coxB2 start codon, that was detected only with the RNA from diazotrophically grown filaments or from isolated heterocysts (not shown). Because the distance from this transcription start point to the end of coxC2 is 3.61 kb, the 3.7 kb mRNA would represent a polycistronic transcript covering the whole cox2 gene cluster. Based on the observed nitrogen regulation and possible heterocyst location of the cox2 transcript, we tested the accumulation of this transcript in Anabaena mutants carrying inactivated versions of the ntcA (strain CSE2) or hetR (strain DR884a) genes. Because these mutants do not grow diazotrophically, induction experiments were carried out in which ammonium-grown filaments were incubated for 24 h in medium lacking combined nitrogen. Expression of the cox2 operon was undetectable in these mutants by means of either RNA/DNA hybridization (Fig. 3B) or primer extension (not shown) analyses.

image

Figure 3. Expression analysis of the cox2 gene cluster.

A. RNA preparations from Anabaena sp. PCC 7120 whole filaments grown with ammonium, nitrate or N 2 as the nitrogen source, or from isolated heterocysts (H), were probed with DNA fragments of coxB2 (upper panel) or, as a loading and transfer control, rnpB .

B. RNA isolated from filaments of strains PCC7120 (wild type, WT), CSE2 ( ntcA ) or DR884a ( hetR ) grown with ammonium and incubated for the indicated periods of time (9, 12 and 24 h) in medium lacking combined nitrogen was hybridized with a coxB2 probe (upper panel) or, as a loading and transfer control, an rnpB probe. The size in kb of the largest detected transcript (marked by a triangle) is indicated.

Download figure to PowerPoint

The pattern of expression of the cox3 gene cluster was similar to that observed for cox2. RNA/DNA hybridization analysis using a coxB3 probe indicated a much higher level of expression in diazotrophically grown filaments than in filaments grown with combined nitrogen (Fig. 4A), and levels detected in heterocysts were again about 10-fold higher than in N2-grown whole filaments (Fig. 4B). Additionally, no indication of cox3 transcription was observed with RNA isolated from the ntcA and hetR mutants after incubation for up to 24 h without combined nitrogen (Fig. 4C). Therefore, transcription of cox3 also appears to be restricted to heterocysts, and signals observed with RNA from nitrate-grown cells (Fig. 4A) may originate from the low percentage of heterocysts that are formed in this cyanobacterium in the presence of nitrate. Although some signals possibly resulting from hybridization with some degraded RNA were observed, a hybridization band of about 5.6 kb was evident with the coxB3 probe. A mRNA of about 5.6 kb could correspond to a polycistronic transcript including, in addition to the three cox3 genes, two ORFs that are located immediately upstream from coxB3 (alr2729 and alr2730), as the distance from the beginning of alr2729 to the end of coxC3 is 4880 nucleotides (Fig. 1). To test whether ORFs alr2729 and alr2730 could really represent two additional genes of the cox3 operon, RNA preparations from ammonium-grown filaments or from filaments grown with ammonium and incubated without any source of combined nitrogen for 18 and 24 h were hybridized with a probe of ORF alr2729. A hybridization band of about 5.6 kb was again observed with RNA isolated from cells incubated in the absence of combined nitrogen (Fig. 4D), consistent with co-transcription of alr2729 (and therefore also of alr2730) and the coxBAC3 genes. The products of ORFs alr2729 and alr2730 show homology to no protein of known function, but they are homologous to each other and predicted to bear transmembrane segments.

image

Figure 4. Expression analysis of the cox3 gene cluster.

A and B. RNA preparations from Anabaena sp. PCC 7120 whole filaments grown with ammonium, nitrate or N2 as the nitrogen source, or from isolated heterocysts (H), were probed with DNA fragments of coxB3 (upper panels) or, as a loading and transfer control, rnpB.

C. RNA isolated from filaments of strains PCC7120 (wild type, WT), CSE2 ( ntcA ) or DR884a ( hetR ) grown with ammonium and incubated for the indicated times (9, 12 and 24 h) in medium lacking combined nitrogen was hybridized with a coxB3 probe (upper panel) or, as a loading and transfer control, an rnpB probe.

D. RNA isolated from Anabaena sp. PCC 7120 whole filaments grown with ammonium (0) or grown with ammonium and incubated for the indicated times (18 or 24 h) in medium lacking combined nitrogen was hybridized with an alr 2729 probe (upper panel) or, as a loading and transfer control, an rnpB probe. The size in kb of the largest detected transcript (marked by a triangle) is indicated.

Download figure to PowerPoint

The time-course experiments presented in Figs 3B and 4C also show that, after nitrogen step-down, induction of cox2 and cox3 gene clusters had already started by about 9 h and peaked by about 12 h. However, induction of cox3 appeared somewhat retarded as compared to that of cox2.

Mutants of cox2 and cox3

The results shown above indicate that the Anabaena sp. PCC 7120 cox2 and cox3 operons are induced under N2-fixing conditions and, likely, specifically in the heterocysts. In contrast, the cox1 operon is expressed constitutively with regard to the nitrogen source, but it appears to be repressed in heterocysts. In order to test a possible role of their products in Anabaena diazotrophic growth, we isolated mutants with defects in the cox2 and cox3 operons.

Two Anabaena cox2 mutants were generated by insertion of the gene-cassettes C.S3 and C.K3 into the coxB2 gene, and one cox3 mutant was isolated after insertion of the C.S3 cassette into coxA3 (see Fig. 1 and Experimental procedures for details). Whereas the cox2 mutants (strains CSAV139 and CSAV140) showed growth characteristics similar to those of the wild type in both liquid and solid media, the cox3 mutant (strain CSAV135) was somewhat impaired in diazotrophic growth in solid, but not in liquid, medium (Fig. 5 and Table 1). To test if growth on N2 could be due to the selection of wild-type chromosomes that might be present at a low frequency in the otherwise apparently segregated mutants, the cox2 and cox3 genome regions were analysed by PCR using as template DNA isolated from cultures of strains CSAV135, CSAV139 and CSAV140 grown in liquid BG110 medium (Fig. 6). In every case, PCR amplification generated only a band corresponding to the inactivated cox gene, coxA3::C.S3 (strain CSAV135), coxB2::C.K3 (strain CSAV139) and coxB2::C.S3 (strain CSAV140), confirming that the coxB2 and coxA3 mutants could grow diazotrophically. We therefore sought the isolation of a cox2 cox3 double mutant. Plasmid pCSAV135 (coxA3::C.S3) was transferred to Anabaena sp. strain CSAV139 (coxB2::C.K3), and a clone homozygous for the two mutations, strain CSAV141 (coxB2::C.K3, coxA3::C.S3), was isolated (Fig. 6). In contrast to the single mutants, strain CSAV141 was specifically impeded in diazotrophic growth in both solid and liquid media (Fig. 5 and Table   1).

image

Figure 5. Growth of Anabaena sp. PCC 7120 and some cox mutants on solid medium. Samples of ammonium-grown filaments containing the amount of chlorophyll a indicated were spotted on the surface of BG11 0 medium solidified with 1% agar and supplemented with nitrate (not shown; growth similar to that shown for ammonium-containing medium), ammonium or no combined nitrogen. The plates were incubated under culture conditions for 12 days and photographed. WT, wild-type strain PCC 7120; 135, CSAV135 ( coxA3 ::C.S3); 139, CSAV139 ( coxB2 ::C.K3); 140, CSAV140 ( coxB2 ::C.S3); 141, CSAV141 ( coxB2 ::C.K3, coxA3 ::C.S3).

Download figure to PowerPoint

Table 1. . Growth rate constants and cytochrome c oxidase activities of Anabaena sp. PCC 7120 and coxB2 and coxA3 mutant strains.
StrainGenotypeGrowth rate constant, µ (day−1)Cytochrome oxidase activity (nmol min−1 mg−1 of Chl)
Nitrogen source: AmmoniumN2Nitrogen source: NitrateN2
  1. Each figure is the mean and standard deviation of the results of three independent experiments. ND, Not determined.

PCC7120Wild type0.53 ± 0.070.35 ± 0.04209.3 ± 8.4350.7 ± 15.1
CSAV140 coxB2 ::C.S3 0.32 ± 0.040.28 ± 0.17167.8 ± 6.4180.7 ± 7.1
CSAV139 coxB2 ::C.K3 0.49 ± 0.080.36 ± 0.05225.9 ± 18.2236.2 ± 2.4
CSAV135 coxA3 ::C.S3 0.49 ± 0.070.31 ± 0.10352.8 ± 15.1352.1 ± 25.3
CSAV141 coxB2 ::C.K3, coxA3 ::C.S3 0.49 ± 0.01<0.02239.5 ± 14.4ND
image

Figure 6. Analysis of the coxB2 and coxA3 genomic regions in Anabaena sp. PCC 7120 and mutant strains CSAV139 ( coxB2 ::C.K3), CSAV140 ( coxB2 ::C.S3), CSAV135 ( coxA3 ::C.S3), and CSAV141 ( coxB2 ::C.K3, coxA3 ::C.S3). DNA was isolated from cultures of strains PCC 7120, CSAV139, CSAV140 and CSAV135 grown for about six to seven generations in BG11 0 medium, or from filaments of strain CSAV141 incubated for 6 days in BG11 0 medium (note that this strain does not grow in this medium), and subjected to PCR amplification of coxB2 and coxA3 . Note that, whereas insertion of the 2 kb gene-cassette C.S3 into coxA3 did not involve any deletion, incorporation of the 1.5 kb gene-cassette C.K3 or the 2 kb gene-cassette C.S3 into coxB2 was accompanied by a 0.62 kb deletion (see Fig. 1 ). S, size standards (λ DNA digested with ClaI).

Download figure to PowerPoint

The cytochrome c oxidase activity of the wild-type strain PCC 7120 and of the coxB2 and coxA3 mutants was tested in filaments grown using nitrate (BG11 medium) or N2 (BG110 medium) as the nitrogen source, and that of the cox2 cox3 double mutant in filaments grown with nitrate. Consistent with the presence of constitutive and inducible cox genes in Anabaena sp. PCC 7120, the wild-type strain exhibited an appreciable cytochrome c oxidase activity when grown in the presence of nitrate that was increased by about 68% when grown fixing N2 (Table 1). The coxB2 mutants, strains CSAV140 and CSAV139, when grown with nitrate, exhibited activities that were similar to that of the wild type (80% and 108%, respectively, of the wild-type activity), and these activities were only marginally increased (<8%) by diazotrophic growth. Because cox3 is induced under the latter conditions, the scarce induction of cytochrome c oxidase activity in strains CSAV140 and CSAV139 indicates that the cox3 products contribute little (if anything) to the cytochrome c oxidase activity assay. In contrast, because cox2 expression appears to be restricted to heterocysts (see above), Cox2-dependent cytochrome c oxidase activity can be considered very significant. Inactivation of coxA3, on the other hand, had the unexpected effect of increasing the cytochrome c oxidase activity observed in BG11 medium to a level similar to that exhibited by BG110-grown filaments (see strain CSAV135 in Table 1). Lack of such an increase in the double mutant (strain CSAV141, Table 1) suggested that the increase in cytochrome c oxidase activity observed in strain CSAV135 is related to Cox2 rather than to Cox1. Whether this effect is the result of an upregulation of the expression of cox2 or of an alteration of the activity of Cox2 is currently unknown.

Incubation of Anabaena sp. PCC 7120 filaments for 24 h in the light under an atmosphere of argon/CO2 permitted subsequent observation of dark nitrogenase activity under aerobic conditions, presumably depending upon carbohydrate accumulated in the filaments (Rippka and Waterbury, 1977). Whereas the dark nitrogenase activity was also observed in the single cox mutants, no activity was observed with the cox2 cox3 double mutant (Table 2). The nitrogenase activity of the wild type was significantly increased in the light, an increase that, though to a limited extent in the case of CSAV135, was also observed in the single mutants (Table 2). The cox2 cox3 double mutant exhibited a very low nitrogenase activity (8% of the wild-type level) even in the light.

Table 2. . Nitrogenase activity of Anabaena sp. PCC 7120 and coxB2 and coxA3 mutant strains.
Experiment typeStrainGenotypeNitrogenase activity (nmol ethylene min−1µg−1 of Chl)
DarknessLight
  1. Experiments type 1, filaments grown diazotrophically prior to incubation under argon/CO2 for 24 h; experiments type 2, ammonium-grown filaments subjected to a 24 h induction in combined nitrogen-free medium before incubation under argon/CO2 for an additional 24 h period. Afterwards, the filament suspensions were treated as described in Experimental procedures for determination of nitrogenase activity in darkness on in the light. Each figure is the mean and standard deviation of the results obtained in two independent determinations for each of three independent cultures.

1PCC7120Wild type0.35 ± 0.030.86 ± 0.11
CSAV140 coxB2 ::C.S3 0.28 ± 0.060.86 ± 0.24
CSAV139 coxB2 ::C.K3 0.32 ± 0.020.95 ± 0.23
CSAV135 coxA3 ::C.S3 0.44 ± 0.020.60 ± 0.23
2PCC7120Wild type0.49 ± 0.001.37 ± 0.15
CSAV141 coxB2 ::C.K3, coxA3 ::C.S3 0.000.11 ± 0.02

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Three putative cox (cytochrome c oxidase) gene clusters, which according to the size of their transcripts likely constitute operons (Figs 2–4), are found in the genomic sequence of the heterocyst-forming cyanobacterium Anabaena sp. PCC 7120 (Kaneko et al., 2001). Expression of two of these operons, cox1 and cox2, has been recently analysed also by Jones and Haselkorn (2002). One of these clusters, which we have named cox1, encodes polypeptides with a high degree of identity to the subunits of the CoxBAC aa3-type cytochrome c oxidase described in A. variabilis as essential for chemoheterotrophic growth (Schmetterer et al., 2001). Because of this homology, and because it is not induced by growth of Anabaena sp. PCC 7120 in media lacking combined nitrogen being expressed only in vegetative cells, we suggest that cox1 encodes the cytochrome c oxidase involved in vegetative growth or maintenance in this cyanobacterium. Indeed, BG11-grown filaments of strain PCC 7120 (in which cox1 is the only expressed cox gene cluster) exhibit a high cytochrome c oxidase activity, which is about twofold that observed in A. variabilis grown chemoheterotrophically with fructose (Schmetterer et al., 2001). We have found that constitutive expression of the cox1 operon in vegetative cells takes place from a transcription start point, which is located 75 nucleotides upstream from the coxB1 start codon, that is preceded by sequences similar to those of σ70-type promoters showing recognizable − 10 (GACCAT, centred at − 9.5) and − 35 (TCGAAA, centered at − 35.5) hexamers. The A. variabilis coxBAC operon is also expressed from a recognizable σ70-type promoter, although in this cyanobacterium further expression of coxBAC takes place from a fructose-inducible promoter (Schmetterer et al., 2001).

The two Anabaena sp. PCC 7120 cox operons, cox2 and cox3, whose putative products exhibit a lower degree of identity to A. variabilis CoxBAC are induced under N2-fixing conditions in an NtcA-dependent manner (Figs 3 and 4). A putative cox2 transcription start point, located 174 nucleotides upstream from the coxB2 start codon, has been identified in this work, but it is not preceded by a canonical NtcA-activated promoter (described in Herrero et al., 2001). On the other hand, it has recently been suggested that the transcription start point for the cox2 operon is located between 300 and 410 nucleotides upstream of the coxB2 start codon, a region where putative NtcA-binding sites are found (Jones and Haselkorn, 2002). Further characterization of the cox2 promoter is obviously necessary, as is also the case for the cox3 promoter that remains to be identified. The high abundance of cox2 and cox3 transcripts in heterocyst RNA preparations suggests that the expression of these operons takes place predominantly in heterocysts, and this conclusion is corroborated by the effect of the hetR mutation, which abolishes cox2 and cox3 expression. We therefore suggest that cox2 and cox3 are expressed specifically in the heterocysts. The same conclusion has recently been reached for cox2 by Jones and Haselkorn (2002) who, with the use of a GFP reporter, localized the expression of cox2 to the heterocysts. Whereas, under our experimental conditions, mature heterocysts appear after about 20–24 h of nitrogen deprivation, induction of the cox2 and cox3 operons has already started by about 9 h and transcript accumulation is highest by 12 h. These observations suggest that the Cox2 and Cox3 complexes may contribute to the process of protoplast maturation that takes place during heterocyst development (Ernst et al., 1992). In contrast to the observed expression of cox2 and cox3 in the heterocysts, we did not detect hybridization of heterocyst RNA with a coxB1 probe indicating that cox1 is not expressed in mature heterocysts (Fig. 2B). Although Jones and Haselkorn (2002) claimed that cox1 is expressed in all cells of the filament, these authors analysed GFP expression from a coxB1::gfp transcriptional fusion 12 h after nitrogen step-down, a time at which repression might have not yet occurred or that may be too short to allow disappearance of GFP after repression of coxB1::gfp. Repression of cox1 and induction of cox2 and cox3 indicates that, rather than just adding oxidases to the developing heterocyst, the developmental program determines substitution of the cytochrome c oxidase-type terminal respiratory oxidase complement of the cells. Such substitution may be linked to the membrane reorganization known to take place during heterocyst development (Wolk et al., 1994).

Disruption of neither coxB2 nor coxA3 abolishes diazotrophic growth, as is also the case for disruption of coxA2 (Jones and Haselkorn, 2002). However, inactivation of both coxB2 and coxA3 results in the inability of Anabaena sp. PCC 7120 to grow diazotrophically under aerobic conditions. These results indicate that both Cox complexes, Cox2 and Cox3, contribute to respiration in the heterocysts, thus generating a microenvironment adequate for nif gene expression and/or nitrogenase operation in these differentiated cells. Indeed, nitrogenase activity both in the light and in darkness was drastically hampered in the cox2 cox3 double mutant. However, our results also indicate that, under certain laboratory conditions, the presence of either Cox2 or Cox3 is sufficient to allow the Anabaena filaments to exhibit a high nitrogenase activity and to grow on N2. This represents an interesting observation given the different properties of these two Cox complexes, because Cox2 is a standard cytochrome c oxidase whose mutation impairs the in vitro-determined cytochrome c oxidase activity (Table 1), whereas Cox3 is similar to Synechocystis ARTO. In this context, it is also interesting to note that the cox3 operon appears to include, in addition to the common coxBAC genes, two genes whose products might represent additional subunits for the Cox3 complex. Nonetheless, from what is known from the similar Synechocystis complexes (Pils and Schmetterer, 2001), both Cox2 and Cox3 are expected to be bioenergetically active (i.e. to generate a proton gradient). This and the fact that both Cox2 and Cox3 can support dark nitrogenase activity suggest that both are able to provide energy for N2 fixation in heterocysts of Anabaena sp. PCC 7120. On the other hand, the low nitrogenase activity observed in the light in the double cox2 cox3 mutant may be supported by photosystem I-dependent photophosphorylation. The fact that also under light conditions one of either Cox2 or Cox3 is required to attain high nitrogenase activities indicates a role of these cytochrome oxidases also in O2 consumption for protection of nitrogenase. In this context, the comparatively low nitrogenase activity of strain CSAV135 in the light (Table 2) may imply a relatively more important role of Cox3 than Cox2 in protection of nitrogenase against O2. Establishing the role of each oxidase in heterocyst metabolism will, however, require further work.

The results presented in this work demonstrate that terminal respiratory oxidases have importance for heterocyst function, and indicate that oxidases specifically induced in the heterocysts, Cox2 and Cox3, are essential for nitrogenase activity and diazotrophic growth. On the other hand, Cox1 is not the only terminal respiratory oxidase present in vegetative cells of Anabaena sp. PCC 7120. Genes cydA (all4024) and cydB (all4023) putatively encoding a cytochrome bd-type quinol oxidase are present in the Anabaena genome (Kaneko et al., 2001), and we have found that expression of these genes takes place independently of the nitrogen source and that a cydAB mutant can grow on N2 (C. Wilken, A. Valladares, A. Herrero and E. Flores, unpublished results). The available genomic sequence of another heterocyst-forming cyanobacterium, Nostoc punctiforme ATCC 29133 (Meeks et al., 2001), contains two putative cox gene clusters each of which bears three ORFs that encode proteins homologous to those of Cox1 or Cox2 (the one in Contig 493 is most similar to Cox1, the one in Contig 509 is most similar to Cox2) and two clusters (one in Contig 473, the other split into two Contigs, 419 and 387) each bearing five ORFs encoding proteins homologous to those of Cox3, but appears not to bear genes for a cytochrome bd-type quinol oxidase. We have also recently cloned a cox2-like gene cluster from A. variabilis (D. Pils and G. Schmetterer, GeneBank acc. no. AJ296086). Therefore, both types of N2 fixation-related Cox complexes, Cox2 and Cox3, may be generally found in heterocyst-forming cyanobacteria.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Organisms

This study was carried out with the heterocyst-forming cyanobacterium Anabaena sp. strain PCC 7120 and two Het derivatives, strain CSE2, an insertional mutant of the ntcA gene (Frías et al., 1994), and strain DR884a, an insertional mutant of the hetR gene (Black et al., 1993).

RNA isolation and analysis

Cells were grown in the light (75 µE m−2 s−1) in BG11 liquid medium (Rippka et al., 1979), BG110 medium (medium BG11 without NaNO3) or BG110 medium containing 8 mM NH4Cl and 16 mM TES-NaOH buffer (pH 7.5). All cultures were supplemented with 10 mM of NaHCO3 and bubbled with a mixture of CO2 and air (1% v/v). Strain CSE2 was grown in the presence of 2 µg ml−1 of streptomycin and 2 µg ml−1 of spectinomycin. Strain DR884a was grown in the presence of 5 µg ml−1 of neomycin. Cultures for heterocyst isolation were grown in ammonium-containing media until they reached a chlorophyll a concentration of 3–5 µg ml−1. Filaments were then washed with, and resuspended in, nitrogen-free medium (BG110) and further incubated until mature heterocysts were observed (19 h in the case of the experiments shown in Figs 2, 3 and 4). Heterocysts were isolated as described (Golden et al., 1991). For RNA isolation, filaments growing in NH4Cl-containing medium were harvested at room temperature and either used directly or washed with, and resuspended in, BG110 medium and further incubated under culture conditions for the number of hours indicated in each experiment. Alternatively, filaments growing exponentially with N2, ammonium or nitrate were used. The RNA from whole filaments or from isolated heterocysts was isolated in the presence of ribonucleoside-vanadyl complex as previously described (Muro-Pastor et al., 2002).

Primer extension analysis of the cox transcripts was carried out as previously described (Muro-Pastor et al., 1999) with 32P-labelled oligonucleotides CB1-1 (6 in Fig. 1) (complementary to nucleotides 61–42 with respect to the translation start of coxB1), CB1-2 (5 in Fig. 1) (complementary to nucleotides 3 to − 16 with respect to the translation start of coxB1), CB2-1 (13 in Fig. 1) (complementary to nucleotides 40–21 with respect to the translation start of coxB2), and CB2-7 (12 in Fig. 1) (complementary to nucleotides − 36 to − 55 with respect to the translation start of coxB2).

For Northern analysis, 40–70 µg of RNA were loaded per lane and subjected to electrophoresis in 1% agarose denaturing formaldehyde gels. Transfer and fixation to Hybond-N+ membranes (Amersham Biosciences) were carried out using 0.1 M NaOH. Hybridization was performed at 65°C according to the recommendations of the manufacturer of the membranes. The cox probes were internal fragments of these genes amplified by PCR, using plasmid pCSAV23 [containing the coxBAC1 genes cloned in vector pIC20R (Marsh et al., 1984)] as a template and oligonucleotides CB1-4 (1 in Fig. 1) (corresponding to nucleotides 100–121 with respect to the translation start of coxB1)/CB1-5 (2 in Fig. 1) (complementary to nucleotides 906–885 with respect to the translation start of coxB1) and oligonucleotides CA1-2 (3 in Fig. 1) (corresponding to nucleotides 3–22 with respect to the translation start of coxA1)/CA1-1 (4 in Fig. 1) (complementary to nucleotides 1725–1706 with respect to the translation start of coxA1) in the case of coxB1 and coxA1 probes respectively. The coxB2 probe was amplified using oligonucleotides CB2-4 (9 in Fig. 1) (corresponding to nucleotides 19–39 with respect to the translation start of coxB2) and CB2-5 (10 in Fig. 1) (complementary to nucleotides 1126–1107 with respect to the translation start of coxB2) and plasmid pCSAV128 (containing the coxB2 gene in vector pGEM-T from Promega). The coxB3 probe was amplified using oligonucleotides CB3-6 (16 in Fig. 1) (corresponding to nucleotides 46–67 with respect to the translation start of coxB3) and CB3-7 (17 in Fig. 1) (complementary to nucleotides 912–891 with respect to the translation start of coxB3) and plasmid pCSAV132. The alr2729 probe was amplified using oligonucleotides alr2729–3 (14 in Fig. 1) (corresponding to nucleotides 46–65 with respect to the translation start of alr2729) and alr2729–4 (15 in Fig. 1) (corresponding to nucleotides 468–450 with respect to the translation start of alr2729) and, as template, a plasmid containing an alr2729 fragment (cloned by PCR using the same oligonucleotides) in vector pGEM-T. All probes were 32P-labelled with a Ready to GoTM DNA labelling kit (Amersham Biosciences) using [α-32P]-dCTP. Images of radioactive filters and gels were obtained and quantified with a Cyclone storage phosphor system and OptiQuant image analysis software (Packard). Quantification was performed using windows covering all hybridization signals for each sample. The data obtained were then normalized using the rnpB signal. The rnpB gene encodes a stable RNA (Vioque, 1997) used as an RNA loading and transfer control. All the Northern analyses were performed with four or five independent RNA preparations (two independent preparations in the case of heterocyst RNA), and a representative hybridization is shown in each case.

DNA isolation and analysis

Total DNA from Anabaena sp. PCC 7120 and its derivatives was isolated as previously described (Cai and Wolk, 1990). For sequencing ladders used in primer extension analysis, sequencing was carried out by the dideoxy chain-termination method, using a T7·SequencingTM kit (Amersham Biosciences) and [α-35S]-thio-dATP. DNA fragments were purified from agarose gels with the Geneclean II kit (BIO 101). Plasmid isolation from Escherichia coli, transformation of E. coli, digestion of DNA with restriction endonucleases, ligation with T4 ligase, and PCR were performed by standard procedures (Sambrook et al., 1989; Ausubel et al., 2001).

Mutant construction

Plasmid pCSAV131 was generated to inactivate the coxB2 gene. This plasmid contains a 1.89 kb DNA fragment from the coxB2 region of Anabaena sp. PCC 7120 amplified by PCR using oligonucleotides CB2-6 (7 in Fig. 1) (corresponding to nucleotides − 770 to − 750 with respect to the translation start of coxB2) and CB2-5 (10 in Fig. 1) (see above) and genomic DNA from strain PCC 7120 as template. Polymerase chain reaction products were cloned in vector pGEM-T (Promega). For the generation of a coxB2 mutant, a 0.62 kb EcoNI fragment from pCSAV131 was substituted, after Klenow-enzyme filling in of the vector ends, by EcoRV-SmaI-ended 2 kb SmR SpR gene cassette C.S3 (Elhai and Wolk, 1988a) and by SmaI-ended 1.5 kb KmR gene-cassette C.K3 (Elhai and Wolk, 1988a) rendering plasmids pCSAV138 and pCSAV137 respectively. Then, the PvuII fragment from pCSAV138 was ligated to the sacB-vector pRL278 (Black et al., 1993) digested with XbaI and filled in with Klenow enzyme rendering pCSAV140. The SpeI-ScaI fragment from pCSAV137, filled with Klenow enzyme, was ligated to the sacB-vector pRL277 digested with BglII and filled with Klenow enzyme rendering pCSAV139. The sacB gene determines sensitivity to sucrose and can be counterselected for in Anabaena sp., allowing positive selection for double recombinants (Cai and Wolk, 1990).

Plasmid pCSAV132 was generated to inactivate gene coxA3. This plasmid contains a 2.53 kb DNA fragment from the coxA3 region of Anabaena sp. PCC 7120 amplified by PCR using oligonucleotides CB3-6 (16 in Fig. 1) (see above) and CA3-2 (19 in Fig. 1) (complementary to nucleotides 1643–1624 with respect to the translation start of coxA3) and genomic DNA from strain PCC 7120 as template. Polymerase chain reaction products were cloned in vector pGEM-T (Promega). The C.S3 gene-cassette excised with EcoRV and SmaI was inserted into the Eco47III site that is present in the Anabaena DNA insert of pCSAV132, to generate plasmid pCSAV133. The PvuII fragment from pCSAV133, containing the interrupted coxA3 gene, was ligated to the sacB-vector pRL278 digested with XbaI and filled in with Klenow enzyme rendering pCSAV135.

In vitro -generated constructs bearing a gene-cassette inserted into coxB2 or coxA3 and cloned in sacB -vectors were transferred by conjugation ( Elhai and Wolk, 1988b ) to Anabaena sp. to generate strains bearing mutations in the coxB2 or/and coxA3 genomic regions. For generation of strains CSAV135, CSAV139, CSAV140 and CSAV141, E. coli HB101 containing plasmid pCSAV135, pCSAV139, pCSAV140 or pCSAV135, respectively, and helper plasmids pRL528 ( Elhai and Wolk, 1988b ) and pRL591-W45 ( Elhai et al., 1994 ) was mixed with E. coli ED8654 carrying the conjugative plasmid pRL443 and thereafter with Anabaena sp. Exconjugants were isolated ( Elhai and Wolk, 1988b ), and double recombinants were identified as clones resistant to the antibiotic for which resistance was encoded in the inserted gene cassette, resistant to sucrose, and sensitive to the antibiotic for which the resistance determinant was present in the vector portion of the transferred plasmid. Segregation of the mutant chromosomes was tested by PCR analysis, and homozygous mutant clones were selected for study. Oligonucleotides used as PCR primers for the coxB2 region were CB2-9 (corresponding to nucleotides − 567 to − 549 with respect to the translation start of coxB2 ; 8 in Fig. 1 ) and CA2-1 (complementary to nucleotides + 6 to − 17 with respect to the translation start of coxA2 ; 11 in Fig. 1 ). Oligonucleotides used as PCR primers for the coxA3 region were CA3-1 (corresponding to nucleotides 2–21 with respect to the translation start of coxA3 ; 18 in Fig. 1 ) and CA3-2 (see above; 19 in Fig. 1 ).

Cytochrome c oxidase activity

Isolated membranes from strain PCC 7120 and some mutants for determination of in vitro cytochrome c oxidase activity were prepared from 200 ml cultures of OD730 ≈ 1 grown in a shaker in an environment of 0.25% (v/v) CO2 in air. After harvesting, the filaments were resuspended in 10 ml Hepes buffer (10 mM 2-[4-(2-hydroxyethyl)-1-piperazinyl]ethanesulphonic acid (Hepes) pH 7.4; 5 mM NaCl) supplemented with 20% (mass/vol) sucrose and 100 mg lysozyme, and incubated for 30 min at 37°C. The cells were spun down, resuspended in 6 ml of ice-cold Hepes buffer per gram wet weight, and incubated on ice for 10 min. After addition of 1 mM phenylmethylsulfonyl fluoride (PMSF; 100 mM in ethanol) and 0.005% (mass/vol) DNase I, the cells were passed through a French press at 11000 p.s.i., three consecutive times. The resulting extracts were centrifuged in an SW41Ti rotor at 0°C, 40000 r.p.m. for 50 min, and the supernatant discarded. Membrane pellets were then resuspended on ice in Hepes buffer in a Potter homogenizer to a chlorophyll a concentration of 0.1–0.2 mg ml−1. The assay medium contained, in a total volume of 3 ml, membranes at about 10 µg chlorophyll a ml−1 in Hepes buffer and 10 or 20 µM horse heart cytochrome c pre-reduced with ascorbic acid. The difference between A550 and A540 was followed in a stirred cuvette at 30°C in a Varian Cary 5 spectrophotometer. Rates were determined using an E550 of 29.5 and 8.4 cm−1 mM−1 for reduced and oxidized cytochrome c respectively (Van Gelder and Slater, 1962). Chlorophyll a (Chl) content was determined by the method of Mackinney (1941).

Nitrogenase activity

Filaments from wild-type strain PCC 7120 and from its derivatives CSAV135, CSAV139 and CSAV140 were grown in BG110 medium supplemented with 10 mM NaHCO3 and bubbled with a mixture of CO2 and air (1% v/v) until they reached a chlorophyll a concentration of 3–5 µg ml−1. Then, the cultures were bubbled for 24 h in the light with a mixture of 1% CO2 in argon. After this period, a sample of 30 ml from each culture was incubated bubbling with a mixture of CO2 and air (1% v/v) in the light or in darkness for 1 h. Then, the nitrogenase activity of the filament suspensions was determined via the acetylene reduction assay under light (150 µE m−2 s−1) or dark conditions, respectively, in an atmosphere of 14% acetylene in air. Ethylene production was found to be linear under these conditions for at least 1 h. For investigation of the double mutant CSAV141, filaments of this strain and of the wild type, used as a control, were grown in BG110 medium supplemented with 8 mM NH4Cl, 16 mM TES-NaOH (pH 7.5) and 10 mM of NaHCO3 and bubbled with a mixture of CO2 and air (1% v/v) until they reached a chlorophyll a concentration of 3–5 µg·ml−1. Filaments were then washed with, and resuspended in, combined nitrogen-free medium as above and further incubated for about 24 h (until heterocysts were observed). Then, they were bubbled for 24 h with a mixture of 1% CO2 in argon and treated as described above.

Growth tests

Cultures for growth tests were grown at 30°C in the light (75 µE m−2 s−1), with shaking (80–90 r.p.m) for liquid cultures. Growth rates were estimated from the increase of protein concentration of cultures inoculated with an amount of cells containing ∼ 0.2 µg ml−1 chlorophyll a and allowed to grow until growth became light-limited (∼ three to four generations). Protein concentration was determined by a modified Lowry procedure (Markwell et al., 1978) in 0.2 ml aliquots withdrawn periodically from the cultures. The growth rate constant (µ) corresponds to ln2/td, where td represents the doubling time.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

We thank Silvia Picossi, Elvira de Olmedo and Alicia M. Muro-Pastor for help with some RNA preparations. The use of Nostoc punctiforme preliminary sequence data from the DOE Joint Genome Institute is acknowledged. This work was supported in part by Austria-Spain Acciones Integradas. Work in Seville was further supported by grants PB98-0481 and BMC2001-0509 from the Ministerio de Ciencia y Tecnología, Spain.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., and Struhl, K. (2001) Current Protocols in Molecular Biology. New York: Greene/Wiley-Interscience.
  • Black, T.A., Cai, Y., and Wolk, C.P. (1993) Spatial expression and autoregulation of hetR, a gene involved in the control of heterocyst development in Anabaena. Mol Microbiol 9: 7784.
  • Buikema, W.J., and Haselkorn, R. (1991) Characterization of a gene controlling heterocyst differentiation in the cyanobacterium Anabaena 7120. Genes Devel 5: 321330.
  • Cai, Y., and Wolk, C.P. (1990) Use of a conditionally lethal gene in Anabaena sp. strain PCC 7120 to select for double recombinants and to entrap insertion sequences. J Bacteriol 172: 31383145.
  • Elhai, J., and Cai, Y., and Wolk, C.P. (1994) Conduction of pEC22, a plasmid coding for MR.EcoT22I, mediated by a resident Tn3-like transposon, Tn5396. J Bacteriol 176: 50595067.
  • Elhai, J., and Wolk, C.P. (1988a) A versatile class of positive-selection vectors based on the nonviability of palindrome-containing plasmids that allows cloning into long polylinkers. Gene 68: 119138.
  • Elhai, J., and Wolk, C.P. (1988b) Conjugal transfer of DNA to cyanobacteria. Methods Enzymol 167: 747754.
  • Ernst, A., Black, T., Cai, Y., Panoff, J.-M., Tiwari, D.N., and Wolk, C.P. (1992) Synthesis of nitrogenase in mutants of the cyanobacterium Anabaena sp. strain PCC 7120 affected in heterocyst development or metabolism. J Bacteriol 174: 60256032.
  • Fay, P. (1992) Oxygen relations of nitrogen fixation in cyanobacteria. Microbiol Rev 56: 340373.
  • Frías, J.E., Flores, E., and Herrero, A. (1994) Requirement of the regulatory protein NtcA for the expression of nitrogen assimilation and heterocyst development genes in the cyanobacterium Anabaena sp. PCC 7120. Mol Microbiol 14: 823832.
  • Golden, J.W., Whorff, L.L., and Wiest, D.R. (1991) Independent regulation of nifHDK operon transcription and DNA rearrangement during heterocyst differentiation in the cyanobacterium Anabaena sp. strain PCC 7120. J Bacteriol 173: 70987105.
  • Häfele, U., Scherer, S., and Böger, P. (1988) Cytochrome aa3 from heterocysts of the cyanobacterium Anabaena variabilis: isolation and spectral characterization. Biochim Biophys Acta 934: 186190.
  • Herrero, A. Muro-Pastor, A.M., and Flores, E. (2001) Nitrogen control in cyanobacteria. J Bacteriol 183: 411425.
  • Howitt, C.A., and Vermaas, W.F. (1998) Quinol and cytochrome oxidases in the cyanobacterium Synechocystis sp. PCC 6803. Biochemistry 37: 1794417951.
  • Jones, K.M., and Haselkorn, R. (2002) Newly identified cytochrome c oxidase operon in the nitrogen-fixing cyanobacterium Anabaena sp. strain PCC 7120 specifically induced in heterocysts. J Bacteriol 184: 24912499.
  • Kaneko, T., Sato, S., Kotani, H., Tanaka, A., Asamizu, E., Nakamura, Y., et al. (1996) Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC6803. II. Sequence determination of the entire genome and assigment of potential protein-coding regions. DNA Res 3: 109136.
  • Kaneko, T., Nakamura, Y., Wolk, C.P., Kuritz, T., Sasamoto, S., Watanabe, A., et al. (2001) Complete genomic sequence of the filamentous nitrogen-fixing cyanobacterium Anabaena sp. strain PCC 7120 DNA Res 8: 205213.
  • Mackinney, G. (1941) Absorption of light by chlorophyll solutions. J Biol Chem 140: 315322.
  • Markwell, M.A.K., Hass, S.M., Bieber, L.L., and Tolbert, N.E. (1978) A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal Biochem 87: 206210.
  • Marsh, J.L., Erfle, M., and Wykes, E.J. (1984) The pIC plasmid and phage vectors with versatile cloning sites for recombinant selection by insertional inactivation. Gene 32: 481485.
  • Meeks, J.C., Elhai, J., Thiel, T., Potts, M., Larimer, F., Lamerdin, J., Predki, P., and Atlas, R. (2001) An overview of the genome of Nostoc punctiforme, a multicellular, symbiotic cyanobacterium. Photosynthesis Res 70: 85106.
  • Muro-Pastor, A.M., Valladares, A., Flores, E., and Herrero, A. (1999) The hetC gene is a direct target of the NtcA transcriptional regulator in cyanobacterial heterocyst development. J Bacteriol 181: 66646669.
  • Muro-Pastor, A.M., Valladares, A., Flores, E., and Herrero, A. (2002) Mutual dependence of the expression of the cell differentiation regulatory protein HetR and the global nitrogen regulator NtcA during heterocyst development. Mol Microbiol 44: 13771385.
  • Pils, D., Gregor, W., and Schmetterer, G. (1997) Evidence for in vivo activity of three distinct respiratory terminal oxidases in the cyanobacterium Synechocystis sp. strain PCC6803. FEMS Microbiol Lett 152: 8388.
  • Pils, D., and Schmetterer, G. (2001) Characterization of three bioenergetically active respiratory terminal oxidases in the cyanobacterium Synechocystis sp. strain PCC 6803. FEMS Microbiol Lett 203: 217222.
  • Rippka, R., and Waterbury, J.B. (1977) The synthesis of nitrogenase by non-heterocystous cyanobacteria. FEMS Microbiol Lett 2: 8386.
  • Rippka, R., Deruelles, J., Waterbury, J.B., Herdman, M., and Stanier, R.Y. (1979) Generic assignments, strain stories and properties of pure cultures of cyanobacteria. J Gen Microbiol 111: 161.
  • Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. 2nd edn. Cold Spring Harbor, NY: Cold Spring harbor Laboratory.
  • Schmetterer, G. (1994) Cyanobacterial respiration . In The Molecular Biology of Cyanobacteria. Bryant, D.A. (ed.). The Netherlands: Kluwer Academic Publishers, pp. 409435.
  • Schmetterer, G., Alge, D., and Gregor, W. (1994) Deletion of cytochrome c oxidase genes from the cyanobacterium Synechocystis sp. PCC 6803: evidence for alternative respiratory pathways. Photosynthesis Res 42: 4350.
  • Schmetterer, G., Valladares, A., Pils, D., Steinbach, S., Pacher, M., Muro-Pastor, A.M., Flores, E., and Herrero, A. (2001) The coxBAC operon encodes a cytochrome c oxidase required for heterotrophic growth in the cyanobacterium Anabaena variabilis strain ATCC 29413. J Bacteriol 183: 64296434.
  • Van Gelder, B.F., and Slater, E.C. (1962) The extinction coefficient of cytochrome c. Biochim Biophys Acta 58: 593595.
  • Vioque, A. (1997) The RNase P from cyanobacteria: short tandemly repeated repetitive (STRR) sequences are present within the RNase P RNA gene in heterocyst-forming cyanobacteria. Nucleic Acids Res 25: 34713477.
  • Wei, T.-F., Ramasubramanian, T.S., and Golden, J.W. (1994) Anabaena sp. strain PCC 7120 ntcA gene required for growth on nitrate and heterocyst development. J Bacteriol 176: 44734482.
  • WoIk., C.P. (2000) Heterocyst formation in Anabaena . In Prokaryotic Development. Brun, Y.V., and Shimkets, L.J. (eds). Washington, DC: American Society for Microbiology, pp. 83104.
  • Wolk, C.P., Ernst, A., and Elhai, J. (1994) Heterocyst metabolism and development. In The Molecular Biology of Cyanobacteria. Bryant, D.A. (ed.). The Netherlands: Kluwer Academic Publishers, pp. 769823.