Expression of the nodularin synthetase genes in the Baltic Sea bloom-former cyanobacterium Nodularia spumigena strain AV1


  • Editor: Patricia Sobecky

Correspondence: Rehab El-Shehawy, Department of Botany, Stockholm University, Lilla Frescativägen 5, SE-106 91 Stockholm, Sweden. Tel.: +46 8 163918; fax: +46 8 165525; e-mail:


Cyanobacterial blooms in the Baltic Sea are a common phenomenon and are formed by the heterocystous, filamentous species Nodularia spumigena. The toxicity of these blooms is attributed to the hepatotoxin nodularin, produced by N. spumigena. Little is known regarding the regulatory mechanisms or environmental signaling that control nodularin production. Here we report the characterization of the transcriptional expression pattern of the nodularin synthetase gene cluster (nda) during phosphate depletion, and nitrogen supplementation. Real-time PCR analysis of these genes revealed that while cells continuously expressed the nda cluster, the expression of all nda genes increased when cells were subjected to phosphate depletion, and decreased in the presence of ammonium. In contrast to the shifts in expression, the intracellular and extracellular nodularin concentrations did not vary significantly during the treatments.


Efficient management of the Baltic Sea requires a deep understanding of the biology of the organisms that have a significant impact on the ecology of this habitat, such as Cyanobacteria. Cyanobacteria form blooms in the Baltic Sea, which are dominated by members of two genera: Aphanizomenon and Nodularia (Stal et al., 2003). The toxicity of the cyanobacterial blooms in the Baltic Sea (Sivonen et al., 1989a, b) causes economic losses to the surrounding societies due to disturbance in tourism and death of fish and domestic animals drinking the water. The blooms typically occur in the summer but predicting when the blooms will develop has proved difficult.

In addition to environmental conditions, phosphate (P-PO4) availability is considered to be one of the most important factors for growth and nitrogen (N2) fixation of the bloom-forming Cyanobacteria in the Baltic Sea (Degerholm et al., 2006). Phosphate is not only required for the synthesis of ATP and nucleotides but also for the functional regulation of protein activity via phosphorylation.

Cyanobacteria living in marine habitats have developed several strategies to cope with the low concentrations of dissolved inorganic phosphate (DIP) and dissolved organic phosphate (DOP). Such strategies include the synthesis of high-affinity transport systems and the production of enzymes that hydrolyze DOP. In the unicellular Cyanobacteria Synechococcus, Prochlorococcus, and Crocosphaera watsonii, the gene cluster pstSCAB encoding the high-affinity phosphate transport was characterized (Moore, 2005; Su et al., 2005; Dyhrman & Haley, 2006). In the unicellular cyanobacterium Synechocystis sp. strain 6803, the expression of pstS was shown to increase under phosphate-deficient conditions (Suzuki et al., 2004). We therefore selected pstS as a molecular marker for phosphate starvation.

The ability of several cyanobacterial genera, such as Nodularia spumigena, to fix atmospheric nitrogen (N2) confers on them a competitive advantage over nonfixing coexisting organisms in aquatic waters. Nitrogen fixation is an energy-demanding process, and nitrogen-fixing Cyanobacteria are able to alternatively use other nitrogen sources, such as ammonia and nitrate, as available in the environment.

Nitrogenase is the enzyme responsible for nitrogen fixation, and nifH, the gene encoding the dinitrogenase reductase subunit of nitrogenase, is often used as a marker for nitrogen-fixation activity (El-Shehawy et al., 2003; Boström et al., 2007; Short & Zehr, 2007).

Phylogenetic analyses have demonstrated that bloom toxicity is restricted to the phytoplankton N. spumigena (Laamanen et al., 2001; Moffitt & Neilan, 2001; Lyra et al., 2005), while the benthic Nodularia harveyana and Nodularia sphaerocarpa were demonstrated to be nontoxic (Lyra et al., 2005), with the exception of N. harveyana PCC7804 (Beattie et al., 2000; Saito et al., 2001). The toxicity of N. spumigena is due to the production of nodularin (Sivonen et al., 1989a, b), which is a potent hepatotoxin and putative tumor promoter (Honkanen et al., 1995; Sivonen & Jones, 1999).

The biological function of nodularin in Cyanobacteria is not understood, as are the conditions that promote its production. A number of studies have investigated the effects of nutrients, temperature, light, and salinity on Cyanobacteria growth and nodularin production (Lehtimäki et al., 1994, 1997; Blackburn et al., 1996; Repka et al., 2001; Moisander et al., 2002; Stolte et al., 2002; Hobson & Fallowfield, 2003; Mazur-Marzec et al., 2005). In general, nodularin production has been found to be highest under optimal growth conditions (Lehtimäki et al., 1994, 1997). However, the data from many of these studies are controversial due to the use of batch cultures, different methods of analyses, and quantification of toxin in relation to cell constituents (Long et al., 2001).

Nodularin shares similarities in structure and toxicology with microcystin, which is a heptapeptide produced by Microcystis aeruginosa as well as other cyanobacterial genera. It is a cyclic pentapeptide consisting of 3-amino-9-methoxy-2,6,8-trimethyl-10-phenyl-4,6-decadienoic acid (Adda), d-glutamic acid (d-Glu), N-methyldehydrobutyrine (MeDhb), d-erythro-β-methylaspartic acid (d-MeAsp), and l-arginine (l-Arg) (Rinehart et al., 1988; Sivonen et al., 1989a, b).

Nodularin is synthesized nonribosomally by a multifunctional enzyme complex consisting of peptide synthetase (PS), polyketide synthases (PKS) modules, and tailoring enzymes (Moffitt & Neilan, 2001, 2004). The PS-PKS is encoded by the nda gene cluster, which is composed of nine genes situated upstream and downstream of a bidirectional promoter (Moffitt & Neilan, 2004). The ndaCDEF cluster encodes enzymes that are responsible for the biosynthesis of Adda, and the ndaFGHAB cluster encodes the enzymes responsible for peptide synthesis, cyclization, and transport (Moffitt & Neilan, 2004).

In general, very few studies have addressed the molecular biology of the nda cluster. Analysis of the transcriptional pattern of these genes is critical in elucidating the regulatory factors affecting toxin biosynthesis, the upstream-signaling pathways that target toxin synthesis, and the molecular mechanism(s) by which toxin production is controlled (Dittmann & Wiegand, 2006).

In this study, we present for the first time the transcriptional pattern of the nine nda genes as determined by real-time reverse transcriptase (RT)-PCR in continuous cultures of N. spumigena strain AV1 under conditions that mimic environmental shifts (phosphate starvation and ammonia supplementation). We combined the expression studies with quantification of toxin production to utilize a biphasic approach that enables the understanding of the effect of environmental fluctuations on nodularin synthesis. In parallel with the studies under laboratory conditions, this study also presents corresponding data from environmental samples collected from blooms of summer 2005.

Materials and methods

Growth conditions

Continuous cultures of axenic N. spumigena strain AV1, previously isolated from the Baltic Sea, were grown in 1-L PC flasks (VWR) in a growth chamber with 16 h of white fluorescent light at 45 μmol m−2 s−2, and 20 °C. The cultures were grown in continuously supplied Z8XN0 medium (Sivonen et al., 1989a, b) (medium without added nitrogen referred to as N0) and bubbled with filtered air provided by an aquarium pump. The speed at which the growth medium was added to the culture was varied to sustain a constant cell density. Cell density was determined by chlorophyll a (chla) measurements according to a previous report (Meeks & Castenholz, 1971).

Phosphate depletion and ammonia supplementation

When the cultures reached a steady state of cell growth, as indicated by chla measurements, phosphate depletion was achieved by supplying the cultures with Z8XN0 containing 1/50 of the original phosphate concentration (3.6 μM K2PO4·3H2O). After 16 days of phosphate starvation, the cultures were supplied with phosphate by replenishing with fresh Z8XN0.

Ammonium supplementation was achieved by supplying the steadily growing cultures with Z8X containing 1 mM NH4Cl. When cells stopped fixing nitrogen, the cultures were filtered and resuspended into Z8XN0.

Each experiment was repeated four times.

Field samples

Samples were collected on four occasions during summer 2005, at a site near station B1 outside Askö, an island located in the archipelago south of Stockholm, Sweden (Fig. 1). Cyanobacteria were collected from surface drifts tows using a Hydrobios net (0.5-m diameter, 90-μm mesh) at noon. The samples were immediately filtered through Whatman filters (mesh-size 8.0 μm), and the Cyanobacteria washed off the filters were stored in RLT buffer, supplied with the RNeasy Plant Mini Kit (Qiagen), in liquid nitrogen until RNA extraction and further processing was performed as described for cultured samples.

Figure 1.

 Location of field station B1. (a) Himmerfjärden bay area with the small island Askö and the sampling station B1. (b) The southern part of Stockholm archipelago. (c) Map of the Baltic Sea.

RNA isolation

Cell samples were collected by filtration (Whatman, 8.0 μm) and in duplicate. Samples were then treated with RNAlater RNA stabilization buffer (Qiagen) and stored in RLT buffer at −80 °C. Before RNA extraction, acid-washed glass beads (Sigma, 212–300 μm) were added to each sample, and cells were homogenized in a FastPrep FP120 (Thermo Electron Corporation) at speed 6 for 6 × 20 s, with 1-min interval cooling on ice. Lysis of each sample was verified by light microscopy, and the glass beads were removed by centrifugation at 12 000 g for 5 min. RNA isolation was then carried out using the RNeasy Plant Mini Kit according to the manufacturer's instructions (Qiagen). During RNA isolation, the samples were treated with RNAse-free DNAse I (Qiagen) to remove contaminating DNA and analyzed using real-time PCR.

Primer design

Primers (Table 1) were designed using the primer 3 software ( The PCR products were sequenced (DNA Technology) and the resulting sequences were blasted against the genome of N. spumigena strain 9414, available on the NCBI database to verify the specificity of the PCR reaction.

Table 1.   Description of primers used in the real-time PCR reactions
Accession numberTarget sequencePrimer namePrimer sequence (5′–3′)Annealing
temperature (°C)
  • *

    Forward primer.

  • Reverse primer.

EF521176pstSpstS 1159(FP)* CTG GAA CCG TTG ACT TTG GT56
EF521177pstSpstS 1102(FP)* GCA GCA ATA CAA TTG CCA GA56
EF08798916S rRNA gene16S(Fp)* AAG CAT CGG CTA ACT56

Real-time RT-PCR

From each sample, 500 ng RNA, quantified with a spectrophotometer (NanoDrop Technologies), were used for cDNA synthesis using the iScript cDNA Synthesis Kit (Bio-Rad) as per the manufacturer's instructions. Real-time PCR was performed (Huggett et al., 2005; Wong & Medrano, 2005) in duplicate using the iCycler real-time PCR machine (Bio-Rad) and the iQ SYBR® Green Supermix (Bio-Rad). A 10-fold dilution series of N. spumigena strain AV1 genomic DNA [extracted according to the method described previously by Wilson (1998)] was used to construct standard curves. The cDNA quantities in the samples (hence mRNA) were measured relative to the standard curve and normalized to the cDNA quantities of 16S rRNA gene.

The efficiency of all real-time PCR reactions was 98±4%. The real-time PCR program was as follows: an initial denaturing step of 3 min at 95 °C, and 40 cycles of 30 s at 94 °C, 30 s at 56 °C, and 30 s at 72 °C. An end-point melt-curve analysis was generated after each run and analyzed to assure the absence of nonspecific PCR products. Melting curves generated during real-time PCR runs with field samples are presented in Fig. 2.

Figure 2.

 Expression of the nda cluster, pstS copies, and nodularin concentration during phosphate treatments. (a) mRNA expression pattern of the nda cluster, pstS 1102, and pstS 1159 in Nodularia spumigena strain AV1. SQ=starting quantity in relative arbitrary units (for details, please see the section of Materials and methods). (b) Nodularin concentrations produced by N. spumigena strain AV1. • marks the intracellular nodularin concentrations, and ○ marks the extracellular nodularin concentrations. Error bars indicate SE.

Determination of extra- and intracellular nodularin concentration

Samples for nodularin quantification were collected in duplicate. Cells were collected by filtration (PC, Whatman, 8.0 μm). Each filtrate was further passed through a 0.2-μm filter (Pall Gilman Laboratory) and used to determine extracellular nodularin concentration by HPLC/MS.

Filtered cells were resuspended in 1.2 mL of 75% methanol and then disrupted in a FastPrep FP120 at speed 5 for 3 × 20 s. Tubes were transferred to a water bath sonicator (Sonorex Superlop, Germany) and sonicated at maximum speed for 30 min at room temperature. The FastPrep/sonication step was repeated once more before centrifugation at 20 000 g for 10 min. The intracellular nodularin concentration was determined by HPLC/MS as detailed in Koskenniemi et al. (2007).

Nodularin was quantified according to a known nodularin standard, and the nodularin concentration was normalized to the chla concentration.

Other analyses

Phosphorus (SRP) concentration was determined using the molybdenum-blue assay (Strickland & Parsons, 1972). Ammonium concentration in the samples was determined using the phenol method (Eaton et al., 1995). Nitrogen-fixation activity was measured using the acetylene reduction (AR) method according to Capone (1993), with modifications described by El-Shehawy et al. (2003).

Statistical analysis

anova test was performed in order to detect any significant differences (P<0.05) among data obtained from nodularin quantification in the samples of each individual experiment.

Results and discussion

Expression of the nda cluster and concentration of nodularin during phosphate starvation

As shown in Fig. 2, both pstS copies were induced in response to phosphate starvation, and the elevation in expression was the highest after 9–13 days of starvation when the phosphate level in the medium was undetectable. When the cultures were replenished with Z8XN0 at day 16, the expression of both copies declined to their prestarvation levels of expression by day 21.

The nda cluster in the N. spumigena strain AV1 was continuously expressed under these conditions (Figs 2a and 3a). The expression of the nda cluster responded to phosphate starvation (Fig. 2a), and the observed increase in expression of the nine nda genes was 2.73-fold ±0.53 (n=8). Although this increase is relatively small, it was reproducible and consistently observed. The results established an effect of phosphate availability on the expression of the nda cluster in the N. spumigena strain AV1, and demonstrated that the nine genes exhibit an identical phosphate-dependent pattern of expression, likely due to transcription from the single bidirectional promoter. The results also showed that individual genes had different transcriptional levels, because the PCR efficiencies of all the nda genes were similar (see Materials and methods); this indicates that transcription from alternative promoters probably occurs similar to what was shown previously for microcystin (Kaebernick et al., 2002).

Figure 3.

 Expression of the nda cluster, nifH, and nodularin concentration during ammonium treatments. (a) mRNA expression pattern of the nda and cluster and nifH in Nodularia spumigena strain AV1. SQ=starting quantity in relative arbitrary units (for details, please see Materials and methods). (b) Nodularin concentrations produced by N. spumigena strain AV1. • marks the intracellular nodularin concentrations, and ○ marks extracellular nodularin concentrations. Error bars indicate SE.

Analysis of nodularin concentration under these experimental conditions revealed that the intracellular and extracellular concentrations of nodularin did not vary significantly in response to phosphate starvation (Fig. 2b). These results are in accord with the results of Repka et al. (2001), which demonstrated that in chemostat cultures of Nodularia strain GR8b, the phosphate concentration did not have a statistically significant effect on nodularin production rate measured either as nodularin cell quotas, or normalized to dry weight or to protein concentration.

Our results suggest that the cells maintain a threshold level of intracellular nodularin concentration, and that any excess nodularin (resulting from the continuous expression of the nda cluster and biosynthesis of nodularin) above the threshold level is downregulated possibly by either intracellular degradation or excretion from cells by means of a transport system, rather than leakage due to cell lysis. However, no intracellular degradation mechanism has been discovered for nodularin or microcystin thus far. In addition, the possibility that there is an unknown transcription-independent regulatory mechanism acting on the enzyme-activity level and regulating the biosynthesis and/or the maturation of nodularin cannot be excluded.

Expression of the nda cluster and toxin concentration during ammonium supplementation

As shown in Fig. 3a, expression of nifH decreased in response to ammonium supplementation in Z8X, reaching an undetectable level at day 6. When the cells were collected, washed, and replenished with Z8XN0, the nitrogen-fixation activity resumed, as seen by the resulting increase in nifH expression.

The expression of the nda cluster also varied in response to ammonium supplementation (Fig. 3a). The expression decreased by a factor of 2.91±0.9 (n=8) when nitrogen-fixation activity was undetectable, and recovered when the nitrogen-fixation activity resumed. These results demonstrate a sensitivity of expression of the nda cluster in response to ammonium supplementation.

The intracellular and extracellular concentrations of nodularin did not vary significantly in response to the treatment with ammonium-rich media (Fig. 3b), further confirming that the cells maintain a threshold intracellular level of nodularin, and pinpointing the possible involvement of a regulatory process controlling the intracellular level of nodularin.

Expression pattern of ndaF and pstS in environmental samples collected from the summer blooms of 2005

The level of available phosphate in the Baltic Sea peaks after winter and reaches the annual minimum after diatom spring blooms. In summer, the onset of warmer weather and the establishment of a thermocline promote the growth of nitrogen-fixing Cyanobacteria. When growth reaches its maximum, a phytoplankton bloom occurs that decreases the phosphate levels to even lower levels (Walve, 2002; Nausch et al., 2004; Hajdu et al., 2007). Water qualities in terms of nutrient levels and chla concentrations are published on the following webpage;

Analysis of melting curves from all PCR products generated during the real- time PCR reactions showed a sharp peak at expected Tm of the products (Fig. 4). These results indicate that all real-time PCR reactions specifically amplified the target DNA.

Figure 4.

 Melting curve analysis of ndaF, nifH, pstS 1159, pstS 1102, and 16S rRNA on field samples using Sybr green detection. (a) Melting curve analysis on ndaF. (b) Melting curve analysis on nifH. (c) Melting curve analysis on pstS 1159. (d) Melting curve analysis on pstS 1102. (e) Melting curve analysis on 16S rRNA gene.

The expression patterns of pstS, nifH, and ndaF were analyzed in bloom samples of N. spumigena collected during the summer of 2005 (Fig. 5). The data (normalized to the expression of 16S rRNA gene, see Materials and methods) showed that near the end of the bloom season, when nutrients are at their minimum levels, the cells responded by inducing the expression of both pstS copies and nifH. As expected, ndaF also increased in level of expression, confirming our above-presented laboratory data.

Figure 5.

 mRNA expression pattern of ndaF, pstS 1102, pstS 1159, and nifH in field samples collected during summer 2005. □ marks the relative expression of nifH, ▪ marks the relative expression of pstS 1159, and • mark the relative expression of pstS 1102. Columns mark the relative expression of ndaF. SQ=starting quantity in relative arbitrary units (for details, please see Materials and methods). Error bars indicate SE.

The our results presented here demonstrated several key points. First, phosphate depletion and nitrogen supplementation affected the expression levels of the nodularin synthetase genes. Second, the changes in nda expression were not accompanied by an effect on the biosynthesis of the toxin normalized to chla. Third, the cells continuously expressed the nodularin synthetase genes and maintained a threshold level of nodularin intracellularly.

Rohrlack & Hyenstrand (2007) argued that microcystin is neither subjected to export nor to intracellular breakdown. However, the authors subjected the cells to extended periods of continuous darkness and light; both are unnatural stress conditions that might have an effect on the cellular metabolism.

Codd (1995) argued that the production of toxins as major intracellular components [nodularin may compromise 1.8% of cellular dry components (Komarek et al., 1993)] is retained throughout evolution for a significant biological function, which was further supported by finding of the coupling of toxin concentration to growth rate. Orr & Jones (1998) hypothesized that microcystin is not a secondary metabolite, as generally perceived, but an essential primary intracellular nitrogen compound.

Our results are in support of Codd (1995) and Orr & Jones (1998), suggesting that nodularin, like microcystin, might be an important primary intracellular nitrogen compound retained for a certain biological function.


This study was supported by grants from the Swedish Research Council (FORMAS), The Royal Swedish Academy of Sciences (KVA), and Granholms Stiftelse to R.E.-S. and from the Academy of Finland to K.S. We thank Mr Matti Wahlsten for determining the nodularin concentrations.