Introduction

Cyanobacterial blooms occur in eutrophic lakes, ponds, and water reservoirs worldwide, causing an environmental problem (Carmichael 1992) that affects water quality by the production and release of cyanotoxins. These compounds are secondary metabolites that can produce toxic effects on microalgae, invertebrates, fish, birds, plants, and mammals (Neilan et al. 1999; Briand et al. 2008) and disturb the community structure in aquatic environments (Giaramida et al. 2013).

Microcystins (MCs) are the most frequently reported cyanotoxins in the freshwater species of the genera Microcystis, Planktothrix, Anabaena, Nostoc, Umezakia, and Synechocystis (Carmichael 1992; Codd et al. 1999; Pflugmacher et al. 1999; Pflugmacher and Wiegand 2001). These toxins are inhibitors of ser/thr protein phosphatases (PP-1 and PP-2) in eukaryotes and promote hepatic and gastric cancer in mammals (Falconer and Humpage 1996; Carmichael et al. 2001; Dittmann and Wiegand 2006; Moreira et al. 2013).

MCs are synthesized by a non-ribosomal pathway through a multienzyme complex known as microcystin synthetase (mcy), integrated by 10 genes arranged in two different operons (mcyA-C and mcyD-J). Transcription of the mcy gene in M. aeruginosa occurs through a central promoter located between the mcyA and mcyD genes (Tillett et al. 2000).

Numerous Microcystis species are the main producers of MCs in toxic blooms (Carmichael 1992; Sivonen and Jones 1999; Bittencourt-Oliviera et al. 2011). Various studies have demonstrated that environmental factors, such as temperature, light intensity, pH, nutrient concentration (NO−, NH4 +, PO4 3−, Fe2+), salinity, CO2 concentration, turbidity, consumers (e.g., daphnids), and xenobiotics, affect the permanence of many species of cyanobacteria in blooms, as well as the dominance of toxigenic strains and the increase in MCs production (Wicks and Thiel 1990; Rapala et al. 1997; Rapala and Sivonen 1998; Chorus and Bartram 1999; Kaebernick et al. 2000; Oh et al. 2000; Kaebernick and Neilan 2001; Long et al. 2001; Amé et al. 2003; van Gremberghe et al. 2009).

Despite the efforts to understand the possible ecological role of MCs and how the environmental factors affect their production, the available information is contradictory (Pearson et al. 2010; Sevilla et al. 2011), or ambiguous (Wood et al. 2011). Currently available information is not conclusive about the factors that affect MCs synthesis (Boopathi and Ki 2014); this could be explained, at least partially, by the fact that MCs are usually quantified using enzymatic (enzyme-linked immunosorbent assay—ELISA—or inhibition of protein phosphatases 1 and 2A) or analytical (HPLC, MS, MALDI-TOF) methods, which hinder the transcriptional evaluation of these toxins. Additionally, other studies have indicated that MCs synthesis occurs in the stationary phase of population growth (Kaebernick et al. 2000; Lyck and Christoffersen 2003). At this stage, high cell density and environmental conditions related to scum formation increase the synthesis and release of MCs, as observed by Wood et al. (2011, 2012), but we hypothesize that synthesis may occur also at the beginning of the cyanobacteria population growth and not until cell concentrations are high and scums can be formed. MCs synthesis can vary depending on the rate of population growth (Orr and Jones 1998), the studied strain (Blackburn et al. 1996; Carmichael 1997), and the differences in critical environmental factors. Moreover, Wood et al. (2012) documented responses few hours after inducing an experimental increase in cell density in a mesocosm system.

In view of the disperse and, in some cases, opposing results in the investigation of the effects of environmental factors involved in MCs production in cyanobacteria, this study was aimed at evaluating, in two strains of M. aeruginosa (one reference strain and one locally obtained through isolation), the effects of different environmental factors frequently recognized as important in cyanotoxins production, i.e. temperature, light intensity, and nutrient concentration (NaNO3, K2PO4, and FeCl3). In addition, we assessed the effect of varying the proportion of two important macronutrients (nitrogen and phosphorus), expressed as the N:P ratio, on the mcyA gene expression at 5, 10, and 15 days of growth. The purpose was to determine whether one of the tested treatments had a greater influence on the assessed endpoints and whether it could be identified as either a stress-causing or a favorable condition to induce cyanotoxins production in Microcystis.

Materials and methods

Test organisms and culture conditions

We used two MCs-producing strains of Microcystis aeruginosa (Arzate-Cárdenas et al. 2010): the wild strain Ch10 isolated from Lago Menor of the Bosque de Chapultepec (urban lake) in Mexico City, Mexico, and the reference strain UTEX LB2385, a microcystin producer strain frequently used as reference strain. When M. aeruginosa Ch10 was isolated from bloom samples, it was growing as dense colonies, but the growth under laboratory conditions is as single cells; the strain UTEX LB2385 grows as small colonies (without mucilage) and also as single cells.

Both strains were grown in 150-mL Erlenmeyer flasks with Z8 medium (5 mM NaNO3, 0.18 mM K2HPO4, 0.1 mM MgSO4, 0.25 mM CaCl2, 0.2 mM Na2CO3, 0.01 mM disodium EDTA, 0.01 mM FeCl3, and micronutrient solution at pH 10) (Kotai 1972). All cultures were incubated in an environmental chamber at 25 ± 1 °C with constant aeration and illuminated with fluorescent “daylight” lamps (25 μmol of photons m−2 s−1 and a photoperiod of 16 h:8 h (light/darkness)). These nutrient concentrations and incubation settings are termed as the normal conditions.

Two independent experiments were performed to evaluate the effect of different environmental factors and nutrient concentrations on the mcyA gene expression. Additionally, we evaluated the effect of these factors on the growth of M. aeruginosa and the MCs production as control measures to associate the gene expression with cellular growth and toxin synthesis.

Effect of temperature, illumination, and nutrient (N, P, and Fe) concentrations

Based on the N, P, and Fe concentrations of the Z8 medium and the temperature and illumination conditions of the control treatment (C), a lower and a higher value for each factor were tested. For nutrients, two concentrations were assessed: one corresponding to 0.25 (low) and the other to four times the one contained in the Z8 medium. Regarding temperature, we tested two temperatures, 20 and 30 °C, that is, one was 5 °C less and the other was 5 °C more than the normal incubation temperature (25 °C). For light, half or double the normal value of the incubation illumination was applied. The factors were changed one at a time, keeping constant the other conditions (the same as for the control), with a total of 10 treatments plus the control. Table 1 shows the values of treatments and the symbols that are used to identify them henceforth.

Table 1 Values of environmental factors and nutrient concentrations tested in strains Ch10 and UTEX LB2385 of Microcystis aeruginosa. Treatments consisted of lower (−) and higher (+) values, taking as reference the magnitudes given in the column of the control (C) series
Full size table

Each experiment was performed in duplicate, and six flasks were prepared per treatment, testing two of them at 5 days (incipient growth phase), two at 10 days (logarithmic phase), and two at 15 days of growth (stationary phase). The sampling days were determined based on the growth kinetics, previously calculated for each strain (data not shown). The biomass of these cultures was used to obtain total RNA and to determine population growth parameters (cell count and chlorophyll a concentration); the samples used for cell counting were fixed in 1 % Lugol’s iodine. Furthermore, MCs concentrations in the 15-day cultures were determined, using a commercial ELISA kit, as described below; this time was selected based on many published studies that report the stationary phase of growth as that when MCs are synthesized.

The control and treatments were inoculated with 2 × 105 M. aeruginosa cells mL−1, in a test volume of 80 mL, incubated at 25 ± 1 °C, with constant aeration, and illuminated with fluorescent “daylight” lamps (25 μmol of photons m−2 s−1 and a photoperiod of 16 h:8 h (light/darkness)).

Effect of N:P ratio

Three NaNO3 and three K2HPO4 concentrations were assayed, in all possible combinations of both nutrients according to a bi-factorial design, to obtain a total of nine treatments, whose values and nomenclature are shown in Table 2. The test values were established based on the N and P concentrations of the Z8 medium (5 mM NaNO3 and 0.18 mM K2HPO4), which has a normal ratio of 28:1 (control treatment), without changing the concentration of the other nutrients of the Z8 medium. One higher concentration, equivalent to four times the Z8 value, and one lower concentration, equivalent to one-fourth of the normal concentration, of both nutrients were selected. According to this design, the normal ratio (28:1) was maintained in one treatment having four times the concentration and the other with 0.25 of the normal concentration. To differentiate from the normal concentrations in all the ratios, the sign (+) or (−) was added to indicate proportions with high or low N and P concentrations, respectively. The number of replicates per treatment and the processing of samples were as described in the previous experiment.

Table 2 Nitrogen (NaNO3) and phosphorus (K2HPO4) concentrations and N:P ratios evaluated in Microcystis aeruginosa strains Ch10 and UTEX LB2385
Full size table

Growth of M. aeruginosa

Cell density was determined at 5, 10, and 15 incubation days counting in a Neubauer chamber, after breaking up the M. aeruginosa colonies by heating at 60 °C for 20 min according to Bernard et al. (2004). In addition, we determined the amount of chlorophyll a (Chl a) per cell through extraction with 90 % acetone using a tissue homogenizer to break the cells, maintaining the homogenates at 4 °C for 24 h protected from light. Afterward, the samples were centrifuged at 5800g to eliminate cell debris and the absorbance of the extracts was measured at 664, 645, and 630 nm. The Chl a concentration was calculated according to Strickland and Parsons (1972); resulting values were divided by the cell density to determine the content per cell.

RNA extraction and cDNA synthesis

The biomass contained in 20 mL of culture in each flask per treatment was separated by centrifugation and handled as a compound sample for total RNA extraction. The pellet was macerated with 200 µL of TRI Reagent® solution (Applied Biosystems, Foster City, CA, USA) with a Teflon pestle in 1.5-mL vials; afterward, 800 µL of TRI Reagent® solution was added, and three cycles of freezing with liquid nitrogen, and thawing at room temperature were performed to disrupt the cells. Finally, complete RNA extraction was accomplished with the RiboPureTM kit (Ambion® by Life Technologies, Austin, TX, USA). RNA integrity and non-DNA contamination were verified through electrophoresis in a 1.5 % agarose gel; the RNA concentration was quantified by means of the A260/A280 ratio. Synthesis of the RNA complementary chain (cDNA) was achieved with the High Capacity RNA-to-cDNA kit (Applied Biosystems, Foster City, CA, USA) following the manufacturer’s protocol and using 2 µg of total RNA in a 20-µL reaction. RNA and cDNA samples were stored at −70 °C until their further processing.

RT-qPCR

The primers and the hydrolysis probe (TaqMan®) of the genes mcyA (Pineda-Mendoza et al. 2014) and PC (reference gene) (Kurmayer and Kutzenberger, 2003) are shown in Table 3.

Table 3 Primers and TaqMan® probes used in qPCR
Full size table

The reaction mixture was set to a final volume of 20 µL, containing 10 µL of TaqMan® universal PCR Master Mix II, no UNG (Applied Biosystems) and 5 µL of the diluted cDNA. For the reference gene, the primers were adjusted to a concentration of 300 nM and the probe to 100 nM, whereas for the target gene (mcyA) adjustments were 900 and 250 nM for the primers and the probe, respectively. RT-qPCR for each treatment was performed in triplicate.

The qPCR were performed in a StepOne® real-time thermocycler using StepOne version 2.1 software (Applied Biosystems) (Supplementary data S1) and the standard ramp speed.

The calibration curves for the reference and target genes were constructed with five cDNA dilutions (1.0, 0.33, 0.11, 0.037, and 0.012 µg µL−1) starting from the initial concentration with a dilution factor of 3. The efficiency of RT-qPCR and the validation of both genes were assessed using a linear regression analysis from the average values obtained in three replicates of the quantification cycles (Cq or CT). The efficiency value was −3.33, and R2 (determination coefficient) was 0.99, indicating an almost perfect fit in the regression line. To assess the specificity of the primers, the amplicons were analyzed by electrophoresis on 1.5 % agarose gels.

Relative gene expression values for each treatment were determined with the C T (\(2^{{ - {\Delta \Delta }C_{\text{T}} }}\)) method (Livak and Schmittgen 2001).

Quantification of microcystins

Cyanotoxins were extracted from 40 mL samples obtained from the 15-day growth cultures, through centrifugation at 12,000g for 15 min; 10 mg of the pellet was weighed, macerated on dry ice, resuspended in 1 mL of 70 % methanol, vigorously vortexed for 1 min, and centrifuged at 13,000g for 5 min to eliminate cellular debris. This procedure was repeated five times for each sample; finally, methanol was evaporated and the extracts were resuspended in deionized water to obtain the MCs concentration contained in 1 mg biomass per mL.

The MC concentrations in each treatment were determined in the extracts by means of ELISA with the QuantiPlate kit for Microcystins (Envirologix™), following the manufacturer’s instructions. This method quantifies five different cyanotoxins (microcystin −LR, −LA, −RR, and −YR, and nodularin), but results are expressed as MC–LR equivalents. The intracellular MCs concentration was divided by the cell density to express results as amount of MC–LR per cell (fg cell−1).

Statistical analyses

To determine significant differences in mcyA gene expression, two-way analysis of variance (ANOVA) was applied, using as factors the treatments and the time of exposure. When ANOVA was significant, Dunnett’s and Tukey’s post hoc comparisons were performed (Supplementary data S2). Values of \(2^{{ - {\Delta \Delta }C_{\text{T}} }}\) were transformed to log2 for their graphical representation and for statistical analysis. SigmaPlot software version 12.0 was used for all the analyses.

Results

Effect of temperature, illumination, and nutrient (N, P, and Fe) concentrations

The cell densities of both M. aeruginosa strains, recorded at days 5, 10, and 15, are shown in Supplementary data S3; with these data, population growth curves were constructed. Ascending trends are observed in most cases, but in some treatments stabilization was observed on day 15. Furthermore, differences in the response of both strains to the same factors are observed. The greatest growth was observed in the UTEX LB2385 strain with treatment −L (72,750,000 cell mL−1) and was higher than the maximum achieved in strain Ch10 (treatment −P: 54,500,000 cell mL−1). Although no clear trends in population growth related to the tested environmental factors are depicted, differences in the response of both strains to the same factor were observed. Notwithstanding the changes in nutrients concentration were not measured during the experiment, the comparatively reduced growth in the lower N treatments (−N) was not provoked by N depletion, because formulation of the Z8 medium has an excess of nutrients; rather, this result could be explained as an unbalance in the nutrient proportions.

Chl a content per cell is shown in Fig. 1; average values for the control (±standard error limits) are shown as horizontal lines to ease comparisons. In general, values were variable for both the treatments and the control at the three times observed, so no clear tendencies can be recognized. For example, compared to the control, Ch10 strain cells showed higher Chl a concentrations per cell on day 5 in treatment −Fe; on day 10, in +P and −L, Chl a concentrations were higher; whereas, on day 15, in +N and +F, Chl a concentrations were higher than in the control. Noteworthy was the gradual reduction in Chl a content per cell as the cultures matured.

Fig. 1
figure 1

Chlorophyll a determined at 0, 5, 10, and 15 days of incubation in strains Ch10 and UTEX LB2385 of Microcystis aeruginosa under different values of some environmental conditions and nutrient concentrations. Average values ± standard error limits are shown. The average value of the control treatment is shown as solid line, and the ± standard error bars are shown as dotted line

Full size image

For the UTEX LB2385 strain, higher concentration of Chl a per cell as compared to the control was attained on days 5 for +P: At day 10, all the treatments had lower values, and at day 15, all the treatments induced higher Chl a concentrations (Fig. 1). Results indicate that no correlation between cell density and Chl a content can be established for both strains.

Responses of both M. aeruginosa strains were greatly variable regarding population growth and chlorophyll content per cell. This can be explained as a consequence of the intrinsic variability of the measured endpoints, but sampling at only three times did not enable us to determine variations along growth kinetics. Because these results are not conclusive, they are important and were obtained as additional support to the main aim of the present study, which was to determine how these factors could influence the expression of one of the main genes involved in microcystins synthesis. These results also support the assumption that different strains of M. aeruginosa are affected in different ways by the same environmental factors.

Expression of the mcyA gene and microcystin concentrations

The relative expression of mcyA gene of strain Ch10 (Fig. 2) revealed up-regulation in all the treatments in which illumination conditions and nutrient concentrations were varied at all the observed times (except for +F at day 10), whereas down-regulation was produced by both temperature conditions. Two-way ANOVA revealed that mcyA gene expression was significantly affected by the different environmental factors tested (F (10,346) = 5563, P < 0.001) and also by the incubation times (F (3,346) = 4500, P < 0.001), but because the interaction between both factors was also significant (F (30,346) = 1076, P < 0.001), no multiple comparison was possible.

Fig. 2
figure 2

mcyA gene expression (mean ± SE) recorded on days 0, 5, 10, and 15 of incubation of Microcystis aeruginosa (Ch10 and UTEX LB2385 strains), under different environmental conditions and nutrient concentrations. mcyA gene expression was normalized using the reference gene PC

Full size image

The mcyA gene of strain UTEX LB2385 was also up-regulated in most of the treatments, mainly at 10 and 15 days incubation, as shown in Fig. 2. Down-regulation was observed in some treatments; specifically for temperature, no clear pattern is depicted, and this situation is quite different from that observed for strain Ch10. Two-way ANOVA evidenced significant differences in mcyA gene expression caused by environmental factors (F (10,346) = 145, P < 0.001) and also by incubation periods (F (3,346) = 1176, P < 0.001), but significant interaction between both factors was detected also (F (30,346) = 551, P < 0.001) so no multiple comparison analysis was possible.

As shown in Table 4, MCs concentration in the control treatments was significantly higher in Ch10 (62.8 fg cell−1), and in this strain, values ranged from 27.2 to 122.2 fg cell−1; only in treatment −L, concentration was significantly lower than the respective control. In strain UTEX LB2385, MC values ranged from 26.6 to 138.6 fg cell−1, but only in the treatments –N, +L, and –T, concentrations higher than the control were recorded. No pattern was detected, but varying conditions from that established for the controls modify MCs production, and differences in strain responses were also observed. Furthermore, mcyA gene expression at day 15 was related to MCs production, but exceptions were documented.

Table 4 Intracellular content of microcystins (fg cell−1 as MC–LR equivalent) quantified in Microcystis aeruginosa (strains Ch10 and UTEX LB2385) at 15-day exposure to different values of environmental factors and nutrient concentrations
Full size table

Effect of N:P ratio

The cell density values are shown in Supplementary data S3. Results show again that the strains responded differently to the same treatments, but in most cases, cell density at day 15 was lower than that recorded for the control. Maximum growth was promoted by different N:P ratios, involving increment or reduction of phosphorus. For example, in strain Ch10, the maximum cell density was observed when P concentration was reduced to a quarter of that contained in the control [treatment 111:1 (−)]; whereas for strain UTEX LB2385, the highest value was for treatment containing four times the P concentration in the control [treatment 7:1 (+)].

Cellular chlorophyll a concentration reduced from day 5 to 15 in both strains as shown in Fig. 3; this condition could indicate aging in cultures, certainly not related to nutrients consumption but to limits in growth imposed by the culture system (batch). As a general trend, Chl a concentration was significantly lower than in the control in most of the N:P ratios at the three observed times; but, noteworthily, the effects of treatments were different in both strains (Fig. 3).

Fig. 3
figure 3

Chlorophyll a concentration determined at 0, 5, 10, and 15 days of incubation in Ch10 and UTEX LB2385 strains of Microcystis aeruginosa at different NaNO3 and K2HPO4 concentrations (N:P ratio). Average values ± standard error bars are shown. The average value of the control treatment is shown as solid line, and the ± standard error bars are shown as dotted line

Full size image

Expression of the mcyA gene and microcystins concentration

The relative expression graph shows that the mcyA gene in strain Ch10 was up-regulated at the three exposure times in most of the treatments (Fig. 4); in the control (28:1), the mcyA gene expression was up-regulated at day 5, but down-regulated at days 10 and 15. Two-way ANOVA revealed significant differences in expression levels between treatments (F (8,285) = 1487, P < 0.001) and incubation times (F (3,285) = 302, P < 0.001), but the interaction between these two factors was also significant (F (24,285) = 73, P < 0.001) so no multiple comparisons test was possible.

Fig. 4
figure 4

mcyA gene expression (mean ± SE) recorded on days 0, 5, 10, and 15 during development of Ch10 and UTEX LB2385 strains of Microcystis aeruginosa under different NaNO3 and K2HPO4 concentrations and N:P ratios. mcyA gene expression was normalized using the reference gene PC

Full size image

The expression of mcyA gene of the UTEX LB2385 strain was up-regulated at all exposure times in most of the treatments and in the control; down-regulated expression was only observed at day 5 in some N:P ratios (Fig. 4). Two-way ANOVA revealed significant differences between treatments (F (8,285) = 524, P < 0.001) and observation times (F (3,285) = 5405, P < 0.001), as well as in the interaction between these two factors (F (24,285) = 418, P < 0.001).

In strain Ch10, the MC concentrations ranged from 27.2 to 220.9 fg cell−1 (Table 5); significantly different concentrations of MCs with respect to the control were recorded only in the treatments 28:1 (−) and 111:1 (−). In strain UTEX LB2385, the values ranged from 41.8 to 195.9 fg cell−1, but the MCs concentration was significantly higher with respect to the control only in the ratio 28:1 (−). Again, strain-specific differences were observed.

Table 5 Intracellular content of microcystins (fg cell−1 as MC–LR equivalent) quantified in Microcystis aeruginosa (strains Ch10 and UTEX LB2385) at 15-day exposure to different nitrogen/phosphorus ratios
Full size table

Discussion

MCs production by cyanobacteria dates back to their phylogenetic origin and could be recognized as a response that has allowed them to persist and dominate in different environments (Rantala et al. 2004). Many efforts have been made to understand the intra- and inter-cellular mechanisms associated with the production of these metabolites, but no clear results have been obtained. From an ecological point of view, it has been noted that diverse environmental, nutritional, and biological factors can regulate and positively or negatively control the production of these toxins (Kotak et al. 2000). Our results indicate that variation in environmental factors, nutrient concentrations, and N:P ratios affects differently the mcyA gene expression and microcystins production, depending on the M. aeruginosa strain and the incubation time. Despite that no general trends could be established, all the variables here tested produced changes in the response of the tested toxigenic strains.

Previous studies suggested that an increase in nitrogen concentration promotes MCs production during blooms (Gobler et al. 2007) and that reduced concentrations of this nutrient limit the population growth of cyanobacteria (Sevilla et al. 2010), as well as microcystins production (Gobler et al. 2007). The latter was partially observed in the present study. The growth of M. aeruginosa Ch10 strain in low nitrogen concentration was lower than in the control, but up-regulation of the mcyA gene and the MCs concentration per cell were higher in the Ch10 and UTEX LB2385 strains, suggesting that this acted as a stimulatory stress factor to produce toxins. These results are contrary to those reported by Gobler et al. (2007), but agree with the observation by Ginn et al. (2010), who found that transcription of the mcyB and ntcA genes increases in nitrogen stress conditions.

Nitrogen limitation in N:P ratios affected the growth of this cyanobacterium, but did not affect MCs synthesis. These results are in agreement with Smith (1983) and Fujimoto et al. (1997), who reported that N availability does not affect MCs synthesis, and with Sevilla et al. (2010), who reported that the transcription of mcyD gene in M. aeruginosa is not disturbed. Differences observed between strains in this study could be explained as a strain-specific response, indicating also the high plasticity observed in this cyanobacterium under diverse conditions.

Phosphorus is the limiting nutrient in primary productivity in freshwater lakes (Kaebernick and Neilan 2001); hence, small changes in the concentration of this nutrient influence significantly the growth of cyanobacteria, as it has been observed in blooms dominated by Microcystis (Vézie et al. 2002). High P concentrations induced an increase in MC’s synthesis in Microcystis strains under laboratory (Sivonen 1990; Vézie et al. 2002) and natural (Rinta-Kanto et al. 2009) conditions. Our results show that the mcyA gene expression was up-regulated independently from the P concentration; however, the highest MC values in both strains were recorded in the treatments with the lowest P and N concentrations. Oh et al. (2000) reported that phosphorus limitation decreases the growth of M. aeruginosa, but increases the MCs content. Certainly, P deficiency affects the growth of any photosynthesizing organisms (Dai et al. 2008); however, its effect on MCs production is not clear.

Apparently, MCs production is not regulated by the environmental concentration of this nutrient, because these toxins can be produced in different strains of M. aeruginosa, Anabaena, and Oscillatoria, in both high and low P concentrations (Utkilen and Gjølme 1995; Chorus and Bartram 1999; Jacoby et al. 2000; Vézie et al. 2002; Kameyama et al. 2002). Furthermore, no clear association between mcy-cluster genes expression and toxins production has been established; however, our results show that higher MCs production is related to an increase in mcyA gene expression. This is a partial result because we did not analyze the production of toxins at all expression times, and only one gene of the mcy-cluster was assessed. Jonasson et al. (2008) demonstrated in the Nodularia spumigena strain AV1 that phosphate starvation has a differential effect on the expression of the nine genes of the nda cluster, but not on the intracellular and extracellular concentration of nodularin. Hence, further studies should be done to evaluate the expression of other mcy-cluster genes and the MCs production in order to establish any association between them in cyanobacteria.

Regarding the influence of micronutrients, Kosakowska et al. (2007) reported that 10 µM of iron is required to maintain Chl a synthesis and a positive growth rate in M. aeruginosa. Nevertheless, we observed that the strains Ch10 and UTEX LB2385 grew adequately under lower Fe concentrations. In addition, the induction of the mcyA gene and intracellular MC values were higher in low Fe concentrations than in high concentration. This result agrees with previous studies indicating that Fe-limiting conditions increase the transcription rate of mcy genes and the production of cyanotoxins (Lukaĉ and Aegerter 1993; Lyck et al. 1996; Amé et al. 2003; Alexova et al. 2011a), resulting from the dissociation of the FurA protein, whose function is to repress the expression of mcy genes in the presence of Fe (Martin-Luna et al. 2006, 2011). Notwithstanding, other studies report an increase in MCs synthesis associated with a high Fe content in the medium (Utkilen and Gjølme 1995; Bickel and Lyck 2001; Amé and Wunderlin 2005); but, for a complete comparison, the specific condition of the other environmental factors should be reviewed.

Our results also indicate that expression of the mcyA gene can change during population growth. Whereas the Ch10 strain showed a marked up-regulation from the start of the exponential phase (5 days), similar to that observed by Rueckert and Cary (2009) and Alexova et al. (2011b) for the mcyE gene, in strain UTEX LB2385, the mcyA gene was up-regulated in the late exponential phase (10 days), similar to the results reported by Welker et al. (2006). Likewise, the difference in these results can be attributed to the metabolic plasticity of Microcystis strains documented by Chorus (2001).

It is not known whether temperature activates or deactivates the genes or proteins that regulate microcystin synthetase activity; however, our results suggest that the effect of temperature is not tightly related to mcyA gene expression, intracellular MCs production, or the population growth of both strains. Gene expression was low at the two temperatures studied, whereas the intracellular MCs concentration was comparatively higher at the lower temperature, which could be expected as a less favorable condition for the growth of both strains, at least in tropical and subtropical latitudes. At low temperature, Sivonen (1990), Watanabe (1996), and Lehman et al. (2008) documented high MC concentrations in different cyanobacteria. These results coincide with observations in Microcystis spp. (Wang et al. 2010) and Oscillatoria agardhii (Sivonen 1990), for which the optimal temperature for MCs production was below 23 °C, but the optimal temperature for growth was above 25 °C (Sivonen 1990; Lehman et al. 2008); thus, optimal population growth might not necessarily be related to the synthesis of these secondary metabolites. Despite the aforementioned, and unlike our results, other studies report that relatively high temperatures (26 and 32 °C) favor an increase in the MCs concentration (Amé and Wunderlin 2005; Dziallas and Grossart 2011).

Temperature variations have been related to changes in the composition of MCs; at temperatures below 25 °C, the Anabaena sp. strains studied by Rapala et al. (1997) and Msagati et al. (2006) produced MC–LR, whereas above 25 °C, they produced MC–RR. In contrast, Amé and Wunderlin (2005) observed a higher proportion of MC–RR at 20 °C and of MC–LR at 28 °C in cultures of concentrated, natural M. aeruginosa populations. In our study, we did not detect different MCs because the quantification method used could not discriminate MC isoforms; thus, no comparison is possible. Additional assays in a wide range of temperatures are necessary to positively conclude on the effect of this factor.

The growth diminution of M. aeruginosa observed in this study at the highest light intensity agrees with observations made in Oscillatoria agardhii (Sivonen 1990). This result could be explained by the saturation of the light uptake complex, inducing the generation of free radicals and photo-inhibition (Walsh et al. 1997; Zilliges et al. 2011), which would lead to a reduction in cell growth. We also observed that the highest light intensity induced up-regulation of the gene mcyA in both M. aeruginosa strains, which agrees with previous studies reporting high transcription levels for mcyB (16–68 µmol photons m−2 s−1) (Kaebernick et al. 2000) and mcyD (16–400 µmol photons m−2 s−1) (Kaebernick et al. 2000; Sevilla et al. 2012). It has been suggested that microcystin synthetase is controlled by the amount of light through regulating sequences located in the promoter region between operons mcyA-C and mcyD-J (Kaebernick et al. 2000). The mcyA gene induction here observed explains the high intracellular MC concentrations recorded in both strains with the high light intensity treatment. These results agree with those reported in Microcystis (Utkilen and Gjølme 1995; Rapala et al. 1997; Kaebernick et al. 2000) and Planktothrix agardhii (Tonk et al. 2005), but differ from those obtained with Anabaena and Oscillatoria (Sivonen 1990; Rapala et al. 1997). Additional assays could be required to test whether the light effect may also be a species-specific trait in cyanobacteria.

The effect of the N:P ratio on the growth of cyanobacteria and the synthesis of MCs (mainly in Microcystis) is not conclusive; in some studies, high ratios increase the synthesis of cyanotoxins (Van der Westhuizen and Eloff 1985; Vézie et al. 2002; Downing et al. 2005), whereas other authors report the opposite (Bulgakov and Levich 1999; Jacoby et al. 2000; Havens et al. 2003). The N:P ratio in natural environments can induce a higher production of MCs and, according to Downing and McCauley (1992), the low N:P ratio could be the consequence and not the cause of bloom formation. Here, the mcyA gene expression and the intracellular MCs concentration did not show a clear correlating tendency with the N:P ratio in the two M. aeruginosa strains, concordant with observations by Marinho and Azevedo (2007) on M. aeruginosa and Aulacoseira distans. The higher MCs synthesis recorded in UTEX LB2385, when nitrogen was reduced [28:1 (−) N:P ratio], agrees with the observations of Downing et al. (2005), suggesting that MC synthesis depends on nitrogen concentration. Similar results were reported by Dai et al. (2008) for other strains of M. aeruginosa; this could indicate that N but not P has a major influence on MCs synthesis for some cyanobacterial strains.

The effects on population growth and MCs production in both M. aeruginosa strains observed in the treatments with different nutrients concentrations and N:P ratios led us to conclude that this cyanobacterium displays stress responses at limiting N or P concentration, whereas in other ratios, the effect is clearly stimulatory, depending on the strain. We could expect that different species and different strains of the same species could react in a different manner, confirming the variability of responses to the effects of important environmental factors.

Based on the experimental results here observed and comparing with those obtained by Wood et al. (2011), in natural conditions, the regulation of genes in the cluster mcy is more complex than frequently assumed and cannot be attributed only to one environmental variable. Future research should be aimed at evaluating the expression of the mcy-cluster genes related to the use of MCs as siderophores, quorum sensing, and in the photosynthesis process, and not just as allelopathic substances or as a defense mechanism against predators. The results observed in a recently isolated strain from an urban lake with high predominance of Microcystis spp. blooms (Ch10), compared to the reference strain (UTEX LB2385), which was isolated more than 50 years ago, indicate that toxigenic strains kept their capability to synthesize MCs, despite the general thought that these secondary metabolites are only produced under selective pressure. This fact indicates that cyanotoxins could be involved in other cell metabolic functions.

Conclusions

Under the conditions here studied, the mcyA gene expression in M. aeruginosa responded to light, temperature, and the concentration of some nutrients (N, P, and Fe), but not to N:P ratios. The differential gene expression observed in the two strains suggests that their adaptive strategies to the environment are complex, because, in general, some strains can respond immediately to environmental changes (Ch10) and others do more slowly (UTEX LB2385). This strategy allows understanding, at least partially, the successful permanence and dominance of the genus Microcystis in blooms, because its species can tolerate environmental variability and react differently to the changes in the aquatic environment.

This strategy could be a consequence of the large physiological plasticity developed among the species of this genus. In addition, mcyA gene expression in early growth stages suggests that MCs production could be important for the development of cyanobacteria. The capability of producing cyanotoxins and the way that this synthesis is promoted by environmental factors were different for both strains, which display variable responses under the same culture conditions. In general, reduced concentrations of nitrogen and iron, low temperature, and high light intensity were the conditions in which a significant increase in MCs concentration was observed. Furthermore, it is important to study the factors that affect the microcystins synthesis to elucidate the molecular mechanisms by which these secondary metabolites are regulated and produced.