Skip to main content
NCBI home page
Applied and Environmental Microbiology logoLink to Applied and Environmental Microbiology
. 2019 Oct 16;85(21):e01614-19. doi: 10.1128/AEM.01614-19

Physiological and Metabolic Responses of Freshwater and Brackish-Water Strains of Microcystis aeruginosa Acclimated to a Salinity Gradient: Insight into Salt Tolerance

Maxime Georges des Aulnois a,✉, Pauline Roux a, Amandine Caruana a, Damien Réveillon a, Enora Briand a, Fabienne Hervé a, Véronique Savar a, Myriam Bormans b, Zouher Amzil a
Editor: Haruyuki Atomic
PMCID: PMC6803299  PMID: 31444201

Blooms of Microcystis aeruginosa and the production of microcystins are major issues in eutrophic freshwater bodies. Recently, an increasing number of proliferations of M. aeruginosa in brackish water has been documented. The occurrence of both M. aeruginosa and microcystins in coastal areas represents a new threat for human and environmental health. In order to better describe the mechanisms involved in Microcystis sp. proliferation in brackish water, this study used two M. aeruginosa strains isolated from fresh and brackish waters. High salinity reduced the growth rate and microcystin production rate of M. aeruginosa. In order to cope with higher salinities, the strains accumulated different cyanobacterial compatible solutes, as well as unsaturated lipids, explaining their distinct salt tolerance.

KEYWORDS: acclimation, metabolomic, Microcystis, salinity, sucrose, trehalose, ecophysiology

ABSTRACT

Proliferation of microcystin (MC)-producing Microcystis aeruginosa in brackish waters has been described in several locations and represents a new concern for public and environmental health. While the impact of a sudden salinity increase on M. aeruginosa physiology has been studied, less is known about the mechanisms involved in salt tolerance after acclimation. This study aims to compare the physiological responses of two strains of M. aeruginosa (PCC 7820 and PCC 7806), which were isolated from contrasted environments, to increasing salinities. After acclimation, growth and MC production rates were determined and metabolomic analyses were conducted. For both strains, salinity decreased the biovolume, growth, and MC production rates and induced the accumulation of polyunsaturated lipids identified as monogalactosyldiacylglycerol. The distinct salt tolerances (7.5 and 16.9) obtained between the freshwater (PCC 7820) and the brackish-water (PCC 7806) strains suggested different strategies to cope with the osmotic pressure, as revealed by targeted and untargeted metabolomic analyses. An accumulation of trehalose as the main compatible solute was obtained in the freshwater strain, while sucrose was mainly accumulated in the brackish one. Moreover, distinct levels of glycine betaine and proline accumulation were noted. Altogether, metabolomic analysis illustrated a strain-specific response to salt tolerance, involving compatible solute production.

IMPORTANCE Blooms of Microcystis aeruginosa and the production of microcystins are major issues in eutrophic freshwater bodies. Recently, an increasing number of proliferations of M. aeruginosa in brackish water has been documented. The occurrence of both M. aeruginosa and microcystins in coastal areas represents a new threat for human and environmental health. In order to better describe the mechanisms involved in Microcystis sp. proliferation in brackish water, this study used two M. aeruginosa strains isolated from fresh and brackish waters. High salinity reduced the growth rate and microcystin production rate of M. aeruginosa. In order to cope with higher salinities, the strains accumulated different cyanobacterial compatible solutes, as well as unsaturated lipids, explaining their distinct salt tolerance.

INTRODUCTION

Due to anthropogenic activities and climate change, many freshwater ecosystems are expected to experience an increase in salinity (S) (1). This phenomenon, exacerbated by drought, hydrologic alterations, land use, and sea level rise (2–5), is predicted to favor the development and expansion of freshwater cyanobacteria versus other freshwater phytoplankton (6, 7). As many freshwater cyanobacteria produce a variety of toxins affecting both animal and human health (8), this expansion is particularly problematic, especially concerning the genus Microcystis. Indeed, Microcystis sp. is currently recognized as one of the most pervasive toxic cyanobacteria in freshwater ecosystems, as blooms have been reported on all continents but Antarctica (9). This global occurrence is explained by high genomic and phenotypic plasticity (9, 10). Therefore, the biotic and abiotic factors influencing blooms of toxic Microcystis sp. have been well-studied in freshwater ecosystems (9). However, the influence of salinity on their development was overlooked despite the observation of Microcystis sp. and the hepatotoxic microcystins (MCs) in many coastal areas (11). In most reported studies, the records of Microcystis sp. in coastal areas were due to the transfer of a Microcystis bloom across the freshwater-to-marine continuum (12–17). This transfer to high salinity conditions resulted in cell lysis and the release of soluble MCs in the water (17–19). Such occurrences of Microcystis sp. in coastal areas resulted ultimately in MC accumulation in marine organisms and represented a potential new foodborne route of exposure to human health. As a result, most experimental studies conducted on the impact of salinity on Microcystis sp. focused on salt stress experiments (19–24), especially on Microcystis aeruginosa as the main bloom-forming species among this genus. Those studies defined M. aeruginosa salinity tolerance between freshwater ecosystems to the upper limit of the oligohaline areas (5 < S < 18) (20–22, 24–27). However, salt stress experiments did not allow the determination of the potential of Microcystis sp. to form blooms in brackish waters. Less frequently, blooms of Microcystis sp. have also been reported in brackish waters (11, 28). The first experimental study on the physiology of M. aeruginosa PCC 7806 acclimated to distinct salinities (0 to 17.5) was conducted by Tonk et al. (29). This study concluded that growth and MC cellular quota of M. aeruginosa decreased above a salinity of 10. In contrast, after acclimation of isolated strains of M. aeruginosa, Mazur-Marzec et al. (30) did not find any effect of salinity (0 to 14) on the total biomass of acclimated M. aeruginosa isolates after 3 weeks of culture in controlled conditions, but they also observed a decrease in MC cellular quotas above a salinity of 7.

The salinity tolerance of cyanobacteria is mainly conditioned by the ability of cells to regulate their internal osmosis through the regulation of the intracellular ionic balance and the accumulation of compatible solutes (31). So far, several compounds were clearly identified as compatible solutes in cyanobacteria, namely, sucrose, trehalose, glucosylglycerol, glucosylglycerate, glycine betaine (GB), glutamate betaine, and proline (31). In addition, dimethylsulfoniopropionate (DMSP) has been described as a putative compatible solute in certain microalgae (32, 33). DMSP or its by-product dimethyl sulfide (DMS) has been more rarely detected in cyanobacteria. As far as we know, DMSP was reported in the marine species Microcoleus lyngbyaceus, Microcoleus chtonoplastes, Lyngbya aestuarii, and Trichodesmium erythraeum; the brackish water genus Synechocystis; and the freshwater species Aphanizomenon flos-aquae (33–35). In the genus Microcystis, the glucosylglycerol was the first-identified compatible solute in Microcystis firma (36). Concerning M. aeruginosa PCC 7806 strain specifically, Kolman et al. (37) identified the compatible solute sucrose as the main metabolite explaining its salt tolerance, and Sandrini et al. (38) showed that this strain and its non-MC-producing mutant were less sensitive to cationic stress than other freshwater isolates. The strain M. aeruginosa PCC 7806 was indeed isolated in brackish water and possesses genes involved in osmotic stress response, which are not present in the other freshwater isolates (37, 38). Recently, Tanabe et al. (28) confirmed that the salt tolerance of Microcystis sp. strains isolated from brackish water may be attributable to the presence of sucrose genes that were absent in freshwater isolates. Altogether, those studies led us to hypothesize that discrepancies in salinity tolerance between strains isolated from freshwater and brackish water could be explained by distinct physiological mechanisms involving the synthesis of compatible solutes. Based on strains acclimated to different salinities, this study used a quantitative analysis of MCs and potential compatible solutes and a nontargeted metabolomic approach to provide new insights about the potential mechanisms involved in salt tolerance to better understand the occurrence of M. aeruginosa in brackish waters.

RESULTS

Growth and salinity tolerance at contrasted salinities.

After acclimation, M. aeruginosa PCC 7820 displayed positive growth at salinities between 0.6 and 7.5, while M. aeruginosa PCC 7806 grew over a wider range of salinities up to 16.9. Only at the lowest salinity (S, 0.6), the growth of both strains reached a stationary phase after an exponential-growth phase (4 to 28 days for PCC 7820; 4 to 24 days for PCC 7806) (Fig. 1A and B). During the exponential-growth phase at salinity control, the cell concentration increased from 3.9 × 106 cells ml−1 (± 3.5 × 104) to 5.8 × 107 cells ml−1 (± 1.3 × 106) for PCC 7820 and 2.2 × 106 cells ml−1 (± 9.7 × 104) to 4.7 × 107 cells ml−1 (± 4.1 × 106) for PCC 7806. The cell concentrations of both strains increased more slowly than the control at higher salinities (Fig. 1A and B). Indeed, the growth rates of M. aeruginosa PCC 7820 and PCC 7806 (Fig. 2) differed significantly with increasing salinities (analysis of variance [ANOVA]; PCC 7820, P < 1.5 × 10−5; PCC 7806, P < 6.6 × 10−11). For the strain M. aeruginosa PCC 7820, no statistical difference was observed between control salinity at 0.6 and 3.7, but a significant decrease with control was obtained at 6.4 and 7.5. Between S = 0.6 (μ = 0.37 ± 0.03 day−1) and S = 7.5 (μ = 0.16 ± 0.03 day−1), the growth rate of this strain differed by 2.3. No significant statistical difference was observed between growth rates of M. aeruginosa PCC 7806 up to 8.4. At S = 10.7 and at higher salinities, the growth rate of M. aeruginosa was significantly lower than that of the control (Dunnett’s test, P < 0.05). The growth rate of M. aeruginosa PCC 7806 differed by 2.6 between the control condition (μ = 0.42 ± 0.02 day−1) and the highest salinity, namely, 16.9 (μ = 0.16 ± 0.02 day−1) (Fig. 2).

FIG 1.

FIG 1

Open in a new tab

Cell concentrations over time for both strains at each salinity condition. (A) Microcystis aeruginosa PCC 7820 and (B) Microcystis aeruginosa PCC 7806. Circles represent means and error bars represent the standard deviations from triplicate cultures.

FIG 2.

FIG 2

Open in a new tab

Mean growth rate obtained during the exponential-growth phase, calculated for each salinity and on each triplicate of culture. (A) Microcystis aeruginosa PCC 7820 and (B) Microcystis aeruginosa PCC 7806. *, significantly different from the control condition (salinity, 0.6), P < 0.05 (Dunnett test). Black circles represent means and error bars represent the standard deviations.

Although salinity affected the growth rate of both strains, it did not affect the maximum quantum yield of the photosystem II. The variable fluorescence (Fv)/maximum fluorescence (Fm) values recorded from cultures of M. aeruginosa PCC 7820 remained in a restricted range between 0.43 and 0.55 for all salinity conditions and sampling times. The values for M. aeruginosa PCC 7806 were also stable at each salinity between 0.29 and 0.59 (see Fig. S1 in the supplemental material). The lowest values of Fv/Fm were recorded for both strains at the lowest salinity (S, 0.6).

A comparison of the relative cell size at day 14 expressed as forward scatter fluorescence (FSC) intensity was conducted. For the strain PCC 7820, the FSC value was 6.1 × 105 ± 0.2 × 105 at salinity control and 4.8 × 105 ± 0.1 × 105 at the highest salinity of 7.5. Similarly, for the strain PCC 7806, the FSC value was 8.3 × 105 ± 0.4 × 105 at salinity control and 3.7 × 105 ± 0.04 × 105 at the highest salinity of 16.9. Relative cell size on day 14 at each salinity significantly decreased compared with the control for both strains, with smaller cells at higher salinities (ANOVA; Dunnett’s test; P < 0.05) (see Fig. S2 in the supplemental material).

Toxin profiles and toxin production of Microcystis aeruginosa.

Five different MC variants were quantified in M. aeruginosa PCC 7820 (MC-LR, MC-LY, MC-LW, MC-LF, and dmMC-LR) (Fig. 3A) and two variants in PCC 7806 (MC-LR and dmMC-LR) (Fig. 3B). Cellular quotas for each MC variant of M. aeruginosa PCC 7820 increased during the exponential growth at low and intermediate salinities (S, 0.6 to 6.4), whereas this pattern was not observed at the highest salinity (S, 7.5) where MC quotas decreased over time (Fig. 3A). The MC-LR and dmMC-LR cellular quotas of M. aeruginosa PCC 7806 remained globally constant during the entire monitoring period at all salinities except at the highest one (S, 16.9). On average for all sampling days at S =16.9, the MC-LR and dmMC-LR quotas differed by 2.3 and 2.5, respectively, compared with the control (Fig. 3B). For both strains, MC cellular quotas were significantly affected by salinity, time, and their interaction (two-way ANOVAs, P < 0.05). The proportion of each variant in total MCs for PCC 7820 changed between 1.3- and 2.7-fold over time and as salinity increased. The same range was obtained in PCC 7806 where MC-LR and dmMC-LR proportions changed by 1.2- and 2.5-fold (data not shown).

FIG 3.

FIG 3

Open in a new tab

(A) Microcystin cellular quotas (including MC-LR, MC-LW, MC-LF, MC-LY, and dmMC-LR) over time and salinities for M. aeruginosa PCC 7820 and (B) for M. aeruginosa PCC 7806 (MC-LR and dmMC-LR). Error bars represent the standard deviations (n = 3).

Extracellular MCs for M. aeruginosa PCC 7820 cultures on day 4 represented 11% (± 1.3%) of the total MC quantified for the control salinity but increased sharply with salinity (44%, 48%, and 65% at salinity 3.6, 6.4, and 7.5) (Fig. 4A). For all salinities but 7.5, the proportion of total extracellular MC variants decreased with time and the accumulation of biomass. At salinities 0.6, 3.6, and 6.4, this proportion was reduced to 1.6% (± 0.3%), 3.1% (± 1.1%), and 11% (± 1.7%), respectively, by day 24. However, at salinity 7.5, an increase in the proportion of extracellular MC variants occurred between days 14 and 24, from 42% (± 3.1%) to 86% (± 2.3%) (Fig. 4A). For M. aeruginosa PCC 7806, the proportion of MC retrieved in the culture medium was maximum 4 days after the inoculation between 16% (± 2.7%) and 68% (± 2.3%). This proportion decreased over time at all salinities (Fig. 4B).

FIG 4.

FIG 4

Open in a new tab

(A) Proportions of intracellular and extracellular MCs over time and salinities for M. aeruginosa PCC 7820 and (B) for M. aeruginosa PCC 7806.

The net production rate of total MCs in both strains decreased significantly at higher salinities (PCC 7820, ANOVA one-way, P = 5.41 × 10−8; PCC 7806, Kruskall-Wallis test, P = 0.046; data not shown). For both strains and salinities, growth rate and net MC production rate were linearly correlated according to the Pearson correlation test (PCC 7820, P = 8.8 × 10−4; PCC 7806, P = 1.1 × 10−8).

DMSP, GB, and proline cellular concentrations.

In all of the samples, the quantity of DMSP was below the detection limit (50 nM). When detected, GB and proline were quantified in both strains of M. aeruginosa (Fig. 5). In M. aeruginosa PCC 7820, GB cellular concentration increased with the salinity but decreased over time (Fig. 5A). At each sampling time, GB cellular quotas were higher at S = 7.5 than the control concentrations, ranging from 1.2 × 10−3 to 4.4 × 10−3 fmol cell−1 and 2.0 × 10−4 to 2.4 × 10−3 fmol cell−1, respectively (Fig. 5A). Similar results were obtained for PCC 7806 but with lower GB concentrations (Fig. 5B). At the salinity control and 3.6, the cellular quota of GB was below the detection limit (50 nM), while GB was quantified at salinities between 10.7 and 16.9 (Fig. 5B). On average, over the salinities and sampling times, the cellular quota of proline in M. aeruginosa PCC 7820 was three times greater than that for the strain M. aeruginosa PCC 7806 (Fig. 5A and B). For both strains, cellular quotas of proline increased with salinity, but this pattern appeared to be more intense in M. aeruginosa PCC 7820 (Fig. 5A).

FIG 5.

FIG 5

Open in a new tab

(A) GB and proline cellular quotas over time and salinities for M. aeruginosa PCC 7820 and (B) for PCC 7806. Error bars represent the standard deviations (n = 3). A lack of histogram bar means that the amount of compounds was below the detection limit.

Metabolomic analyses.

Metabolomic analyses were performed to evaluate the impact of an increase in salinity on the metabolomes of M. aeruginosa PCC 7820 and PCC 7806 and to highlight putative biomarkers of salinity acclimation for both strains. Based on a matrix of 83 compounds after data filtrations, the resulting principal-component analysis (PCA) score plot is represented in Fig. 6, and the total variance due to the two main axes accounted for 81.3%. The first principal component represented the distinction between the two strains and illustrated the intraspecific variability of M. aeruginosa (59.3%), while the second (22%) allowed for a clustering depending on salinity (Fig. 6). Hence, for each strain, after an acclimation period, the salinity of the medium induced an evolution of the metabolome of the two M. aeruginosa strains. For each strain, two groups of salinity were noted. For M. aeruginosa PCC 7806, only the highest salinity (i.e., S = 16.9) formed a distinct cluster, while the two lowest and the two highest salinities were separately grouped for M. aeruginosa PCC 7820.

FIG 6.

FIG 6

Open in a new tab

PCA score plot obtained from UPLC-HRMS profiles for both strains and all tested salinities.

In order to pinpoint metabolites that were differently expressed, we used partial least squares-discriminant analysis (PLS-DA), which is a well-known dimension reduction method for biomarker discovery in metabolomics (39, 40). It constructs a set of orthogonal components that maximize the sample covariance between the response and the linear combination of the predictor variables, thus focusing on class separation (41). Using a PLS-DA and variable importance in projection (VIP) scores, it was possible to highlight 35 and 31 compounds that significantly contributed to the discrimination of the different salinity classes for M. aeruginosa PCC 7820 and PCC 7806, respectively. Among these potential biomarkers, 16 were common (i.e., for variables having VIP scores of >1) to both strains, while 19 were specific to M. aeruginosa PCC 7820 and 15 to M. aeruginosa PCC 7806.

Only some VIPs could be identified (Fig. 7). Among those showing a decreased relative expression with increasing salinity, we observed dmMC-LR for both strains (i.e., expression pattern in agreement with targeted analysis) (Fig. 7A and B) and cyanopeptolin C but only in PCC 7806 (Fig. 7B). Concerning the biomarkers having an increased relative expression with increasing salinity, we notably confidently identified the second most important VIP in M. aeruginosa PCC 7820 as trehalose (Fig. 7A). However, the VIP corresponding to the same exact mass and retention time in M. aeruginosa PCC 7806 was sucrose. Trehalose and sucrose, having the same mass, were discriminated by their fragmentation patterns compared with pure standards. Indeed, they yielded the specific ion ratios 365.1/203.05, 365.1/183.04, and 203.05/183.04, which equaled 0.8, 2.3, and 2.7 for sucrose and 4.5, 102, and 23 for trehalose, and spectra for PCC 7806 and 7820 matched accordingly (i.e., ratios 0.7, 1.9, and 2.7 versus 3.9, 99, and 25, respectively). However, trace presence of trehalose in PCC 7806 and sucrose in PCC 7820 cannot be totally excluded. In addition, several monogalactosyldiacylglycerols (MGDGs) were putatively identified, namely, MGDG (34:2 and 34:3) in PCC 7806 and MGDG (36:3) in both strains. Tandem mass spectrometry (MS/MS) spectra were characteristic of galactolipids (e.g., galactose-derived neutral loss of 162 or 179, acylium ions, and neutral losses corresponding to fatty acid moieties).

FIG 7.

FIG 7

Open in a new tab

Details about the VIPs detected for PCC 7820 (A) and PCC 7806 (B). For each VIP, the VIP score, the ratio m/z, the retention time on the HPLC column, the putative identification (ID), and the relative cellular quota as a heatmap representation are presented. NA, compound not identified. A green color represents expression of a relatively lower cellular quota of a compound per sample. Levels 1 and 2 refer to Sumner et al. (77) and provide a confidence level for the ID of the VIP.

The relative cellular abundance of the VIPs was in general proportional to the salinities in both strains (Fig. 7).

DISCUSSION

Salinity tolerance of M. aeruginosa.

As M. aeruginosa has been reported to form blooms in brackish waters (11, 28), studying the response of acclimated strains of M. aeruginosa to higher salt concentrations allowed us to better assess the salinity tolerance of M. aeruginosa. After acclimation, the two strains presented distinct salinity tolerance up to 7.5 and 16.9 for M. aeruginosa PCC 7820 and PCC 7806, respectively, with these salinities corresponding to freshwater environments up to mesohaline zones (5 < S < 18) (26). The strain PCC 7806 presented a wider range of tolerance in our culture conditions than the strain PCC 7820. This result could be explained by the origin of the strain (28, 38).

Our results are in agreement with field studies on the salinity tolerance of Microcystis sp. (3, 20, 21, 30, 42). In our experiment, an increase in salinity led to a decrease of the growth rate for both strains, as well as a reduced cell size. Such a biovolume decrease has already been reported as a response to both gradually increasing salinity (29) and salt stress (31) for cyanobacteria. Tonk et al. (29) considered this phenomenon a consequence of the plasmolysis during their acclimation process. However, in our case, the biovolume decrease was observed even after acclimation. It was a long-term response compared to plasmolysis.

Impact of salinity on MC production.

Factors influencing MC production as well as the inherent role of MCs are still under debate (43). With the exception of one study (44), reports on the impact of salinity variation concluded that there was a decrease of MC cellular quotas under higher salinity conditions (23, 24, 29, 30). A decrease in MC cellular quota is frequently associated with an increase in dissolved MC because of cell lysis (19, 23, 29). Our study also showed that the MC cellular quota was decreased with increasing salinity for both strains. In addition, a linear correlation was found between growth rate and net production rate of MCs. This relation shows that MC cellular quotas in both strains are strongly coupled to cell division, as already evidenced by Orr et al. (45). Furthermore, as explained by Orr and Jones (46), this correlation illustrates that salinity only indirectly influences the production of MCs through its effect on growth. Also, as a comparison, Amé and Wunderlin (47) and Van de Waal et al. (48) showed that temperature and nutrients changed the MC congener ratios by up to an 82-fold change. Our results suggested that salinity did not seem to alter the MC ratios in M. aeruginosa in the same order of range. The extracellular MC fractions decreased with time, indicating no significant cell lysis phenomenon. At high salinities, the higher proportion of extracellular MCs resulted from the larger volume of inoculum. These results suggest that no active excretion of MCs occurred in response to higher salinities. Thus, the growth conditions were convenient for cell membrane integrity as well, as shown by the stability of the maximum photochemical efficiency of photosystem II over time. As a comparison, an active release of saxitoxin and gonyautoxins 2 and 3 can be induced by sodium and potassium increase in Raphidiopsis sp., another toxic-bloom-forming cyanobacterium (49). Based on previous studies (24, 29) and our results, the MCs did not appear to be accumulated in M. aeruginosa cells in response to higher salinity. Overall, the growth of the two M. aeruginosa strains was not restricted to freshwater conditions after acclimation. Therefore, our results confirm that this genus could be considered a salt-tolerant cyanobacterium (6). Based on laboratory experiments, our results also support the recommendation that the intensification of M. aeruginosa blooms in brackish waters should be considered by the cyanobacterial harmful algal bloom monitoring programs worldwide (6, 28).

Targeted and nontargeted analyses of metabolites at different salinities.

Using a metabolomic approach, this study describes the potential metabolites involved in sustaining M. aeruginosa growth at high salinities. M. aeruginosa is considered an organism using the “salt-out” strategy in response to an osmolarity of the external medium (51, 52). The salt-out strategy requires energy and organic matter for the accumulation of compatible solutes, which are used instead of inorganic ions to balance the osmotic potential and to maintain turgor (52). Compatible solutes can be accumulated in high molar amounts in the cells without negative impacts on the metabolism (53) and have a protective action on macromolecules (52). Here, the metabolomic approach was applied to follow the evolution of the metabolome and identify biomarkers of salt acclimation. First, both targeted and untargeted approaches showed that M. aeruginosa PCC 7820 and PCC 7806 possessed distinct metabolic profiles. Despite a nonnegligible strain specificity, it was possible to point out an evolution of metabolite profiles in response to the acclimation to distinct salinities. The consequences of the salt-out strategy as an energy-costing strategy were illustrated in this study by a lower growth rate and a distinct metabolic profile at higher salinity than the control.

Among the diversity of metabolites detected in our experiment, the relative cellular quota of some compounds was overexpressed at a higher salinity, suggesting that these compounds could be assimilated to compatible solutes. Two compounds identified in the literature as cyanobacterial osmolytes were detected, namely, sucrose (mainly produced in PCC 7806) and trehalose (mainly produced in PCC 7820) (31). Kolman et al. (37) found sucrose and the respective genes involved in its biosynthesis in M. aeruginosa PCC 7806. A sucrose accumulation in the cells was noticed after a hypoxic or salt stress in PCC 7806 (37). After recent work with new isolates of M. aeruginosa strains from brackish water, Tanabe et al. (28) confirmed the importance of sucrose synthesis in the brackish strains, compared with the lack of sucrose genes and synthesis in the freshwater isolates. We identified sucrose as a VIP in M. aeruginosa PCC 7806, but we cannot exclude the presence of trehalose based on our liquid chromatography high-resolution mass spectrometry (LC-HRMS) method (i.e., same retention time and mass). Indeed, only one study did report the accumulation of trehalose in this strain in response to light intensity stress (54). On the other hand, trehalose but not sucrose (or only trace) was identified as a VIP in the less salt tolerant strain M. aeruginosa PCC 7820. The occurrence of trehalose is poorly documented in M. aeruginosa. Here, we pointed out a distinct response of these M. aeruginosa strains in response to salt acclimation. As reviewed by Hagemann (31) for several other genera, the discrepancies in salt tolerance of M. aeruginosa strains could be partly explained by the accumulation of trehalose or sucrose.

As far as we know, a restricted number of studies reported the analyses of other identified compatible solutes using targeted analyses in Microcystis sp. (36). DMSP is described as a constitutive compatible solute in some microalgae (32) and was found in small amounts in a few genera of cyanobacteria (33). Based on our analyses, DMSP was not detected in M. aeruginosa, corroborating the absence of the potentially released by-product DMS observed by Steinke et al. (35). Targeted analyses of GB and proline led us to quantify these compounds in M. aeruginosa PCC 7820 and PCC 7806. GB and proline have been described as compatible solutes in other genera of cyanobacteria (31). Salt stress-induced proline is described for Escherichia coli, diatoms, and plants (55), while GB accumulation has long been described in response to salt increase in heterotrophic bacteria, cyanobacteria, and plants (55). Our results confirmed that GB acts as a compatible solute for M. aeruginosa. Both strains exhibited a decline of GB after 10 days, but this may be compensated by other compatible solutes, as GB in Microcystis sp. belongs to a larger pool of compatible solutes. Nevertheless, the accumulation of GB appeared to be specific to more halotolerant species than M. aeruginosa, e.g., species living in a hypersaline environment (56).

Metabolomic analyses led us to putatively identify cyanopeptolin C as a VIP of low salinity in PCC 7806, while no trace of this compound was found in PCC 7820. Cyanopeptolin C is a secondary metabolite already reported to be produced by PCC 7806 (57, 58). Besides evidence of inhibition activity on diverse human enzymes and toxicity for grazers (59), limited data are available about the physiological role of cyanopeptolin C in Microcystis cells. Abiotic factors, such as temperature, light intensity, and phosphorus limitation, influenced the amount of cyanopeptolin C per cell in M. aeruginosa (60), but our experiment suggested that salinity could also be an influencing factor. Using untargeted analyses led us to putatively identify polyunsaturated glycolipids as VIP in both strains. MGDG (36:3) in PCC 7820 and MGDG (34:2 and 34:3) in PCC 7806 were relatively overexpressed with increasing salinity compared with the control conditions. MGDGs are prevalent lipids in cyanobacterial cell and thylakoid membranes (61). As one of the major glycerolipids found in cyanobacteria (61), MGDG plays an important role in structural stabilization and the function of membranes. Membrane fluidity is strongly dependent on the lipid unsaturation degree and composition, and its evolution is recognized as one of the mechanisms involved in response to environmental fluctuation (62). Allakhverdiev et al. (63) studied the salt tolerance of Synechocystis sp. strain PCC 6803 by comparing wild and mutant strains, and the latter did not contain polyunsaturated lipids. They concluded that the degree of unsaturation of fatty acids is involved in the tolerance of the photosynthetic machinery to salt stress. Moreover, this study pointed out that the activity and synthesis of the Na+/H+ antiporter system might be suppressed under salt stress conditions. This effect appeared to be counteracted by the accumulation of unsaturated lipids (63). Until now, the presence of genes coding for Na+/H+ antiporter in M. aeruginosa genomes appeared to be strain dependent and correlated to salt tolerance (38), but no data are available about the lipid composition of M. aeruginosa in response to salt stress. Identification of MGDGs as VIP in M. aeruginosa in salt-acclimated strains suggested that these glycerolipids could be considered salt-acclimated biomarkers and involved in the tolerance of M. aeruginosa to elevated salinity. Besides the different VIPs identified in this experiment, several others presented an expression pattern close to the definition of compatible solutes. However, exhaustive identification of the detected compounds in the metabolomic analyses remains challenging. Hence, this study pointed out a potential diversity of compatible solutes in the genus Microcystis which could explain the discrepancies observed in salt tolerance among the two strains.

Here, we showed that salinity only affected MC production indirectly through a decrease in the growth rate of both strains. Intraspecific variability in salinity tolerances is reflected in metabolite profile changes between the two strains. It was possible to identify different metabolites overexpressed in response to an increase in salinity. These two strains did not rely on the same mechanisms in response to elevated salinity, which illustrates the plasticity of Microcystis aeruginosa in long-term adaptation to brackish conditions. Our data provide new insights about the metabolites involved in Microcystis sp. acclimation to higher salinity. Also, short-term experiments at higher salinity would determine whether similar mechanisms are involved in M. aeruginosa. Further experiments based on a genetic approach and specific quantification of supplementary osmolytes in several strains of Microcystis spp. are needed to better understand the diversity of osmolytes in this genus. At last, further investigations are required to confirm another hypothesis made by Orr et al. (21) suggesting that toxic phenotypes could survive longer at higher salinities than the nontoxic ones.

MATERIALS AND METHODS

Organisms and culture conditions.

Two axenic and toxic M. aeruginosa strains, available from the Pasteur Culture collection of Cyanobacteria (PCC; https://webext.pasteur.fr/cyanobacteria/), were studied, namely, the PCC 7820 and PCC 7806 strains. Both strains were isolated from contrasting environments, namely, the freshwater lake Balgaries (Scotland) and the brackish water of Braakman Reservoir (Netherlands) for PCC 7820 and PCC 7806, respectively. Cells were routinely grown in BG110*, which is a BG110 medium (64) supplemented with NaNO3 (2 mM) and NaHCO3 (10 mM), at a constant temperature of 22 ± 0.6°C under a 12:12-h light:dark cycle using cool-white fluorescent tubes (Philips) with 35 μmol photon m−2 s−1 illumination. Artificial seawater was made with the addition of NaCl (450 mM), KCl (10 mM), CaCl2 (9 mM), MgCl2 (6H2O) (30 mM), and MgSO4 (7H2O) (16 mM) before nutrient enrichment to achieve the BG110 + salts* preparation (65). Hence, the different salinities used in this study were obtained by mixing both medium BG110* and salt-enriched BG110 + salts*. Salinity was checked using a conductivity meter, Cond 3110 Set 1 (WTW).

Acclimation.

Before starting the experiment, strains were progressively acclimated to different salinities by transferring the culture after 20 days of growth (i.e., during the exponential-growth phase) into fresh medium of higher salinity. After each new transfer, the salinity was increased by 2 units. This process was repeated until reaching the likely maximal salinity tolerance of the two strains. At the end of the acclimation process, cultures were maintained between 25 and 50 generations at each salinity. At least three constant growth rates confirmed the acclimation of the strains, in agreement with MacIntyre and Cullen (66). Hence, M. aeruginosa PCC 7820 and PCC 7806 were acclimated to 3 salinities (3.7, 6.4, and 7.5) and 7 salinities (3.6, 6.0, 8.4, 10.7, 12.2, 14.8, and 16.9), respectively, in addition to the control culture condition in BG110* (S, 0.6).

Experiment.

Once acclimated to each salinity, both strains were maintained in batch culture and transferred to new fresh medium during exponential-growth phase. Therefore, triplicate batch cultures in 300-ml flasks were inoculated at 1 × 106 cells ml−1 with exponential growing cells and maintained under temperature and light conditions described above for 30 days.

Sampling and analyses.

(i) Growth rate. Cell concentration was monitored every 2 days. Samples (2 ml) were fixed with glutaraldehyde (0.1%) (Sigma-Aldrich) and stored at –80°C until analyzed on an Accuri C6 flow cytometer (Becton, Dickinson) (67). Growth rate (μ) was calculated during exponential growth using the least-squares regression method (68). Furthermore, the forward scatter fluorescence (FSC) was used to compare the mean relative cell size between salinities on sampling day 14 only (i.e., during mid-exponential-growth phase).

(ii) Maximum photochemical efficiency of photosystem II. Maximum quantum efficiency of the photosystem II (Fv/Fm) was measured after 4, 10, 14, and 24 days using an Aquapen-C 100 fluorimeter (Photon Systems Instruments). Before fluorescence analyses, samples (3 ml) were incubated in the dark for 15 min (69).

(iii) Microcystin analyses. To determine the intra- and extracellular toxin contents after 4, 10, 14, and 24 days, samples (15 ml) were harvested and centrifuged at 4,248 × g for 15 min at 4°C. The supernatant was directly stored at –80°C, and the remaining cell pellet was quenched in liquid nitrogen and stored at –80°C until extraction. For the extraction of intracellular toxins, 250 mg of glass beads (0.15 to 0.25 mm; VWR) and 1 ml of methanol (MeOH; Honeywell) were added to the pellet before mechanical grinding using a mixer mill (MM400; Retsch) for 30 min at 30 Hz. Tubes were centrifuged at 13,000 × g, and supernatants were filtered through a 0.2-μm filter (Nanosep MF; Pall). Extracellular MCs were extracted by solid-phase extraction (200 mg; Bond Elut C18 cartridges; Agilent) to remove salts and concentrate the MCs. After conditioning with 3 ml of MeOH and 3 ml of Milli-Q water, 15 ml of cell-free medium was loaded, rinsed with 2.4 ml of MeOH/Milli-Q water (5:95, vol/vol), and eluted with 4 ml of MeOH. Subsequently, intracellular and extracellular MC analyses were performed by ultrafast liquid chromatography (Nexera; Shimadzu), coupled to a triple quadrupole mass spectrometer (5500 QTrap; ABSciex) equipped with a TurboV electrospray ionization source (liquid chromatography-tandem mass spectrometry [LC-MS/MS]). The analytical chromatographic column was a Kinetex XB C18 (100 × 2.1 mm; 2.6 μm; Phenomenex) equipped with a suited guard column. The column and sample temperatures were 25°C and 4°C. The flow rate was 0.3 ml min−1, and the volume of injection was 5 μl. The binary gradient consisted of water (A) and acetonitrile (B), both containing 0.1% formic acid (vol/vol). The elution gradient was 30% to 80% B (0 to 5 min), 80% B (5 to 6 min), and 30% B (6.5 to 10.5 min). MS/MS detection was performed in positive ionization mode using multiple reaction monitoring (MRM) with two transitions per toxin (see Table S1 in the supplemental material). Nine certified MC standards (Novakits) were used to quantify the MC content in both intracellular and extracellular fractions, using an external 6-point calibration curve. Acquisition and data processing were performed using Analyst 1.6.3 (ABSciex) software. Total recoveries (i.e., combining extraction recovery and remaining matrix effect when spiking before extraction) were evaluated for intracellular and extracellular MC quantifications (see Tables S3 and S4 in the supplemental material). Intracellular MC recoveries were higher than 93%, and extracellular MC recoveries were similar between a medium with and without salt (with differences of 3% to 9%) (Tables S3 and S4). Therefore, no corrective factor was applied. The net production rate of each MC variant was calculated according to the method described by Orr et al. (45) by application of first-order rate kinetics.

(iv) DMSP, GB, and proline analyses. DMSP, GB, and proline quantifications were measured by LC-MS/MS on intracellular cell extracts prepared for toxin analysis on an LC system (model UFLC XR; Shimadzu) coupled to a triple quadrupole mass spectrometer (4000 QTrap; ABSciex). Chromatography was performed with a Hypersil GOLD HILIC column (150 × 3.0 mm; 3 μm; ThermoScientific) with a suited guard column, based on Curson et al. (70). The binary gradient consisted of water-acetonitrile (90:10 [vol/vol]) containing 4.5 mM ammonium formate (A) and water (5:95 [vol/vol]) containing 5 mM ammonium formate (B). The flow rate was 0.4 ml min−1, and injection volume was 5 μl. The column and sample temperatures were 30°C and 4°C, respectively. The elution gradient was 90% B (0 to 1 min), 90% to 45% B (1 to 8 min), 45% B (8 to 12 min), and 90% B (12 to 15 min). The LC-MS/MS system was used in positive ionization mode and MRM, with the two transitions per compounds (see Table S2 in the supplemental material). Compounds were quantified using external 5-point calibration curves of standards (Sigma-Aldrich) solubilized in methanol, with concentrations from 50 nM to 5,000 nM. Acquisition and data processing were performed using Analyst 1.6.3 (ABSciex) software. Total extraction recoveries were evaluated between strains and salt conditions (see Table S5 in the supplemental material). As total recoveries did not vary more than 20% between control condition (S, 0.6) and maximum conditions for both strains (S, 7.5 and 16.9), the quantification results were not corrected (Table S5).

Metabolomic analysis.

(i) Data acquisition. Metabolomic analysis was only conducted on intracellular extracts prepared for toxin analysis at day 14 (i.e., during mid-exponential-growth phase). Metabolomic profiles were acquired by ultraperformance liquid chromatography–high-resolution mass spectrometry (UPLC-HRMS). The instrumentation consisted of a ultra-high-performance liquid chromatography (UHPLC) system (1290 Infinity II; Agilent) coupled to a quadrupole-time of flight mass spectrometer (QTOF 6550; Agilent) equipped with a Dual Jet Stream electrospray ionization (ESI) interface. Analyses were carried out using an analytical core-shell phenyl-hexyl column (150 × 2.1 mm; 1.7 μm; Phenomenex) with a suited guard column. The column and sample temperatures were 40°C and 4°C, respectively. The flow rate was 0.5 ml min−1, and the injection volume was 5 μl. Mobile phases consisted of water (A) and acetonitrile (B), both containing 0.1 formic acid (vol/vol). The following elution gradient was used: 5% B (0 to 2 min), 5% to 100% B (2 to 10 min), 100% B (10 to 14 min), and 5% B (14 to 18 min). Mass spectra were recorded in full-scan mode from m/z of 100 to 1,700 (500-ms scan time) at a mass resolving power of 25,000 full width at half-maximum (fwhm; m/z, 922.0099) and an acquisition rate of 2 spectra s−1. A calibration check was performed continuously over the entire run time using reference masses m/z 121.0509 (purine) and m/z 922.0099 (hexakis phosphazene). Samples were randomly injected. Qualitative control samples (QC) were prepared and injected 10 times at the beginning of the batch sequence and then every 4 samples (including blank). Blanks were prepared as cell pellets (i.e., MeOH and glass beads in a polypropylene tube). To identify specific compounds of interest, tandem mass spectrometry analyses were carried out at an acquisition rate of 5 spectra s−1 for MS (m/z 100 to 1,700; 200-ms scan time) and 15 spectra s−1 for MS/MS (m/z 50 to 1,700; 66.7-ms scan time). Collision-induced dissociation was performed on the 3 most intense ions above an absolute threshold of 2,000 counts. A collision energy value proportional to the m/z of the selected precursor ion, an isolation window of 1.3 amu, and a dynamic exclusion time of 30 s were used. Annotation was conducted first with exact masses and freely available databases (e.g., HMDB, LipidMaps, Metlin, KEGG, and CEU Mass Mediator [http://ceumass.eps.uspceu.es/; 71]), then with fragmentation patterns, and using a molecular networking approach (https://gnps.ucsd.edu/; 72). Acquisition and data processing were performed using MassHunter Workstation software (version B.06.01 and B.07; Agilent).

(ii) Data preprocessing and filtering. LC-MS raw data (.d) were converted to .mzXML format using MS-Convert (Proteowizard 3.0) (73) and processed with the XCMS package (74) under the R 3.2.3 environment. Peak picking was performed with the “centWave” algorithm, retention time correction with the “obiwarp” algorithm, peak grouping with “bw” = 5 and “mzwid” = 0.015, and peak filling with the default parameters.

Three successive filtering steps using in-house scripts on R were applied to remove variables with low intensities (exclusion of variables with a signal-to-noise ratio of <10 compared to a blank), to remove signals showing high variability (exclusion of variables with a coefficient of variation of >20% in QC samples), and to suppress redundancy (exclusion of all variables but the most intense one when the coefficient of autocorrelation is >80% at the same retention time). Preprocessing led to 298 variables, and 83 variables remained after the filtrations, on which statistical analyses were performed.

Statistical analyses.

Statistical analyses of the data were performed using R software (R Core Team). Data are presented as mean and standard deviation (SD). After checking the homoscedasticity and normal distribution of residuals using Bartlett and Shapiro-Wilk tests, the effect of salinity on growth was tested using one-way analysis of variance (ANOVA), and the post hoc Dunnett’s test was applied to check a difference between growth rates (in comparison to the control). Otherwise, if the hypothesis of normal distribution and homoscedasticity of residuals were not verified, a nonparametric Kruskal-Wallis test was performed. The impact of salinity and time on the individual MC congener quotas was checked by performing two-way ANOVAs. Two-way ANOVAs on repeated measures with time as the within-subjects factor and the salinity as the between-subjects factor were conducted. The Pearson’s correlation coefficient was calculated between growth rates and net production rates of MCs. For metabolomics, multivariate statistical analyses were performed using MetaboAnalyst 4.0. (75). First, as different amounts of cells were extracted depending on salinity conditions, the data were normalized to cell concentration as a sample-specific normalization. Then, data were log-transformed and Pareto-scaled based on the recommendation of van den Berg et al. (76). Principal-component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) were used. The PLS-DA models were validated using a permutation test (measuring group separation distance using the sum-of-squares-between/sum-of-squares-within ratio; 1,000 class-assignment permutations; P < 0.007) and allowed us to highlight discriminant metabolites using the variable importance in projection (VIP) score commonly used for biomarker selection, assuming that a VIP score of >1 corresponded to a biomarker.

Data availability.

Metabolomics data were deposited in DataRef (https://doi.org/10.12770/414f1f4c-5f1c-4e98-a187-c66216393d6e).

Supplementary Material

Supplemental file 1
AEM.01614-19-s0001.pdf (227.9KB, pdf)

ACKNOWLEDGMENTS

We acknowledge Ifremer and the Regional Council of the Région des Pays de la Loire for the Ph.D. funding of Maxime Georges des Aulnois.

Footnotes

Supplemental material for this article may be found at https://doi.org/10.1128/AEM.01614-19.

REFERENCES

  • 1.O’Neil JM, Davis TW, Burford MA, Gobler CJ. 2012. The rise of harmful cyanobacteria blooms: the potential roles of eutrophication and climate change. Harmful Algae 14:313–334. doi: 10.1016/j.hal.2011.10.027. [DOI] [Google Scholar]
  • 2.Nielsen DL, Brock MA, Rees GN, Baldwin DS. 2003. Effects of increasing salinity on freshwater ecosystems in Australia. Aust J Bot 51:655–665. doi: 10.1071/BT02115. [DOI] [Google Scholar]
  • 3.Verspagen JMH, Passarge J, Jöhnk KD, Visser PM, Peperzak L, Boers P, Laanbroek HJ, Huisman J. 2006. Water management strategies against toxic Microcystis blooms in the Dutch delta. Ecol Appl 16:313–327. doi: 10.1890/04-1953. [DOI] [PubMed] [Google Scholar]
  • 4.White E, Kaplan D. 2017. Restore or retreat? Saltwater intrusion and water management in coastal wetlands. Ecosyst Heal Sustain 3:e01258. doi: 10.1002/ehs2.1258. [DOI] [Google Scholar]
  • 5.Huisman J, Codd GA, Paerl HW, Ibelings BW, Verspagen JMH, Visser PM. 2018. Cyanobacterial blooms. Nat Rev Microbiol 16:471–483. doi: 10.1038/s41579-018-0040-1. [DOI] [PubMed] [Google Scholar]
  • 6.Paerl HW, Huisman J. 2009. Climate change: a catalyst for global expansion of harmful cyanobacterial blooms. Environ Microbiol Rep 1:27–37. doi: 10.1111/j.1758-2229.2008.00004.x. [DOI] [PubMed] [Google Scholar]
  • 7.Paerl H. 2014. Mitigating harmful cyanobacterial blooms in a human- and climatically-impacted world. Life 4:988–1012. doi: 10.3390/life4040988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Funari E, Manganelli M, Buratti FM, Testai E. 2017. Cyanobacteria blooms in water: Italian guidelines to assess and manage the risk associated to bathing and recreational activities. Sci Total Environ 598:867–880. doi: 10.1016/j.scitotenv.2017.03.232. [DOI] [PubMed] [Google Scholar]
  • 9.Harke MJ, Steffen MM, Gobler CJ, Otten TG, Wilhelm SW, Wood SA, Paerl HW. 2016. A review of the global ecology, genomics, and biogeography of the toxic cyanobacterium, Microcystis spp. Harmful Algae 54:4–20. doi: 10.1016/j.hal.2015.12.007. [DOI] [PubMed] [Google Scholar]
  • 10.Humbert J-F, Barbe V, Latifi A, Gugger M, Calteau A, Coursin T, Lajus A, Castelli V, Oztas S, Samson G, Longin C, Medigue C, de Marsac NT. 2013. A tribute to disorder in the genome of the bloom-forming freshwater cyanobacterium Microcystis aeruginosa. PLoS One 8:e70747. doi: 10.1371/journal.pone.0070747. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Preece EP, Hardy FJ, Moore BC, Bryan M. 2017. A review of microcystin detections in estuarine and marine waters: environmental implications and human health risk. Harmful Algae 61:31–45. doi: 10.1016/j.hal.2016.11.006. [DOI] [Google Scholar]
  • 12.Robson BJ, Hamilton DP. 2003. Summer flow event induces a cyanobacterial bloom in a seasonal Western Australian estuary. Mar Freshwater Res 54:139–151. doi: 10.1071/MF02090. [DOI] [Google Scholar]
  • 13.Takahashi T, Umehara A, Tsutsumi H. 2014. Diffusion of microcystins (cyanobacteria hepatotoxins) from the reservoir of Isahaya Bay, Japan, into the marine and surrounding ecosystems as a result of large-scale drainage. Mar Pollut Bull 89:250–258. doi: 10.1016/j.marpolbul.2014.09.052. [DOI] [PubMed] [Google Scholar]
  • 14.Preece EP, Moore BC, Hardy FJ. 2015. Transfer of microcystin from freshwater lakes to Puget Sound, WA and toxin accumulation in marine mussels (Mytilus trossulus). Ecotoxicol Environ Saf 122:98–105. doi: 10.1016/j.ecoenv.2015.07.013. [DOI] [PubMed] [Google Scholar]
  • 15.Martínez de la Escalera G, Kruk C, Segura AM, Nogueira L, Alcántara I, Piccini C. 2017. Dynamics of toxic genotypes of Microcystis aeruginosa complex (MAC) through a wide freshwater to marine environmental gradient. Harmful Algae 62:73–83. doi: 10.1016/j.hal.2016.11.012. [DOI] [PubMed] [Google Scholar]
  • 16.Lehman PW, Kurobe T, Lesmeister S, Baxa D, Tung A, Teh SJ. 2017. Impacts of the 2014 severe drought on the Microcystis bloom in San Francisco Estuary. Harmful Algae 63:94–108. doi: 10.1016/j.hal.2017.01.011. [DOI] [PubMed] [Google Scholar]
  • 17.Bormans M, Amzil Z, Mineaud E, Brient L, Savar V, Robert E, Lance E. 2019. Demonstrated transfer of cyanobacteria and cyanotoxins along a freshwater-marine continuum in France. Harmful Algae 87:101639. doi: 10.1016/j.hal.2019.101639. [DOI] [PubMed] [Google Scholar]
  • 18.Paerl HW, Otten TG, Kudela R. 2018. Mitigating the expansion of harmful algal blooms across the freshwater-to-marine continuum. Environ Sci Technol 52:5519–5529. doi: 10.1021/acs.est.7b05950. [DOI] [PubMed] [Google Scholar]
  • 19.Ross C, Warhurst BC, Brown A, Huff C, Ochrietor JD. 2019. Mesohaline conditions represent the threshold for oxidative stress, cell death and toxin release in the cyanobacterium Microcystis aeruginosa. Aquat Toxicol 206:203–211. doi: 10.1016/j.aquatox.2018.11.019. [DOI] [PubMed] [Google Scholar]
  • 20.Otsuka S, Suda S, Li R, Watanabe M, Oyaizu H, Matsumoto S, Watanabe MM. 1999. Characterization of morphospecies and strains of the genus Microcystis (Cyanobacteria) for a reconsideration of species classification. Phycological Res 47:189–197. doi: 10.1046/j.1440-1835.1999.00162.x. [DOI] [Google Scholar]
  • 21.Orr PT, Jones GJ, Douglas GB. 2004. Response of cultured Microcystis aeruginosa from the Swan River, Australia, to elevated salt concentration and consequences for bloom and toxin management in estuaries. Mar Freshw Res 55:277–283. doi: 10.1071/MF03164. [DOI] [Google Scholar]
  • 22.Bartolomé MC, D’ors A, Sánchez-Fortún S. 2009. Toxic effects induced by salt stress on selected freshwater prokaryotic and eukaryotic microalgal species. Ecotoxicology 18:174–179. doi: 10.1007/s10646-008-0269-y. [DOI] [PubMed] [Google Scholar]
  • 23.Black K, Yilmaz M, Phlips EJ. 2011. Growth and toxin production by Microcystis aeruginosa PCC 7806 (Kutzing) Lemmerman at elevated salt concentrations. J Environ Prot Sci 2:669–674. doi: 10.4236/jep.2011.26077. [DOI] [Google Scholar]
  • 24.Martín-Luna B, Sevilla E, Bes MT, Fillat MF, Peleato ML. 2015. Variation in the synthesis of microcystin in response to saline and osmotic stress in Microcystis ruginosa PCC7806. Limnetica 34:205–214. [Google Scholar]
  • 25.Reed RH, Walsby AE. 1985. Changes in turgor pressure in response to increases in external NaCl concentration in the gas-vacuolate cyanobacterium Microcystis sp. Arch Microbiol 143:290–296. doi: 10.1007/BF00411252. [DOI] [Google Scholar]
  • 26.Mitsch WJ, Gosselink JG. 2000. Wetlands. John Wiley & Sons, Inc, New York, NY. [Google Scholar]
  • 27.Zhang Y, Xu Q, Xi B. 2013. Effect of NaCl salinity on the growth, metabolites, and antioxidant system of Microcystis aeruginosa. J Freshwater Ecol 28:477–487. doi: 10.1080/02705060.2013.782579. [DOI] [Google Scholar]
  • 28.Tanabe Y, Hodoki Y, Sano T, Tada K, Watanabe MM. 2018. Adaptation of the freshwater bloom-forming cyanobacterium Microcystis aeruginosa to brackish water is driven by recent horizontal transfer of sucrose genes. Front Microbiol 9:1150. doi: 10.3389/fmicb.2018.01150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Tonk L, Bosch K, Visser P, Huisman J. 2007. Salt tolerance of the harmful cyanobacterium Microcystis aeruginosa. Aquat Microb Ecol 46:117–123. doi: 10.3354/ame046117. [DOI] [Google Scholar]
  • 30.Mazur-Marzec H, Forycka K, Kobos J, Pliński M, Browarczyk-Matusiak G. 2010. Morphological, genetic, chemical and ecophysiological characterisation of two Microcystis aeruginosa isolates from the Vistula Lagoon, southern Baltic. Oceanologia 52:127–146. doi: 10.5697/oc.52-1.127. [DOI] [Google Scholar]
  • 31.Hagemann M. 2011. Molecular biology of cyanobacterial salt acclimation. FEMS Microbiol Rev 35:87–123. doi: 10.1111/j.1574-6976.2010.00234.x. [DOI] [PubMed] [Google Scholar]
  • 32.Stefels J. 2000. Physiological aspects of the production and conversion of DMSP in marine algae and higher plants. J Sea Res 43:183–197. doi: 10.1016/S1385-1101(00)00030-7. [DOI] [Google Scholar]
  • 33.Oren A. 2007. Diversity of organic osmotic compounds and osmotic adaptation in cyanobacteria and algae, p 639–655. In Seckbach J. (ed), Algae and cyanobacteria in extreme environments. Cellular origin, life in extreme habitats and astrobiology, vol 11. Springer, Dordrecht, The Netherlands. [Google Scholar]
  • 34.Bucciarelli E, Ridame C, Sunda WG, Dimier-Hugueney C, Cheize M, Belviso S. 2013. Increased intracellular concentrations of DMSP and DMSO in iron-limited oceanic phytoplankton Thalassiosira oceanica and Trichodesmium erythraeum. Limnol Oceanogr 58:1667–1679. doi: 10.4319/lo.2013.58.5.1667. [DOI] [Google Scholar]
  • 35.Steinke M, Hodapp B, Subhan R, Bell TG, Martin-Creuzburg D. 2018. Flux of the biogenic volatiles isoprene and dimethyl sulfide from an oligotrophic lake. Sci Rep 8:630. doi: 10.1038/s41598-017-18923-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Hagemann M, Erdmann N, Wittenburg E. 1987. Synthesis of glucosylglycerol in salt-stressed cells of the cyanobacterium Microcystis firma. Arch Microbiol 148:275–279. doi: 10.1007/BF00456704. [DOI] [Google Scholar]
  • 37.Kolman MA, Torres LL, Martin ML, Salerno GL. 2012. Sucrose synthase in unicellular cyanobacteria and its relationship with salt and hypoxic stress. Planta 235:955–964. doi: 10.1007/s00425-011-1542-5. [DOI] [PubMed] [Google Scholar]
  • 38.Sandrini G, Huisman J, Matthijs HCP. 2015. Potassium sensitivity differs among strains of the harmful cyanobacterium Microcystis and correlates with the presence of salt tolerance genes. FEMS Microbiol Lett 362:fnv121. doi: 10.1093/femsle/fnv121. [DOI] [PubMed] [Google Scholar]
  • 39.Szymańska E, Saccenti E, Smilde AK, Westerhuis JA. 2012. Double-check: validation of diagnostic statistics for PLS-DA models in metabolomics studies. Metabolomics 8:3–16. doi: 10.1007/s11306-011-0330-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Favre L, Ortalo-Magné A, Pichereaux C, Gargaros A, Burlet-Schiltz O, Cotelle V, Culioli G. 2018. Metabolome and proteome changes between biofilm and planktonic phenotypes of the marine bacterium Pseudoalteromonas lipolytica TC8. Biofouling 34:132–148. doi: 10.1080/08927014.2017.1413551. [DOI] [PubMed] [Google Scholar]
  • 41.Lê Cao K-A, Boitard S, Besse P. 2011. Sparse PLS discriminant analysis: biologically relevant feature selection and graphical displays for multiclass problems. BMC Bioinformatics 12:253. doi: 10.1186/1471-2105-12-253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Atkins R, Rose T, Brown RS, Robb M. 2001. The Microcystis cyanobacteria bloom in the Swan River—February 2000. Water Sci Technol 43:107–114. doi: 10.2166/wst.2001.0518. [DOI] [PubMed] [Google Scholar]
  • 43.Omidi A, Esterhuizen-Londt M, Pflugmacher S. 2018. Still challenging: the ecological function of the cyanobacterial toxin microcystin—what we know so far. Toxin Rev 37:87–105. doi: 10.1080/15569543.2017.1326059. [DOI] [Google Scholar]
  • 44.Chen L, Mao F, Kirumba GC, Jiang C, Manefield M, He Y. 2015. Changes in metabolites, antioxidant system, and gene expression in Microcystis aeruginosa under sodium chloride stress. Ecotoxicol Environ Saf 122:126–135. doi: 10.1016/j.ecoenv.2015.07.011. [DOI] [PubMed] [Google Scholar]
  • 45.Orr PT, Willis A, Burford MA. 2018. Application of first order rate kinetics to explain changes in bloom toxicity—the importance of understanding cell toxin quotas. J Ocean Limnol 36:1063–1074. doi: 10.1007/s00343-019-7188-z. [DOI] [Google Scholar]
  • 46.Orr PT, Jones GJ. 1998. Relationship between microcystin production and cell division rates in nitrogen-limited Microcystis aeruginosa cultures. Limnol Oceanogr 43:1604–1614. doi: 10.4319/lo.1998.43.7.1604. [DOI] [Google Scholar]
  • 47.Amé MV, Wunderlin DA. 2005. Effects of iron, ammonium and temperature on microcystin content by a natural concentrated Microcystis aeruginosa population. Water Air Soil Pollut 168:235–248. doi: 10.1007/s11270-005-1774-8. [DOI] [Google Scholar]
  • 48.Van de Waal DB, Verspagen JMH, Lürling M, Van Donk E, Visser PM, Huisman J. 2009. The ecological stoichiometry of toxins produced by harmful cyanobacteria: an experimental test of the carbon-nutrient balance hypothesis. Ecol Lett 12:1326–1335. doi: 10.1111/j.1461-0248.2009.01383.x. [DOI] [PubMed] [Google Scholar]
  • 49.Soto-Liebe K, Méndez MA, Fuenzalida L, Krock B, Cembella A, Vásquez M. 2012. PSP toxin release from the cyanobacterium Raphidiopsis brookii D9 (Nostocales) can be induced by sodium and potassium ions. Toxicon 60:1324–1334. doi: 10.1016/j.toxicon.2012.09.001. [DOI] [PubMed] [Google Scholar]
  • 50.Reference deleted.
  • 51.Galinski EA. 1995. Osmoadaptation in bacteria. Adv Microb Physiol 37:272–328. [PubMed] [Google Scholar]
  • 52.Hagemann M. 2013. Genomics of salt acclimation: synthesis of compatible solutes among cyanobacteria, p 27–55. In Chauvat F, Cassier-Chauvat C (ed), Genomics of cyanobacteria. Academic Press, Amsterdam, The Netherlands. [Google Scholar]
  • 53.Brown AD. 1976. Microbial water stress. Bacteriol Rev 40:803–846. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Meissner S, Steinhauser D, Dittmann E. 2015. Metabolomic analysis indicates a pivotal role of the hepatotoxin microcystin in high light adaptation of Microcystis. Environ Microbiol 17:1497–1509. doi: 10.1111/1462-2920.12565. [DOI] [PubMed] [Google Scholar]
  • 55.Hagemann M. 2016. Coping with high and variable salinity: molecular aspects of compatible solute accumulation, p 359–372. In Borowitzka MA, Beardall J, Raven JA (ed), The physiology of microalgae. Springer International Publishing, Basel, Switzerland. [Google Scholar]
  • 56.Klähn S, Hagemann M. 2011. Compatible solute biosynthesis in cyanobacteria. Environ Microbiol 13:551–562. doi: 10.1111/j.1462-2920.2010.02366.x. [DOI] [PubMed] [Google Scholar]
  • 57.Martin C, Oberer L, Ino T, König WA, Busch M, Weckesser J. 1993. Cyanopeptolins, new depsipeptides from the cyanobacterium Microcystis sp. PCC 7806. J Antibiot (Tokyo) 46:1550–1556. doi: 10.7164/antibiotics.46.1550. [DOI] [PubMed] [Google Scholar]
  • 58.Briand E, Bormans M, Gugger M, Dorrestein PC, Gerwick WH. 2016. Changes in secondary metabolic profiles of Microcystis aeruginosa strains in response to intraspecific interactions. Environ Microbiol 18:384–400. doi: 10.1111/1462-2920.12904. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Janssen E. 2019. Cyanobacterial peptides beyond microcystins—a review on co-occurrence, toxicity, and challenges for risk assessment. Water Res 151:488–499. doi: 10.1016/j.watres.2018.12.048. [DOI] [PubMed] [Google Scholar]
  • 60.Tonk L, Welker M, Huisman J, Visser PM. 2009. Production of cyanopeptolins, anabaenopeptins, and microcystins by the harmful cyanobacteria Anabaena 90 and Microcystis PCC 7806. Harmful Algae 8:219–224. doi: 10.1016/j.hal.2008.05.005. [DOI] [Google Scholar]
  • 61.Wada H, Murata N. 1998. Membrane lipids in cyanobacteria BT—lipids in photosynthesis: structure, function and genetics, p 65–81. In Siegenthaler P-A, Murata N (ed), Lipids in photosynthesis: structure, function and genetics. Springer, Dordrecht, The Netherlands. [Google Scholar]
  • 62.Mikami K, Murata N. 2003. Membrane fluidity and the perception of environmental signals in cyanobacteria and plants. Prog Lipid Res 42:527–543. doi: 10.1016/S0163-7827(03)00036-5. [DOI] [PubMed] [Google Scholar]
  • 63.Allakhverdiev SI, Nishiyama Y, Suzuki I, Tasaka Y, Murata N. 1999. Genetic engineering of the unsaturation of fatty acids in membrane lipids alters the tolerance of Synechocystis to salt stress. Proc Natl Acad Sci U S A 96:5862–5867. doi: 10.1073/pnas.96.10.5862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Rippka R, Deruelles J, Waterbury JB, Herdman M, Stanier RY. 1979. Generic assignments, strain histories and properties of pure cultures of cyanobacteria. Microbiology 111:1–61. doi: 10.1099/00221287-111-1-1. [DOI] [Google Scholar]
  • 65.Le Bail A, Charrier B. 2013. Plant organogenesis. Humana Press, Totowa, NJ. [Google Scholar]
  • 66.MacIntyre HL, Cullen JJ. 2005. Using cultures to investigate the physiological ecology of microalgae, p 287–326. In Anderson RA. (ed), Algal culturing techniques. Elsevier, Amsterdam, The Netherlands. [Google Scholar]
  • 67.Marie D, Simon N, Vaulot D. 2005. Phytoplankton cell counting by flow cytometry, p 1–61. In Anderson RA. (ed), Algal culturing techniques. Elsevier, Amsterdam, The Netherlands. [Google Scholar]
  • 68.Wood AM, Everroad RC, Wingard LM. 2005. Measuring growth rates in microalgal cultures, p 269–285. In Anderson RA. (ed), Algal culturing techniques. Elsevier, Amsterdam, The Netherlands. [Google Scholar]
  • 69.Kromkamp JC, Forster RM. 2003. The use of variable fluorescence measurements in aquatic ecosystems: differences between multiple and single turnover measuring protocols and suggested terminology. Eur J Phycol 38:103–112. doi: 10.1080/0967026031000094094. [DOI] [Google Scholar]
  • 70.Curson ARJ, Williams BT, Pinchbeck BJ, Sims LP, Martínez AB, Rivera PPL, Kumaresan D, Mercadé E, Spurgin LG, Carrión O, Moxon S, Cattolico RA, Kuzhiumparambil U, Guagliardo P, Clode PL, Raina J-B, Todd JD. 2018. DSYB catalyses the key step of dimethylsulfoniopropionate biosynthesis in many phytoplankton. Nat Microbiol 3:430–439. doi: 10.1038/s41564-018-0119-5. [DOI] [PubMed] [Google Scholar]
  • 71.Gil-de-la-Fuente A, Godzien J, Saugar S, Garcia-Carmona R, Badran H, Wishart DS, Barbas C, Otero A. 2019. CEU Mass Mediator 3.0: a metabolite annotation tool. J Proteome Res 18:797–802. doi: 10.1021/acs.jproteome.8b00720. [DOI] [PubMed] [Google Scholar]
  • 72.Wang M, Carver JJ, Phelan VV, Sanchez LM, Garg N, Peng Y, Nguyen DD, Watrous J, Kapono CA, Luzzatto-Knaan T, Porto C, Bouslimani A, Melnik AV, Meehan MJ, Liu W-T, Crüsemann M, Boudreau PD, Esquenazi E, Sandoval-Calderón M, Kersten RD, Pace LA, Quinn RA, Duncan KR, Hsu C-C, Floros DJ, Gavilan RG, Kleigrewe K, Northen T, Dutton RJ, Parrot D, Carlson EE, Aigle B, Michelsen CF, Jelsbak L, Sohlenkamp C, Pevzner P, Edlund A, McLean J, Piel J, Murphy BT, Gerwick L, Liaw C-C, Yang Y-L, Humpf H-U, Maansson M, Keyzers RA, Sims AC, Johnson AR, Sidebottom AM, Sedio BE, et al. 2016. Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking. Nat Biotechnol 34:828–837. doi: 10.1038/nbt.3597. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Chambers MC, Maclean B, Burke R, Amodei D, Ruderman DL, Neumann S, Gatto L, Fischer B, Pratt B, Egertson J, Hoff K, Kessner D, Tasman N, Shulman N, Frewen B, Baker TA, Brusniak M-Y, Paulse C, Creasy D, Flashner L, Kani K, Moulding C, Seymour SL, Nuwaysir LM, Lefebvre B, Kuhlmann F, Roark J, Rainer P, Detlev S, Hemenway T, Huhmer A, Langridge J, Connolly B, Chadick T, Holly K, Eckels J, Deutsch EW, Moritz RL, Katz JE, Agus DB, MacCoss M, Tabb DL, Mallick P. 2012. A cross-platform toolkit for mass spectrometry and proteomics. Nat Biotechnol 30:918–920. doi: 10.1038/nbt.2377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 74.Smith CA, Want EJ, O'Maille G, Abagyan R, Siuzdak G. 2006. XCMS: processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification. Anal Chem 78:779–787. doi: 10.1021/ac051437y. [DOI] [PubMed] [Google Scholar]
  • 75.Chong J, Soufan O, Li C, Caraus I, Li S, Bourque G, Wishart DS, Xia J. 2018. MetaboAnalyst 4.0: towards more transparent and integrative metabolomics analysis. Nucleic Acids Res 46:W486–W494. doi: 10.1093/nar/gky310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.van den Berg RA, Hoefsloot HCJ, Westerhuis JA, Smilde AK, van der Werf MJ. 2006. Centering, scaling, and transformations: improving the biological information content of metabolomics data. BMC Genomics 7:142. doi: 10.1186/1471-2164-7-142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 77.Sumner LW, Amberg A, Barrett D, Beale MH, Beger R, Daykin CA, Fan T-M, Fiehn O, Goodacre R, Griffin JL, Hankemeier T, Hardy N, Harnly J, Higashi R, Kopka J, Lane AN, Lindon JC, Marriott P, Nicholls AW, Reily MD, Thaden JJ, Viant MR. 2007. Proposed minimum reporting standards for chemical analysis. Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics 3:211–221. doi: 10.1007/s11306-007-0082-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental file 1
AEM.01614-19-s0001.pdf (227.9KB, pdf)

Data Availability Statement

Metabolomics data were deposited in DataRef (https://doi.org/10.12770/414f1f4c-5f1c-4e98-a187-c66216393d6e).

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES