Occurrence of microcystins and anatoxin-a in eutrophic lakes of Saint Petersburg, Northwestern Russia

The present study was conducted in two shallow, humic, eutrophic lakes with a different nutrient status – Lake Sestroretskij Razliv (Razliv) and Lower Suzdal Lake (Suzdal) – located in Saint Petersburg region, Russia, and used for recreational purposes. Razliv is a large artificial reservoir with a surface area of 1100 ha and average depth of 1.6 m. The surface area of Suzdal is 97 ha and the mean depth is 3.0 m.

Surface water samples (1 l) were collected from these lakes with plastic bottles and phytoplankton biomass was collected with a net (mesh size of 85 μm). Sampling was performed every 2 weeks from May to October in 2010-2012. Altogether, 45 water samples and 40 phytoplankton samples were collected from the two lakes. The water samples in 1 l bottles were fixed with a Lugol-formalin solution. Cyanobacterial material from net samples was transported to a laboratory in screw-cap tubes (Axigen, California, USA) and immediately frozen at –20°C until freeze-drying.

Qualitative and quantitative analyses of phytoplankton were carried out under a light microscope (Mikromed-3, LOMO, Saint Petersburg, Russia). The volume of the counting chamber Uchinskaya was 0.1 ml. The biovolume of algae and cyanobacteria was calculated using species-specific geometric formulas (Olenina et al. 2006). The phytoplankton biomass was determined from the total volume of algae according to the counted cell density and the measured average cell density.

Microcystin standards (MC-LR, MC-RR, and MC-YR) were purchased from Sigma-Aldrich (St. Louis, MO, USA) and (+/-) anatoxin-a fumarate was obtained from Tocris Bioscience (Bristol, UK). Acetonitrile (HPLC-grade) and methanol (LiChrosolv hypergrade for LC-MS) were purchased from Merck (Darmstadt, Germany). Water was purified to 18.2 MΩ cm⁻¹ in a Millipore Direct-Q water purification system (Bedford, MA, USA). Formic acid was obtained from Fluka Chemika (Buchs, Switzerland). Membrane filter discs (nylon 66, 47-mm diameter) were obtained from Supelco (Bellefonte, PA).

Water samples (1 l) for the analysis of extracellular cyanotoxins (AN and MCs) were filtered on a Supelco mobile-phase filtration apparatus (nylon-66 membrane filter discs, 47-mm diameter, Supelco, Bellefonte, PA). Solid phase extraction (SPE) with Oasis HLB cartridges (60 or 200 mg, Waters, Milford, Massachusetts, USA) was used. The toxins were eluted from the cartridges with 10 ml methanol. The collected extracts were dried using a rotary evaporator and stored at –20°C until further analysis. The dried extracts were reconstituted in 1 ml of 80% aqueous methanol and centrifuged (CM-50 centrifuge, ELMI, Riga, Latvia) at 14 000 rpm for 10 min prior to analysis by high-performance liquid chromatography/high-resolution tandem mass spectrometry (LC-HR-MS/MS).

For the extraction of biomass-bound (intracellular) MCs and anatoxin-a, the dry lyophilized cells (10-30 mg) were treated twice with 1 ml of 75% aqueous methanol by sonication in an ultrasound bath (US) for 15 min, and then centrifuged at 14 000 rpm for 15 min. The obtained extracts were combined.

The LC-MS experiments were carried out in the Accela HPLC system (Thermo Finnigan, San Jose, CA, USA) coupled with a Hybrid Ion Trap-Orbitrap Mass Spectrometer – LTQ Orbitrap XL (Thermo Fisher Scientific, San Jose, USA) with the electrospray interface. Separation of the toxins was performed on a Thermo Hypersil Gold RP C18 column (100 × 3 mm, 3 μm) with a Hypersil Gold drop-in guard column (Thermo Fisher Scientific). The mobile phase consisted of water (solvent A) and acetonitrile (solvent B), both containing 0.05% formic acid. The gradient program started at 5% B (held for 10 min), then the content of phase B increased to 75% over 18 min (held for 3 min), and then increased to 95% B over 1 min (held for 5 min). The flow rate was 0.2 ml min⁻¹. The column temperature was 40°C and the injection volume was 25 μl for water sample extracts and 10 μl for biomass sample extracts.

The MS conditions using electrospray ionization were as follows: Ion Spray voltage 3.2 kV, capillary Voltage 20 V, ion-transfer capillary temperature 320°C, tube lens 130 V. The mass spectrometer operated in the positive ion mode at a resolving power of 30 000. A scan function (m/z 500–1200) for the detection of all MC variants was used. The identification of target compounds, whose standards were available (AN and MC-LR, -RR, -YR), was based on the accurate mass measurement of [M+H]⁺ or [M+2H]²⁺ ions (accuracy within 5 ppm), the collected fragmentation spectrum of the ions and the retention times (coincidence with the analytical standard within 0.3 min). The detection of anatoxin-a was based on the presence of a peak with m/z at 166.12 and retention time corresponding to the anatoxin-a standard. The structure of the neurotoxin was confirmed using fragmentation spectra (MS/MS) of the precursor ion with m/z 166.12 covering m/z 110-170. Other MC congeners, whose standards were not available, were identified based on the exact m/z values for described brutto formulas (Mayumi et al. 2006; Furey 2008), calculated using the NIST IsoForm program ver. 1.02 (NIST Formula and Isotopic Pattern Generator, NIST, USA) and fragmentation spectra of their pseudomolecular ions.

For quantitative analysis of cyanotoxins, the calibration curves for standards solutions of AN and MC-LR, -RR, and -YR were determined (r² = 0.996-0.998) within a calibration range of 5-500 ng ml⁻¹. Procedural LODs were 3-6 ng l⁻¹ for water samples (SPE extraction) and 0.08-0.30 μg g⁻¹ DW for freeze-dried biomass (ultrasonic extraction), respectively. Procedural LOQs were 9-20 ng l⁻¹ for natural water samples and 0.27-0.90 μg g⁻¹ for freeze-dried biomass, respectively.

Measurements of the concentrations of the singly charged MC variants, whose standards were not available, were based on the standard curve for MC-LR. As there were large variations in the MS response of the singly charged MCs and the doubly charged ions of different MC-RR congeners, the concentrations of their desmethylated (dm) and methylated counterparts were determined on the basis of the curve drawn for m/z 519.79 of the MC-RR standard.

Major hydrochemical parameters for both lakes were similar. The maximum gradients for ammonia nitrogen (0.6-2.0 mg N l⁻¹) and phosphates (0.002-0.600 mg P l⁻¹) were typical of eutrophic lakes of Northwestern Russia. During the study period, the pH values in the lakes changed within the range of 6.2-8.5 in Razliv, and within the range of 7.6-9.0 in Suzdal.

The seasonal cyanobacterial dynamics in Razliv and Suzdal was determined based on the analysis of 41 phytoplankton samples. The total number of phytoplankton taxa identified in the collected samples was 138. The identified species belonged to nine taxonomic groups of algae, with the dominance of green algae (41% of the total number of species), cyanobacteria (18%), euglenoids (13%), and diatoms (11%). Among green algae, the family Chlorococcoaceae was represented by the largest number of species. The data on the abundance and biomass of the observed taxonomic groups are presented in supplementary tables (Supplementary tables S1A-S1F).

Both lakes were described as β-mezosaprobic. During the study period, the phytoplankton biomass varied significantly within and between the seasons and lakes (Fig. 1). Among the identified species, the following were most numerous: Aphanizomenon flos-aquae, Planktothrix agardhii (Cyanobacteria), Aulacoseira ambigua, Aulacoseira islandica, Aulacoseira italica (Bacillariophyta), and Cryptomonas rostrata (Cryptophyta).

Figure 1

The highest mean seasonal cyanobacterial biomass was recorded in 2012 in Suzdal (10.97 mg l⁻¹) and in 2011 in Razliv (15.05 mg l⁻¹). The contribution of cyanobacteria to the total phytoplankton biomass varied from 0.4% (0.02 mg l⁻¹) to 99% (61.94 mg l⁻¹) in Razliv and from 0.1% (0.01 mg l⁻¹) to 41% (26.03 mg l⁻¹) in Suzdal (Fig. 1, Supplementary tables S1A-S1F).

In 2010 and 2012, Aph. flos-aquae and Dolichospermum flos-aquae dominated in the plankton community in Razliv at the beginning of the sampling seasons. In summer, the dominance shifted to representatives of the genus Microcystis (M. aeruginosa, M. wesenbergii and M. viridis) (Fig. 1 A, C). In 2011, Microcystis species and Aph. flos-aquae co-occurred almost throughout the whole summer season (Fig. 1B). In Suzdal, the variability in the dominant species composition was not as significant as that in Razliv. In 2010 and 2012, the dominance of P. agardhii and Aph. flos-aquae in the cyanobacterial community at the beginning of the warm season shifted in August-September to Microcystis spp. (Fig. 1D, F). During 2011, the quantitative dominance of P. agardhii over Aph. flos-aquae was observed (Fig. 1 E).

The studies carried out in 2010-2012 showed the occurrence of cyanotoxins both in Razliv and in Suzdal. The concentrations of microcystins exceeding the value of ≥ 0.1 μg l⁻¹ for the extracellular fraction, or 0.1 mg g⁻¹ on a dry-weight (DW) basis for the intracellular fraction, were determined in 54 water and biomass samples (59% of all the collected samples). Extracellular microcystins were detected in 25 of 46 water samples (54% of the water samples) (Tables 1, 2, Supplementary tables S2-S7). AN was present only in 5 water samples from Suzdal (Table 2, Supplementary tables S5, S6).

The median seasonal and maximum concentrations of the total MCs, extracellular and intracellular MCs concentrations, and the concentrations of the three main toxins identified in the two water bodies by LC-MS/MS in each sampling period are shown in Tables 1 and 2. The sampling periods are highlighted in the tables according to the change in the dominant species. The total number of detected MCs variants during our three-year study was 14 in Razliv, and 9 in Suzdal. The number and structures of microcystin congeners identified in the lakes are presented in Table 3.

Although cyanobacterial biomass was not high in 2010 (Fig. 1A, 1D), the median values for total extracellular toxins reached their maxima in Razliv (1.61 μg l⁻¹) and Suzdal (0.44 μg l⁻¹) (Tables 1, 2). The maximum extracellular MCs concentrations in each water body for the whole study period were recorded in the same season. In Razliv, the maximum extracellular MCs concentration (41.37 μg l⁻¹ mainly due to 38.54 μg l⁻¹ of desmethyl-MC-RR) was determined in August 2010 (Table 1, Supplementary tables S2). The maximum concentration of extracellular MCs in Suzdal was noted in September 2010 and it amounted to 2.89 μg l⁻¹ (Table 2, Supplementary tables S5). Concentrations of intracellular MCs were low during this sampling season in both lakes, their median values were 132 μg g⁻¹ DW in Razliv and 18 μg g⁻¹ DW in Suzdal (Tables 1, 2).

In 2011, in spite of higher cyanobacterial biomass and higher contribution of cyanobacteria to the phytoplankton community than in 2010 (Fig. 1 B, E), the values of median concentration of extracellular MC were low in both lakes (Tables 1 and 2). The maximum dissolved MCs concentrations recorded this season were also low in both lakes and amounted to 0.70 μg l⁻¹ in Razliv (Table 1, Fig. 1B, and Supplementary Table S3) and 0.40 μg l⁻¹ in Suzdal (Table 2, Fig. 1F, and Supplementary table S6). On the other hand, the intracellular MC content was high and reached 2095 μg g-1 in Razliv and 698 μg g⁻¹ in Suzdal.

The highest median concentrations of intracellular MCs in both lakes were recorded in 2012. At the beginning of August 2012, the maximum concentration of intracellular MCs for the entire study period (2294 μg g⁻¹ DW) was recorded in Razliv. In Suzdal, peak values of intracellular MCs concentrations (1075-1796 μg g⁻¹ DW) were recorded in September (Table 2, Supplementary table S7).

Neurotoxic AN was detected only in Suzdal. It was found in water samples collected from late June to early August 2010 and 2011 (Table 2, Supplementary tables S5, S6). Extracellular concentration of AN in filtered water samples was in the range of 0.16-0.54 μg l-1 (2 samples) in 2010 and 0.14-0.27 μg l⁻¹ (3 samples: two samples in the sampling period June-July and one in August) in 2011 (Table 2, Supplementary table S5). Intracellular AN (30 μg g⁻¹ DW) was recorded only in one biomass sample at the beginning of July. The presence of P.agardhii and Aph.flos-aquae was observed in all anatoxin-positive samples, Dolichospermum spp. was present only in a trace amount (Table 2, Supplementary table S6).