Oceanological and Hydrobiological Studies

Transcription

1 Oceanological and Hydrobiological Studies International Journal of Oceanography and Hydrobiology Vol. XXXVII, No.4 Institute of Oceanography ISSN X (3-21) 2008 University of Gdańsk eissn DOI /v Original research paper Received: Accepted: June 14, 2008 November 17, 2008 Cyanobacterial hepatotoxins, microcystins and nodularins, in fresh and brackish waters of the Pomeranian Province, northern Poland Hanna Mazur-Marzec 1,3, Lisa Spoof 2, Justyna Kobos 1, Marcin Pliński 1, Jussi Meriluoto 2 1 Department of Marine Biology and Ecology, University of Gdańsk Al. Marszałka Piłsudskiego 46, Gdynia, Poland 2 Department of Biochemistry and Pharmacy, Åbo Akademi University Tykistökatu 6A, Turku, Finland Key words: Cyanobacterial blooms, cyclic peptides, LC-MS/MS Abstract Microcystins (MCs) and structurally related nodularins (NODs) are hepatotoxic cyclic peptides produced by bloom-forming cyanobacteria. These toxins have been implicated in the deaths of wild and domestic animals as well as in incidents of human illness. Cyanobacterial toxins occurring in the fresh and brackish waters of the Pomeranian Province, northern Poland were characterized in this study. Water samples collected from seven lakes in August and September 2005 were analysed by high performance liquid chromatography (HPLC), enzyme linked immunosorbent assay (ELISA) and protein phosphatase inhibition assay (PPIA). Cyanobacterial toxins present in field samples and in an isolated strain of Planktothrix agardhii were also characterized by liquid chromatography electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS). In most of the fresh water samples MC-LR, MC-RR and MC- 3 Corresponding author. Tel.: , fax: , biohm@univ.gda.pl

2 4 H. Mazur-Marzec, L. Spoof, J. Kobos, M. Pliński, J. Meriluoto YR dominated. In the lakes where P. agardhii was most abundant demethylated microcystin variants tentatively identified as [D-Asp 3 ]MC-LR, [D-Asp 3 ]MC-YR and [D-Asp 3 ]MC-RR, were found. Total concentrations of the toxins measured by HPLC ranged from 0.1 μg l -1 to μg l -1. Nodularia spumigena bloom samples were collected from brackish waters of the Gulf of Gdańsk, southern Baltic, and LC-ISP-MS/MS of extract from these revealed the presence of two geometrical isomers of linear nodularin and nodularin variant with aspartic acid methyl ester [MeAsp 1 (OMe)]NOD. INTRODUCTION The Pomeranian Province in northern Poland, situated by the Baltic Sea, is characterized by a variety of landscapes, with sandy beaches, forests, hills and many post-glacial lakes. In summer this area is frequented by holidaymakers. Many local people make a living out of tourism, so the water quality and health of the environment are matters of great concern. Unfortunately, excessive nutrient loading from agricultural, industrial and urban runoff has lead to increased eutrophication of many water bodies in the region. As a consequence, in the summer months, cyanobacterial blooms are recurrent. In the fresh waters blooms are primarily caused by species belonging to the genera Microcystis, Anabaena and Planktothrix. In the brackish waters of the Baltic Sea Nodularia, Aphanizomenon and Anabaena predominate (Finni et al. 2001, Stal et al. 2003, Mazur-Marzec et al. 2006a). Some of the cyanobacterial species produce hepatotoxic secondary metabolites belonging to microcystins (MCs) or nodularins (NODs). Microcystins are monocyclic heptapeptides that contain two variable L-amino acids and two unusual amino acids: 3-amino-9-methoxy- 2,6,8-trimethyl-10-phenyldeca-4(E),6(E)- dienoic acid (Adda) and N- methyldehydroalanine (Mdha) (Rinehart et al. 1988, Rinehart et al. 1994). The general structure of the commonly occurring and most toxic MC-LR (LD 50 = 50 μg kg -1 body weight) is cyclic (-D-Ala 1 -L-Leu 2 -D-MeAsp 3 (iso-linkage)-l-arg 4 - Adda 5 -D-Glu 6 (iso-linkage)-mdha 7 -), where MeAsp is D-erythro-β-methyl aspartic acid. At present, about 80 structural microcystin variants with different toxicity have been characterized (Chorus and Bartram 1999, Meriluoto and Codd 2005). In some microcystins leucine (L) or arginine (R) can be replaced by other L-amino acids, for example tyrosine (Y), homotyrosine (Hty), phenylalanine (F), methionine (M), tryptophan (W) or homoarginine (Har). Other modifications in microcystin structure include replacement of Mdha by dehydrobutyrine (Dhb), N-methylserine (MeSer), serine (Ser) or alanine (Ala). Nodularins are pentapeptides of general structure cyclo (-Adda-D-Glu-Mdhb- MeAsp-L-Arg-) where Mdhb is N-methyldehydrobutyric acid (Rinehart et al. 1988). So far, ten nodularins have been characterized from bloom and cultured cyanobacterial samples of the genus Nodularia. Both microcystin and nodularin variants differ with respect to the number of methyl groups at various locations

3 Cyanobacterial hepatotoxins, microcystins and nodularins 5 in the peptides. 6(Z)-Adda stereoisomers as well as 9-hydroxy and 9-acetoxy derivatives of Adda can also be biosynthesized (Rinehart et al. 1994, Namikoshi et al. 1994). Microcystins and nodularins are inhibitors of eukaryotic protein phosphatases and tumour promoters. Incidents of human toxicity through recreational activities in the area of cyanobacterial bloom, or consumption of contaminated water, have been described by Chorus and Bartram (1999) and Falconer (2005). It has been reported that up to 12 microcystin (Diehnelt et al. 2006) and 8 nodularin (Mazur-Marzec et al. 2006b) variants can be produced by a single strain of cyanobacterium. The toxins are synthesised nonribosomally by multifunctional enzyme complexes consisting of peptide synthetases and polyketide synthases (Tillett et al. 2000). Profile of the toxins, and the rate of their synthesis, can vary with environmental factors, and light conditions in particular (Sivonen 1990a, Hobson and Fallowfield 2001, Oberholster et al. 2004, Tonk et al. 2005). Since individual variants of microcystins and nodularins differ significantly in toxicity, an exact identification of their structures is required for a reliable risk assessment. In the current study the presence of hepatotoxic cyclic peptides, microcystins and nodularins, in both fresh and brackish waters of the Pomeranian Province, northern Poland were examined in the year For the purpose of the study waters with regular incidents of cyanobacteria blooms were selected. Identification of cyanobacterial hepatotoxins and elucidation of their structures were performed by liquid chromatography electrospray ionization mass spectrometry (LC-ESI-MS/MS). MATERIALS AND METHODS Chemicals The nodularin standard was purchased from Calbiochem (LaJolla, CA, USA). Microcystins (dmmc-rr, MC-RR, MC-YR, MC-LR, MC-LY, MC-LF and MC-LW) were from methanolic extracts of Microcystis strains: NIES 107 (National Institute for Environmental Studies, Tsukuba, Japan) and PCC 7820 (Pasteur Culture Collection, Paris, France). HPLC gradient grade acetonitrile was purchased from Merck (Darmstadt, Germany) and HPLC grade methanol from Baker (Deventer, the Netherlands). Trifluoroacetic acid (TFA) of protein sequencing grade was acquired from Fluka (Buchs, Switzerland), and formic acid (p.a.) from Riedel-de Haën (Seelze, Germany). ELISA kits were obtained from EnviroLogix (Portland, Main, USA). Protein phosphatase 1 (PP1) was acquired from New England Biolabs, USA. Bovine serum albumin (BSA) was obtained from Sigma-Aldrich (St. Louis, MO, USA). Dithiothreitol (DTT),

4 6 H. Mazur-Marzec, L. Spoof, J. Kobos, M. Pliński, J. Meriluoto MgCl 2 6H 2 O, MnCl 2 4H 2 O, Na 2 SO 4, p-nitrophenyl phosphate (p-nnp), tris- (hydroxymethyl)-aminomethane (Tris) were of analytical grade. Water was purified to 18.2 MΩ cm (MilliQ water) on a Milli-Q plus PF system from Millipore (Millipore, Molsheim, France). Sampling and sample sites On 31 August and 19 September 2005 surface water samples were collected in 1-l plastic bottles from seven lakes in Pomeranian Province (shore-line of Jasień, Karczemne, Karlikowskie, Klasztorne Małe, Klasztorne Duże, Sianowskie and Tuchomskie) (Fig. 1, Table 1). All these lakes are classified as eutrophic. Samples were passed through a Whatman GF/C (Maidstone, UK) glass microfibre filter disc within four hours of collection. The volumes of the samples ranged from 5 to 100 ml, depending on the intensity of the bloom, as determined by visual inspection. The filters with cyanobacterial material were immediately frozen and kept at 20 C until analysis. Additionally, concentrated bloom samples were collected with a 100-μm mesh plankton net from coastal waters of the Gulf of Gdańsk off Sopot (54 30'N, 18 47'E) on 14 July This material was placed in a Petri dish and lyophilised. Further, 100-ml water samples were fixed in 1% Lugol s solution. Phytoplankton community composition was determined using a Nikon E-200 Eclipse (Tokyo, Japan) light microscope. Fig. 1 Location of the sampled lakes in Pomeranian Province.

5 Cyanobacterial hepatotoxins, microcystins and nodularins 7 Table 1 Physical parameters of the sampled lakes in Pomeranian Province. Lake ϕ λ Area Max. depth Average depth [km 2 ] [m] [m] Karczemne Klasztorne Duże Klasztorne Małe ; Karlikowskie Jasień Sianowskie Tuchomskie Culture conditions The clonal Planktothrix agardhii strain PKLD 0205 was isolated from a sample collected at Lake Klasztorne Duże ( 'N, 'E) on August 31, The cyanobacterium was grown in 300-ml flasks with liquid BG-11 medium (Rippka 1988) based on MilliQ water. Cultures were incubated in in vitro chambers at 25 C with a photoperiod of 12 hours and 30 μe m 2 s -1 light intensity. P. agardhii cells were harvested at the end of exponential growth phase (three weeks) by filtration through Whatman GF/C filters. Sample extraction and high performance liquid chromatography (HPLC) The filters bearing freshwater cyanobacteria were extracted with 1 ml of 90% methanol and 0.1% TFA for 15 min in an ultrasonic bath and for 1 min in Branson Sonifier II W-250 ultrasonic homogenizer (Danbury, CT, USA) equipped with a microtip probe (30% duty cycle). After centrifugation (10,000 g, 10 min) 500 μl of the supernatant was evaporated to dryness under a stream of argon and the dry residue was redissolved in 100 μl of 30% methanol. The samples were centrifuged again and subjected to further analyses. N. spumigena cells (1 g), collected with plankton net and lyophilised, were extracted with 25 ml of 30% methanol for 20 min in an ultrasonic bath followed by 1 min probe sonication with ultrasonic disrupter HD 2070 Sonopuls (20 khz, Bandelin, Berlin, Germany) equipped with MS 72 probe (25% duty cycle). After centrifugation at 10,000 g for 15 min, the supernatants were rotary evaporated at 35 C. The dry residue was dissolved in 30% methanol and centrifuged again. All cyanobacterial extracts were analysed with HPLC-DAD (Agilent 1100 series, Waldbronn, Germany) using an Ascentis RP-Amide column (100 mm 4.6 mm I.D., 3 μm) from Supelco (Bellefonte, PA, USA). The mobile phase

6 8 H. Mazur-Marzec, L. Spoof, J. Kobos, M. Pliński, J. Meriluoto consisted of water (A) and acetonitrile (B) both with 0.05% TFA. Gradient elution conditions were as follows: 25% B at 0 min, 70% B at 7 min, then 70% B for 3 min and from 70% B to 25% B in 0.1 min. Injection interval was 15 min, sample volume 10 μl, flow rate 1 ml min -1 and column temperature 40 C. Toxins were identified by their retention time and UV spectrum. As all microcystins and nodularins have the same chromophore, a conjugated double bound in the Adda residue, it was assumed that their molar absorption coefficients are equal to that of MC-LR. Microcystin was prepared according to the procedure described by Meriluoto and Codd (2005) and the standard curve for the toxin was constructed. The linear regression equation obtained was y = 0.78x, where y is ng MC-LR per 10 μl injection and x is the peak area. The y- axis intercept, b, was neglected as its value was lower than 0.3. Chlorophyll-a (Chl a) measurements Filter discs with cyanobacterial cells were extracted in 1.0 ml of 90% methanol for 15 min in an ultrasonic bath and for 1 min in the ultrasonic homogenizer equipped with a microtip. The samples were clarified by centrifugation for 15 min at g. Chl a concentrations in the extracts were determined spectrophotometrically by absorbance measurements at 665 nm (A 665 ) and 750 nm (A 750 ) using the equation C chla = (A 665 A 750 ) [μg ml -1 ] (Mackinney 1941). Enzyme-linked immunosorbent assay (ELISA) and protein phosphatase inhibition assay (PPIA) Extracts of freshwater cyanobacteria collected on filters were diluted with MilliQ water ,000 times to give final concentrations within the standard range of ng ml -1 for the ELISA and ng ml -1 for the PPIA. The ELISAs were performed using the EnviroLogix microcystin microplate kits according to the manufacturers instructions. The PPIAs were carried out on 96- well microplates according to the method described by An and Carmichael (1994) and Rapala et al. (2002). The substrate and enzyme buffers were prepared immediately before the test. 2.5 catalytic subunits of commercially available enzyme (PP1) were diluted in 1.5 ml of the enzyme buffer. Subsequently 10 μl of standard solution or sample were added to the well and mixed with 10 μl of PP1 in buffer. After 5 min incubation, 200 μl of p-npp in buffer solution was added to each well. The contents of the wells were mixed by swirling the plate horizontally. After 2 h incubation at 37 C the absorbances of the solutions were measured. The ELISA and PPIA plates were read on microplate reader (Bio-Rad, Benchmark Microplate Reader, Hercules, CA, USA) at 450 nm (ELISA) and 405 nm (PPIA).

7 Cyanobacterial hepatotoxins, microcystins and nodularins 9 Mass spectrometric analyses Microcystins present in fresh water cyanobacterial extracts were analyzed with an ion trap instrument Agilent 1100 series LC/MSD SL (Agilent Technologies, Waldbronn, Germany) mass spectrometer. Separation was performed on a Merck Purospher STAR RP-18e (30 mm 4 mm I.D., 3 μm) column protected by a RP-18e 4 mm 4 mm guard column, also from Merck. The mobile phase consisted of 0.5% formic acid (A) : acetonitrile (B); the gradient run was from 25% B to 70% B in 10 min. Injection volume was 10 μl, flow rate 0.5 ml min -1 and column temperature 40 C. Electrospray ionization (ESI) was used and ions were detected in the positive mode. The ionisation parameters were: HV capillary 4 kv, capillary exit offset 89.1 V, trap drive 55.1, dry temperature 350 C, dry gas 12 l min -1, nebulizer gas pressure 40 p.s.i (1 p.s.i. = Pa). The scan range was m/z N. spumigena extracts were analysed with Hybride quadrupole-time-offlight liquid chromatography/tandem mass spectrometer (TOF LC-MS/MS) QStar XL (Applied Bioscience Sciex Instruments, USA). Separation was performed on a Symetry C18 column (5 μm, 150 mm 4.6 mm I.D.) using the Agilent 1100 system. Isocratic elution with acetonitrile : water (32:68), both containing 0.1% formic acid, at a flow rate 0.5 ml min -1, was employed. A mass range of m/z was covered with a scan time of 1 s, the instrument was operated in positive ion mode. Ion Spray voltage was 5.5 kv with the nebulizer gas air pressure and curtain gas nitrogen pressure set at 30 and 20 p.s.i., respectively. Fragmentation was achieved with nitrogen collision gas at a pressure of 6 p.s.i. and collision energy of 64 ev. RESULTS AND DISCUSSION Since 2000, a routine survey of numerous water bodies in the Pomeranian Province has been performed to study the dynamics of cyanobacterial blooms and toxin production (Mazur-Marzec et al. 2006a, Kobos et al. 2005). In most of the studies HPLC with diode array detectors were used to detect and quantify the toxins. This method, however, does not allow for definite identification of all microcystin or nodularin variants; its use being limited to a few toxins either commercially available or purified in-house. Modern LC-MS/MS instruments are more specific and sensitive tools for cyanobacterial toxin analyses. In the current study the microcystins and nodularins detected by UV at 238 nm were identified from the ESI-MS/MS spectra.

8 10 H. Mazur-Marzec, L. Spoof, J. Kobos, M. Pliński, J. Meriluoto Cyanobacteria and microcystins in lake samples The results presented in Table 2 show that in most of the samples collected from the lakes of the Pomeranian Province cyanobacteria belonging to the genera Planktothrix, Microcystis and Anabaena co-dominated. Anabaena was present in all fresh waters studied, although the abundance of this genus usually did not exceed 20% of the phytoplankton. The only exception was the sample collected from Lake Sianowskie on 31 August where A. spiroides f. crassa and A. planctonica made up ca. 85%. In lakes Klasztorne Duże and Klasztorne Małe P. agardhii comprised ca. 50% 80% of the phytoplankton. Microcystis was the main cyanobacterial component in lakes Jasień (40%), Tuchomskie (75%) and Karczemne (70% 98%). Woronichinia was most abundant in Lake Karlikowskie (50%), with low numbers also present in Lake Sianowskie. Lake Jasień was characterized by the most complex phytoplankton community composition with green algae and cyanobacteria of the genera Microcystis, Planktothrix, Anabaena, Gomphosphaera and Phormidium present. Generally, the cyanobacterial pattern in fresh waters of the Pomeranian Province was typical of the northern part of Europe (Meriluoto et al. 1989, Sivonen et al. 1990b, Luukkainen et al. 1993, Fastner et al. 1999). In Polish water bodies studied by Jurczak et al. (2004) in the years , cyanobacterial blooms were mainly formed by Microcystis aeruginosa, with Aphanizomenon flos-aquae or Nostoc spp. co-occurring in some samples. According to Mankiewicz et al. (2005), cyanobacteria of the genera Microcystis constituted a main phytoplankton component only in one out seven lakes in northern Poland, whereas Planktothrix was found in three lakes and Aphanizomenon in four lakes. Generally, cyanobacterial community composition, and the proportion of toxic and non-toxic strains, depend on environmental conditions such as nutrient level, intensity of solar radiation and water temperature. All of these vary temporarily and spatially having an impact on local toxin profiles and concentrations. Microcystins were present in all lakes of the Pomeranian Province examined in this study (Table 3). Total concentration of the toxins in seston measured by HPLC ranged from 0.1 μg l -1 to μg l -1. Neither HPLC nor LC-MS/MS of the samples collected in 2005 showed any evidence of compounds with chromatographic or spectral properties typical of anatoxin-a or cylindrospermopsin. In 2005, the highest concentrations of MC-LR and MC-RR (109.2 and μg l -1, respectively) were recorded in samples collected on 31 August from Lake Karczemne. MC-YR and demethylated microcystin variants were observed in lakes Klasztorne Duże and Klasztorne Małe, which were characterized by the dominance of P. agardhii. Cell-bound concentrations of the

9 Cyanobacterial hepatotoxins, microcystins and nodularins 11 Table 2 Dominant species of phytoplankton and concentration of chlorophyl a in the lakes of Pomeranian Province. Lakes Date Dominants of phytoplankton assemblage Chl a [µg l -1 ] Karczemne Klasztorne Duże Klasztorne Małe Karlikowskie Jasień Sianowskie Tuchomskie Microcystis 70% (M. aeruginosa 90% ; M. flos-aquae 10%,) Planktothrix 30%, (P. agardhii) Anabaena 8% (A.. flos-aquae 45%, A. lemmermannii 40%, A. spiroides 5%) other Microcystis 98% (M. aeruginosa 90%, M. flos-aquae 10%) Anabaena 1% other Planktothrix 80% (P. agardhii) Microcystis 10% (M. aeruginosa 45%, M. wesenbergii 45%, M. flos-aquae 10%) Anabaena 8% (A. flos-aquae) other Planktothrix 80% (P. agardhii) Phormidium sp. 8% Microcystis 5% (M. aeruginosa 50%, M. wesenbergii 50%) Anabaena 2% (A. flos-aquae) other Planktothrix 60% (P. agardhii) Microcystis 30% (M. aeruginosa 90%, M. wesenbergii 5%, M. flos-aquae 3%, M. botrys 2%) Anabaena 15% (A. flos-aquae 95%, A. spiroides 5%) other Planktothrix 50% (P. agardhii) Microcystis 5% (M. aeruginosa 95%, M. wesenbergii 5%) Phormidium sp. 3% Anabaena 2% (A..flos-aquae) other Woronichinia 50% (W. naegelianii) Microcystis 20% (M. aeruginosa 50%, M. ichtioblabe 50%) Anabaena 20% (A. lemmermannii 90%, A. planctonica 5%, A. circinalis 3%, A. affinis 2%) Planktothrix 5% (P. agardhii) other Phormidium sp. 30% Anabaena 15% (A. flos-aquae) Microcystis 10% (M. aeruginosa 95%, M. wesenbergii 5%) other 45% Microcystis 40% (M. aeruginosa 40%, M. wesenbergii 40%, M. botrys 13%,M. flos-aquae 5%, M. viridis 2%) Planktothrix 32% (P. agardhii) Anabaena 5% (A. sp. (f. straight) 80%, A. flos-aquae 10%, A. spiroides f. crassa 10%) other 8% Anabaena 85% (A. spiroides f. crassa 50%, A. planctonica 50%) Microcystis 5% (M. aeruginosa 80%, M. flos-aquae 20%) Woronichinia 2%, Aphanizomenon 3%, other Microcystis 75% (M. viridis 80%, M. aeruginosa 15%, M. wesenbergii 5%, M. ichtioblabe 3%, M. botrys 2%) Anabaena 10% (A. flos-aquae 50%, A. spiroides f. crassa 50%) other 10% toxins ranged from 0.5 μg l -1 to 4.8 μg l -1 (MC-YR) and from 0.8 μg l -1 to 17.3 μg l -1 (dmmc) (Table 3). In the total ion chromatogram (TIC) of the bloom sample extracts from Lake Klasztorne Duże, peaks of the ions at m/z 513 (3.8 min), 981 (5.6 min),

10 12 H. Mazur-Marzec, L. Spoof, J. Kobos, M. Pliński, J. Meriluoto Table 3 Main toxins identified in the lakes of Pomeranian Province. Lakes Date dmmc-rr Concentration of microcystins by HPLC-DAD [µg l -1 ] MC-RR MC-YR dmmc-lr MC-LR dmmc-yr Main toxins identified by LC-MS/MS HPLC Total concentration of microcystins [µg l -1 ] ELISA PPIA Karczemne MC-RR MC-LR MC-YR Karczemne MC-LR Klasztorne Duże Klasztorne Duże Klasztorne Małe Klasztorne Małe Karlikowskie dmmc-rr dmmc-lr MC-YR dmmc-yr dmmc-rr dmmc-lr MC-YR dmmc-yr dmmc-rr dmmc-lr MC-LR MC-YR dmmc-yr dmmc-rr dmmc-lr MC-YR dmmc-yr MC-RR MC-LR Karlikowskie MC-LR 1031 (6.3 min) and 1015 (6.4) were recorded. Retention times and the m/z values of protonated ions indicated that the ions could have derived from demethylated variants of MC-RR, MC-LR, MC-YR and MC-FR, respectively. Fragmentation spectra of [M+H] + ions were studied to elucidate their structures and determine the demethylation sites. An intense ion peak at m/z 513 can be derived from a doubly charged [M+2H] 2+ ion of one of the three possible dmmc-rr variants: [D-Asp 3 ]MC- RR, [Dha 7 ]MC-RR or [D-Asp 3,(E)-Dhb 7 ]MC-RR (Chorus and Bartram 1999, Sano and Kaya 1995). Doubly charged ions are typical of microcystins with two Arg residues (Edwards et al. 1993, Yuan et al. 1999). Some product of the 1024 [M+H] + ion were observed at m/z 599 [Arg-Adda-Glu+H] +, 873 [M+H-134- NH 3 ] + and 890 [M+H-134] + (Table 4). Ions at m/z 213 [Glu-Mdha+H] +, 311 [Mdha-Ala-Arg+H] +, 446 [C 11 H 15 O-Glu-Mdha-Ala] + and 343 [Ala-Arg-D- Asp+H] suggested the presence of N-methyldehydroalanine and indicated a lack

11 Cyanobacterial hepatotoxins, microcystins and nodularins 13 Table 4 Fragment ions of the [M+H] + of demethylated microcystin variants in cyanobacterial extract from Lake Klasztorne Duże. Fragment ions m/z dmmc-rr dmmc-lr dmmc-yr [M+H] [M+2H] [M+H H 2 O] [M+H - 134] [M+H NH 3 ] [M+H - Adda] [Asp-Arg-Adda-Glu-Mdha (Dhb) + H] [Arg-Adda-Glu-Mdha (Dhb) + H] [Mdha (Dhb)-Ala-Leu-Asp-Arg + H] [Arg-Adda-Glu + H] [C 11 H 15 O-Glu-Mdha (Dhb)-Ala] [C 11 H 15 O-Glu-Mdha (Dhb)] [Ala-Arg-Asp + H] [Mdha (Dhb)-Ala-Arg + H] [Glu-Mdha (Dhb) + H] of methyl group on Asp. Based on the fragmentation pattern, the structure of the compound was presumed to be [D-Asp 3 ]MC-RR. Since Mdha and Dhb have exactly the same mass, it is possible that [D-Asp 3, Dhb 7 ]MC-RR was produced in the lakes. Dhb-containing MC-RR variants have been rarely reported, although they were identified in P. agardhii by Sano and Kaya (1995). According to Fastner et al. (1999) [D-Asp 3, Mdha 7 ]MC-RR was most abundant in bloom samples containing high proportions of P. agardhii, whereas [D-Asp 3, Dhb 7 ]MC-RR dominated in blooms of P. rubescens. In the latter cyanobacterial species [D-Asp 3, (E)-Dhb 7 ]MC-HilR was also identified (Sano et al. 2004). Dominance of P. agardhii in the lakes in the present study suggested that the demethylated MC-RR detected might contain Mdha rather than Dhb amino acid residues. In order to definitively identify the microcystin structure NMR and amino acid analyses must be undertaken. The ion at m/z = 981 could correspond to one out of five known demethylated MC-LR variants: [D-Asp 3 ]MC-LR, [DMAdda 5 ]MC-LR, [Dha 7 ]MC-LR, [D-Asp 3, (E)-Dhb 7 ]MC-LR and [D-Asp 3, (Z)-Dhb 7 ]MC-LR (Chorus and Bartram 1999; Sano and Kaya 1998a,b). ESI-MS/MS analysis of cyanobacterial bloom extracts from lakes Klasztorne Duże and Klasztorne Małe revealed that the m/z 981 ion gave fragment ions at m/z 847 [M+H 134] +, 830 [M+H 134 NH 3 ] + and 599 [Arg-Adda-Glu+H] + (Table 4). These ions proved the presence of methoxyl group in the Adda residue. Moreover, fragment ions at

12 14 H. Mazur-Marzec, L. Spoof, J. Kobos, M. Pliński, J. Meriluoto m/z 446 [C 11 H 15 O-Glu-Mdha/Dhb-Ala] + and 682 [Arg-Adda-Glu- Mdha/Dhb+H] + indicated the presence of Mdha or Dhb. These ions, as well as ions at m/z 539 [Arg-Asp-Leu-Ala-Mdha/Dhb+H] + and 797 [Asp-Arg-Adda- Glu-Mdha/Dhb+H] +, show that the structure of the microcystin is either [D- Asp 3 ]MC-LR or [D-Asp 3, dhb 7 ]MC-LR. The ion peak at m/z 1031 corresponds to three microcystin structures [D- Asp 3,Dha 7 ]MC-HtyR, [Dha 7 ]MC-YR or [D-Asp 3 ]MC-YR. Product scan of the m/z 1031 ion gave ions at m/z 1013 and 897, which probably posses the structures of [M+H-H 2 O] + and [M+H-134] +, respectively (Table 4). A fragment ion at m/z 375 [C 11 H 14 O-Glu-Mdha] + suggested that N-methyldehydroalanine is present in the molecule, hence the structure of the compound is likely to be [D-Asp 3 ]MC-YR. The fragmentation pathway of the m/z 1015 ion did not confirm the structure of the compound to be dmmc-fr. Production of dmmc-fr was reported by Hummert et al. (2001) and Jurczak et al. (2004) in Vietnamese and Polish waters, respectively. LC-MS/MS analysis of P. agardhii strain PKLD 0205, isolated from Lake Klasztorne Duże, showed the presence of [D-Asp 3 ]MC-LR and [D-Asp 3 ]MC- YR. Based on this finding it might be concluded that P. agardhii is the source of the two demethylated microcystin variants in the lake. Cyanobacterial species belonging to the genera Planktothrix (P. agardhii and P. rubescence) commonly occur in the lakes of northern Europe. They are efficient producers of microcystins, especially variants with different degrees and sites of demethylation (Laub et al. 2002, Spoof et al. 2003, Barco et al. 2004). In one strain of Planktothrix isolated from Lake Maxsee (Brandenburg, Germany) Welker et al. (2004) detected 11 microcystins, all of them lacking a methyl group on the Asp residue. Studies by Lindholm and Meriluoto (1991) and Luukkainen et al. (1993) also revealed high content of dmmc vatiants in P. agardhii. Maximum abundances of the cyanobacterium were found in the metalimnion of stratified lakes. It was postulated that light intensity has a crucial effect on growth and toxin production by the cyanobacterium (Sivonen 1990b, Tonk et al. 2005). Generally, the profiles of dominant microcystin variants determined in the present study are similar to the profiles observed in other parts of Poland by Jurczak et al. (2004). In 25 mg samples of lyophilised cyanobacterial material the authors detected MC-LR, MC-RR, MC-YR, dmmc-lr and minor amounts of eight other microcystin variants (dmdmmc-rr, dmmc-rr, dmdmmc-yr, dmmc-yr, dmdmmc-lr, MC-AR, MC-(H 4 )YR and dmmc-fr). Since the phytoplankton materials were collected from water samples of ml, it is possible that less abundant microcystins might have been below detection limits.

13 Cyanobacterial hepatotoxins, microcystins and nodularins 15 Most of the HPLC results were comparable with those obtained by ELISA, but did not always correspond to the measurements by PPIA (Table 3). The differences in the observed results obtained by the three methods might be attributed to the presence of microcystin variants with different antibody specificity or/and different abilities to inhibit protein phosphatase. It was revealed that the MeAsp-Arg-Adda-Glu region is essential for reactivity of the toxins (An and Carmichael 1994, Rapala et al. 2002). Microcystins with a modified Adda side chain, i.e. [(6Z)Adda 5 ]MC-LR and [DMAdda 5 ]MC-LR, did not cross-react with antibodies specific for MC-LR whereas hydrophobic microcystins without an Arg residue (LW, LF), have a low affinity to the antibodies. In PPIA the demethylated microcystin variants [D-Asp 3 ]MC-LR and [D-Asp 3 ]MC-RR show much lower inhibition effects than MC-LR (An and Carmichael 1994, Rapala et al. 2002). Lower concentrations of microcystins measured in lakes Klasztorne Duże and Klasztorne Małe by PPIA, compared to HPLC data, could be attributed to the presence of [D-Asp 3 ] microcystins in the samples. Generally, the values obtained by ELISA were the highest, with PPI results, apart from Lake Jasień samples, the lowest. Further, it is possible that in the lake sample extracts non-microcystin components cross-reacted with the antibodies (ELISA) or showed protein phosphatase activity (PPIA). Lakes Karczemne, Klasztorne Duże and Klasztorne Małe are highly eutrophicated water bodies situated in the vicinity of the town Kartuzy. Chl a concentrations in this study ranged between μg l -1. Data systematically collected in 2006, and also earlier measurements, showed the highest Chl a and microcystin concentrations between September and October. On 5 October 2006, in lakes Klasztorne Duże and Karczemne, Chl a concentrations reached 25.1 μg l -1 and 28,208.8 μg l -1 with MCs concentrations of 8.64 μg l -1 and 11, μg l -1, respectively (Mazur-Marzec personal communication). It has been postulated that cyanobacterial growth and microcystin production is stimulated by the seasonal reduction in light intensity (Lindholm and Meriluoto 1991). Recently, attempts have been made to reduce the amount of available phosphorus and limit the development of cyanobacterial blooms in the lakes of the Pomeranian Province. In our future studies we are going to examine the efficacy and impacts of these efforts on phytoplankton community structure and cyanotoxin production. Nodularins in the coastal waters of the Gulf of Gdańsk Phytoplankton was collected during an N. spumigena bloom in 2005, when measured cell-bound nodularin concentrations in coastal waters of the Gulf of Gdańsk ranged from 1.9 3,964 μg l -1. At that time, N. spumigena was the major phytoplankton component in the gulf (approx. 95%), but a few bundles of

14 16 H. Mazur-Marzec, L. Spoof, J. Kobos, M. Pliński, J. Meriluoto Aphanizomenon flos-aquae and single filaments of Anabaena sp. were also present. TOF-LC/MS/MS with ionspray (ISP) and collision-induced dissociation (CID) used for analyses of N. spumigena bloom sample extract showed the presence of several nodularin variants previously reported by Namikoshi et al. (1994) and Mazur Marzec et al. (2006b). Additionally, a small amount of a compound characterized by [M+H] + ion, with m/z at 839, was detected (Fig. 2). In the selected ion record chromatogram at m/z 839 a low signal at 5.73 min was present. At collision-energy (CE) 64 ev, the m/z 839 product ion spectrum gave some fragment ions that are typical of nodularin (Fig. 2). The base peak at m/z 135 has been assigned to [PhCH 2 CH(OMe)] +, a fragment ion resulting from Fig. 2. Selected ion recording chromatogram of m/z 839 and ISP-MS/CID mass spectrum of [MeAsp 1 (OMe)]NOD (at 5.73 min, marked with an arrow).

15 Cyanobacterial hepatotoxins, microcystins and nodularins 17 cleavage of the methoxy group of Adda. Other ions were observed at m/z 227 [Glu-Mdhb+H] +, 253 [CO-Glu-Mdhb H] +, 389 [C 11 H 15 O-Glu-Mdhb] + and 404 [Adda-Glu-Mdhb 135] +. There were also some ions that occurred at 14 units higher m/z than in the NOD spectrum. They were recorded at m/z 283, 380, 397, 480, 569, and 688. This finding suggested that an additional methyl group might be located either in the Arg or MeAsp residue, hence the ions might be attributed to 283 [MeAsp-(+CH 2 )-Arg+H NH 3 ], 380 [Mdhb-MeAsp-(+CH 2 )- Arg + H NH 3 ] +, 397 [Mdhb-MeAsp-(+CH 2 )-Arg +H] +, 480 [C 11 H 15 O-NH 2 - Arg-(+CH 2 )-MeAsp +2H] +, 569 [MeAsp-(+CH 2 )-Arg-Adda+H CO 2 ] +, or 688 [C 11 H 15 O-Glu-Mdhb-MeAsp-(+CH 2 )-Arg] +. Nodularin [L-Har 2 ]NOD, with an additional methyl group in Arg, was identified as a main toxin variant in Nodularia PCC 7804 (Pasteur Culture Collection) isolated from a thermal spring in France (Beattie et al. 2000, Saito et al. 2001), but this toxin has never been detected in genetically different N. spumigena. Therefore, it seems probable that the additional methyl group is located in Asp hence the new nodularin detected is Asp methyl ester [MeAsp 1 (OMe)]NOD. So far, many demethylated or methylated cyanobacterial toxin peptides have been characterized (Chorus and Bartram 1999, Meriluoto and Codd 2005). Rinehart et al. (1994), with reference to Choi personal communication, identified nodularin with glutamic acid methyl ester [Glu 4 (OMe)]NOD. This compound was also found in N. spumigena collected from the Gulf of Gdansk (Mazur-Marzec et al. 2006b). To our knowledge microcystin or nodularin analogues with methyl Asp methyl ester have never been reported. Since water : methanol (70:30) was used to prepare the N. spumigena cell extract, the ester could not have been formed during the extraction procedure. Additionally, in the total ion chromatogram (TIC) of the N. spumigena bloom extract there were three peaks of the m/z 843 ions (Fig. 3). One of them (3.93 min) was identified as a linear nodularin (Mazur-Marzec et al. 2006b) that is presumed to be a biogenic precursor of the cyclic peptide. In the spectra of the other two ions with m/z 843 (in the TIC at 3.69 and 4.34 min), the same fragmentation patterns were recorded. We concluded that they are geometrical isomers of the linear nodularin. Rinehart et al. (1988) determined the stereochemistry of Adda as 2S, 3S, 8S, 9S. The geometrical isomers of microcystins and nodularins, (6Z)Adda and (4Z)Adda, can be formed under sunlight irradiation (Namikoshi et al. 1994, Tsuji et al. 1995). It has been estimated that toxic blooms usually contain 5-15% of geometrical isomers of cyclic microcystins (Harada et al. 1990). Much higher proportions of linear nodularin isomers found in N. spumigena bloom samples might be explained by a lack of steric hindrances and higher flexibility of the linear peptide.

16 18 H. Mazur-Marzec, L. Spoof, J. Kobos, M. Pliński, J. Meriluoto Fig. 3. Selected ion recording chromatogram of m/z 843 and ISP-MS/CID mass spectrum of the linear NOD form (at 3.93 min, marked with an arrow). CONCLUSIONS In cyanobacterial bloom samples from freshwaters of the Pomeranian Province, MC-LR is the most common cyanotoxin detected. In the samples, microcystins MC-RR, MC-YR and demethylated variants of MC-LR, MC-YR and MC-RR were also identified. Production of demethylated MCs was attributed mainly to the presence of Planktothrix agardhii. New nodularin variants detected in Nodularia spumigena bloom samples from the Gulf of Gdańsk were characterized as: two geometrical isomers of linear nodularin and the cyclic peptide with Asp methyl ester [MeAsp 1 (OMe)]NOD. ACKNOWLEDGEMENTS We thank Agata Błaszczyk for assistantance with sample collection. This work was financially supported by the Office of the Marshal of the Pomeranian Voivodeship in Poland (grant number 34/UM/DPS/2007). Nodularin mass spectra were obtained on an instrument of the Pomeranian Science and Technology Park in Gdynia. LS and JM acknowledge financial support from the Academy of Finland (RC for Biosciences and Environment), decision number Dr. Olli Sjövall is thanked for his help with ion trap LC-MS.

17 Cyanobacterial hepatotoxins, microcystins and nodularins 19 REFERENCES An J., Carmichael W.W., 1994, Use of a colorimetric protein phosphatase inhibition assay and enzyme linked immunosorbent assay for the study of microcystins and nodularins, Toxicon, 32, Barco M., Flores C., Rivera J., Caixach J., 2004, Determination of microcystin variants and related peptides present in a water bloom of Planktothrix (Oscillatoria) rubescens in a Spanish drinking water reservoir by LC/ESI-MS, Toxicon, 44, Beattie K.A., Kaya K., Codd G.A., 2000, The cyanobacterium Nodularia PCC 7804, of freshwater origin, produces [L-Har 2 ] nodularin, Phytochemistry, 54, Chorus I., Bartram J., 1999, Toxic Cyanobacteria in Water. A Guide to their Public Health Consequences, Monitoring and Management, London, E & FN Spon. Diehnelt C.W., Dugan N.R., Peterman S.M., Budde W.L., 2006, Identification of microcystin toxins from a strain of Microcystis aeruginosa by liquid chromatography introduction into a hybrid linear ion trap-fourier transform ion cyclotron resonance mass spectrometry, Anal. Chem., 78, Edwards C., Lawton L.A., Beattie K.A., Codd G.A., Pleasance S., Dear G.J., 1993, Analysis of microcystins from cyanobacteria by liquid chromatography with mass spectrometry using atmospheric pressure ionization. Rapid Commun, Mass Spectrom, 7, Falconer J., Cyanobacterial Toxins of Drinking Water Supplies. Cylindrospermopsins and Microcystins, London, CRC Press,. Fastner J., Erhard M., Carmichael W.W., Sun F., Rinehart K.L., Rönicke H., Chorus I., 1999, Characterization and diversity of microcystins in natural blooms and strains of the genera Microcystis and Planktothrix from German freshwaters, Arch. Hydrobiol., 145, Finni T., Kononen K., Olsonen R., Wallström K., 2001, The history of cyanobacterial blooms in the Baltic Sea, Ambio, 30, Harada K.-I., Ogawa K., Matsuura K., Murata H., Suzuki M., Watanabe M.F., Itezono Y., Nakayama N., 1990, Structural determination of geometrical isomers of microcystin-lr and RR from cyanobacteria by two-dimensional NMR spectroscopic techniques, Chem. Res. Toxicol., 3, Hobson P., Fallowfield H.J., 2001, Effect of total dissolved solids and irradiance on growth and toxin production by Nodularia spumigena, J. Appl. Phycol., 11, Hummert Ch., Dahlmann J., Reinhardt K., Dang H. Ph., Dang D.K., Luckas B., 2001, Liquid chromatography-mass spectrometry identification of microcystins in Microcystis aeruginosa strain from Lake Thanh Cong, Hanoi, Vietnam, Chromatogr., 54, Jurczak T., Tarczyńska M., Karlsson K., Meriluoto J., 2004, Characterization and diversity of cyanobacterial hepatotoxins (microcystins) in blooms from Polish freshwaters identified by liquid chromatography-electrospray ionization mass spectrometry, Chromatogr., 59, Kobos J., Mazur-Marzec H., Dietmar M., Witek B., Pliński M., 2005, Toxic cyanobacterial blooms in the Kociewskie Lakes (Northern Poland), Oceanol. Hydrobiol. Stud., 34, Laub J., Henriksen P., Brittain S.M., Wang J., Carmichael W.W., Rinehart K.L., Moestrup Ø., 2002, [ADMAdda 5 ]-microcystins in Planktothrix aghardii strain PH-123 (cyanobacteria) importance for monitoring of microcystin in the environment, Environ. Toxicol., 17, Lindholm T., Meriluoto J.A.O., 1991, Recurrent depth maxima of the hepatotoxic cyanobacterium Oscillatoria agardhii,. Can. J. Fish Aquat. Sci., 48,

18 20 H. Mazur-Marzec, L. Spoof, J. Kobos, M. Pliński, J. Meriluoto Luukkainen R., Sivonen K., Namikoshi M., Färdig M., Rinehart K.L., Niemelä S.I., 1993, Isolation and identification of eight microcystins from 13 Oscillatoria agardhii strains and structure of a new microcystin, Appl. Environ. Microbiol., 59, Mackinney G., 1941, Absorption of light by chlorophyll solutions, J. Biol. Chem., 140, Mankiewicz J., Komárkova J., Izydorczyk K., Jurczak T., Tarczyńska M., Zalewski M., 2005, Hepatotoxin cyanobacterial blooms in the lakes of northern Poland, Environ. Toxicol., 20, Mazur-Marzec H., Krężel A., Kobos J., Pliński M., 2006a, Toxic Nodularia spumigena blooms in the coastal waters of the Gulf of Gdańsk: a ten-year survey. Oceanologia, 48, Mazur-Marzec H., Meriluoto J., Pliński M., Szafranek J., 2006b, Characterization of nodularin variants from the Baltic Sea using liquid chromatography/mass spectrometry/mass spectrometry, Rapid Commun. Mass Spectrom, 20, Meriluoto J.A.O., Sandström A., Eriksson J.E., Remaud G., Craig A.G., Chattopadhyaya J., 1989, Structure and toxicity of a peptide hepatotoxin from the cyanobacterium Oscillatoria agardhii, Toxicon, 27, Meriluoto J., Codd A.G., Toxic cyanobacterial monitoring and cyanotoxin analyses. Acta Acad. Aboensis B 65 (1). Ǻbo, Ǻbo Akademi University Press. Namikoshi M., Choi B.W., Sakai R., Sun F., Rinehart K.L., Carmichael W.W., Evans W.R., Cruz P., Munro M.H.G., Blunt J.W., 1994, New nodularin: A general method for structure assignment, J. Org. Chem., 59, Oberholster P.J., Botha A.-M., Grobbelaar J.U., 2004, Microcystis aeruginosa: source of toxic microcystins in drinking water, Afr. J. Biotechnol., 3, Rapala J., Erkomaa K., Kekkonen J., Sivonen K., Lahti K., 2002, Detection of microcystins with protein phosphatase inhibition assay, high-performance liquid chromatography-uv detection and enzyme-linked immunosorbent assay, comparison of methods, Anal. Chim. Acta, 466, Rinehart K.L., Harada K.-I., Namikoshi M., Chen C., Harvis C.A., Munroe M.H.G., Blunt J.W., Mulligan P.E., Beasley V.R., Dahlem A.M., Carmichael W.W., 1988, Nodularin, microcystin and the configuration of Adda, J. Am. Chem.Soc., 110, Rinehart K.L., Namikoshi M., Choi W., 1994, Structure and biosynthesis of toxins from bluegreen algae (cyanobacteria), J. Appl. Phycol., 6, Rippka R., 1988, Isolation and purification of cyanobacteria, Metods Enzymol., 167, Saito K, Konno A, Ishii H, Saito H, Nishida F, Abe T., Chen Ch., 2001, Nodularin-Har: a new nodularin from Nodularia, J. Nat. Prod., 64, Sano T., Kaya K., 1995, A 2-amino-2-butenoic acid (Dhb)-containing microcystin isolated from Oscillatoria agardhii, Tetrahedron Lett., 36, Sano T., Kaya K., 1998a, Two new (E)-2-amino-2-butenoic acid (Dhb)-containing microcystins isolated from Oscillatoria agardhii, Tetrahedron, 54, Sano T., Beattie K.A., Codd G.A., Kaya K., 1998b, Two (Z)-dehydrobutyrine-containing microcystins from a hepatotoxic bloom of Oscillatoria agardhii from Soulseat Loch, Scotland, J. Nat. Prod., 61, Sano T., Takagi H., Kaya K., 2004, A Dhb-microcystin from the filamentous cyanobacterium Planktothrix rubescens, Phytochem., 65, Sivonen K., Niemelä S.I., Niemi R.M., Lepistö L., Luona T.H., Räsänen L.A., 1990a, Toxic cyanobacteria (blue-green algae) in Finnish fresh and coastal waters, Hydrobiologia, 190, Sivonen K., 1990b, Effect of light, temperature, nitrate, orthophosphate, and bacteria on growth of the hepatotoxin production by Oscillatoria agardhii strains, Appl. Environ. Microbiol., 56,

19 Cyanobacterial hepatotoxins, microcystins and nodularins 21 Spoof L., Vesterkvist P., Lindholm T., Meriluoto J., 2003, Screening for cyanobacterial hepatotoxins, microcystins and nodularin in environmental water samples by reversed-phase liquid chromatography-electrospray ionisation mass spectrometry, J. Chromatogr. A, 1020, Stal L.J., Albertano P., Bergman B., Bröckel K., Gallon J.G., Hayes P.K., Sivonen K., Walsby A.E., 2003, BASIC: Baltic Sea cyanobacteria. An investigation of the structure and dynamics of water blooms of cyanobacteria in the Baltic Sea responses to a changing environment, Continental Shelf Res., 23, Tillett D., Dittmann E., Erhard M., von Döhren H., Börner T., Neilan B.A., 2000, Structural organization of microcystin biosynthesis in Microcystis aeruginosa PCC7806: an integrated peptide-polyketide synthetase system, Chem. Biol., 7, Tonk L., Visser P.M., Christiansen G., Dittmann E., Snelder E., Wiedner C., Mur L.R., Huisman J., 2005, The microcystin composition of the cyanobacterium Planktothrix agardhii changes toward a more toxic variant with increasing light intensity, Appl. Environ. Microbiol., 71, Tsuji K., Watanuki T., Kondo F., Watanabe M.F., Suzuki S., Nakazawa H., Suzuki M., Uchida H., Harada K.-I., 1995, Stability of microcystins from cyanobacteria- II. Effect of UV light on decomposition and isomerization, Toxicon, 33, Welker M., Christiansen G., von Döhren H., 2004, Diversity of coexisting Planktothrix (cyanobacteria) chemotypes deduced by mass spectral analysis of microcystins and other oligopeptides, Arch. Microbiol., 182, Yuan M., Namikoshi M., Otsuki A., Watanabe M.F., Rinehart K.L., 1999, Electrospray ionization mass spectrometric analysis of microcystins, cyclic heptapeptide hepatotoxins: modulation of charge states and [M+H] + to [M+Na] + ratio, J. Am. Soc. Mass Spectrom. 10,