LC-MS Analysis with Thiol Derivatization to Differentiate [Dhb7]- from

a hepatotoxic bloom of Oscillatoria agardhii from Soulseat Loch, Scotland J. Nat. ..... rubescens as compared to different microcystins Toxicon 20...
0 downloads 0 Views 508KB Size
Article pubs.acs.org/est

LC-MS Analysis with Thiol Derivatization to Differentiate [Dhb7]from [Mdha7]‑Microcystins: Analysis of Cyanobacterial Blooms, Planktothrix Cultures and European Crayfish from Lake Steinsfjorden, Norway Christopher O. Miles,†,* Morten Sandvik,† Sigrid Haande,‡ Hezron Nonga,†,§ and Andreas Ballot†,‡ †

Norwegian Veterinary Institute, P.O. Box 750 Sentrum, N-0106 Oslo, Norway Norwegian Institute for Water Research, Gaustadalléen 21, N-0349 Oslo, Norway § Sokoine University of Agriculture, P.O. Box 3021, Morogoro, Tanzania ‡

S Supporting Information *

ABSTRACT: Kinetic studies showed that [Asp3, Dhb7]MC-RR reacted with mercaptoethanol hundreds of times more slowly than MC-RR and a range of other [Mdha7]-containing microcystin congeners. The difference in reaction rate was sufficiently large that derivatization of microcystin-containing samples with mercaptoethanol, followed by LC-MS analysis, clearly discriminated between microcystins containing the isobaric [Dhb7]- and [Mdha7]-groups. Application of this approach, using LC-MS with both-ion trap and triplequadrupole mass spectrometers, to water samples and Planktothrix cultures from Lake Steinsfjorden, Norway, demonstrated the presence of [Asp3, Dhb7]MC-RR (5), [Asp3]MC-RY (14), and [Asp3]MC-LY (16), as well as analogues tentatively identified as [Asp3]MC-RR (4), [Asp3, DMAdda5, Dhb7]MC-LR (6), [Asp3, Dhb7]MC-HtyR (8), [Asp3]MC-HtyR (9), [Asp3, Dhb7]MC-LR (10), [Asp3]MC-LR (11), [Asp3, Dhb7]MC-RY (15), and [Asp3, Dhb7]MC-LY (17), together with low levels of several other analogues. This is the first use of this thiol-based LC-MS approach to identify Dhb-containing microcystins, and allowed identification of LC-MS peaks in a mixture of [Mdha7]- and [Dhb7]-congeners of [Asp3]MC-RR (4, 5), -RY (14, 15), and -LY (16, 17) in the samples from L. Steinsfjorden. This is also the first report of MC-RY-congeners outside of Africa, or in Planktothrix spp. Analysis of European crayfish (Astacus astacus) taken from L. Steinsfjorden revealed the presence of only trace levels of microcystins in the edible parts.



INTRODUCTION

RR) by LC-MS in cases where standards are not readily available, making reliable identification difficult. Large quantities of Planktothrix spp. have been observed in shallow areas and on banks of L. Steinsfjorden in spring after ice melting and in the summer months of July and August.1 The accumulated Planktothrix biomass is a potential food source for the omnivorous European crayfish (Astacus astacus) population in L. Steinsfjorden. Toxins, presumed to be microcystins, have been detected in hepatopancreas and muscle tissue of European crayfish (A. astacus) from L. Steinsfjorden during a study in 1997.8 Efficient reaction of the [Mdha7]-moiety of MC-LR, -YR, and -RR with the thiol groups of glutathione and cysteine has been described, along with thorough characterization of the resulting derivatives.9 A recent method describes derivatization of microcystins with thiols under weakly basic conditions as an aid to LC-MS analysis of microcystin analogues, including details of the chromatographic and mass spectral characteristics

Lake Steinsfjorden is a dimictic and mesotrophic lake in southeastern Norway. The lake has a surface area of 13.9 km2 and a maximum depth of 24 m, and is used for agricultural irrigation and recreational purposes, for example, fishing and water sports.1 It is the most important crayfish locality in Norway and accounts for approximately 25% of the annual harvest.2,3 For decades, L. Steinsfjorden has been subject to regular cyanobacterial blooms, often dominated by the microcystinproducing cyanobacterial species Planktothrix agardhii and Planktothrix rubescens.1,4 A range of microcystins (Figure 1) have been reported in L. Steinsfjorden, or in cultures of Planktothrix spp. isolated from L. Steinsfjorden, including desmethyl-MC-LR, desmethyl-MC-RR, desmethyl-MC-YR, and desmethyl-MC-HtyR, although the site of demethylation has not been reported.4−6 A range of German, Austrian, and Swiss lakes affected by Planktothrix blooms were recently shown7 to contain [Asp3]MC-RR, [Asp3, Dhb7]MC-RR, [Asp3]MC-HtyR, and [Asp3]MC-LR. However it is difficult to differentiate microcystins containing the isobaric amino acids Mdha and Dhb at position-7 (e.g., [Asp3]MC-RR and [Asp3, Dhb7]MC© 2013 American Chemical Society

Received: Revised: Accepted: Published: 4080

December March 22, March 26, March 26,

19, 2012 2013 2013 2013

dx.doi.org/10.1021/es305202p | Environ. Sci. Technol. 2013, 47, 4080−4087

Environmental Science & Technology

Article

Figure 1. Structures of microcystin analogues discussed in the text, with the seven amino acid residues labeled 1−7, and showing the Addafragmentation (m/z 135 and [MH−134]+). Amino acid variations in the structures are indicated with standard 3- or 4-letter amino acid abbreviations. Compound numbers refer to analogues identified in the samples (Table 2, Figure 2), with the remaining analogues being either standards or unidentified analogs.

of the thiol derivatives.10 This study also showed that nodularin, which is similar to MC-LR but is a pentapeptide containing N-methyl-dehydrobutyrine (Mdhb) at position-5, did not react at a detectable rate with mercaptoethanol under these conditions, and it was suggested that this approach might be able to discriminate between microcystins containing the isobaric [Dhb7]- and [Mdha7] residues.10 However, no data on the relative reaction rates of [Dhb7]- and [Mdha7]-containing microcystins is available to confirm this hypothesis. Here we report measurement of the relative rates of reaction with mercaptoethanol of a range of microcystins, including an authenticated [Dhb7]-containing congener, and show that the [Dhb7]-congener reacted several hundred-fold more slowly than Mdha-containing microcystins. We then applied this method to confirm the presence of a range of [Dhb7]- and [Mdha7]-containing microcystin analogues by LC-MS in cyanobacterial bloom material from L. Steinsfjorden, Norway, including congeners of MC-RY. Analysis of extracts from cultures isolated from L. Steinsfjorden, and European crayfish harvested from L. Steinsfjorden, revealed the presence of some of the same microcystin analogues.

standards were from Alexis Biochemicals (Grünberg, Germany), [Asp3, (E)-Dhb7]MC-RR was from Cyano Biotech GmbH, Berlin, Germany, and [Dha7]MC-LR was from IMB NRC, Halifax, NS, Canada. An extract of a Microcystis aeruginosa culture (NIVA-CYA548) from Lake Naivasha, Kenya, containing [Asp3]MC-RY and [Asp3]MC-LY as the major microcystin congeners, was obtained as described elsewhere.10 An extract from a cyanobacterial bloom containing MC-RY was obtained from Lake Victoria, Tanzania, in 2010. Specimens of purified MC-RY,11 [Asp3]MC-RY and [Asp3]MC-LY were also available from these sources (C. O. Miles et al., unpublished). Water. Sample 1. A concentrated phytoplankton sample was derived from a net sample (25 μm mesh) taken from L. Steinsfjorden, Norway, on 31 August 2011. Several net hauls from 13 m depth up to the surface were combined to give 800 mL of concentrated water sample which was stored frozen at −20 °C. The sample was thawed, ultrasonicated, and filtered (Whatman #1 filter paper, Whatman Ltd., Maidstone, UK). The filter was rinsed with distilled water (100 mL) and the washings added to the filtrate. Activated HP-20 resin (4 g) (DIAION HP-20, Mitsubishi Chemical Corporation, Tokyo, Japan) was added and the sample was shaken for 24 h. The resin was recovered by filtration through nylon netting (200 μm mesh), rinsed with water, and eluted slowly with 40 mL MeOH. The eluate was analyzed qualitatively by LC-MS2



MATERIALS AND METHODS Chemicals and Cyanobacterial Samples. Mercaptoethanol was from Sigma−Aldrich, Oslo, Norway. Microcystin (MCRR, MC-LR, MC-YR, MC-LA, MC-LY, MC-LF, MC-LW) 4081

dx.doi.org/10.1021/es305202p | Environ. Sci. Technol. 2013, 47, 4080−4087

Environmental Science & Technology

Article

(including the minor stereoisomer,10 when detectable), were measured and used to calculate the percent conversion of each analogue at each analysis time. The data was fitted to pseudo first-order kinetics to give the half-life of the reaction for each analogue in the mixture (Table 1).

(methods A1 and A2) and quantitatively by LC-MS/MS (method B). Sample 2. Quantitative water samples were taken on 31 August 2011 at 1 m intervals with a 3.4 L Limnos water sampler (Limnos Oy, Turku, Finland) to determine the phytoplankton composition. Volume-weighted samples for the epi- (0−7m) and metalimnion (8−14 m) were prepared according to the bathymetry of L. Steinsfjorden in the laboratory the same day and fixed with Lugol’s solution. Phytoplankton taxa were counted in sedimentation chambers (Hydro-Bios Apparatebau GmbH Kiel, Germany) using an inverted microscope (Leica DM IRB, Leica Microsystems, Wetzlar, Germany) according to Utermöhl.12 Planktothrix species were identified as described by Halstvedt et al.1 Samples (100 mL) from each depth (0−14 m) were filtered on a cellulose nitrate membrane filter (40 mm diameter, 0.45 μm pore size, Sartorius AG, Gö ttingen, Germany). Filament lengths, unit counts, and the mean diameter were used to determine the biovolume concentration of Planktothrix spp. Sample 3. An integrated water sample (1 L) from 0−14 m was collected from L. Steinsfjorden on 26 September 2012. The sample was stored frozen until required for analysis, then thawed, ultrasonicated, and shaken for 24 h with activated HP20 resin (3 g). The resin was recovered as above and eluted slowly with 30 mL MeOH. The eluate was concentrated under vacuum and redissolved in 2 mL MeOH−H2O for LC-MS2 analysis (methods A1 and A2). Crayfish. Ten European crayfish (A. astacus) were collected from L. Steinsfjorden on August 2011, and frozen. Crayfish were briefly cooked in boiling water (5 min) to facilitate separation of the flesh from the shells, cooled, and the head/ thorax separated from the tails. The fleshy tissue was removed from both samples and homogenized separately. Homogenized tissue and MeOH (1:4, w/v) were ultrasonicated together for 5 min then shaken for 15 min, centrifuged, and an aliquot of the supernatant filtered (Spin-X, 0.2 μm) for LC-MS2 analysis (methods A1 and A2). Cultures. Four P. aghardii and three P. rubescens strains isolated from L. Steinsfjorden between 1984 and 2004 were purchased from the NIVA culture collection (NIVA-CYA56/3, 137, 406, 408, 532, 537 and 544), and grown in Z8 medium.4,13 Samples were frozen until analysis, then thawed, ultrasonicated, and aliquots (0.5 mL) added to MeOH (0.5 mL) and filtered (Spin-X, 0.2 μm) for LC-MS2 analysis (method A2). To the remaining culture was added 0.5 g activated HP-20 resin. After shaking overnight, the resin was recovered as above, eluted slowly with MeOH (15 mL), and evaporated to dryness in a stream of nitrogen. The residue was taken up in 1:1 MeOH− H2O (1 mL) for LC-MS2 analysis (method A2). Kinetics of Thiol Derivatization. A mixed standard containing MC-RR, MC-LR, MC-YR, MC-LA, MC-LY, MCLF, MC-LW, prepared as described elsewhere, 10 was supplemented with standards of [Dha7]MC-LR, and [Asp3, Dhb7]MC-RR such that all analogues were present at ca. 0.4−1 μg/mL in MeOH−H2O (1:1). To the mixed standard (200 μL) was added sodium carbonate buffer (0.2 M, pH 9.7, 50 μL) and the vial was allowed to equilibrate to the tray temperature of the autosampler (set to 5 °C) before addition of mercaptoethanol (1 μL) with brief vortex-mixing. Progress of the reaction was followed by LC-MS2 analysis (method A1) for 24 h. Peak areas (for [MH]+ or [M+2H]2+, as appropriate to the microcystin congener, without MS fragmentation) for each microcystin congener, and its mercaptoethanol derivative

Table 1. Measured Half-Lives for Reaction of Microcystin Congeners with Mercaptoethanol at 5 °C and pH 9.7 compound

m/z [MH]+

t1/2 (h)

MC-RR [Asp3, Dhb7]MC-RR (5) MC-YR MC-LR [Dha7]MC-LR MC-LA MC-LY MC-LW MC-LF

1038 1024 1045 995 981 910 1002 1025 986

0.34 170 0.31 0.42 0.31 0.71 0.64 0.84 0.70

Toxin Analysis. LC-MS2 (Method A1). Liquid chromatography (5 μL sample injected) was performed on a Symmetry C18 column (3.5 μm, 100 × 2.1 mm; Waters, Milford, MA), using a Surveyor MS Pump Plus and a Surveyor Auto sampler Plus (Finnigan, Thermo Electron Corp., San Jose, CA) eluted with a linear gradient of acetonitrile (A) and water (B) each containing 0.1% formic acid and a mobile phase flow of 300 μL min−1. The gradient was from 22.5% to 42.5% A over 4 min, then to 75% A at 10 min, to 95% A at 11 min (1 min hold) followed by a return to 22.5% A with a 3-min hold to equilibrate the column. The HPLC system was coupled to a Finnigan LTQ ion trap mass spectrometer (Finnigan Thermo Electron Corp., San Jose, CA, USA) operated in full-scan positive ion ESI mode (m/z 500−1600). The ion injection time was set to 100 ms with a total of three microscans. ESI parameters were a spray voltage of 6 kV, a capillary temperature of 375 °C, a sheath gas rate of 55 units N2 (ca. 550 mL/min) and an auxiliary gas rate of 5 units N2 (ca. 50 mL/min). ESI settings were optimized while continuously infusing (syringe pump) 0.1 μg/mL of the MC-RR (3, m/z 1038.5) standard at 10 μL/min. MS2 spectra were acquired using the same chromatographic conditions for specified m/z values, but with scanning up to m/z 1150, isolation width 2.0, normalized collision energy 50, Activation Q 0.250, and activation time 0.25 ms. For lake water and culture extracts, mercaptoethanolderivatization and control samples were held at ca. 20 °C for ca. 2 h before LC-MS2 analysis to allow essentially complete reaction of [Mdha7]- and [Dha7]-microcystins. LC-MS2 (Method A2). Analysis was performed as for method A1, except that a Phenomenex Kinetex C18 column (2.6 μm, 75 × 2.1 mm, 100 Å) was used with a linear gradient from 20% to 44% A over 40 min, then to 95% A for 2 min and return to 20% A (3 min hold). LC-MS/MS (Method B). Liquid chromatography was performed on a Luna C18 column (100 × 2.0 mm, 3 μm, 100 Å; Phenomenex, Torrance, CA), using an Accela pump equipped with an autosampler (Thermo Scientific, San Jose, CA) with a mobile phase flow of 300 μL min−1, and injection volume of 10 μL. Separation was achieved using a linear gradient starting with acetonitrile−water (22:78, both containing 0.1% formic acid) rising to 55% acetonitrile over 12 min, and then to 100% acetonitrile over 3 min followed by a 5 min hold before returning to the starting eluent and re-equilibrated 4082

dx.doi.org/10.1021/es305202p | Environ. Sci. Technol. 2013, 47, 4080−4087

Environmental Science & Technology

Article

Table 2. Microcystin Congeners Detected by LC-MS2 (Method A2, with and without Mercaptoethanol Derivatization), And Their Relative Abundance, In Water Samples and Planktothrix Isolates from L. Steinsfjorden water Samples (%)c

microcystin

status

[MH]+ m/za

desmethyl[Asp3] MC-RR (1) desmethyl[Asp3, Dhb7]MC-RR (2) [Asp3, Dha7]MCRR (3) [Asp3]MC-RR (4) [Asp3, Dhb7]MCRR (5) [Asp3, DMAdda5, Dhb7]MC-LR (6) [Asp3, DMAdda5] MC-RY (7) [Asp3, Dhb7]MCHtyR (8) [Asp3]MC-HtyR (9) [Asp3, Dhb7]MCLR (10) [Asp3]MC-LR (11) [Asp3, Dhb7]MCAhaR (12) [Asp3, Dha7]MCRY (13) [Asp3]MC-RY (14) [Asp3, Dhb7]MCRY (15) [Asp3]MC-LY (16) [Asp3, Dhb7]MCLY (17)

unidentified

1010f

4.11

fast

unidentified

1010f

4.19

slow

0.3

trace

tentative

1010f

4.94

fast

0.3

trace

tentative confirmed

1024f 1024f

5.06 5.19

fast slow

3.9 79.8

20.9 65.9

tentative

967

7.07

slow

0.4

0.8

tentative

1017

10.22

fast

tentative

1045

12.84

slow

tentative

1045

12.99

fast

tentative

981

14.12

slow

tentative

981

14.31

fast

tentative

995

16.27

slow

tentative

1017

18.64

fast

confirmed

1031

19.41

fast

1.0

tentative

1031

20.23

slow

0.6

confirmed

988

28.16

fast

trace

tentative

988

29.12

slow

trace

Rt (min)b

thiol reactiond

sample 1 (2011)

cultures (%)c

sample 3 (2012)

CYA 56/3 P. aga.e

CYA137 P. aga.e

0.4

0.4

0.6

0.8

64.7

66.9

0.1 1.2

CYA408 P. rub.e

0.2

0.2

0.4

0.2

87.3

38.6

0.3

0.8

CYA537 P. aga.e

CYA544 P. rub.e

0.4

0.5

0.6

0.9

0.8

0.9

60.7

98.0

98.0

0.7

0.5

18.9

0.1

10.3

CYA532 P. aga.e

0.3 0.5

0.1 12.4

CYA406 P. rub.e

0.1 11.7

42.1

2.4

2.5

5.4

0.6

0.5

0.7

28.8

27.0

28.0

2.4

1.4

3.7

0.1

1.4 trace

a Nominal value. bUsing LC-MS2 (method A2). Under this system, the microcystin standards had the following retention times: [Asp3, Dhb7]MCRR (5), 5.19 min; MC-RR, 6.09 min; MC-YR, 12.86 min; MC-LR, 14.17 min; [Dha7]MC-LR, 14.43 min, [Asp3]MC-RY (14), 19.41 min; MC-RY, 20.89 min; MC-LA, 26.58 min; [Asp3]MC-LY (16), 28.16 min; MC-LY, 29.41; MC-LW, 36.34 min; MC-LF, 37.59 min. cPercentage of the total microcystins identified in each sample. dWith mercaptoethanol; Fast = complete reaction within 2 h, Slow = no detectable reaction within 2 h. e NIVA accession number and species name (P. aga. = P. agardhii, P. rub. = P. rubescens), from Rohrlack et al.4. fDetected as doubly charged [M +2H]2+ ions (m/z 606.1 or 613.1) using methods A2 and B. At least two other minor microcystins with [M+2H]2+ ions at m/z 606.1 also eluted at 4−5 min.

(5 min). The HPLC was coupled to a TSQ quantum access tandem quadrupole mass spectrometer (Thermo Scientific) operating with an electrospray interface (ESI) in positive mode. The mass analyzer was operated in multiple reaction monitoring (MRM) mode for the transitions listed below. The parameters for the ESI as well as optimum collision energies were set using automatic and semiautomatic tuning procedures while tuning solutions of MC-RR, -LR, and -LA standards (35−701 ng mL−1) were continuously infused into the mobile phase at starting condition for the gradient. The parameters for the ESI interface were adjusted as follows: spray voltage 4.5 kV, capillary temperature 180 °C, skimmer offsets of 0, 8, and 10 for MC-RR, -LR, and -LA respectively, and sheath gas and auxiliary gas flows of 55 and 5 arbitrary units, respectively. The argon pressure in Q2 was set to 1.3 mTorr, and divert valve was set to waste during the first 2 min of the run. Transitions monitored were 519.7→135.0, 512.7→135.0 (collision energy 32 for microcystins with two Arg residues);

1045.7→135.0, 995.5→135.0, 981.5→135.0, 1031.5→135.0, 995.5→135.0, 981.5→135.0 (collision energy 63, for microcystins with one Arg residue); 910.5→776.5, 1002.5→868.5, 1025.5→891.2, 986.2→852.2 (collision energy 20, for microcystins without Arg residues); and 558.5→135.0 (mercaptoethanol derivative of MC-RR). Microcystins were quantified using external four-point calibration curves in the concentration range 0.1−250 ng/mL. Where authentic standards were not available, quantitation of microcystins containing two, one or no arginine groups was performed against standards of MC-RR, [Dha7]MC-LR, and MC-LY, respectively. The limits of detection (LOD) and the limits of quantification (LOQ), defined as the minimum concentration resulting in two diagnostic MRM traces with signal-to-noise ratios (S/N) of 3 and 10, respectively, were MC-RR, 3 and 11; MC-YR, 13 and 43; MC-LR, 27 and 90; and MC-LA, 5 and 17 pg/mL. 4083

dx.doi.org/10.1021/es305202p | Environ. Sci. Technol. 2013, 47, 4080−4087

Environmental Science & Technology

Article

Figure 2. LC-MS2 (method A2) chromatograms (base peak mode for m/z 500−570 + 900−1180) of sample 1 from L. Steinsfjorden (2011) treated with carbonate buffer: (A) without added mercaptoethanol, and; (B) 2 h after addition of mercaptoethanol, with m/z values for [M+2H]2+ (MC-RR congeners 1−5) or [M+H]+ (microcystins 6−17) and compound numbers (Figure 1 and Table 2) shown for prominent microcystin peaks. Compound numbers for microcystin congeners derivatized with mercaptoethanol (major isomer) in chromatogram B are appended with an “a” (e.g., 4a is the mercaptoethanol derivative of [Asp3]MC-RR (4)), and the vertical scale is modified for segments Rt 6−17 min (0−55%) and 17−30 min (0−7%) to facilitate display of minor congeners.



RESULTS AND DISCUSSION A previous study with the pentapeptide nodularin indicated that it might be possible to discriminate between isobaric [Mdha7]and [Dhb7]-containing microcystins by exploiting their difference in reactivity toward thiols in combination with LC-MS analysis.10 However, it is necessary to confirm this with authentic [Dhb7]-containing microcystins before this approach can be used with confidence on natural samples, as nodularin is a pentapeptide and contains an N-methyl-Dhb-group (rather than a Dhb-group), both of which could influence its reactivity toward thiols. A commercial standard of [Asp3, Dhb7]MC-RR (5) was added to a mixed standard containing a range of microcystins and treated with mercaptoethanol under weakly basic conditions at 5 °C, and the progress of the reactions was followed by LC-MS2 (method A1). This temperature was chosen so as to reduce the reaction rate of the [Mdha7]containing analogues, as in the previous study10 these analogues reacted too quickly for accurate rate measurements by LC-MS2 at pH 9.7. The kinetic study was conducted with a mixture so that the rate of reaction of all analogues could be measured under identical reaction conditions. Results of this study (Table 1) showed that [Asp3, Dhb7]MC-RR reacted with mercaptoethanol 500-times more slowly than MC-RR, and 200-times more slowly than the slowest-reacting [Mdha7]-containing microcystin tested. At room temperature, the reaction of [Mdha7]-containing microcystins was essentially complete within 2 h (half-life ca. 15 min),10 at which time less than 3% of any corresponding [Dhb7]-congeners would be expected to have reacted (based on equations for first-order kinetics and an expected half-life ≥ ca. 50 h). This approach can therefore be used with confidence to differentiate between [Mdha7]- and [Dhb7]-containing microcystins as long as the reaction time is kept relatively short. L. Steinsfjorden is regularly affected by cyanobacterial blooms dominated by Planktothrix spp. Numerous cultures of Planktothrix strains isolated from L. Steinsfjorden have been analyzed by LC-MS/MS and reported to contain desmethylanalogues of MC-RR, MC-LR, MC-YR, and MC-HtyR.4−6

However, it is not uncommon for Planktothrix-dominated blooms to contain [Dhb7]-congeners of microcystins.7 During analysis of a sample collected from L. Steinsfjorden in 2011, we observed several peaks in the LC-MS2 chromatograms (method A1) that possessed similar masses to common microcystins, but which appeared not to react with mercaptoethanol (Table 2). MS2 spectra of these peaks were consistent with microcystin analogues (e.g., prominent fragments10,11,14 at m/z 599, 440, 426, or corresponding to [MH− 134]+), suggesting the presence of [Dhb7]microcystins. A slower gradient LC-MS2 method (method A2) was therefore used to obtain better separation of [Mdha7]- and [Dhb7]containing microcystin congeners in order to study the microcystin composition of the samples in more detail (Figure 2). A second water sample from L. Steinsfjorden, taken in 2012, and a small selection of microcystin-producing Planktothrix cultures isolated from L. Steinsfjorden that had been investigated in an earlier4 study, were analyzed using the same LC-MS2 method (A2) as the 2011 L. Steinsfjorden sample and their toxin profiles compared to that of the original sample. Investigation of the two water samples (from 2011 and 2012) from L. Steinsfjorden by LC-MS2 confirmed the presence of a range of major and minor peaks that displayed microcystin-like MS2 spectra (Supporting Information). Treatment of the samples with mercaptoethanol in weak base10 revealed that some of the putative microcystins reacted rapidly, whereas the others reacted only very slowly. As in the previous study,10 no changes in the abundance of Na+-adduct ions in the mass spectra of microcystins were observed during LC-MS2 analysis (methods A1 and A2) of the buffered thiol-reactions for any of the samples, presumably due to the small injection volume and use of an acidic mobile phase. The retention time, MS2 spectrum and lack of thiol reactivity of the major peak (Rt 5.19 min, m/z 513; Table 2) showed that it was identical with the standard of [Asp3, Dhb7]MC-RR (5). Minor peaks (Rt 19.41 min, m/z 1031, and Rt 28.16 min, m/z 988; Table 2) similarly corresponded to [Asp3]MC-RY (14) and [Asp3]MCLY (16) identified in a culture of African origin.10 The remainder of the microcystin peaks were tentatively identified 4084

dx.doi.org/10.1021/es305202p | Environ. Sci. Technol. 2013, 47, 4080−4087

Environmental Science & Technology

Article

from their fragmentation patterns10 observed during LC-MS2, and were assumed to be [Mdha7]-, or [Dhb7]-congeners on the basis of their reactivity, or lack of reactivity (respectively), with mercaptoethanol. Results of this analysis are presented in Table 2. Several of these compounds, such as [Asp3, Dhb7]MC-RR (5), [Asp3, Dhb7]MC-LR (10), and [Asp3, Dhb7]MC-HtyR (8),15−18 have been reported from Planktothrix spp. previously, although a number of the compounds have either not been reported previously (e.g., [Asp3, Dhb7]MC-RY (15) and [Asp3, Dhb7]MC-LY (17)) or have not been reported from Planktothrix (e.g., [Asp3]MC-RY (14) and [Asp3]MC-LY (16)). Interestingly, although essentially the same fragments were present in equivalent pairs of [Mdha7]- and [Dhb7]microcystin analogues, the ratios of some of the fragments showed characteristic changes as a result of this difference in residue-7 (e.g., m/z 682 in 10 and 11, Supporting Information) which could potentially be helpful for structure determination by mass spectrometry. When a mixture of MC-RR, MC-LR, [Dha7]MC-LR and MC-LA standards (found, 68.0, 37.9, 32.7, and 36.5 ng/mL, respectively) was analyzed by LC-MS/MS with a triplequadrupole instrument in MRM mode (method B), mercaptoethanol derivatization resulted in essentially complete disappearance of the parent toxins (2.0 ng/mL of residual MC-RR, with MC-LR, [Dha7]MC-LR and MC-LA not detectable). Furthermore, a peak corresponding to the mercaptoethanol derivative of MC-RR was detected after reaction, with a concentration (71.2 ng/mL) very close to that of MC-RR in the original sample (68.0 ng/mL, assuming the same response factor). A similar LC-MS/MS analysis of the L. Steinsfjorden water from 2011 (sample 1) (method B) gave results in close agreement with those shown in Table 2 obtained with an ion trap instrument (method A2). Although pairs of [Mdha7]- and [Dhb7]-congeners were not separated chromatographically in method B due to the rapid gradient, clear differences in peak intensities for congener-pairs were observed upon treatment with mercaptoethanol that were consistent with the [Mdha7]:[Dhb7] ratios observed with method A2 (Table 2). Thus, the measured concentrations in the extract before, and after, thiol derivatization were 4+5, 119 and 105 ng/mL; 8+9, 2.05 and 0.27 ng/mL; 10+11, 7.70 and 8.08 ng/mL; 12, 0.10 and not detected; and 14+15, 0.55 and 0.15 ng/mL. These preliminary results suggest that, with further development, the thiol derivatization approach could potentially be extended to quantitative determination of mixtures of [Mdha7]- and [Dhb7]-microcystins with LC-MS/ MS using triple-quadrupole instrumentation in MRM mode, even in the absence of chromatographic separation of congener pairs. It should be noted, however, that the possible influence of matrix effects, which could vary between the thiol-derivatized and underivatized samples, would need to be evaluated as part of method validation. The microcystins found in the L. Steinsfjorden water sample in 2011 can be related to the presence of P. rubescens and P. agardhii, which are both known microcystin producers.4,5,17−20 P. agardhii was the dominant cyanobacterium, comprising 68% and 46% of the cyanobacterial biomass in the epilimnion (0−7 m) and metalimnion (8−14 m), respectively (Table 3). The biomass of P. rubescens was much lower and comprised 13% of the cyanobacterial biomass in the epilimnion and 7% of the cyanobacterial biomass in the metalimnion. Other groups present were Bacillariophyceae, Cryptophyceae, Chrysophyceae, Chlorophyceae, Dinophyceae, and Euglenophyceae. The

Table 3. Phytoplankton Biomass Composition in L. Steinsfjorden on 31 August 2011 (Sample 2) biomass wet weight (mg/L) taxon

epilimnion (0−7 m)

metalimnion (8−14 m)

P. agardhii P. rubescens other cyanobacteria bacillariophyceae chlorophyceae chrysophyceae cryptophyceae dinophyceae euglenophyceae total

0.26 0.05 0.07 0.17 0.02 0.06 0.06 0.14 0.002 1.02

0.23 0.04 0.24 0.40 0.04 0.38 0.33 0.85 0.01 2.76

biomasses of all groups were generally higher in the metalimnion (Table 3). Small numbers of the potentially microcystin-producing cyanobacteria Anabaena spp. and Microcystis spp. were also observed. The presence of both [Mdha7]- and [Dhb7]-microcystin congeners in the two water samples from L. Steinsfjorden prompted us to examine a selection of Planktothrix isolates from the NIVA Culture Collection of Algae that had been identified as microcystin producers in an earlier study.4 The majority of the microcystin analogues that were identified in the natural water samples were also detected in the selected Planktothrix cultures (Table 2). However, one notable feature of the isolates was that they produced essentially either [Mdha7]-microcystins, or [Dhb7]-microcystins, but not both. That this should be the case was proposed by Kurmayer et al. based on correlations between genotype and production of [Asp3]- and [Asp3, Dhb7]MC-RR (4 and 5) in Planktothrix cultures,21 but analytical methods available at that time could not differentiate [Asp3]- and [Asp3, Dhb7]-variants of other microcystin congeners. All congeners detected in the cultures from L. Steinsfjorden, as with the L. Steinsfjorden water samples, were Asp3-containing microcystins. As in an earlier study of Hartbeespoort Dam (A. Ballot, M. Sandvik, T. Rundberget, C. J. Botha and C. O. Miles, unpublished), minor amounts of [DMAdda5]-congeners of the major microcystin variants were also apparent in several of the samples, possibly suggesting low-level of incomplete O-methylation during Adda biosynthesis. Another interesting feature was the presence of MC-RY congeners 14 and 15 in both of the water samples and in several of the cultures. Curiously, no MC-YR analogues were detected although MC-HtyR congeners (8 and 9) were present. None of the tested cultures produced [Asp3, Dhb7]MC-RY (15) or -LY (17), minor components of the L. Steinsfjorden water samples, possibly because strains producing these compounds constituted only a small proportion of the Planktothrix community in the lake when they were isolated. What these finding mean in terms of the amino acid specificity of Planktothrix’ microcystin synthase complexes is not at this stage clear, but the results highlight the importance of determining whether Mdha or Dhb is present at position-7. The reported mouse i.p. LD50 of [Asp3, Dhb7]MC-LR (10) (70 μg/kg)18 is less than half that of [Asp3]MC-LR (11) (160− 300),22 although their IC50 values for PP2A inhibition are rather similar;23 similarly, the mouse i.p. LD50 of [Asp3, Dhb7]MC-RR (5) (250 μg/kg)18 is less than half that of MCRR (500−800 μg/kg).24 The limited data suggests that the presence of Dhb, instead of Mdha, at position-7, may lead to a 4085

dx.doi.org/10.1021/es305202p | Environ. Sci. Technol. 2013, 47, 4080−4087

Environmental Science & Technology



modest increase in the acute mammalian toxicity of microcystin congeners.25 However, the much lower reactivity of [Dhb7]microcystins toward thiols could be expected to influence their in vivo uptake, distribution, metabolism and excretion due to their reduced ability to form conjugates with thiol-containing biomolecules. L. Steinsfjorden is a major fishery for the European crayfish (A. astacus) in Norway,3 and earlier studies using mouse bioassays indicated the presence of low levels of toxins, possibly microcystins, in the digestive glands of crayfish from this lake.8 Samples of crayfish taken from L. Steinsfjorden at about the same time as water sample 1 were cooked and the edible and nonedible soft tissues extracted and analyzed for microcystins by LC-MS2 (method A1). While 10 μg/kg of [Asp3, Dhb7]MCLR (10) together with traces of [Asp3]MC-RR (4) and [Asp3, Dhb7]MC-RR (5) were detected in tissue from the head/ thorax, only traces (less than 1 μg/kg) of these compounds were detectable in the edible tissues from the tail. The provisional TDI (tolerable daily intake) of MC-LR for humans26 is 0.04 μg kg−1 d−1. To exceed the TDI of microcystins, a human of 70 kg bodyweight would have to consume more than 2.8 kg crayfish tail tissue per day. However, the presence of toxins in crayfish from L. Steinsfjorden suggests that a more extensive study of the uptake and depuration of microcystins, and of the possible effects of the toxins on the health of this population of this locally threatened species, is warranted. In summary, [Dhb7]microcystins were shown to be more than 2 orders of magnitude less reactive than [Mdha7]microcystins toward mercaptoethanol. A recently developed mercaptoethanol-derivatization/LC-MS procedure10 was found to be a convenient and powerful tool for discriminating between microcystins containing the isobaric amino acids Mdha and Dhb at position-7, both in Planktothrix cultures and Planktothrix-containing cyanobacterial blooms. The in situ derivatization is very simple, takes only a few minutes to perform and, while it is necessary to wait 1−2 h (depending10 on temperature and pH) for complete reaction, this time can be used to analyze the corresponding underivatized (control) samples. This approach revealed a variety of [Mdha7]- and [Dhb7]-containing microcystins in Planktothrix-dominated blooms and P. rubescens and P. agardhii cultures from L. Steinsfjorden, including congeners of MC-RY (14 and 15) for the first time outside of Africa, and the first [Dhb7]-congener of MC-RY (15) and MC-LY (17). Some of these microcystins were detected in European crayfish taken from L. Steinsfjorden, but only at very low concentrations.



ACKNOWLEDGMENTS

We thank the Norwegian Institute for Water Research (NIVA) for providing the Planktothrix cultures, and IMB NRC, Halifax, NS, Canada for NMR-quantitated cyanotoxin standards, and Jonathan Puddick for helpful discussions. This study was supported by grant 196085/V10 (Monitoring of Cyanotoxins in Southern Africa) from The Research Council of Norway, and by The Norwegian Programme for Development, Research and Higher Education (NUFU PRO 07/10224) and SIDA SAREC: VICRES Endocrine disruptors project (SUA).



ABBREVIATIONS 3-amino-9-methoxy-2,6,8-trimethyl-10-phenyl-4,6decadienoic acid Aha aminoheptanoic acid Dha dehydroalanine Dhb dehydrobutyrine DMAdda 9-desmethylAdda ESI electrospray ionization Hty homotyrosine MC microcystin Mdha N-methydehydroalanine Mdhb N-methyldehydrobutyrine MRM multiple reaction monitoring TDI tolerable daily intake Adda



REFERENCES

(1) Halstvedt, C. B.; Rohrlack, T.; Andersen, T.; Skulberg, O.; Edvardsen, B. Seasonal dynamics and depth distribution of Planktothrix spp. in Lake Steinsfjorden (Norway) related to environmental factors. J. Plankton Res. 2007, 29, 471−482. (2) Skurdal, J.; Garnås, E.; Taugbøl, T. Management strategies, yield and population development of the noble crayfish Astacus astacus in Lake Steinsfjorden. Bull. Fr. Pêche Piscic. 2002, 367, 845−860. (3) Johnsen, S. I. Nasjonal Overvåking av EdelkrepsPresentasjon av Overvåkingsdata og Bestandsstatus; NINA Rapport 492; Norsk institutt for naturforskning: Lillehammer, 2010; 100 pp. (4) Rohrlack, T.; Edvardsen, B.; Skulberg, R.; Halstvedt, C. B.; Utkilen, H. C.; Ptacnik, R.; Skulberg, O. M. Oligopeptide chemotypes of the toxic freshwater cyanobacterium Planktothrix can form subpopulations with dissimilar ecological traits. Limnol. Oceanogr. 2008, 53, 1279−1293. (5) Hormazábal, V.; Østensvik, Ø.; Underdal, B.; Skulberg, O. M. Simultaneous determination of the cyanotoxins anatoxin a, microcystin desmethyl-3-RR, LR, RR and YR in water using liquid chromatography−mass spectrometry. J. Liq. Chromatogr. Relat. Technol. 2000, 23, 3155−3164. (6) Halstvedt, C. B.; Rohrlack, T.; Ptacnik, R.; Edvardsen, B. On the effect of abiotic environmental factors on production of bioactive oligopeptides in field populations of Planktothrix spp. (Cyanobacteria). J. Plankton Res. 2008, 30, 607−617. (7) Ostermaier, V.; Kurmayer, R. Application of real-time PCR to estimate toxin production by the cyanobacterium Planktothrix sp. Appl. Environ. Microbiol. 2010, 76, 3495−3502. (8) Aune, T.; Ramstad, H.; Skulberg, O. M.; Underdal, B.; Yndestad, M.; Østensvik, Ø. Cyanotoksiner og EdelkrepsToksinproduserende Blågrønnalger i Steinsfjorden Sommeren 1997; Norges Veterinærhøgskole, Institutt for Farmakologi, Mikrobiologi og Næringsmiddelhygiene: Oslo, 1997; 17 pp. (9) Kondo, F.; Ikai, Y.; Oka, H.; Okumura, M.; Ishikawa, N.; Harada, K.-i.; Matsuura, K.; Murata, H.; Suzuki, M. Formation, characterization, and toxicity of the glutathione and cysteine conjugates of toxic heptapeptide microcystins. Chem. Res. Toxicol. 1992, 5, 591−596. (10) Miles, C. O.; Sandvik, M.; Nonga, H. E.; Rundberget, T.; Wilkins, A. L.; Rise, F.; Ballot, A. Thiol derivatization for LC-MS

ASSOCIATED CONTENT

S Supporting Information *

LC-MS2 spectra for [MH]+ and/or [M+2H]2+ of 1−17 and from the available microcystin standards obtained using method A2. This material is available free of charge via the Internet at http://pubs.acs.org.



Article

AUTHOR INFORMATION

Corresponding Author

*Phone: +47 2321 6226; fax: +47 2321 6201; e-mail: chris. [email protected]. Notes

The authors declare no competing financial interest. 4086

dx.doi.org/10.1021/es305202p | Environ. Sci. Technol. 2013, 47, 4080−4087

Environmental Science & Technology

Article

identification of microcystins in complex matrices. Environ. Sci. Technol. 2012, 46, 8937−8944. (11) Miles, C. O.; Sandvik, M.; Nonga, H. E.; Rundberget, T.; Wilkins, A. L.; Rise, F.; Ballot, A. Identification of microcystins in a Lake Victoria cyanobacterial bloom using LC-MS with thiol derivatization. Toxicon 2012, in press (http://dx.doi.org/10.1016/ j.toxicon.2013.03.016). (12) Utermö hl, H. Zur Vervollkommnung der quantitativen Phytoplankton-Methodik. Mitt. Int. Verein. Limnol. 1958, 9, 1−38. (13) Kotai, J. Instructions for Preparation of Modified Nutrient Solution Z8 for Algae; Norwegian Institute for Water Research: Oslo, Norway, 1972. (14) Mayumi, T.; Kato, H.; Imanishi, S.; Kawasaki, Y.; Hasegawa, M.; Harada, K.-I. Structural characterization of microcystins by LC/MS/ MS under ion trap conditions. J. Antibiot. 2006, 59, 710−719. (15) Meriluoto, J. A. O.; Sandström, A.; Eriksson, J. E.; Remaud, G.; Craig, A. G.; Chattopadhyaya, J. Structure and toxicity of a peptide hepatotoxin from the cyanobacterium Oscillatoria agardhii. Toxicon 1989, 27, 1021−1034. (16) Sano, T.; Beattie, K. A.; Codd, G. A.; Kaya, K. Two (Z)dehydrobutyrine-containing microcystins from a hepatotoxic bloom of Oscillatoria agardhii from Soulseat Loch, Scotland. J. Nat. Prod. 1998, 61, 851−853. (17) Sano, T.; Kaya, K. A 2-amino-2-butenoic acid (Dhb)-containing microcystin isolated from Oscillatoria agardhii. Tetrahedron Lett. 1995, 36, 8603−8606. (18) Sano, T.; Kaya, K. Two new (E)-2-amino-2-butenoic acid (Dhb)-containing microcystins isolated from Oscillatoria agardhii. Tetrahedron 1998, 54, 463−470. (19) Christiansen, G.; Yoshida, W. Y.; Blom, J. F.; Portmann, C.; Gademann, K.; Hemscheidt, T.; Kurmayer, R. Isolation and structure determination of two microcystins and sequence comparison of the McyABC adenylation domains in Planktothrix species. J. Nat. Prod. 2008, 71, 1881−1886. (20) Sano, T.; Takagi, H.; Kaya, K. A Dhb-microcystin from the filamentous cyanobacterium Planktothrix rubescens. Phytochemistry 2004, 65, 2159−2162. (21) Kurmayer, R.; Christiansen, G.; Gumpenberger, M.; Fastner, J. Genetic identification of microcystin ecotypes in toxic cyanobacteria of the genus Planktothrix. Microbiology 2005, 151, 1525−1533. (22) Harada, K.; Ogawa, K.; Kimura, Y.; Murata, H.; Suzuki, M.; Thorn, P. M.; Evans, W. R.; Carmichael, W. W. Microcystins from Anabaena flos-aquae NRC 525−17. Chem. Res. Toxicol. 1991, 4, 535− 540. (23) Ikehara, T.; Imamura, S.; Sano, T.; Nakashima, J.; Kuniyoshi, K.; Oshiro, N.; Yoshimoto, M.; Yasumoto, T. The effect of structural variation in 21 microcystins on their inhibition of PP2A and the effect of replacing cys269 with glycine. Toxicon 2009, 54, 539−544. (24) Namikoshi, M.; Rinehart, K. L.; Sakai, R.; Stotts, R. R.; Dahlem, A. M.; Beasley, V. R.; Carmichael, W. W.; Evans, W. R. Identification of 12 hepatotoxins from a Homer Lake bloom of the cyanobacteria Microcystis aeruginosa, Microcystis viridis, and Microcystis wesenbergii: nine new microcystins. J. Org. Chem. 1992, 57, 866−872. (25) Blom, J. F.; Robinson, J. A.; Jüttner, F. High grazer toxicity of [D-Asp3,(E)-Dhb7]microcystin-RR of Planktothrix rubescens as compared to different microcystins. Toxicon 2001, 39, 1923−1932. (26) Kuiper-Goodman, T.; Falconer, I.; Fitzgerald, J. Human health aspects. In Toxic Cyanobacteria in Water: A Guide to their Public Health Consequences, Monitoring and Management, Chapter 4; Chorus, I., Bartram, J., Eds.; F & FN Spon for WHO: London, 1999; pp 114− 149.



NOTE ADDED AFTER ASAP PUBLICATION There was an error in Figure 1 in the version of this paper published April 12, 2013. The correct version published May 7, 2013.

4087

dx.doi.org/10.1021/es305202p | Environ. Sci. Technol. 2013, 47, 4080−4087