Chem. Res. Toxicol. 2004, 17, 1423-1433
1423
A Novel Pectenotoxin, PTX-12, in Dinophysis Spp. and Shellfish from Norway Christopher O. Miles,*,†,‡ Alistair L. Wilkins,†,§ Ingunn A. Samdal,† Morten Sandvik,† Dirk Petersen,| Michael A. Quilliam,⊥ Lars J. Naustvoll,# Thomas Rundberget,† Trine Torgersen,∇ Peter Hovgaard,O Dwayne J. Jensen,@ and Janine M. Cooney@ National Veterinary Institute, PB 8156 Dep., N-0033 Oslo, Norway, AgResearch Ltd., Ruakura Research Centre, Private Bag 3123, Hamilton, New Zealand, Chemistry Department, The University of Waikato, Private Bag 3105, Hamilton, New Zealand, Chemistry Department, University of Oslo, P.O. Box 1033, N-0315, Oslo, Norway, Institute for Marine Biosciences, National Research Council of Canada, 1411 Oxford Street, Halifax, NS B3H 3Z1, Canada, Flødevigen Research Station, Institute of Marine Research, Flødevigen, N-4817 His, Norway, Norwegian School of Veterinary Science, P.O. Box 8146 Dep., N-0033, Oslo, Norway, Sogn og Fjordane University College, N-6851 Sogndal, Norway, and HortResearch Ltd., Ruakura Research Centre, Private Bag 3123, Hamilton, New Zealand Received May 12, 2004
Two novel pectenotoxins (PTXs) were detected by LC-MS in solid phase extracts of net hauls taken at Flødevigen, Norway, in June 2002 that were dominated by Dinophysis acuminata and Dinophysis norvegica. The new compounds were isolated as minor components from a large collection of a Dinophysis acuta-dominated bloom obtained from Skjer, Sognefjorden, Norway, in October 2002. LC-MS and NMR analyses revealed that the new components, 36SPTX-12 and 36R-PTX-12, occurred as a pair of equilibrating diastereoisomers differing from PTX-2 in that they contained an exocylic olefinic methylene rather than a methyl group at C-38. Analyses of shellfish extracts revealed that PTX-12 accumulated in Norwegian blue mussels (Mytilus edulis) and cockles (Cerastoderma edule), along with PTX-12 seco acids occurring as a complex mixture of diastereoisomers. LC-MS analysis of algal cells picked from the net haul from Flødevigen revealed that PTX-12 predominated in D. acuta and D. norvegica, whereas PTX-2 was the predominant pectenotoxin in D. acuminata. Preliminary observations indicate that the relative contents of PTX-2 and PTX-12 vary between sites and years in Norway, even within a single species of Dinophysis. Our data also suggest that heterotrophic dinoflagellates may accumulate toxins from their prey.
Introduction Pectenotoxins (PTXs;1 Figure 1) are a group of toxins originating from Dinophysis species throughout the world (1). PTX-2 has been identified in Dinophysis fortii from Japan and Italy (2-4) and Dinophysis acuta from New Zealand and Ireland (5-8). However, Lee et al. (4) did not detect PTX-2 in D. acuta, Dinophysis acuminata, Dinophysis rotundata, Dinophysis tripos, Dinophysis mitra, or Dinophysis norvegica collected from various locations including Japan, Spain, France, and Norway. * To whom correspondence should be addressed. Fax: +47-23216226 or +64-7-838-5189. E-mail:
[email protected] or chris.miles@ agresearch.co.nz. † National Veterinary Institute. ‡ AgResearch Ltd. § The University of Waikato. | University of Oslo. ⊥ National Research Council of Canada. # Institute of Marine Research. 3 Norwegian School of Veterinary Science. O Sogn og Fjordane University College. @ HortResearch Ltd. 1 Abbreviations: DTX, dinophysistoxin; gs-HSQC, gradient-selected homonuclear single quantum coherence; ME, methyl ester; MeCN, acetonitrile; MeOH, methanol; NOESY, nuclear Overhauser enhancement spectroscopy; OA, okadaic acid; PTX, pectenotoxin; SA, seco acid; SPE, solid phase extraction; TOCSY, total correlation spectroscopy.
PTX-2 is hydrolyzed enzymatically to PTX-2 seco acid (SA) in some species of mussel and scallop (9-11), a reaction that appears to be a detoxification mechanism (9). In addition, PTX-2 can be oxidized at C-43 in some scallop species, to afford hydroxy (PTX-1), aldehyde (PTX3), and carboxy (PTX-6) derivatives (3). Recently, PTX11 (34S-hydroxyPTX-2) has been identified as a natural component of D. acuta from New Zealand at concentrations comparable to that of PTX-2 (8, 12). PTX-11 is much more resistant to enzymatic hydrolysis than is PTX-2 (12) and, therefore, accumulates to a greater extent in Greenshell mussels (Perna canaliculus) (7). In addition to these analogues, there are a number of PTXs isomeric at C-7 and in ring B, many of which appear to be of artifactual origin (8). PTXs appear to be highly toxic by intraperitoneal injection (9, 12-14), leading to positive responses in the mouse bioassay for lipophilic marine biotoxins. Because of this, the maximum permitted level of PTXs (PTX-1 and PTX-2) in European shellfish has been set at 160 µg/kg of okadaic acid (OA) equivalents (15). However, PTXs appear to be of low toxicity orally and, unlike OA, are not diarrheogenic, even at doses as high as 5000 µg/kg (9, 12). Here, we report finding 36S-PTX-12 and 36R-PTX-12 in algae and mussels from Norway and the determination
10.1021/tx049870a CCC: $27.50 © 2004 American Chemical Society Published on Web 10/22/2004
1424
Chem. Res. Toxicol., Vol. 17, No. 11, 2004
Miles et al.
Figure 1. Structures of PTX-1, PTX-2, PTX-3, PTX-6, PTX-11, 36S-PTX-12, 36R-PTX-12, PTX-2SA and its ME (PTX-2SA ME), PTX-12SA and its ME (PTX-12SA ME), OA, OA ME, DTX-1, and DTX-2.
of their structures by NMR and MS analyses. In addition, we report the identity of the algal sources of the toxin in a mixed bloom of Dinophysis spp. by LC-MS analyses of cells picked from the bloom and on the widespread occurrence of PTX-12 and its SAs in Norwegian shellfish.
Experimental Procedures Plankton from Flødevigen. Net hauls (20 µm) were collected at Flødevigen on June 17, 18, and 20, 2002. The net hauls from June 18 and 20 were used for solid phase extraction (SPE) after subsamples (15 and 10 mL, respectively) were taken and fixed with Lugol’s solution for counting of larger dinoflagellates at 10× magnification with a Nikon Optiphot microscope. Varying volumes (50 or 240 mL) of net hauls from June 18 and 20 were applied under gentle vacuum to C-18 SPE cartridges (500 mg; Varian). Small wads of cotton wool were packed on top of the columns to reduce clogging by algal cells (Yasumoto, T. Personal communication). After sample application, the columns were washed with freshwater (25 mL) and eluted with methanol (MeOH) (5 mL). The MeOH was concentrated under a stream of nitrogen at 40 °C, and the residue was dissolved in 80% MeOH (1 mL) and analyzed by LC-MS. The samples were subsequently stored at -20 °C except when used for further LCMS analyses. The net haul sample from June 17 was used to pick single cells as described by Samdal et al. (16). The cells were stored frozen in glass vials until processing. Plankton from Skjer. Plankton was harvested at Skjer, in the Sognefjord (Figure 2), on October 9-24, 2002, by lowering a 1 in. rubber hose connected to a centrifugal pump to the desired depth and pumping water to the surface, where it was filtered through a standard water filter cartridge (polypropylene, 10 µm) in a plastic casing. The maximum flow (ca 50 L/min) was obtained for at least an hour after which it gradually declined to ca. 5 L/min after 5-6 h. Assuming a linear decline
Figure 2. Map of Norway showing sites for collections of algal (Flødevigen and Skjer) and shellfish (Stryn, Verpeneset, Kvinnherad, Lindesnes, Flødevigen, Risør, Kragerø, and Hvaler) samples containing PTX-12. in flow, it was estimated that up to 5 m3 of seawater was filtered in 4 h before the cartridge was replaced. With 2000 cells/L of D. acuta, the total catch for one cycle amounted to 10-60 × 106 cells per day. During the 14 day sampling period, it was estimated that about 700-800 × 106 D. acuta cells were collected, together with smaller amounts of D. norvegica, D.
Pectenotoxin-12 from Norwegian Algae and Shellfish rotundata, Ceratium spp., Protoperidinium spp., and Prorocentrum micans (Supporting Information). After the plankton were harvested, the filter cartridge was placed in a cylinder filled with MeOH and left for several hours with intermittent gentle shaking. Most of the dinoflagellates fell off and settled to the bottom of the cylinder. The filter was gently brushed on the outside to remove adhering cells. The methanolic extract was decanted and concentrated on a rotary evaporator to ca. 500 mL. The residual cells were collected, and under the microscope, it was observed that it consisted of both intact and broken cells. Both fractions were stored at ca. 4 °C until analysis for algal toxins by regulatory LC-MS. Both fractions were then recombined, sonicated for 10 min, and filtered under vacuum, and the filter cake washed with MeOH (200 mL) and the residual MeOH removed under vacuum. The dark brown aqueous residue (ca. 200 mL) was then extracted successively with diethyl ether (2 × 150 mL) and dichloromethane (2 × 150 mL), and the ethereal extract was used for isolation of PTX-12 (see below). Purification of PTX-12. Purification of fractions was monitored by LC-MS. The ethereal extract of the plankton from Skjer (above) was evaporated in vacuo to afford a yellow-brown oil (198 mg). This was purified [modified after Yasumoto et al. (17)] by application to a column of basic alumina (20 g, activity III; Merck) in CHCl3-MeOH (1:1, 5 mL) and elution with CHCl3MeOH (1:1, 60 mL) (fraction 1), MeOH (30 mL) (fraction 2), and 1% aqueous NH3-MeOH (1:1, 80 mL) (fraction 3). These were evaporated to dryness in vacuo immediately after elution. LCMS indicated that ca. 62% of the total PTX-2 and 18% of the total PTX-12 were in fraction 1, whereas 25% of the PTX-2 and 71% of the PTX-12 were in fraction 3. Fraction 3 was purified by flash column chromatography on LiChroprep RP-18 (17 cm × 2 cm; 40-63 µm; Merck, Germany). The column was eluted with a stepwise gradient of MeOH in water (50, 60, 70, 80, 90, and 100%; 100 mL each), and 20 mL fractions were collected. Fractions 17-21, which contained most of the PTX-12 along with PTX-2, were combined and applied to a column of LiChroprep RP-18 (32 cm × 1 cm; 40-63 µm; Merck) connected to an HPLC system (Pharmacia, Sweden). The column was eluted (5 mL/min) with a continuous gradient of acetonitrile (MeCN) in water (60% hold for 40 mL; 60-68% over 160 mL; and 100% hold from 200 mL), and 10 mL fractions were collected. Fractions 6-8 and 11-19, which contained most of the PTX-12 by LC-MS, were combined, while fractions 9-10 (which contained mostly PTX-2) were retained separately. Flash chromatographic purification of PTX-12 was repeated in a similar manner on this chromatography system using varying linear gradients (typically from 55 to 65%) of MeCN in water, until complete separation of PTX-12 (ca. 200 µg) from PTX-2 and other contaminants had been achieved. All fractions containing PTX-2 were retained for subsequent purification. Recrystallization of these fractions from MeOH-water afforded PTX-2 (13.0 mg) as colorless needles. Chemical Analyses of Phytoplankton. LC-MS analysis was performed with a model P4000 pump, model AS3000 autosampler (Thermo Separation Products, San Jose, CA), and LCQ classic mass spectrometer (Finnigan MAT, San Jose, CA) in positive ion electrospray mode and a scan range of m/z 2001000 without the use of a splitter. Separation was achieved on a Symmetry C-18 column (4.6 mm × 150 mm; Waters, Milford, MA) with linear gradient elution (40-100% B over 15 min and then 5 min hold at 100%) at 0.6 mL/min and 10 µL injections. Eluent A was water, and eluent B was MeCN (containing 0.5% water to dissolve the buffer), each containing 2 mM ammonium formate and 0.01% formic acid. LC-MS3 was performed on an LCQ Deca ion trap mass spectrometer fitted with an ESI interface (ThermoQuest, Finnigan) and coupled to a Surveyor HPLC and PDA detector. The column was a Prodigy 5 µm ODS(3) 100 Å, 150 mm × 2 mm (Phenomenex, Torrance, CA). A 0.2 µm in-line filter (Alltech, Deerfield, IL) was installed before the analytical column, and the temperature of the column oven was maintained at 35 °C. Gradient elution was performed using
Chem. Res. Toxicol., Vol. 17, No. 11, 2004 1425 Table 1. Cell Concentration (Cells/mL) of Larger Dinoflagellates in Net Hauls Taken June 18 and 20, 2002, at Flødevigen Bay, Norway cell concentration (cells/mL) species
June 18
June 20
D. acuta D. acuminata D. norvegica D. rotundata D. dens Protoceratium reticulatum Protoperidinium spp. Prorocentrum micans Ceratium fusus Ceratium tripos Ceratium macroceros Ceratium lineatum Ceratium longipes Ceratium furca unidentified dinoflagellates
880 3210 8640 1440
270 1830 975 75 30 2 45 30 525 30 180 30 14 45 45
5880 3080 80 40
MeOH-0.1% ammonium formate (3:17) containing 0.1% formic acid (solvent A) and 100% MeOH (solvent B). Linear gradients were run from 70 to 100% B over 10 min, held for 3 min, and then reset to the initial conditions. The flow rate was 200 µL/ min, the injection volume was 10 µL, and the PDA detector scanned from 200 to 600 nm. MS data were acquired in both positive and negative modes using a fully automated datadependent LC-MSn method with dynamic exclusion enabled and a repeat count of two. This method isolated and fragmented the most intense parent ion to give MS2 data, then isolated and fragmented the most intense daughter ion to give MS3 data. Each ion interrogated was then added to an exclusion list for a period of 12 s to enable mass spectral fragmentation data to be obtained from less intense ions. The ESI voltage, capillary temperature, sheath gas pressure, and auxiliary gas were set at 4 kV, 275 °C, 35 psi, and 0 psi, respectively. Chemical Analysis of Shellfish. Shellfish analyses were based on the method of Aasen et al. (18). Homogenized digestive gland (1 g) from blue mussels (Mytilus edulis) or homogenized whole meat from cockles (Cerastoderma edule) (1 g) was extracted with MeOH (4 mL) by mixing for 10 min. After centrifugation, 800 µL of extract was mixed with 50 mM ammonium acetate (200 µL) and filtered through 0.22 µm filters (Spin-X, Costar, United States). Separation of toxins was achieved on a Varian Omnispher 5 C18 column (5 µm, 150 mm × 2.0 mm) with a mobile phase consisting of MeCN-water (65: 35) with 2 mM ammonium formate and 50 mM formic acid, at 300 µL/min and 20 µL injections. The effluent was directly coupled to an API-2000 triple quadrupole mass spectrometer (PE-SCIEX, Concorde, ON, Canada) without splitting. Selected ion monitoring (SIM) was used to record the signals from the [M + NH4]+ ions at m/z 874.5 (PTX-12), 876.5 (PTX-2), 892.5 (PTX-1 and PTX-12SA), 894.5 (PTX-2SA), 906.5 (PTX-6), and 910.5 (PTX-1SA). A second aliquot of the extract (1 mL) was treated with methanolic NaOH (1 M, 600 µL) at room temperature for 1 h to hydrolyze esters of DSP toxins [OA and dinophysistoxin (DTXs)]. After the addition of HCl (1 M, 700 µL) and water (1 mL), the hydrolyzate was washed with 2 mL of hexane (2 mL) and extracted with dichloromethane (2 × 2 mL). The dichloromethane was evaporated under a stream of nitrogen, and the residue was dissolved in water-MeOH (1:4, 500 µL) and filtered through a 0.22 µm filter. The sample was analyzed as above except that the mobile phase consisted of MeOH-water (4:1) with 10 mM ammonium acetate at 300 µL/ min and SIM was used to record the [M - H]- ions at m/z 803.5 (OA and DTX-2) and 817.5 (DTX-1). All chromatographic equipment (degasser, binary pump, autosampler, and column oven) was from the PE 200 series (Perkin-Elmer). Concentrations of toxins in whole blue mussels were calculated from the LC-MS analyses by assuming that the entire toxin burden was located in the hepatopancreas and that this organ constituted 20 wt % of the whole mussel (19). Quantitation of OA, DTX-1,
1426
Chem. Res. Toxicol., Vol. 17, No. 11, 2004
Table 2. 1H and
13C
NMR Assignments for 36S-PTX-12, 36R-PTX-12, and PTX-2 in CD3OD; PTX-2 NMR Assignments Are from Ref 9 PTX-2
13Ca
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
172.1 49.2 77.4 30.5 22.7 35.4 108.7 33.4 22.9 81.9 76.1 82.6 44.7 214.6 80.2 72.0 36.9 81.6 34.4 28.9 110.2 80.2 30.1 38.2 86.1 51.1 31.5 141.1 131.8 136.1 122.3 84.1 76.1 33.8 82.6 98.6 71.7 30.6 28.2 61.8 15.9 23.3 26.1 26.7 23.8 12.7 17.9
Miles et al.
36S-PTX-12 1Hb
2.32 3.46 1.52, 1.16 1.83, 1.57 1.68, 1.68 2.54, 1.55 2.07, 1.63 4.27 4.02 2.86, 1.97 3.80 4.25 2.08, 1.31 1.92, 1.68 2.19, 1.99 3.85 2.02, 1.67 1.69, 1.42 1.70, 1.54 2.58 5.26 6.48 5.42 4.78 5.47 2.22, 2.10 4.50 3.29 2.12 1.66, 1.25 3.98, 3.68 1.09 1.19 1.34 1.22 0.96 1.70 0.95
13Cc
d 49.3 77.4 30.5 22.8 35.1 d 33.4 22.9 81.9 76.1 d 44.5 d 80.2 72.1 36.9 d 34.6 29.0 d 80.2 30.2 38.2 d 51.1 31.7 141.3 d 136.2 d 84.0 76.4 d 82.1 d 74.2 d 62.4 d 15.9 23.4 26.2 26.7 23.9 12.7 113.3
1Hb
2.32 3.46 1.52, 1.16 1.83, 1.57 1.68, 1.68 2.54, 1.55 2.07, 1.63 4.28 4.02 2.86 (ax), 1.98 (eq) 3.80 4.26 2.08, 1.31 1.92, 1.68 2.19, 1.99 3.85 2.02, 1.67 1.69, 1.42 1.70, 1.54 2.58 5.27 (br d, 10.0 Hz) 6.47 (dd, 15.7, 1.5 Hz) 5.42 (dd, 15.7, 3.3 Hz) 4.77 5.47 2.24, 2.11 4.55 3.79 (eq) 2.75 (ax), 2.05 (eq) 3.94, 3.78 1.09 1.19 1.34 1.22 0.973 (d, 6.9 Hz) 1.70 4.98, 4.96
36R-PTX-12 13Cc
d 49.3 77.4 30.5 22.8 35.1 d 33.4 22.9 81.9 76.1 d 44.5 d 80.2 72.1 36.9 d 34.6 29.0 d 80.2 30.2 38.2 d 51.1 31.7 141.3 d 136.2 d 84.0 76.4 d 82.1 d 73.2 d 62.4 d 15.9 23.4 26.2 26.7 23.9 12.7 108.4
1Hb
2.32 3.46 1.52, 1.16 1.83, 1.57 1.68, 1.68 2.54, 1.55 2.07, 1.63 4.28 4.02 2.86 (ax), 1.98 (eq) 3.80 4.26 2.08, 1.31 1.92, 1.68 2.19, 1.99 3.85 2.02, 1.67 1.69, 1.42 1.70, 1.54 2.58 5.28 (br d, 10.5 Hz) 6.45 (dd, 15.7, 1.5 Hz) 5.42 (dd, 15.7, 3.3 Hz) 4.75 5.47 2.62, 2.04 4.56 4.08 (ax) 2.39 (ax), 2.33 (eq) 3.86, 3.65 1.09 1.19 1.34 1.22 0.977 (d, 6.7 Hz) 1.70 5.14, 4.96
a 13C NMR shifts determined relative to CD OD ) 49.0 ppm. b 1H NMR shifts determined relative to CHD OD ) 3.31 ppm. 3 2 detected 13C NMR signals ((0.5 ppm) determined in an HSQC NMR experiment. d Signal not detected.
DTX-2, PTX-2, and PTX-2SA was conducted relative to authentic standards, whereas PTX-12 and PTX-12SA were quantitated relative to PTX-2 and PTX-2SA by assuming equal response factors. Authentic OA and DTX-1 were obtained from Calbiochem (San Diego, CA), DTX-2 was a generous gift from T. Yasumoto (Japan Food Research Laboratories, Tokyo, Japan), and PTX-2 and PTX-2SA were prepared as described by Miles et al. (9). For the derivatization reaction to demonstrate the presence of a carboxylic acid group, an excess of diazomethane in EtOAc was added to 1 mL of mussel extract and the mixture was allowed to stand for 30 min (9). The solution was then evaporated under a stream of nitrogen, and the residue was dissolved in 1 mL of 80% MeOH and analyzed by LC-MS3 in tandem with the underivatized extract. NMR Spectroscopy. NMR spectra were obtained from a mixture of 36S-PTX-12 and 36R-PTX-12 (ca. 200 µg) in 0.5 mL of CD3OD (99.8 at % D, Cambridge Isotope Laboratories Inc., MA). The spectra were acquired on a Bruker Avance DRX 500 MHz NMR spectrometer with a 5 mm TXI (1H/13C,15N-2H) triple resonance inverse probe and on a Bruker Avance AV 600 MHz NMR spectrometer with a 5 mm BBI (1H/BB-2H) broad band
c 1H-
inverse probe, both probes equipped with Z-gradient coils. The data were processed using Bruker XWIN NMR (version 3.5) software. NMR assignments (Table 2) are inferred from 1H, 1H-double solvent-suppressed, total correlation spectroscopy (TOCSY), nuclear Overhauser enhancement spectroscopy (NOESY), and gradient-selected homonuclear single quantum coherence (gs-HSQC) NMR spectral data. Chemical shifts, determined at 298 K, are reported relative to internal CHD2OD (3.31 ppm) and CD3OD (49.0 ppm). Molecular modeling was performed with Chem3D Ultra version 8.0 (CambridgeSoft, Cambridge, MA) using the supplied MM2 parameters, and the crystal structure of PTX-1 (14) was used to generate starting conformations for 36R- and 36S-PTX-12. Analysis of Picked Algal Cells. The samples were transported and stored at -20 °C. MeCN (100 µL) was added with gentle swirling immediately after removal from the freezer, and then the solvent was carefully evaporated to dryness under a gentle stream of dry N2 at ca. 40 °C. This treatment was designed to assist extraction of toxins from the cells and to azeotropically remove water to prevent enzymatic hydrolysis of PTXs during storage (9, 12). The sealed vials were stored dry
Pectenotoxin-12 from Norwegian Algae and Shellfish at
