Neurotoxins in axenic oscillatorian cyanobacteria: coexistence of anatoxin-a and homoanatoxin-a determined by ligand-binding assay and GC/MS

Rómulo Aráoz1,{dagger}, Hoàng-Oanh Nghiêm2, Rosmarie Rippka1, Nicolae Palibroda3,{ddagger}, Nicole Tandeau de Marsac1 and Michael Herdman1

1 Unité des Cyanobactéries (CNRS URA 2172), Institut Pasteur, 25-28 rue du Docteur Roux, 75724 Paris Cedex 15, France
2 Unité des Récepteurs et Cognition (CNRS URA 2182), Institut Pasteur, 25-28 rue du Docteur Roux, 75724 Paris Cedex 15, France
3 Laboratoire de Chimie Structurale des Macromolécules, Institut Pasteur, 25-28 rue du Docteur Roux, 75724 Paris Cedex 15, France

Correspondence
Michael Herdman
mherdman{at}pasteur.fr


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Two neurotoxic alkaloids, anatoxin-a and its homologue homoanatoxin-a, were purified from the filamentous cyanobacteria Oscillatoria sp. strain 193 (PCC 9240) and Oscillatoria formosa NIVA CYA-92 (PCC 10111), respectively, and characterized by mass spectrometry. Biological activity was determined by examining the capacity of the toxins to competitively inhibit the binding of radiolabelled bungarotoxin to acetylcholine receptors, using post-synaptic membrane fractions of Torpedo electric tissue. Inhibition was concentration dependent, with a Ki of 5·4±1·1x10–8 M for anatoxin-a and 7·4±0·9x10–8 M for homoanatoxin-a. Their high affinities for the nicotinic cholinergic receptors were exploited to adapt the radioligand-binding assay for routine detection of this class of neurotoxins directly in low-molecular-mass cell extracts of cyanobacteria. Confirmation of the results and toxin identification were achieved by coupled gas chromatography-mass spectrometry (GC/MS). Seventy-six axenic strains, representative of 13 genera, were analysed. Five strains of the genus Oscillatoria, hitherto unknown for their toxicity, inhibited bungarotoxin binding. GC/MS revealed that Oscillatoria sp. strains PCC 6407, PCC 6412 and PCC 9107 synthesized exclusively anatoxin-a, whereas both anatoxin-a and homoanatoxin-a were produced by strain PCC 9029. Oscillatoria sp. strain PCC 6506, an isolate co-identic with strain PCC 9029, also produced both neurotoxins, but their respective presence depended upon growth conditions. The latter results suggest that regulatory differences in at least some of the cyanobacterial strains may account for the preferential synthesis of only one of the two neurotoxins or for their simultaneous occurrence.


Abbreviations: ANTX, anatoxin-a; Bgx, bungarotoxin; EI, electron impact ionization; HANTX, homoanatoxin-a; nAChR, nicotinic acetylcholine receptors

{dagger}Present address: Unité de Génétique Moléculaire Bactérienne, Institut Pasteur, 28 rue du Docteur Roux, 75724 Paris Cedex 15, France.

{ddagger}Present address: Institute of Isotopic and Molecular Technologies, 71-103 rue Donath, PO Box 700, RO-3400 Cluj-Napoca 5, Romania.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Worldwide proliferation of cyanobacterial blooms constitutes a serious environmental and economic problem that menaces wildlife and human health. Bloom development, a phenomenon related to water eutrophication, is favoured by the ability of planktonic cyanobacteria to synthesize gas vesicles that allow them to float to the surface layers of the water column. The massive growth and accumulation of cyanobacteria greatly impair food-web dynamics and the physico-chemical factors inside a given aquatic ecosystem. Moreover, many cyanobacteria are able to produce potent hepatotoxins such as microcystin, cylindrospermopsin and nodularin, and/or neurotoxins such as anatoxin-a, homoanatoxin-a, anatoxin-a(s) and saxitoxin (Carmichael, 1994; Sivonen & Jones, 1999).

Anatoxin-a (ANTX) (2-acetyl-9-azabicyclo[4.2.1]non-2-ene) and its methylene homologue homoanatoxin-a (HANTX) (2-(propan-1-oxo-1-yl)-9-azabicyclo[4.2.1]non-2-ene) (Devlin et al., 1977; Skulberg et al., 1992), are low-molecular-mass bicyclic secondary amines synthesized by some planktonic and benthic strains of the genera Anabaena, Oscillatoria, Aphanizomenon and Cylindrospermum (Sivonen et al., 1989, 1990). At nanomolar levels, ANTX is a specific cholinergic agonist whose potency relies upon its high affinity for the nicotinic acetylcholine receptors (nAChRs). The binding of ANTX to nAChRs induces conformational changes in the post-synaptic receptor ion channel complex that lead to a blockage of neuromuscular depolarization (Aronstam & Witkop, 1981; Swanson et al., 1985). The LD50 for ANTX and HANTX ranges between 200 and 250 µg kg–1 i.p., mouse (Carmichael & Gorham, 1978; Skulberg et al., 1992). Animal poisonings by ingestion of ANTX/HANTX-producing cyanobacteria in lakes and ponds have been well documented, and are generally caused by planktonic species of Anabaena or Aphanizomenon (Bruno et al., 1994; Sivonen & Jones, 1999). However, good evidence for a similar incident due to a benthic filamentous cyanobacterium of the genus Oscillatoria has also been obtained (Edwards et al., 1992). The aim of the present study was to examine the toxin production of axenic isolates of diverse filamentous cyanobacteria in the Pasteur Culture Collection (PCC), with special emphasis being given to benthic members of the genus Oscillatoria. For the survey of ANTX and HANTX, the receptor radioligand-binding assay was improved for high-throughput screening directly on cell extracts.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Strains and culture conditions.
Axenic strains (Table 1) were from the Pasteur Culture Collection of Cyanobacteria (PCC). Purity was checked with each transfer as previously (Rippka et al., 1979), but with one modification: liquid Luria–Bertani medium (Sambrook et al., 1989) was added to a final concentration of 5 % (v/v) to BG11, in addition to 0·2 % (w/v) glucose and 0·02 % (w/v) Difco Casamino acids. Purification of Oscillatoria formosa NIVA CYA-92 to give the axenic strain PCC 10111 was performed by single-filament isolation as described by Rippka (1988).


View this table:
[in this window]
[in a new window]
 
Table 1. Axenic cyanobacteria of the PCC analysed for ANTX and HANTX by the radioligand receptor-binding assay

PCC numbers indicated in bold type refer to strains shown to produce ANTX and/or HANTX (this study). Bold type in the medium, temperature and yield columns indicates large-scale cultivation. ND, Not determined; (?), source doubtful. Assignment to sections and clusters as in Rippka & Herdman (1992).

 
Cyanobacterial strains were grown without agitation in 500 ml Erlenmeyer flasks containing 200 ml liquid medium, using precultures of 20–40 ml as inocula. For freshwater strains, the media were either BG11, or medium BG11o supplemented with 2 mM NaNO3 (indicated as medium 2N in Table 1). Filamentous heterocystous cyanobacteria were grown in medium BG11o, supplemented with 5 mM NaHCO3 (indicated as BG11o5 in Table 1). Strains of marine origin were cultivated in a mixture of equal volumes of media ASNIII and BG11. The Na2CO3 concentration of all media was increased twofold with respect to the original recipes (Rippka et al., 1979). Mass cultures (10 l carboys) of strains PCC 6506, PCC 7515, PCC 9240, PCC 9029 and PCC 10111 were grown under continuous light at 25 °C and continuous stirring. For these, the media (BG11 or 2N) were supplemented with 10 mM NaHCO3 (indicated respectively as BG1110 or 2N10 in Table 1). The cultures were gassed with CO2-enriched air (1 : 99, v/v).

Cultures incubated at 22 °C (Table 1) received a photosynthetic photon flux density (PPFD) of 10 µmol quanta m–2 s–1 (measured with a LICOR LI-185B quantum/radiometer/photometer equipped with a LI-190SB quantum sensor), provided by Osram Universal White fluorescent tubes over a light/dark regime of 14 h/10 h. Lyophilized pellets of 13 strains of the genus Oscillatoria were kindly provided by Dr Jennifer Best (Institut Pasteur), and resulted from cultures grown at 28 °C (Table 1) under continuous light with a PPFD of 5 µmol quanta m–2 s–1. Cyanobacterial cells were harvested by centrifugation (5000 g for 15 min at 22 °C) after 2–6 weeks, prior to reaching the stationary phase of growth. The pellets were washed once in sterile H2O, or with 150 mM NaCl for marine strains, and lyophilized. The lyophilized cell material was stored at –20 °C.

Toxin purification.
Axenic cultures of Oscillatoria sp. strain PCC 9240 and Oscillatoria formosa PCC 10111 were used for ANTX and HANTX purification, respectively. Both toxins were purified as described for ANTX (Harada et al., 1993). Briefly, 1 g of lyophilized cells was ground with a mortar and pestle in 100 ml 50 mM acetic acid (AcOH). The homogenate was incubated in the dark at room temperature for 4 h with constant stirring and then centrifuged at 12 000 g for 15 min, at 15 °C. The supernatant was passed through a 0·1 µm filter (Millex-VV, Millipore) and the pH of the solution was adjusted to 9 with NH4OH. Solid-phase extraction was performed using a C18 column (5g Bakerbond resin) previously conditioned with 20 ml methanol (MeOH) and 60 ml H2O. The sample was loaded by gravity and the column was washed successively with 60 ml H2O and 20 ml 10 % (v/v) MeOH/H2O. The toxin was eluted with 50 ml MeOH. The solvent was evaporated in an analytical evaporator at 40 °C with a stream of N2 (Organomation). The toxins were dissolved in 0·1 % (v/v) AcOH/H2O and loaded onto a µBonda Pack column (150x3·9 mm, particle size 5 µm, 300 Å, Waters). The column effluent was monitored using a medium-pressure liquid chromatograph (Biologic System, Bio-Rad) with a fixed-wavelength detector (254 nm). The flow rate was 2 ml min–1. The mobile phase consisted of 0·1 % (v/v) AcOH/H2O (A) and aqueous 80 % (v/v) acetonitrile (CH3CN), 0·1 % (v/v) AcOH (B). The elution profile was: isocratic 5 % B, 0–10 min; gradient 5–25 % B, 10–30 min. The collected fractions were analysed by GC/MS. ANTX/HANTX-containing fractions were evaporated as described and stored at –20 °C. The concentration of ANTX and HANTX was estimated by UV-spectroscopy ({lambda}max 227 nm in ethanol, {varepsilon}=12 000 l mol–1 cm–1) (Harada et al., 1989). 1H NMR spectra of anatoxin-a were recorded on a Bruker AC-300 spectrometer operating at 300 MHz with D2O as a solvent.

GC/MS analysis of ANTX and HANTX.
MS was performed with a Varian 3400 gas chromatograph coupled to a Saturn 2000 ion-trap mass spectrometer. The GC column, an Rtx-5 Sil MS fused silica capillary column (30 mx0·25 mmx0·25 µm) (Hewlett Packard), was operated in the splitless mode. The injector temperature was 200 °C. The carrier gas (He) flow was 1 ml min–1. Prior to GC/MS analysis, all samples were resuspended in an appropriate volume of CH3CN and evaporated under N2 using an analytical evaporator as described above. This step was repeated twice. Samples of 1 µl in CH3CN were injected into the GC/MS. The GC temperature profile was: isothermal 50 °C, 1 min; temperature gradient 50–165 °C (5 °C min–1); isothermal 165 °C, 5 min; temperature gradient 165–290 °C, (5 °C min–1); isothermal 290 °C, 2 min. Mass spectra in the electron impact ionization (EI) mode were recorded at 70 eV with an ionization current of 20 µA; the ion trap temperature was 140 °C and the transfer line temperature was set to 170 °C. Mass analysis in the positive chemical ionization (CI) mode was done using CH3CN as reagent gas. A thermal regeneration step was intercalated between each run to avoid contamination by carry-over of residual analytes.

Toxin stability.
One millilitre of a 400 µg ml–1 solution of ANTX in 50 mM AcOH (pH 3) was incubated in the dark at room temperature. Aliquots (50 µl) were withdrawn at given time points (0, 2, 4 and 6 h, 1, 2 and 5 days), dried with a stream of N2 and resuspended in 50 µl CH3CN. This step was repeated twice and 1 µl was injected into the GC/MS. In one set of experiments, the purified degradation product of ANTX (m/z 179) was added (50 µg ml–1) to the initial solution of ANTX as internal reference.

Receptor-binding assays.
nAChR-rich membranes were purified from Torpedo electric organ as described by Hill et al. (1991). Inhibition by competitive binding of small agonist ligands (nicotine, ANTX, HANTX) to nAChR-rich membranes was performed at a fixed concentration of 10–8 M 125I-{alpha}-bungarotoxin (125I-Bgx) (Amersham Biosciences). One microlitre of the agonist at increasing concentrations was incubated with 100 µl Torpedo membranes (0·08 mg protein ml–1) in PBS (10 mM sodium phosphate buffer, 130 mM sodium chloride, pH 7·0), containing 0·1 % w/v BSA and 0·1 % (v/v) Tween 20, at room temperature for 1 h with constant shaking. Two microlitres of 0·5x10–6 M 125I-Bgx (1·5–2·2 TBq mmol–1) was then added and the whole mixture was further incubated for 5 min. The reactions were stopped by addition of 4 ml washing buffer (PBS, 1 %, w/v, milk powder, pH 7·4). The samples were immediately filtered through glass microfibre filter discs (GF/C, Whatman) and the filters were rinsed twice with the same buffer (4 ml at 4 °C). The membrane-bound 125I-Bgx on the filters was measured with a {gamma} counter (Multigamma II, LKB). Inhibition of Bgx binding to nAChRs was calculated from the reduction in 125I-Bgx radioactive emissions, as compared with control incubations of 125I-Bgx with Torpedo membranes in the absence of agonists. Inhibition affinity constants were determined from the equation Ki=EC50/(1+[ligand]/Kd) (Cheng & Prusoff, 1973), where the competition between the inhibitor and the radioligand is assumed to concern one population of sites and ligand depletion is assumed negligible in our experimental conditions; [ligand] is the concentration of the 125I-Bgx radioligand; Kd is the dissociation constant of the radioligand and corresponds to the concentration at which 125I-Bgx gives 50 % of the maximal binding to the nAChRs in absence of inhibitors; EC50 is the concentration of inhibitor at which 125I-Bgx binding to the nAChRs is inhibited by 50 %. All experiments were performed in triplicate.

Toxin screening of cyanobacterial extracts.
Lyophilized cyanobacterial cells (20 mg) were homogenized in 4 ml 50 mM AcOH with an Ultraturrax T25 homogenizer (Janke and Kunkel), applying 3 bursts of 1 min at maximum speed with 1 min intervals. The homogenates were incubated for 4 h at room temperature in the dark and centrifuged (12 000 g for 15 min, at 15 °C). The supernatants were filtered using Vivaspin 500 µl units, MWCO 5 kDa (Sartorius), to eliminate the majority of contaminating proteins. Aliquots of 20 µl from the resulting low-molecular-mass filtrate, either undiluted or at a 20-fold dilution, were incubated with Torpedo membranes as described, except that the molarity of the phosphate buffer was increased to 20 mM to compensate for the acidity of the extracts. All experiments were performed in triplicate.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Purification of ANTX and HANTX
Toxins were extracted from Oscillatoria sp. strain PCC 9240 and Oscillatoria formosa PCC 10111. The former strain corresponds to the axenic isolate Oscillatoria sp. strain 193, known to synthesize ANTX (Sivonen et al., 1989); the latter was rendered axenic for this study from the impure O. formosa strain NIVA CYA-92, previously reported to produce HANTX (Skulberg et al., 1992). These structurally closely related water-soluble alkaloids were purified by the same protocol (see Methods). For the extraction of ANTX or HANTX, acidified water was chosen, since it avoids co-extraction of solvent-soluble compounds such as chlorophyll a and carotenoids that may interfere with GC/MS analysis. ANTX (m/z 165) eluted from the HPLC column at 12 % B while HANTX (m/z 179) eluted at 16 % B. The extraction yield of ANTX was 0·2 % of the initial dry weight. 1H NMR analysis of ANTX showed a spectrum similar to those published previously (Stevens & Krieger, 1991). ANTX purity was greater than 95 % (data not shown). Similar yields were obtained for HANTX. The UV spectra of ANTX and HANTX overlap. Their {lambda}max at 227 nm ({varepsilon}=12 000 l mol–1 cm–1) due to the {alpha},{beta}-unsaturated ketone is not affected by the methyl group at the carboxyl side chain of HANTX because of the rigidity of the azabicyclononene ring (Swanson et al., 1991).

Purification of the m/z 179 ANTX degradation product
Storage of purified ANTX in CH3CN for 2 weeks at room temperature resulted in the formation of a degradation product of m/z 179. In order to confirm its distinction from HANTX, which has an identical mass, the degradation product m/z 179 was purified and characterized by GC/MS. It eluted from the HPLC column at 10 % B. The UV spectra showed that the {lambda}max of the m/z 179 degradation product was increased by 10 nm compared to ANTX and HANTX (data not shown).

Characterization of ANTX, HANTX and m/z 179 by GC/MS
Inclusion of an isothermal step at 165 °C in the GC temperature profile led to successful separation of ANTX from its degradation product (m/z 179). This modification was adopted since initial application of a thermal linear gradient (5 °C min–1) did not permit optimal separation of these compounds (data not shown). Under these conditions ANTX eluted at the 600th scan. The EI mass spectrum of ANTX (Fig. 1a) showed the following sequential fragmentation of the molecular ion (relative abundance %, ion): m/z 165 (100 %, M+); m/z 150 (39 %, {mic1511263E001}); m/z 136 (76 %, {mic1511263E002}); m/z 122 (85 %, {mic1511263E003}). MS/MS analysis of ANTX using a mass window of 165±1 amu resulted in the fragment ions m/z 150, 136 and 122 (data not shown). The EI pattern of ANTX showed good correlation to collision-induced fragmentation observed by MS/MS. The mass of ANTX was confirmed by operating the GC/MS in the positive chemical ionization (CI) mode. As expected, a protonated molecule of m/z 166 (M+H)+ was obtained that eluted at the 600th scan (data not shown).



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 1. GC/MS analyses of ANTX, HANTX and the degradation product of ANTX. (a) EI spectrum of ANTX. Retention time: 600th scan. (b) EI spectrum of HANTX. Retention time: 654th scan. Base peak: m/z 150. (c) GC chromatogram of the ANTX degradation product m/z 179 after purification by HPLC. Retention time: 670th scan. Inset: EI mass spectrum of m/z 179. Base peak: m/z 136.

 
HANTX eluted from the gas capillary column at the 654th scan. The EI fragmentation pattern of HANTX (Fig. 1b) (relative abundance %, ion) was: m/z 179 (76 %, M+), m/z 164 (29 %, {mic1511263E004}), m/z 150 (100 %, {mic1511263E005}), m/z 136 (39 %, {mic1511263E006}) and m/z 122 (85 %, {mic1511263E007}). Its base-peak ion m/z 150 (Fig. 1b) may result from the loss of the ethyl group from the side chain of the molecular ion. MS/MS analysis of HANTX using a mass window of 179±1 amu confirmed that the major fragments derived from the parental ion m/z 179 had the masses m/z 150, 136 and 122 (data not shown). GC/MS in the CI mode gave a protonated molecule m/z 180 (M+H)+ (data not shown), thus confirming the Mr of 179 for HANTX.

The ANTX degradation product of m/z 179 and HANTX share the same mass but they are two different chemical entities. GC/MS analysis of the degradation product showed a single peak that eluted at the 670th scan (Fig. 1c). The mass fragmentation pattern (Fig. 1c, inset) for the ANTX degradation product (relative abundance %, ion) was: m/z 179 (63 %, M+); m/z 151 (25 %, {mic1511263E008}); m/z 136 (100 %, {mic1511263E009}); m/z 122 (55 %, {mic1511263E010}). The base-peak ion m/z 136 that resulted from the cleavage of the side chain is characteristic for this molecule.

Toxin stability
ANTX was stable for at least 5 days in 50 mM AcOH, at pH 3, even if incubated at room temperature in the dark (data not shown). This was examined particularly carefully in a second experiment, in which the purified degradation product m/z 179 of ANTX was added as an internal reference. As expected, there was no ANTX degradation according to GC/MS analysis, the ratio ANTX/m/z 179 remaining constant throughout the experiment (5 days, data not shown). When stored at –20 °C in 50 mM AcOH, both ANTX and HANTX were stable for several months as verified by GC/MS and inhibition binding assay (data not shown).

Inhibition of 125I-bungarotoxin binding by ANTX and HANTX
Incubation of nAChR-rich Torpedo membranes with increasing amounts of either nicotine, ANTX or HANTX resulted in concentration-dependent inhibition of the binding of 125I-Bgx to the nAChRs (Fig. 2). Both cyanobacterial neurotoxins were inhibitory to 125I-Bgx binding at concentrations about two orders of magnitude lower than that of nicotine (EC50 for nicotine 6·5±0·5x10–6 M, EC50 for ANTX 5·5±0·9x10–8 M and EC50 for HANTX 7·5±0·8x0–8 M). The Ki values for ANTX and HANTX were similar (5·4±1·1x10–8 M and 7·4±0·9x10–8 M, respectively), implying that both toxins are equally potent agonists for nAChRs. These results showed that the radioligand test, though being very sensitive and specific for the two cyanobacterial neurotoxins as a class, could not distinguish between ANTX and HANTX. In contrast, the m/z 179 degradation product of ANTX gave no inhibition of the 125I-Bgx binding even at concentrations of 1x10–4 M (Fig. 2).



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 2. Inhibition of 125I-Bgx binding to nAChRs of Torpedo postsynaptic membranes by various agonists at different concentrations. {blacksquare}, Nicotine; {circ}, ANTX; {square}, HANTX; +, degradation product m/z 179 of ANTX; {bullet}, partially purified ANTX after solid-phase extraction from Oscillatoria sp. strain PCC 9240. The data points for each concentration of agonist represent mean values obtained for assays performed in triplicate.

 
The receptor ligand-binding assay was next used to detect the presence of ANTX in a partially purified ANTX sample from Oscillatoria sp. strain PCC 9240 after solid-phase extraction. The concentration of ANTX in this fraction was calculated by UV spectrometry. The dose–response curve of partially purified ANTX (Fig. 2, solid circles) was similar to that of purified ANTX. As a negative control an extract of Oscillatoria sancta strain PCC 7515 was included that had been fractionated in the same way as strain PCC 9240. According to GC/MS analyses, the latter strain lacked ANTX and HANTX but contained an ion with a mass of m/z 179 that eluted from the gas capillary column at the 560th scan (data not shown), excluding that it corresponded to the degradation product of ANTX (eluting at the 670th scan). No inhibition of 125I-Bgx binding could be detected with the extract of strain PCC 7515 (Fig. 3b).



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 3. Inhibition of 125I-Bgx binding to nAChRs of Torpedo post-synaptic membranes by low-molecular-mass cell extracts of several axenic cyanobacterial strains. (a) Inhibition by low-molecular-mass cell extracts from Oscillatoria sp. strain PCC 9240 at different dilutions. (b) Histogram showing inhibition observed in assays with undiluted extracts of Oscillatoria sp. strains PCC 6407, PCC 6412, PCC 6506, PCC 9029, PCC 9107, PCC 9240, Oscillatoria formosa PCC 10111, Oscillatoria sancta PCC 7515, Anabaena flos-aquae PCC 9349 and Oscillatoria sp. strain PCC 10016. Bars indicate the standard deviation. All assays were performed in triplicate.

 
Detection of ANTX/HANTX in cyanobacterial extracts
In order to adapt the receptor ligand-binding assay for the screening of potential ANTX and or HANTX producers among the cyanobacterial strains of the PCC, 20 µl of cyanobacterial extracts was added to the Torpedo membrane preparation instead of the 1 µl used for purified toxins. The molarity of the phosphate buffer in the reaction was increased from 10 mM to 20 mM to compensate for the acidity of the extracts. These modifications did not affect the kinetics of 125I-Bgx binding as verified in control assays (data not shown). A dose–response curve of inhibition was produced with a series of dilutions of the low-molecular-mass cell extracts of strain PCC 9240. Inhibition of I125-Bgx binding to nAChRs was concentration dependent, and saturation levels of inhibition were observed with up to 25-fold dilutions of the extract (Fig. 3a). The analyses of extracts prepared from 76 axenic strains, representative of 13 genera (Table 1), revealed that in addition to the two strains of Oscillatoria already known for ANTX or HANTX toxicity, a further five members of this genus (strains PCC 6407, PCC 6412, PCC 6506, PCC 9029 and PCC 9107) were highly inhibitory in the receptor radioligand-binding assay (Fig. 3b). In contrast, extracts of all other Oscillatoria strains and the representatives of the other 12 genera examined (Table 1) gave less than 10 % inhibition in this assay (for examples, see Fig. 3b) and were thus considered negative for ANTX and/or HANTX production.

All extracts that inhibited 125I-Bgx binding, and several of those that did not, were analysed by GC/MS to confirm the results and to discriminate between ANTX and HANTX content. As expected, extracts of the control strains PCC 9240 and PCC 10111 contained ANTX and HANTX, respectively (Table 2). No toxin production could be detected in strains that gave negative results in the receptor ligand-binding assay. Of the strains previously unknown for toxicity but inhibitory in the receptor ligand-binding assay, three Oscillatoria sp. strains (PCC 6407, PCC 6412 and PCC 9107) synthesized ANTX (Table 2). Two strains of Oscillatoria sp. (PCC 6506 and PCC 9029), although derived from the same original isolate (see co-identity, Table 1), showed toxin patterns that were not in agreement with their strain records. Strain PCC 6506, grown under standard conditions in non-gassed flasks, in both medium BG11 and 2N, produced exclusively ANTX (Table 2). In contrast, strain PCC 9029 (deposited in the PCC 25 years later than strain PCC 6506), grown under the same conditions in medium 2N, produced mostly HANTX, but a small peak of ANTX eluting at the 600th scan was detected in the same chromatogram (Fig. 4a and insets). A difference spectrum mediated over six scans between scans 600 to 605 and between scans 590 to 595, to eliminate the background noise, confirmed unambiguously the presence of small amounts of ANTX together with the dominant HANTX (compare Fig. 4a, left inset, and Fig. 4b). In order to examine these unexpected results more thoroughly, cultures of larger volumes (10 l) were grown up for toxin analyses of strains PCC 6506 and PCC 9029. Aiming to achieve good growth and large biomass, medium BG11 supplemented with 10 mM NaHCO3 was employed and the vessels were gassed with air/CO2 (Table 2). These modifications of the growth conditions, surprisingly, led to the detection of HANTX instead of ANTX in the extracts of strain PCC 6506 (Table 2), whereas no change in toxin pattern was detected for strain PCC 9029 (Table 2). Similarly, no change in toxin pattern was observed for strains PCC 9240 and PCC 10111 in small- and large-scale cultures grown in media 2N and 2N10, respectively (Table 2), because they do not tolerate the high concentrations of nitrate present in media BG11 and BG1110.


View this table:
[in this window]
[in a new window]
 
Table 2. Growth conditions and yields of Oscillatoria strains in which ANTX and/or HANTX were detected by GC/MS

 


View larger version (29K):
[in this window]
[in a new window]
 
Fig. 4. GC/MS analysis of low-molecular-mass cell extracts of Oscillatoria sp. strain PCC 9029. (a) GC chromatogram. Right inset, EI spectrum of HANTX. Retention time: 654th scan. Left inset, EI spectrum of ANTX. Retention time: 600th scan. (b) Difference mass spectrum mediated over six scans between scans 600 to 605 and 590 to 595 to eliminate the background noise (see Fig. 4a, left inset).

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Since commercial standards of ANTX and HANTX were not available, these toxins were purified from Oscillatoria sp. strain PCC 9240 and O. formosa PCC 10111, respectively. UV spectroscopy and MS analyses (see Results) were comparable to previous data (Himberg, 1989; Skulberg et al., 1992). The limit of detection for ANTX by GC/MS was 1 ng on column (1 µl injection, full-scan mass spectrum, with signal-to-noise level of the order of 20 : 1, or better), corresponding to 6·1x10–6 M ANTX. This is a relatively low level of sensitivity compared to some other methods of detection (Himberg, 1989; Takino et al., 1999; Furey et al., 2003a) but was sufficient to reliably identify ANTX or HANTX in the low-molecular-mass cell extracts (Table 2), even when both toxins were present in the samples (Fig. 4a, b).

ANTX is an unstable molecule that undergoes hydration as well as oxidation reactions under certain conditions of pH and sunlight (Harada et al., 1993). In the present study a third degradation product, the molecule m/z 179, was observed during the HPLC purification step and after prolonged storage of the purified toxin in CH3CN, but not when low-molecular-mass extracts were analysed by GC/MS (data not shown). This product most likely results from the addition of one oxygen atom and the concomitant loss of two hydrogens to form a carbonyl (C=O) in place of a methylene (–CH2–) group (Stevens & Krieger, 1991, and references therein).

Electrophysiological and receptor-binding assays have mainly been used to determine the affinity of ANTX and HANTX for nAChRs and to characterize neuronal and muscular nAChR subtypes (Aronstam & Witkop, 1981; Wonnacott et al., 1991; Adeyemo & Sirén, 1992; Amar et al., 1993; Thomas et al., 1993). We have confirmed the high affinity of purified ANTX and HANTX to nAChRs by inhibition of 125I-Bgx binding to nAChRs present in Torpedo postsynaptic membranes, and their Ki values (see Results) are in the range of previously determined values. The method has a high sensitivity, the limit of detection of about 1x10–8 M being close to the security margin of 1 µg l–1 ({approx}6x10–9 M) proposed for ANTX in drinking water (Fawell et al., 1999).

The radioligand-binding assay permitted the detection of ANTX and/or HANTX in seven of the 76 strains tested (Tables 1 and 2), including five hitherto unknown for toxicity. All seven strains are members of the genus Oscillatoria, demonstrating the relatively high incidence of neurotoxicity in this genus (17 % of the 42 strains screened). In contrast, none of the strains of the other 12 taxa (Table 1), less extensively sampled but including members of the planktonic genera Arthrospira, Planktothrix, Anabaena and Aphanizomenon, as well as three strains whose genomes have been sequenced (Synechocystis sp. strain PCC 6803, Nostoc/Anabaena sp. strain PCC 7120 and Nostoc punctiforme PCC 73102), produced this class of toxins.

Oscillatoria sp. strain PCC 9240 is relatively large in diameter ({approx}8 µm) and synthesizes phycoerythrin, while the other six toxin-producing strains exhibit narrower trichomes (4–5 µm diameter) and lack phycoerythrin. These morphotypes correspond respectively to Oscillatoria clusters 2 and 4 as defined previously (Rippka & Herdman, 1992; Castenholz et al., 2001), frequently found in aquatic and terrestrial environments (R. Rippka, unpublished observations) and of widespread geographical distribution. Therefore, neurotoxicity can be correlated neither with taxonomic entities nor with the geographical origin of the strains, this class of toxins having been detected in isolates from Scandinavia, The Netherlands and the USA (Tables 1 and 2).

Extensive surveys for ANTX and/or HANTX of cyanobacterial blooms in Finland (Sivonen et al., 1989) and Ireland (Furey et al., 2003b) demonstrated that 16–20 % of the bloom samples contained these toxins. Concentrations of ANTX of 12–4360 µg per g freeze-dried material were reported for cyanobacterial blooms in Finland, and estimates for HANTX in Irish lakes were between 1·4 and 34 µg per l of bloom samples (Sivonen et al., 1989; Furey et al., 2003b). Under our growth conditions, 200 ml of an axenic culture of strain PCC 9240 yielded on average 20 mg lyophilized material containing about 80 µg toxin ({approx}4000 µg ANTX per g dry weight). The resulting extract prepared at 20 µg ANTX ml–1 showed 50 % inhibition of 125I-Bgx binding even after 100-fold dilution (Fig. 3a). Therefore, it should be possible to apply our experimental protocol for detection of ANTX and/or HANTX in cyanobacteria from environmental samples, provided that interfering components are lacking in such more complex material. Furthermore, similar to the loss of biological activity of other ANTX degradation compounds (Harada et al., 1993), the ANTX product m/z 179 showed no inhibition in the receptor-binding assay (Fig. 2). Therefore, if detection of inactive forms of the toxins is desired, GC/MS analyses will be required. The urgent need to monitor toxicity in the environment not only on blooms or the water column, but also on cyanobacterial mats in sediments or material attached on shore (Edwards et al., 1992), should be re-emphasized: none of the Oscillatoria strains here shown to produce ANTX and/or HANTX exhibit buoyancy (though some form small amounts of gas vesicles), and thus in nature may occur as benthic populations.

As reported for the heterocystous cyanobacterium Raphidiopsis mediterranea (Namikoshi et al., 2003), Oscillatoria sp. strain PCC 9029 synthesized traces of ANTX in addition to the dominant HANTX (Fig. 4a, b, Table 2), and thus the enzymes necessary for the synthesis of the two related toxins are present and functional. Oscillatoria sp. strain PCC 6506, supposedly the same isolate as Oscillatoria sp. strain PCC 9029 (see Table 1), is also capable of synthesizing ANTX and HANTX, but the respective presence of the toxins, surprisingly, is influenced by the growth conditions. These unexpected results may be due to a mutation in one of the two strains, thus affecting the regulation of synthesis and/or activities of the enzyme(s) that convert the common precursor (Hemscheidt et al., 1995) to either ANTX or HANTX. However, the co-identity of strains PCC 6506 and PCC 9029 needs to be confirmed in order to support this assumption. It should also be considered that all neurotoxin-producing strains, except O. formosa PCC 10111, entered the PCC as axenic cultures 10–35 years ago and, although they have maintained the capacity for toxin synthesis even under laboratory conditions, some changes in response to environmental factors may not be surprising.

In conclusion, we have shown the complementary utility of the receptor radioligand-binding assay and GC/MS for the detection of ANTX and HANTX. The former method permits tentative identification of these neurotoxins by virtue of their biological activity and has a high sensitivity, but does not discriminate between ANTX and HANTX. In contrast, GC/MS has a lower sensitivity of detection, but permits unambiguous identification of the toxins, as well as their respective degradation products. Consequently, if available, GC/MS would be the method of choice when discrimination between ANTX and HANTX is desired. However, for the purpose of environmental surveys, for which distinction between these two equally potent toxins may not be essential, the receptor ligand-binding assay is worthy of consideration as an alternative tool for routine analyses. With such a goal in mind it would be desirable, for the safety of the experimentalist(s), to further improve the receptor ligand-binding assay by developing a non-radioactive method of detection.


   ACKNOWLEDGEMENTS
 
This work was supported by the Institut Pasteur and its ‘Programme Transversal de Recherche’ (PTR 2000/25), by the CNRS (URA 2172 and URA 2182) and the College de France. We thank T. Laurent for harvesting and lyophilization of cyanobacterial material; T. Coursin for culture maintenance; J. Best for providing some freeze-dried samples; J. Ughetto for NMR analysis and N. Bouaïcha for allowing us to compare by HPLC our purified ANTX with his commercial standard. O. Skulberg and K. Sivonen kindly supplied strains Oscillatoria formosa NIVA CYA-92 and Oscillatoria sp. 193, respectively. We are also grateful to I. Guijarro and B. Baron for fruitfull discussions. We are much indebted to M. A. Bloch and B. Robert for their support. H.-O. N. and R. A. thank Professeur J. P. Changeux for his continuous interest and encouragement.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Adeyemo, O. M. & Sirén, A. L. (1992). Cardio-respiratory changes and mortality in the conscious rat induced by (+)- and (±)-anatoxin-a. Toxicon 30, 899–905.[CrossRef][Medline]

Amar, M., Thomas, P., Johnson, C., Lunt, G. G. & Wonnacott, S. (1993). Agonist pharmacology of the neuronal {alpha}7 nicotinic receptor expressed in Xenopus oocytes. FEBS Lett 327, 284–288.[CrossRef][Medline]

Aronstam, R. S. & Witkop, B. (1981). Anatoxin-a interactions with cholinergic synaptic molecules. Proc Natl Acad Sci U S A 78, 4639–4643.[Abstract]

Booker, M. J. & Walsby, A. E. (1979). The relative form resistance of straight and helical blue-green algal filaments. Br Phycol J 14, 141–150.

Bruno, M., Barbini, D. A., Pierdominici, E., Serse, A. P. & Ioppolo, A. (1994). Anatoxin-a and a previously unknown toxin in Anabaena planctonica from blooms found in Lake Mulargia (Italy). Toxicon 32, 369–373.[CrossRef][Medline]

Burger-Wiersma, T., Veenhuis, M., Korthals, H. J., Van de Wiel, C. C. M. & Mur, L. R. (1986). A new prokaryote containing chlorophylls a and b. Nature 320, 262–264.[CrossRef]

Carmichael, W. W. (1994). The toxins of cyanobacteria. Sci Am 270, 78–86.[Medline]

Carmichael, W. W. & Gorham, P. R. (1978). Anatoxins from clones of Anabaena flos-aquae isolated from lakes of western Canada. Mitt Int Ver Limnol 21, 285–295.

Castenholz, R. W., Rippka, R., Herdman, M. & Wilmotte, A. (2001). Subsection III (Formerly Oscillatoriales Elenkin 1934). In Bergey's Manual of Systematic Bacteriology, 2nd edn, vol. 1, the Archaea and the Deeply Branching and Phototrophic Bacteria, pp. 539–562. Edited by D. R. Boone, R. W. Castenholz & G. M. Garrity. New York, Berlin, Heidelberg: Springer.

Cheng, Y. & Prusoff, W. H. (1973). Relationship between the inhibition constant (Ki) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem Pharmacol 22, 3099–3108.[CrossRef][Medline]

Devlin, J. P., Edwards, O. E., Gorham, P. R., Hunter, N. R., Pike, R. K. & Stavric, B. (1977). Anatoxin-a, a toxic alkaloid from Anabaena flos-aquae NRC-44h. Can J Chem 55, 1367–1371.

Edwards, C., Beattie, K. A., Scrimgeour, C. M. & Codd, G. A. (1992). Identification of anatoxin-a in benthic cyanobacteria (blue-green algae) and in associated dog poisonings at Loch Insh, Scotland. Toxicon 30, 1165–1175.[CrossRef][Medline]

Fawell, J. K., Michell, R. E., Hill, R. E. & Everett, D. J. (1999). The toxicity of cyanobacterial toxins in the mouse. II. anatoxin-a. Hum Exp Toxicol 18, 168–173.[CrossRef][Medline]

Furey, A., Crowley, J., Lehane, M. & James, K. J. (2003a). Liquid chromatography with electrospray ion-trap mass spectrometry for the determination of anatoxins in cyanobacteria and drinking water. Rapid Commun Mass Spectrom 17, 583–588.[CrossRef][Medline]

Furey, A., Crowley, J., Shuilleabhain, A. N., Skulberg, O. M. & James, K. J. (2003b). The first identification of the rare cyanobacterial toxin, homoanatoxin-a, in Ireland. Toxicon 41, 297–303.[CrossRef][Medline]

Garcia-Pichel, F., López-Cortés, A. & Nübel, U. (2001). Phylogenetic and morphological diversity of cyanobacteria in soil desert crusts from the Colorado plateau. Appl Environ Microbiol 67, 1902–1910.[Abstract/Free Full Text]

Harada, K.-I., Kimura, Y., Ogawa, K., Suzuki, M., Dahlem, A. M., Beasley, V. R. & Carmichael, W. W. (1989). A new procedure for the analysis and purification of naturally occurring anatoxin-a from the blue-green alga Anabaena flos-aquae. Toxicon 27, 1289–1296.[CrossRef][Medline]

Harada, K.-I., Nagai, H., Kimura, Y., Suzuki, M., Park, H.-D., Watanabe, M. F., Luukkainen, R., Sivonen, K. & Carmichael, W. W. (1993). Liquid chromatography/mass spectrometric detection of anatoxin-a, a neurotoxin from cyanobacteria. Tetrahedron 49, 9251–9260.[CrossRef]

Hemscheidt, T., Rapala, J., Sivonen, K. & Skulberg, O. M. (1995). Biosynthesis of anatoxin-a in Anabaena flos-aquae and homoanatoxin-a in Oscillatoria formosa. J Chem Soc Commun 13, 1361–1362.

Hill, J. A., Nghiêm, H. O. & Changeux, J.-P. (1991). Serine-specific phosphorylation of nicotinic receptor associated 43K protein. Biochemistry 30, 5579–5585.[CrossRef][Medline]

Himberg, K. (1989). Determination of anatoxin-a, the neurotoxin of Anabaena flos-aquae cyanobacterium, in algae and water by gas chromatography-mass spectrometry. J Chromatogr 481, 358–362.[CrossRef][Medline]

Mahmood, N. A. & Carmichael, W. W. (1986). The pharmacology of anatoxin-a(s), a neurotoxin produced by the freshwater cyanobacterium Anabaena flos-aquae NRC 525-17. Toxicon 24, 425–434.[CrossRef][Medline]

Namikoshi, M., Murakami, T., Watanabe, M. F., Oda, T., Yamada, J., Tsujimura, S., Nagai, H. & Oishi, S. (2003). Simultaneous production of homoanatoxin-a, anatoxin-a, and a new non-toxic 4-hydroxyhomoanatoxin-a by the cyanobacterium Raphidiopsis mediterranea Skuja. Toxicon 42, 533–538.[CrossRef][Medline]

Rippka, R. (1988). Isolation and purification of cyanobacteria. Methods Enzymol 167, 3–27.[Medline]

Rippka, R. & Herdman, M. (1992). Pasteur Culture Collection of Cyanobacterial Strains in Axenic Culture: Catalogue & Taxonomic Handbook, vol. I, Catalogue of Strains. Paris: Institut Pasteur.

Rippka, R., Deruelles, J., Waterbury, J. B., Herdman, M. & Stanier, R. Y. (1979). Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J Gen Microbiol 111, 1–61.

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Schlösser, U. G. (1994). SAG – Sammlung von Algenkulturen at the University of Göttingen, Catalogue of Strains 1994. Bot Acta 107, 113–186.

Sivonen, K. & Jones, G. (1999). Cyanobacterial toxins. In Toxic Cyanobacteria in Water: a Guide to their Public Health Consequences, Monitoring and Management, pp. 41–111. Edited by I. Chorus & J. Bartram. London: E. & F. N. Spon.

Sivonen, K., Himberg, K., Luukkainen, R., Niemelä, S. I., Poon, G. K. & Codd, G. A. (1989). Preliminary characterization of neurotoxic cyanobacteria blooms and strains from Finland. Toxic Assess 4, 339–352.

Sivonen, K., Niemelä, S. I., Niemi, R. M., Lepistö, L., Luoma, T. H. & Räsänen, L. A. (1990). Toxic cyanobacteria (blue-green algae) in Finnish fresh and coastal waters. Hydrobiologia 190, 267–275.[CrossRef]

Skulberg, O. M., Carmichael, W. W., Andersen, R. A., Matsunaga, S., Moore, R. E. & Skulberg, R. (1992). Investigations of a neurotoxic oscillatorialean strain (Cyanophyceae) and its toxin. Isolation and characterization of homoanatoxin-a. Environ Toxicol Chem 11, 321–329.

Starr, R. C. & Zeikus, J. A. (1993). UTEX – the culture collection of algae at the University of Texas at Austin. J Phycol 29 (supplement), 1–106.[CrossRef]

Stevens, D. K. & Krieger, R. I. (1991). Stability studies on the cyanobacterial nicotinic alkaloid anatoxin-a. Toxicon 29, 167–179.[CrossRef][Medline]

Sugiyama, J. (1998). Microalgae. In IAM Catalogue of Strains, 2nd edn, pp. 237–264. Edited by the Editorial Board of the IAM Catalogue of Strains. Center for Cellular and Molecular Research, Institute of Molecular and Cellular Biosciences, University of Tokyo. Tokyo: Microbiology Research Foundation.

Swanson, K. L., Allen, C. N., Aronstam, R. S., Rapoport, H. & Albuquerque, E. X. (1985). Molecular mechanisms of the potent and stereospecific nicotinic receptor agonist (+)-anatoxin-a. Mol Pharmacol 29, 250–257.

Swanson, K. L., Aronstam, R. S., Wonnacott, S., Rapoport, H. & Albuquerque, E. X. (1991). Nicotinic pharmacology of anatoxin analogs. I. Side chain structure-activity relationships at peripheral agonist and noncompetitive antagonist sites. J Pharmacol Exp Ther 259, 377–386.[Abstract]

Takino, M., Daishima, S. & Yamaguchi, K. (1999). Analysis of anatoxin-a in freshwaters by automated on-line derivatization-liquid chromatography-electrospray mass spectrometry. J Chromatogr A 862, 191–197.[CrossRef][Medline]

Thomas, P., Stephens, M., Wilkie, G. & 8 other authors (1993). (+)-Anatoxin-a is a potent agonist at neuronal nicotinic acetylcholine receptors. J Neurochem 60, 2308–2311.[Medline]

Tomaselli, L., Giovannetti, L. & Torzillo, G. (1993). Physiology of stress response in Spirulina spp. Bull Inst Océanogr (Monaco) no. spécial 12, 65–75.

Viti, C., Ventura, S., Lotti, F., Capolino, E., Tomaselli, L. & Giovannetti, L. (1997). Genotypic diversity and typing of cyanobacterial strains of the genus Arthrospira by very sensitive total DNA restriction profile analysis. Res Microbiol 148, 605–611.[CrossRef][Medline]

Wonnacott, S., Jackman, S., Swanson, K. L., Rapoport, H. & Albuquerque, E. X. (1991). Nicotinic pharmacology of anatoxin analogs. II. Side chain structure-activity relationships at neuronal nicotinic ligand binding sites. J Pharmacol Exp Ther 259, 387–391.[Abstract]

Received 28 September 2004; revised 30 December 2004; accepted 4 January 2005.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Aráoz, R.
Articles by Herdman, M.
Articles citing this Article
PubMed
PubMed Citation
Articles by Aráoz, R.
Articles by Herdman, M.
Agricola
Articles by Aráoz, R.
Articles by Herdman, M.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2005 Society for General Microbiology.