- Email: [email protected]
A rapid microplate fluorescence method to detect yessotoxins based on their capacity to activate phosphodiesterases
ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 326 (2004) 93–99 www.elsevier.com/locate/yabio
A rapid microplate fluorescence method to detect yessotoxins based on their capacity to activate phosphodiesterases Amparo Alfonso,a Mercedes R. Vieytes,b Takeshi Yasumoto,c and Luis M. Botanaa,* a
Departamento de Farmacologıa, Facultad de Veterinaria, Univesidad de Santiago de Compostela, 27002 Lugo, Spain b Departamento de Fisiologıa, Facultad de Veterinaria, Univesidad de Santiago de Compostela, 27002 Lugo, Spain c Japan Food Research Laboratories, Tama, Tokyo 206-0025, Japan Received 30 September 2003
Abstract This paper describes an easy and fast assay with enough sensitivity to detect yessotoxin (YTX) in shellfish samples. YTX decreases intracellular adenosine 30 ; 50 -cyclic monophosphate (cAMP) levels by increasing the activity of phosphodiesterases (PDEs). Looking for new methods to detect YTXs, we developed a technique based on this effect. We use the fluorescent derivative of cAMP, anthranyloyl-cAMP, whose fluorescence decreases in time by hydrolysis effect of PDEs. The fluorescence fall is quantified in a plate reader. PDEs induce an anthranyloyl-cAMP hydrolysis rate that is increased in the presence of YTX. This effect is dose dependent, and the representation of YTX concentration versus rate of hydrolysis follows a lineal regression. The measurable range of YTX in this assay is 0.1 to 10 lM, while by mouse bioassay, the official method to detect YTXs, the detection limit is 2 lM. We determined by this method the concentration of YTX from alcoholic extracts whose concentrations were first determined by high performance liquid chromatography and the variation of concentration was from 5.26 lM by fluorescence to 6 lM by high performance liquid chromatography and from 3.16 by fluorescence to 3 lM by HPLC. Ó 2003 Elsevier Inc. All rights reserved.
Yessotoxin (YTX)1 is a polycyclic ether toxin produced by phytoplankton and was originally isolated from the digestive gland (hepatopancreas) of the scallop Patinopecten yessoensis by Murata et al. [1]. There are more than 50 natural derivatives and only a few structures have been reported (Scheme 1). Initial YTX episodes were restricted to specific environments; however, in recent years they have been found in several marine areas around the world [2]. YTXs coexist with other phycotoxins in shellfish, namely the diarrhetic shellfish poisoning (DSP) okadaic acid, dinophysistoxins, pectenotoxins, and azaspirazids. These seafood toxins can
*
Corresponding author. Fax: +34-982-252-242. E-mail address: [email protected] (L.M. Botana). 1 Abbreviations used: YTXs, yessotoxins; PDEs, phosphodiestarases; cAMP, adenosine 30 ; 50 -cyclic monophosphate; Ant-cAMP, adenosine 30 ; 50 -cyclic monophosphate; 20 -O-anthraniloyl-, sodium salt; DSP, diarrheic shellfish poisoning; IBMX, 3-isobutyl-1-methylxanthine. 0003-2697/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2003.11.022
cause important health and economic damage, and for this reason it is very important to standardize a method to determine its presence and concentration, since, so far, the only validated method is the mouse bioassay. Although used for a long time, the use of bioassay has been very much stressed out, since recent regulatory modifications in the European Union require the monitoring of all the above-mentioned toxins, and this poses a major problem for the use of the bioassay, which cannot cope with so many different toxin groups. YTXs are not diarrhogenic toxins, and in fact there is no clear evidence that these toxins are toxic to humans, since oral administration does not show any toxicity to rodents. However, YTXs show a high lethality when they are intraperitoneally injected in mice [3]; hence they are a major cause of false positives in the monitoring of DSP toxins. As mentioned, YTXs often coexist with DSP and produce positive results when tested by the conventional mouse bioassay method for detecting DSP toxins [4]. The mouse bioassay is the basic tool in monitoring programs for DSP and YTXs. This assay is
94
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
Scheme 1. Structure of yessotoxin.
limited not only by ethical considerations but also by occasional false positive and negative results, lack of accuracy and specificity, and the extraction/purification procedure. As a consequence the tolerance limits are not equivalent among different countries [2]. Initially the mouse bioassay involved the extraction of all DSP with acetone, drying the extract, dissolving it in Tween 60, and intraperitoneal injection of the suspension; however, fatty acids can also be extracted and could be lethal by intraperitoneal injection [5]. This assay was later modified, some organic extractions were added to eliminate fatty acids, and the time of mice survival was changed from 24 to 6 h [4]. To reveal the mechanism of action of YTXs it is essential to develop new detection assays. It has been reported that YTX does not inhibit phosphatases, as is the case for okadaic acid [2] and it modifies calcium fluxes in human lymphocytes [6,7]. Recently we reported that YTX decreases adenosine 30 ; 50 cyclic monophosphate (cAMP) by activation of cellular phosphodiesterases (PDEs), which points to the PDE system as the intracellular target of YTX [8]. PDEs are enzymes that hydrolyze the cyclic nucleotides cAMP and cGMP to their biologically inactive 50 derivatives. Cyclic nucleotide PDEs comprise 11 families of enzymes (PDE1–PDE11) with different substrate specificity, affinity, sensitivity to inhibitors, and tissue localization. PDE families 5, 6, and 9 are cGMP specific [9], PDE families 1, 3, 4, 7, and 8 hydrolyze cAMP [10,11], and PDE families 2, 10, and 11 hydrolyze both cAMP and cGMP [12,13]. All cAMP PDEs have been localized in the brain, even though in
some cases it is not the tissue where the PDEs are highly expressed [11,12,14,15]. In a previous study [16] we determined the phosphodiesterase activity in rat mast cells by using a fluorescent analog of cAMP, anthranyloyl-cAMP (Ant-cAMP). The fluorescence of this analog decreases in the presence of PDEs that catalyze its hydrolysis. This substrate is very sensitive and the PDE activity can easily be calculated and is usually expressed as Ant-cAMP hydrolysis rate [17]. Due to the YTX effect on PDEs, the rate of hydrolysis should be increased in the presence of the toxin. Taking together all these results, the aim of this study was to develop a YTX assay to detect the toxin in marine samples and eventually replace the need for the bioassay to monitor YTXs.
Materials and methods Chemicals Yessotoxin was purified by Dr. T. Yasumoto, and part of it was used to obtain a YTX extract in methanol from a matrix of dinoflagellate cultures. Ant-cAMP sodium was from Calbiochem. Phosphodiesterase, 30 ; 50 cyclic nucleotide specific from bovine brain, and all other chemicals were from Sigma, Madrid, Spain. The composition of saline solution used was (mM) Tris–HCl 10 mM, CaCl2 1 mM, and the final pH was adjusted to 7.4.
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
Measurement of phosphodiesterase activity PDE activity was checked by using an analog of cAMP, adenosine 30 ; 50 cyclic monophosphate, 20 -Oanthraniloyl-, sodium salt [16]. Ant-cAMP presents a maximum of fluorescence for 330-nm excitation and 445-nm emission; its fluorescence decreases when phosphodiesterases are activated. The PDE activity was measured in a 96-well plate with a final volume of 100 ll. Each condition was performed in triplicate. The fluorescence was read in a fluorescence plate reader (Bio-Tek FL600) by using 350 nm and 445 nm as excitation and emission wavelengths, respectively. A calibration curve was prepared (r > 0:999) with substrate standards covering the range 0–10 lM. The fluorescence values were converted to lmol (or pmol) substrate/ml by using this calibration curve. The enzyme activity is expressed as a rate of substrate hydrolysis lmol-pmol/ml/min. Shellfish extract processing Uncontaminated mussels (Mytilus galloprovincialis Lmk) were opened and the meat was collected and homogenized; 100 g of homogenate was extracted with 300 ml acetone. The meat extract was filtrated by vacuum and extracted again with 300 ml of methanol. Acetone and methanol extract were mixed and vacuum dried. The dried extract was diluted in 100 ml dichloromethane and extracted with 50 ml water. The organic phase with YTX was vacuum dried and dissolved in 4 ml methanol at 90%. This protocol was made following the CEE decision [18]. This is the uncontaminated sample C. When necessary, HPLC determinations of YTX were done according to [19]. Data analysis
95
PDEs (final volume 100 ll) are added and then the fluorescence versus time is recorded. As mentioned under Materials and methods, in parallel we prepared a calibration curve of substrate in the range 0–10 lM where PDEs were not added. With values obtained in this curve the substrate fluorescence is transformed in concentration. As Fig. 1 shows, the amount of AntcAMP decreases from 8 to 6 lM under control conditions without PDE, and in the presence of 0.05 U/ml PDEs the decrease goes from 8 to 1.5 lM. From these data, taking the concentration increase in time intervals, the concentration fall can be shown as rate of hydrolysis, which is lmol/ml substrate hydrolyzed per minute. This is the parameter that will be shown in the following graphics. The effect of PDEs is dose dependent. As Fig. 2 shows, the rate of hydrolysis increases (taken from the slope of the curves) when the amount of PDE is increased. From this experiment the amount of 0.03 U/ml PDEs was chosen as the amount best suited to hydrolyze the dye Ant-cAMP. Next we determined the best time to study the hydrolysis of Ant-cAMP, which is the time where the fluorescence fall is higher. As Fig. 3 shows, in the presence of 0.03 U/ml PDEs, after 5 min the rate of hydrolysis is higher than the rate at 10 or 15 min. However, the variability is lower after 10 or 15 min than after 5 min; therefore the best time frame to check the variation of hydrolysis rate is after 10 or 15 min and, we chose the 10-min interval. We then studied whether the PDE effect on AntcAMP fluorescence could be modified by PDE modulators. We used 3-isobutyl-1-methylxanthine (IBMX), a known PDE inhibitor, imidazole, and YTX as PDE activators. We added PDE modulators 2 min later than PDEs and then calculated the rate of hydrolysis at 5, 10, and 15 min after drug addition. As Fig. 4 shows, 500 lM
All experiments were carried out at least three times in triplicate. Results were analyzed using the Student t test for paired data or the ANOVA test. A probability level of 0.05 or smaller was used for statistical significance. Data were normalized and results expressed as the mean SE.
Results Looking for new methods to detect YTXs, we developed a technique based on the effect of this toxin on the PDE activity recently described [8]. Antraniloyl cAMP is a known fluorescent substrate for PDEs [16]. Its fluorescence decreases when the substrate is hydrolyzed. Initially we checked the decrease of the dye fluorescence with and without PDEs. In a 96well plate saline solution, 8 lM dye, and 0.05 U/ml
Fig. 1. Variation in Ant-cAMP concentration over time under control conditions (closed circles) and in the presence of 0.05 U/ml PDEs. A representative assay is shown.
96
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
Fig. 2. Variation in the rate of hydrolysis of Ant-cAMP in the presence of different concentrations of PDEs: 0.01, 0.02, 0.03, 0.04, and 0.05 U/ ml of PDEs was added and the variation in Ant-cAMP concentration was recorded and transformed to hydrolysis rate. A representative assay is shown.
Fig. 3. Variations in the rate of hydrolysis of Ant-cAMP after 5, 10, and 15 min. Mean SE of three experiments. (*) Significant differences with respect to the control without PDEs.
IBMX inhibits by 40% the hydrolysis rate induced by 0.03 U/ml PDEs after 5, 10, and 15 min; 5 mM imidazole increases by 26% the hydrolysis rate after 5 and 10 min and by 46% after 15 min; 2 lM YTX increases by 60% the hydrolysis rate after 5 and 10 min and by 30% after 15 min. Stock solutions of these drugs were made in dimethyl sulfoxide, and this solvent does not modify the hydrolysis rate (data not shown). All these experiments were done at 37 °C. Next we check the effect of temperature on the hydrolysis rate. As Fig. 5 shows, the hydrolysis rate induced by 0.03 U/ ml PDEs is higher at 20 °C than at 37 °C. However, the increase induced by YTXs is significantly higher at 37 °C. Fig. 5B shows the increase in hydrolysis rate induced by 2 lM YTX from data of Fig. 5A. As shown, at
Fig. 4. Variations in Ant-cAMP hydrolysis rate in the presence of PDE modulators; 0.03 U/ml PDEs was first added and then 2 lM YTX, 5 mM imidazole, or 500 lM IBMX. The variation in Ant-cAMP concentration was checked after 5, 10, and 15 min. Mean SE of three experiments. (*) Significant differences with respect to the PDEs control.
37 °C the increase induced by YTX is 50%; however, at 20 °C it is only 10%. Next we checked the effect of different amounts of YTX in the hydrolysis rate after 10 min. We use YTX in the range 10 to 0.1 lM. As Fig. 6 shows, the hydrolysis rate decreases with the concentration. The representation of these data follows a lineal regression (r > 0:9). Finally, after these results we elaborated a method to detect YTX in contaminated samples. We checked our method with two YTX extracts disolved in methanol and obtained from spiking YTX to dinoflagellate cultures extracts, whose YTX concentrations were determined by HPLC [19], sample A with 6 lM YTX and sample B with 3 lM YTX. We also checked the hydrolysis rate induced by an extract from uncontaminated mussel, sample C, to study the interferences with our technique of the sample matrix obtained according to the CEE extraction method [18]. In addition we studied the interference of methanol on the hydrolysis rate since following other official methods [20] the extract from contaminated mussels can be dissolved in methanol. The experiment was carried out at 37 °C, final volume 100 ll, and the hydrolysis rate was calculated after 10 min. In one approach to do the experiment, in a 96well plate there are 4 control wells: (I) Ant-cAMP calibration wells, with saline solution and substrate in the range 0–10 lM; (II) control wells, with saline solution, 8 lM substrate, and 0.03 U/ml PDEs; (III) YTX calibration wells, with 8 lM substrate, 0.03 U/ml PDEs, and YTX in the range 0.1–10 lM; and (IV) problem sample wells, with saline solution, 8 lM substrate, 0.03 U/ml PDEs, and samples A, B, C or methanol. We add saline solutions and the corresponding amount of dye in all wells and then read the plate for 2 min. After addition in wells II, III, and IV of 0.03 U/ml PDEs, the fluorescence
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
97
Fig. 5. (A) Effect of temperature in the Ant-cAMP hydrolysis rate induced by PDEs in the presence of YTX; 0.03 U/ml PDEs was first added and then 2 lM YTX. The variation in Ant-cAMP concentration was checked after 5, 10, and 15 min. The experiment was done at 20 and 37 °C. Mean SE of three experiments. (*) Significant differences with respect to the PDEs control at any temperature. (B) % of increase with respect to PDE control (from data in A).
Table 1 Concentration of YTX on different samples obtained by HPLC and the fluorescent method
Sample A Sample B Sample C Methanol
HPLC
Fluorescence plate method
6 lM 3 lM 0 0
5.26 1.1 lM 3.16 0.4 lM —
0
Sample A and Sample B, methanolic extracts; Sample C, uncontaminated extract; Methanol, control of methanol. Mean SE of three experiments.
Fig. 6. Variations in Ant-cAMP hydrolysis rate induced by PDEs in the presence of different YTX concentrations; 0.03 U/ml PDEs was first added and then different concentrations of YTX. The variation in Ant-cAMP concentration was checked after 10 min. Each point is the mean of three determinations. Standard error was typically 5–10%.
ternative extraction method based in methanol, as used for the phosphatase inhibition assay [20], can be safely used, as shown above. Therefore, this is a sensitive method to detect YTX in samples.
Discussion is read again for 2 min. Then, YTX or samples are added to wells III and IV and the fluorescence is read back for 10 min. With the calibration curve obtained with wells I, fluorescence values are converted into Ant-cAMP concentrations, and then the concentration is transformed to rate of hydrolysis. From data of wells III it is possible to obtain a rate of hydrolysis for each YTX concentration and then perform a calibration curve. With the equation of this calibration curve (YTX concentration versus hydrolysis rate) it is feasible to transform the rate of hydrolysis of wells IV (samples A, B, C, and methanol) in YTX concentration. As Table 1 shows, with this method, the YTX concentration values obtained were 5.26 lM in sample A and 3.16 lM in sample B, instead of 6 and 3 lM, respectively, and no interference was detected with methanol as solvent. However the uncontaminated mussels extract obtained following the extraction method of CEE is so dark and fatty that it interferes with the detection technique, although an al-
The mouse bioassay currently represents the only method to detect and monitor YTX contaminations. Nevertheless this method has ethical problems, it can show false-positive and false-negative results, and it has no specificity. An important tool to develop new detection techniques to any marine toxin is to know its mechanism of action. Recently, we have shown that YTX activates cAMP hydrolysis by PDEs activation [8]. Therefore, after this confirmation, the aim of this work was, by using this finding, to develop a method to detect YTX. In the present assay we directly measure the effect of YTX by using the intracellular toxin target and a fluorogenic substrate. By using an assay based on the use of the biochemical target, it is feasible to discriminate between toxins present in a sample. Usually YTX appears together with DSP toxins and it interferes in DSP bioassay since YTX is a very toxic substance after
98
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
intraperitoneal injection [4]. As far as we know, there is no other marine toxin that shows an effect on PDEs, and therefore with this method it is possible to check the YTX present in a sample without any special extraction to separate other toxins. Usually to separate DSP and YTX it is necessary to perform several organic extractions that collect fatty extracts that also interfere with the mouse bioassay [21]. The linear response rate of hydrolysis from different concentrations of YTX was between 0.1 and 10 lM; therefore this fluorescent method can detect YTX at concentrations over 0.5 lM, while the limit of detection for the mousse bioassay is 2 lM YTX [2]. Therefore the fluorescent method is more sensitive and allows one to detect and measure the toxin concentration in a sample, while the mouse bioassay provides only information on the presence absence of the toxin. The main problem with our technique seems to be the fatty interference of the extract, but this is a problem that also appears in the mouse bioassay [4], and even if some fatty acids are eliminated with organic extractions, in these processes the toxin can be lost [22,23] and the final extract is inevitably dirty. There are other official methods to perform the extraction to detect YTX and DSP toxins. For example, the direct extraction of toxins from 1 g of digestive mussel glands with 4 ml 80% methanol [20] provides clear extracts where YTX can be detected. This is a method used to isolate DSP toxins due to their lipophilic nature [2] which also provides good recovery results with other detection methods [20]. YTXs are accumulated in digestive glands, so extracting only this gland does not induce toxin losses and the fatty part from the shellfish meat is avoided. In the search for new methods to detect YTX, some cytotoxicity assays were developed. It was shown that YTX induces some morphologic changes in the cell surface; however, these changes are not dose dependent [24]. Hepatocytes were also proposed as a cellular line to discriminate YTX and other DSP toxins; however, confusing results discard this assay [2]. Recently, a functional method measuring E-cadherine detection [25] has been reported. This method is very sensitive and specific, but the time and the complicated process required to obtain the results are important drawbacks; also the use of cell lines is a major issue against validating a detection method, due to its variability. There is also an immunoassay to detect YTX [26]. This assay is a quantitative and fast analytical method but the existence of more than 50 analogs with different specificities is an important problem to overcome. Finally the available instrumental methods (liquid or gas chromatography, capillary electrophoresis, mass spectrometry) are specific and sensitive but there is no reference material for a routine detection [2], and these methods are rather expensive and difficult to implement in a routine laboratory.
In summary, this paper reports a new functional, sensitive, easy, and fast assay to detect YTXs in a marine sample.
Acknowledgments This work was funded by Grants MCYT BMC20000441, SAF2003-08765-C03-02, REN2001-2959-C04-03, REN2003-06598-C02-01, INIA CAL01-068, Xunta PGIDT99INN26101 and PGDIDIT03AL26101PR, and FISS REMA-G03-007.
References [1] M. Murata, M. Kumagai, J.S. Lee, T. Yasumoto, Tetrahedron Lett. 28 (1987) 5869–5872. [2] R. Draisci, L. Lucentini, A. Mascioni, in: L.M. Botana (Ed.), Seafood and Freshwater Toxins: Pharmacology, Physiology and Detection, Marcel Dekker, New York, 2000, pp. 289–324. [3] H. Ogino, M. Kumagai, T. Yasumoto, Nat. Toxins 5 (1997) 255– 259. [4] T. Yasumoto, M. Murata, Y. Oshima, G.K. Matsumoto, J. Glardy, in: E.P. Ragelis, (Ed.), ACS Symp. Series. No 262. Seafood toxins. American Chemical Society, Washington, DC, 1984, pp. 207–214. [5] T. Yasumoto, Y. Oshima, M. Yamaguchi, Bull. Jpn. Soc. Sci. Fish 44 (1978) 1249–1255. [6] L.A. De la Rosa, A. Alfonso, N. Vilari~ no, M.R. Vieytes, L.M. Botana, Biochem. Pharmacol. 61 (2001) 827–833. [7] L.A. De la Rosa, A. Alfonso, N. Vilari~ no, M.R. Vieytes, T. Yasumoto, L.M. Botana, Cell. Signal 13 (2001) 711–716. [8] A. Alfonso, L.A. de la Rosa, M.R. Vieytes, T. Yasumoto, L.M. Botana, Biochem. Pharmacol. 65 (2003) 193–208. [9] D.A. Fisher, J.F. Smith, J.S. Pillar, S.H. St Denis, J.B. Cheng, J. Biol. Chem. 273 (1998) 15559–15564. [10] J. Mukai, T. Asai, M. Naka, T. Tanaka, Br. J. Pharmacol. 111 (1994) 389–390. [11] J.M. Hetman, S.H. Soderling, N.A. Glavas, J.A. Beavo, Proc. Natl. Acad. Sci. USA 97 (2000) 472–476. [12] J.A. Beavo, Physiol. Rev. 75 (1995) 725–748. [13] K. Fujishige, J. Kotera, K. Yuasa, K. Omori, Eur. J. Biochem. 267 (2000) 5943–5951. [14] C.H. Cho, D.H. Cho, M.R. Seo, Y.S. Juhnn, Exp. Mol. Med. 32 (2000) 110–114. [15] A.Z. Zhao, J.N. Huan, S. Gupta, R. Pal, A. Sahu, Nat. Neurosci. 5 (2002) 727–728. [16] A. Alfonso, M. Estevez, M.C. Louzao, M.R. Vieytes, L.M. Botana, Cell. Signal 7 (1995) 513–518. [17] R.L. Kincaid, V.C. Manganiello, Methods Enzymol. 159 (1988) 457–470. [18] 2002/225/EC, D., Official J. Eur. Communities L75 (2002) 62–64. [19] T. Yasumoto, A. Takizawa, Biosci. Biotechnol. Biochem. 61 (1997) 1775–1777. [20] J.C. Gonzalez, F. Leira, O.I. Fontal, M.R. Vieytes, F.F. Arevalo, J.M. Vieites, M. Berm udez-Puente, S. Mu~ niz, C. Salgado, T. Yasumoto, L.M. Botana, Anal. Chim. Acta 466 (2002) 233–246. [21] M.L. Fernandez, A. Mıguez, E. Cacho, A. Martınez, J. Diogene, T. Yasumoto, in: E.A. Sar, M.E. Ferrario, M. Reguera, (Eds.), Floraciones algales nocivas en el cono sur americano. UNESCO y Ministerio de Ciencia y Tecnologıa, Madrid, 2002.
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99 [22] R. Draisci, L. Giannetti, L. Lucentini, C. Marchiafava, K.J. James, A.G. Bishop, B.M. Healy, S.S. Kelly, J. Chromatogr. A 798 (1998) 137–145. [23] R. Draisci, E. Ferretti, L. Palleschi, C. Marchiafava, R. Poletti, A. Milandri, A. Ceredi, M. Pompei, Toxicon 37 (1999) 1187– 1193.
99
[24] T. Aune, T. Yasumoto, E. Engeland, J. Toxicol. Environ. Health 34 (1991) 1–9. [25] S. Pierotti, C. Malaguti, A. Milandri, R. Poletti, P. Rossini, Anal. Biochem. 312 (2003) 208–216. [26] I. Garthwaite, K. Ross, C. Miles, L. Briggs, N. Towers, T. Borrell, P. Busby, J. AOAC Int. 84 (2001) 1643–1648.
A rapid microplate fluorescence method to detect yessotoxins based on their capacity to activate phosphodiesterases Amparo Alfonso,a Mercedes R. Vieytes,b Takeshi Yasumoto,c and Luis M. Botanaa,* a
Departamento de Farmacologıa, Facultad de Veterinaria, Univesidad de Santiago de Compostela, 27002 Lugo, Spain b Departamento de Fisiologıa, Facultad de Veterinaria, Univesidad de Santiago de Compostela, 27002 Lugo, Spain c Japan Food Research Laboratories, Tama, Tokyo 206-0025, Japan Received 30 September 2003
Abstract This paper describes an easy and fast assay with enough sensitivity to detect yessotoxin (YTX) in shellfish samples. YTX decreases intracellular adenosine 30 ; 50 -cyclic monophosphate (cAMP) levels by increasing the activity of phosphodiesterases (PDEs). Looking for new methods to detect YTXs, we developed a technique based on this effect. We use the fluorescent derivative of cAMP, anthranyloyl-cAMP, whose fluorescence decreases in time by hydrolysis effect of PDEs. The fluorescence fall is quantified in a plate reader. PDEs induce an anthranyloyl-cAMP hydrolysis rate that is increased in the presence of YTX. This effect is dose dependent, and the representation of YTX concentration versus rate of hydrolysis follows a lineal regression. The measurable range of YTX in this assay is 0.1 to 10 lM, while by mouse bioassay, the official method to detect YTXs, the detection limit is 2 lM. We determined by this method the concentration of YTX from alcoholic extracts whose concentrations were first determined by high performance liquid chromatography and the variation of concentration was from 5.26 lM by fluorescence to 6 lM by high performance liquid chromatography and from 3.16 by fluorescence to 3 lM by HPLC. Ó 2003 Elsevier Inc. All rights reserved.
Yessotoxin (YTX)1 is a polycyclic ether toxin produced by phytoplankton and was originally isolated from the digestive gland (hepatopancreas) of the scallop Patinopecten yessoensis by Murata et al. [1]. There are more than 50 natural derivatives and only a few structures have been reported (Scheme 1). Initial YTX episodes were restricted to specific environments; however, in recent years they have been found in several marine areas around the world [2]. YTXs coexist with other phycotoxins in shellfish, namely the diarrhetic shellfish poisoning (DSP) okadaic acid, dinophysistoxins, pectenotoxins, and azaspirazids. These seafood toxins can
*
Corresponding author. Fax: +34-982-252-242. E-mail address: [email protected] (L.M. Botana). 1 Abbreviations used: YTXs, yessotoxins; PDEs, phosphodiestarases; cAMP, adenosine 30 ; 50 -cyclic monophosphate; Ant-cAMP, adenosine 30 ; 50 -cyclic monophosphate; 20 -O-anthraniloyl-, sodium salt; DSP, diarrheic shellfish poisoning; IBMX, 3-isobutyl-1-methylxanthine. 0003-2697/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2003.11.022
cause important health and economic damage, and for this reason it is very important to standardize a method to determine its presence and concentration, since, so far, the only validated method is the mouse bioassay. Although used for a long time, the use of bioassay has been very much stressed out, since recent regulatory modifications in the European Union require the monitoring of all the above-mentioned toxins, and this poses a major problem for the use of the bioassay, which cannot cope with so many different toxin groups. YTXs are not diarrhogenic toxins, and in fact there is no clear evidence that these toxins are toxic to humans, since oral administration does not show any toxicity to rodents. However, YTXs show a high lethality when they are intraperitoneally injected in mice [3]; hence they are a major cause of false positives in the monitoring of DSP toxins. As mentioned, YTXs often coexist with DSP and produce positive results when tested by the conventional mouse bioassay method for detecting DSP toxins [4]. The mouse bioassay is the basic tool in monitoring programs for DSP and YTXs. This assay is
94
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
Scheme 1. Structure of yessotoxin.
limited not only by ethical considerations but also by occasional false positive and negative results, lack of accuracy and specificity, and the extraction/purification procedure. As a consequence the tolerance limits are not equivalent among different countries [2]. Initially the mouse bioassay involved the extraction of all DSP with acetone, drying the extract, dissolving it in Tween 60, and intraperitoneal injection of the suspension; however, fatty acids can also be extracted and could be lethal by intraperitoneal injection [5]. This assay was later modified, some organic extractions were added to eliminate fatty acids, and the time of mice survival was changed from 24 to 6 h [4]. To reveal the mechanism of action of YTXs it is essential to develop new detection assays. It has been reported that YTX does not inhibit phosphatases, as is the case for okadaic acid [2] and it modifies calcium fluxes in human lymphocytes [6,7]. Recently we reported that YTX decreases adenosine 30 ; 50 cyclic monophosphate (cAMP) by activation of cellular phosphodiesterases (PDEs), which points to the PDE system as the intracellular target of YTX [8]. PDEs are enzymes that hydrolyze the cyclic nucleotides cAMP and cGMP to their biologically inactive 50 derivatives. Cyclic nucleotide PDEs comprise 11 families of enzymes (PDE1–PDE11) with different substrate specificity, affinity, sensitivity to inhibitors, and tissue localization. PDE families 5, 6, and 9 are cGMP specific [9], PDE families 1, 3, 4, 7, and 8 hydrolyze cAMP [10,11], and PDE families 2, 10, and 11 hydrolyze both cAMP and cGMP [12,13]. All cAMP PDEs have been localized in the brain, even though in
some cases it is not the tissue where the PDEs are highly expressed [11,12,14,15]. In a previous study [16] we determined the phosphodiesterase activity in rat mast cells by using a fluorescent analog of cAMP, anthranyloyl-cAMP (Ant-cAMP). The fluorescence of this analog decreases in the presence of PDEs that catalyze its hydrolysis. This substrate is very sensitive and the PDE activity can easily be calculated and is usually expressed as Ant-cAMP hydrolysis rate [17]. Due to the YTX effect on PDEs, the rate of hydrolysis should be increased in the presence of the toxin. Taking together all these results, the aim of this study was to develop a YTX assay to detect the toxin in marine samples and eventually replace the need for the bioassay to monitor YTXs.
Materials and methods Chemicals Yessotoxin was purified by Dr. T. Yasumoto, and part of it was used to obtain a YTX extract in methanol from a matrix of dinoflagellate cultures. Ant-cAMP sodium was from Calbiochem. Phosphodiesterase, 30 ; 50 cyclic nucleotide specific from bovine brain, and all other chemicals were from Sigma, Madrid, Spain. The composition of saline solution used was (mM) Tris–HCl 10 mM, CaCl2 1 mM, and the final pH was adjusted to 7.4.
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
Measurement of phosphodiesterase activity PDE activity was checked by using an analog of cAMP, adenosine 30 ; 50 cyclic monophosphate, 20 -Oanthraniloyl-, sodium salt [16]. Ant-cAMP presents a maximum of fluorescence for 330-nm excitation and 445-nm emission; its fluorescence decreases when phosphodiesterases are activated. The PDE activity was measured in a 96-well plate with a final volume of 100 ll. Each condition was performed in triplicate. The fluorescence was read in a fluorescence plate reader (Bio-Tek FL600) by using 350 nm and 445 nm as excitation and emission wavelengths, respectively. A calibration curve was prepared (r > 0:999) with substrate standards covering the range 0–10 lM. The fluorescence values were converted to lmol (or pmol) substrate/ml by using this calibration curve. The enzyme activity is expressed as a rate of substrate hydrolysis lmol-pmol/ml/min. Shellfish extract processing Uncontaminated mussels (Mytilus galloprovincialis Lmk) were opened and the meat was collected and homogenized; 100 g of homogenate was extracted with 300 ml acetone. The meat extract was filtrated by vacuum and extracted again with 300 ml of methanol. Acetone and methanol extract were mixed and vacuum dried. The dried extract was diluted in 100 ml dichloromethane and extracted with 50 ml water. The organic phase with YTX was vacuum dried and dissolved in 4 ml methanol at 90%. This protocol was made following the CEE decision [18]. This is the uncontaminated sample C. When necessary, HPLC determinations of YTX were done according to [19]. Data analysis
95
PDEs (final volume 100 ll) are added and then the fluorescence versus time is recorded. As mentioned under Materials and methods, in parallel we prepared a calibration curve of substrate in the range 0–10 lM where PDEs were not added. With values obtained in this curve the substrate fluorescence is transformed in concentration. As Fig. 1 shows, the amount of AntcAMP decreases from 8 to 6 lM under control conditions without PDE, and in the presence of 0.05 U/ml PDEs the decrease goes from 8 to 1.5 lM. From these data, taking the concentration increase in time intervals, the concentration fall can be shown as rate of hydrolysis, which is lmol/ml substrate hydrolyzed per minute. This is the parameter that will be shown in the following graphics. The effect of PDEs is dose dependent. As Fig. 2 shows, the rate of hydrolysis increases (taken from the slope of the curves) when the amount of PDE is increased. From this experiment the amount of 0.03 U/ml PDEs was chosen as the amount best suited to hydrolyze the dye Ant-cAMP. Next we determined the best time to study the hydrolysis of Ant-cAMP, which is the time where the fluorescence fall is higher. As Fig. 3 shows, in the presence of 0.03 U/ml PDEs, after 5 min the rate of hydrolysis is higher than the rate at 10 or 15 min. However, the variability is lower after 10 or 15 min than after 5 min; therefore the best time frame to check the variation of hydrolysis rate is after 10 or 15 min and, we chose the 10-min interval. We then studied whether the PDE effect on AntcAMP fluorescence could be modified by PDE modulators. We used 3-isobutyl-1-methylxanthine (IBMX), a known PDE inhibitor, imidazole, and YTX as PDE activators. We added PDE modulators 2 min later than PDEs and then calculated the rate of hydrolysis at 5, 10, and 15 min after drug addition. As Fig. 4 shows, 500 lM
All experiments were carried out at least three times in triplicate. Results were analyzed using the Student t test for paired data or the ANOVA test. A probability level of 0.05 or smaller was used for statistical significance. Data were normalized and results expressed as the mean SE.
Results Looking for new methods to detect YTXs, we developed a technique based on the effect of this toxin on the PDE activity recently described [8]. Antraniloyl cAMP is a known fluorescent substrate for PDEs [16]. Its fluorescence decreases when the substrate is hydrolyzed. Initially we checked the decrease of the dye fluorescence with and without PDEs. In a 96well plate saline solution, 8 lM dye, and 0.05 U/ml
Fig. 1. Variation in Ant-cAMP concentration over time under control conditions (closed circles) and in the presence of 0.05 U/ml PDEs. A representative assay is shown.
96
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
Fig. 2. Variation in the rate of hydrolysis of Ant-cAMP in the presence of different concentrations of PDEs: 0.01, 0.02, 0.03, 0.04, and 0.05 U/ ml of PDEs was added and the variation in Ant-cAMP concentration was recorded and transformed to hydrolysis rate. A representative assay is shown.
Fig. 3. Variations in the rate of hydrolysis of Ant-cAMP after 5, 10, and 15 min. Mean SE of three experiments. (*) Significant differences with respect to the control without PDEs.
IBMX inhibits by 40% the hydrolysis rate induced by 0.03 U/ml PDEs after 5, 10, and 15 min; 5 mM imidazole increases by 26% the hydrolysis rate after 5 and 10 min and by 46% after 15 min; 2 lM YTX increases by 60% the hydrolysis rate after 5 and 10 min and by 30% after 15 min. Stock solutions of these drugs were made in dimethyl sulfoxide, and this solvent does not modify the hydrolysis rate (data not shown). All these experiments were done at 37 °C. Next we check the effect of temperature on the hydrolysis rate. As Fig. 5 shows, the hydrolysis rate induced by 0.03 U/ ml PDEs is higher at 20 °C than at 37 °C. However, the increase induced by YTXs is significantly higher at 37 °C. Fig. 5B shows the increase in hydrolysis rate induced by 2 lM YTX from data of Fig. 5A. As shown, at
Fig. 4. Variations in Ant-cAMP hydrolysis rate in the presence of PDE modulators; 0.03 U/ml PDEs was first added and then 2 lM YTX, 5 mM imidazole, or 500 lM IBMX. The variation in Ant-cAMP concentration was checked after 5, 10, and 15 min. Mean SE of three experiments. (*) Significant differences with respect to the PDEs control.
37 °C the increase induced by YTX is 50%; however, at 20 °C it is only 10%. Next we checked the effect of different amounts of YTX in the hydrolysis rate after 10 min. We use YTX in the range 10 to 0.1 lM. As Fig. 6 shows, the hydrolysis rate decreases with the concentration. The representation of these data follows a lineal regression (r > 0:9). Finally, after these results we elaborated a method to detect YTX in contaminated samples. We checked our method with two YTX extracts disolved in methanol and obtained from spiking YTX to dinoflagellate cultures extracts, whose YTX concentrations were determined by HPLC [19], sample A with 6 lM YTX and sample B with 3 lM YTX. We also checked the hydrolysis rate induced by an extract from uncontaminated mussel, sample C, to study the interferences with our technique of the sample matrix obtained according to the CEE extraction method [18]. In addition we studied the interference of methanol on the hydrolysis rate since following other official methods [20] the extract from contaminated mussels can be dissolved in methanol. The experiment was carried out at 37 °C, final volume 100 ll, and the hydrolysis rate was calculated after 10 min. In one approach to do the experiment, in a 96well plate there are 4 control wells: (I) Ant-cAMP calibration wells, with saline solution and substrate in the range 0–10 lM; (II) control wells, with saline solution, 8 lM substrate, and 0.03 U/ml PDEs; (III) YTX calibration wells, with 8 lM substrate, 0.03 U/ml PDEs, and YTX in the range 0.1–10 lM; and (IV) problem sample wells, with saline solution, 8 lM substrate, 0.03 U/ml PDEs, and samples A, B, C or methanol. We add saline solutions and the corresponding amount of dye in all wells and then read the plate for 2 min. After addition in wells II, III, and IV of 0.03 U/ml PDEs, the fluorescence
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
97
Fig. 5. (A) Effect of temperature in the Ant-cAMP hydrolysis rate induced by PDEs in the presence of YTX; 0.03 U/ml PDEs was first added and then 2 lM YTX. The variation in Ant-cAMP concentration was checked after 5, 10, and 15 min. The experiment was done at 20 and 37 °C. Mean SE of three experiments. (*) Significant differences with respect to the PDEs control at any temperature. (B) % of increase with respect to PDE control (from data in A).
Table 1 Concentration of YTX on different samples obtained by HPLC and the fluorescent method
Sample A Sample B Sample C Methanol
HPLC
Fluorescence plate method
6 lM 3 lM 0 0
5.26 1.1 lM 3.16 0.4 lM —
0
Sample A and Sample B, methanolic extracts; Sample C, uncontaminated extract; Methanol, control of methanol. Mean SE of three experiments.
Fig. 6. Variations in Ant-cAMP hydrolysis rate induced by PDEs in the presence of different YTX concentrations; 0.03 U/ml PDEs was first added and then different concentrations of YTX. The variation in Ant-cAMP concentration was checked after 10 min. Each point is the mean of three determinations. Standard error was typically 5–10%.
ternative extraction method based in methanol, as used for the phosphatase inhibition assay [20], can be safely used, as shown above. Therefore, this is a sensitive method to detect YTX in samples.
Discussion is read again for 2 min. Then, YTX or samples are added to wells III and IV and the fluorescence is read back for 10 min. With the calibration curve obtained with wells I, fluorescence values are converted into Ant-cAMP concentrations, and then the concentration is transformed to rate of hydrolysis. From data of wells III it is possible to obtain a rate of hydrolysis for each YTX concentration and then perform a calibration curve. With the equation of this calibration curve (YTX concentration versus hydrolysis rate) it is feasible to transform the rate of hydrolysis of wells IV (samples A, B, C, and methanol) in YTX concentration. As Table 1 shows, with this method, the YTX concentration values obtained were 5.26 lM in sample A and 3.16 lM in sample B, instead of 6 and 3 lM, respectively, and no interference was detected with methanol as solvent. However the uncontaminated mussels extract obtained following the extraction method of CEE is so dark and fatty that it interferes with the detection technique, although an al-
The mouse bioassay currently represents the only method to detect and monitor YTX contaminations. Nevertheless this method has ethical problems, it can show false-positive and false-negative results, and it has no specificity. An important tool to develop new detection techniques to any marine toxin is to know its mechanism of action. Recently, we have shown that YTX activates cAMP hydrolysis by PDEs activation [8]. Therefore, after this confirmation, the aim of this work was, by using this finding, to develop a method to detect YTX. In the present assay we directly measure the effect of YTX by using the intracellular toxin target and a fluorogenic substrate. By using an assay based on the use of the biochemical target, it is feasible to discriminate between toxins present in a sample. Usually YTX appears together with DSP toxins and it interferes in DSP bioassay since YTX is a very toxic substance after
98
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99
intraperitoneal injection [4]. As far as we know, there is no other marine toxin that shows an effect on PDEs, and therefore with this method it is possible to check the YTX present in a sample without any special extraction to separate other toxins. Usually to separate DSP and YTX it is necessary to perform several organic extractions that collect fatty extracts that also interfere with the mouse bioassay [21]. The linear response rate of hydrolysis from different concentrations of YTX was between 0.1 and 10 lM; therefore this fluorescent method can detect YTX at concentrations over 0.5 lM, while the limit of detection for the mousse bioassay is 2 lM YTX [2]. Therefore the fluorescent method is more sensitive and allows one to detect and measure the toxin concentration in a sample, while the mouse bioassay provides only information on the presence absence of the toxin. The main problem with our technique seems to be the fatty interference of the extract, but this is a problem that also appears in the mouse bioassay [4], and even if some fatty acids are eliminated with organic extractions, in these processes the toxin can be lost [22,23] and the final extract is inevitably dirty. There are other official methods to perform the extraction to detect YTX and DSP toxins. For example, the direct extraction of toxins from 1 g of digestive mussel glands with 4 ml 80% methanol [20] provides clear extracts where YTX can be detected. This is a method used to isolate DSP toxins due to their lipophilic nature [2] which also provides good recovery results with other detection methods [20]. YTXs are accumulated in digestive glands, so extracting only this gland does not induce toxin losses and the fatty part from the shellfish meat is avoided. In the search for new methods to detect YTX, some cytotoxicity assays were developed. It was shown that YTX induces some morphologic changes in the cell surface; however, these changes are not dose dependent [24]. Hepatocytes were also proposed as a cellular line to discriminate YTX and other DSP toxins; however, confusing results discard this assay [2]. Recently, a functional method measuring E-cadherine detection [25] has been reported. This method is very sensitive and specific, but the time and the complicated process required to obtain the results are important drawbacks; also the use of cell lines is a major issue against validating a detection method, due to its variability. There is also an immunoassay to detect YTX [26]. This assay is a quantitative and fast analytical method but the existence of more than 50 analogs with different specificities is an important problem to overcome. Finally the available instrumental methods (liquid or gas chromatography, capillary electrophoresis, mass spectrometry) are specific and sensitive but there is no reference material for a routine detection [2], and these methods are rather expensive and difficult to implement in a routine laboratory.
In summary, this paper reports a new functional, sensitive, easy, and fast assay to detect YTXs in a marine sample.
Acknowledgments This work was funded by Grants MCYT BMC20000441, SAF2003-08765-C03-02, REN2001-2959-C04-03, REN2003-06598-C02-01, INIA CAL01-068, Xunta PGIDT99INN26101 and PGDIDIT03AL26101PR, and FISS REMA-G03-007.
References [1] M. Murata, M. Kumagai, J.S. Lee, T. Yasumoto, Tetrahedron Lett. 28 (1987) 5869–5872. [2] R. Draisci, L. Lucentini, A. Mascioni, in: L.M. Botana (Ed.), Seafood and Freshwater Toxins: Pharmacology, Physiology and Detection, Marcel Dekker, New York, 2000, pp. 289–324. [3] H. Ogino, M. Kumagai, T. Yasumoto, Nat. Toxins 5 (1997) 255– 259. [4] T. Yasumoto, M. Murata, Y. Oshima, G.K. Matsumoto, J. Glardy, in: E.P. Ragelis, (Ed.), ACS Symp. Series. No 262. Seafood toxins. American Chemical Society, Washington, DC, 1984, pp. 207–214. [5] T. Yasumoto, Y. Oshima, M. Yamaguchi, Bull. Jpn. Soc. Sci. Fish 44 (1978) 1249–1255. [6] L.A. De la Rosa, A. Alfonso, N. Vilari~ no, M.R. Vieytes, L.M. Botana, Biochem. Pharmacol. 61 (2001) 827–833. [7] L.A. De la Rosa, A. Alfonso, N. Vilari~ no, M.R. Vieytes, T. Yasumoto, L.M. Botana, Cell. Signal 13 (2001) 711–716. [8] A. Alfonso, L.A. de la Rosa, M.R. Vieytes, T. Yasumoto, L.M. Botana, Biochem. Pharmacol. 65 (2003) 193–208. [9] D.A. Fisher, J.F. Smith, J.S. Pillar, S.H. St Denis, J.B. Cheng, J. Biol. Chem. 273 (1998) 15559–15564. [10] J. Mukai, T. Asai, M. Naka, T. Tanaka, Br. J. Pharmacol. 111 (1994) 389–390. [11] J.M. Hetman, S.H. Soderling, N.A. Glavas, J.A. Beavo, Proc. Natl. Acad. Sci. USA 97 (2000) 472–476. [12] J.A. Beavo, Physiol. Rev. 75 (1995) 725–748. [13] K. Fujishige, J. Kotera, K. Yuasa, K. Omori, Eur. J. Biochem. 267 (2000) 5943–5951. [14] C.H. Cho, D.H. Cho, M.R. Seo, Y.S. Juhnn, Exp. Mol. Med. 32 (2000) 110–114. [15] A.Z. Zhao, J.N. Huan, S. Gupta, R. Pal, A. Sahu, Nat. Neurosci. 5 (2002) 727–728. [16] A. Alfonso, M. Estevez, M.C. Louzao, M.R. Vieytes, L.M. Botana, Cell. Signal 7 (1995) 513–518. [17] R.L. Kincaid, V.C. Manganiello, Methods Enzymol. 159 (1988) 457–470. [18] 2002/225/EC, D., Official J. Eur. Communities L75 (2002) 62–64. [19] T. Yasumoto, A. Takizawa, Biosci. Biotechnol. Biochem. 61 (1997) 1775–1777. [20] J.C. Gonzalez, F. Leira, O.I. Fontal, M.R. Vieytes, F.F. Arevalo, J.M. Vieites, M. Berm udez-Puente, S. Mu~ niz, C. Salgado, T. Yasumoto, L.M. Botana, Anal. Chim. Acta 466 (2002) 233–246. [21] M.L. Fernandez, A. Mıguez, E. Cacho, A. Martınez, J. Diogene, T. Yasumoto, in: E.A. Sar, M.E. Ferrario, M. Reguera, (Eds.), Floraciones algales nocivas en el cono sur americano. UNESCO y Ministerio de Ciencia y Tecnologıa, Madrid, 2002.
A. Alfonso et al. / Analytical Biochemistry 326 (2004) 93–99 [22] R. Draisci, L. Giannetti, L. Lucentini, C. Marchiafava, K.J. James, A.G. Bishop, B.M. Healy, S.S. Kelly, J. Chromatogr. A 798 (1998) 137–145. [23] R. Draisci, E. Ferretti, L. Palleschi, C. Marchiafava, R. Poletti, A. Milandri, A. Ceredi, M. Pompei, Toxicon 37 (1999) 1187– 1193.
99
[24] T. Aune, T. Yasumoto, E. Engeland, J. Toxicol. Environ. Health 34 (1991) 1–9. [25] S. Pierotti, C. Malaguti, A. Milandri, R. Poletti, P. Rossini, Anal. Biochem. 312 (2003) 208–216. [26] I. Garthwaite, K. Ross, C. Miles, L. Briggs, N. Towers, T. Borrell, P. Busby, J. AOAC Int. 84 (2001) 1643–1648.