A Fluorescent Microplate Assay for Microcystin-LR

A Fluorescent Microplate Assay for Microcystin-LR

O. I. Fontal,* M. R. Vieytes,† J. M. V. Baptista de Sousa,‡ M. C. Louzao,* and L. M. Botana*,1

*Departamento de Farmacologı́a, †Departamento de Fisiologı́a, Facultad de Veterinaria, Universidad de Santiago deCompostela, 27002 Lugo, Spain; and ‡ANFACO-CECOPESCA, Campus Universitario, Lagoas (Marcosende), Vigo, Spain

Received October 1, 1998

A fluorescent enzyme inhibition assay for microcys-tin-LR was developed using a new fluorescent sub-strate of protein phosphatases 1 (PP1) and 2A (PP2A),6,8-difluoro-4-methylumbelliferyl phosphate. The PP1and PP2A inhibition assay for microcystin-LR wasperformed in a microtiter plate and the fluorescenceyielded by the enzymatic hydrolysis of the substratewas quantified in a fluorescence plate reader. The con-centration of microcystin-LR causing 50% inhibitionof PP1 and PP2A activity (IC50) was 0.01 nM for PP1and 0.08 nM for PP2A. The measurable range of micro-cystin-LR was 800 to 0.08 pg/well for both enzymes.The described assay is fast and very sensitive for thedetection of microcystin-LR. Furthermore, this assaycan be successfully applied to the study of toxins thatinhibit PP1 or PP2A. © 1999 Academic Press

Key Words: microcystin; okadaic acid; protein phos-phatase; fluorescent assay; microplate.

Microcystins are a family of cyclic polypeptides pro-duced by different species of cyanobacteria (blue–greenalgae), which can form blooms in lakes and water reser-voirs (1). Their basic structure is a cyclic heptapeptideand their variations give rise to more than 50 types ofmicrocystins known today (1). The most extensively stud-ied form is microcystin-LR that contains L-leucine andL-arginine in the two main variant positions.

Microcystins and related polypeptides are potenthepatotoxins in fish, birds, and mammals (2). The con-sequence of an acute poisoning by these compounds isa rapid desorganization of the hepatic architecture (2,3), leading to massive intrahepatic hemorrhage, oftenfollowed by death of the animals by hypovolemic shockor hepatic insufficiency (4). Matsushima et al. (5) haveseen that microcystins penetrate with difficulty intothe epithelial cells, which reflects tissue specificity, and

their target cell is the hepatocyte. This cellular speci-ficity and organotropism of microcystins is due to theselective transport system, the multispecific bile acidtransport system, present only in hepatocytes (3, 6).

Microcystins are potent inhibitors of protein phos-phatases 1 (PP1)2 and 2A (PP2A) (7–9), which areregulatory enzymes present in the cytosol of the mam-malian cells. This action may explain the effects ofmicrocystins as cancer promoters (10, 11) and the pro-motion of primary liver cancer in humans exposed tolong-term low doses of these cyclic peptide toxinsthrough drinking water (12–14) as well as the cytoskel-etal disruption and formation of plasma membraneblebs (blebbing) in hepatocytes (3).

Since microcystins are potent hepatotoxins for hu-mans and animals, the development of sensitive andreliable detection methods becomes of great impor-tance. The efforts have been aimed at developing moresensitive screening methods to replace the nonspecificmouse bioassay, traditionally used for the identifica-tion of toxic strains of Microcystis.

Thus far, physicochemical techniques have been usedas a sensitive method of analysis (15–17), but this ap-proach relies on the availability of toxin standards forcomparison and is therefore only applicable to knowntoxins. It is also a relatively slow technique and requiresexpensive equipment and appropriate training.

The development of biological methods was first fo-cused on enzyme-linked immunosorbent assays(ELISA). Using polyclonal antibodies, the sensitivitylimit was first ng/mL (18), and later on it was 95 pg/mL(19). The use of a monoclonal antibodies that selec-tively recognized microcystin did enhance this limit to10 pg without complicated cleanup steps and concen-

1 To whom correspondence should be addressed.

2 Abbreviations used: PP1, protein phosphatase 1; PP2A, proteinphosphatase 2A; DiFMUP, 6,8-difluoro-4-methylumbelliferyl phos-phate; 4-MUP, 4-methylumbelliferyl phosphate; DiFMU, 6,8-dif-luoro-4-methylumbelliferone; IC50, 50% inhibitory concentration;ELISA, enzyme-linked immunosorbent assay; pNPP, para-nitrophe-nyl phosphate; OA, okadaic acid; BSA, bovine seroalbumin.

0003-2697/99 $30.00 289Copyright © 1999 by Academic PressAll rights of reproduction in any form reserved.

Analytical Biochemistry 269, 289–296 (1999)Article ID abio.1999.3099, available online at http://www.idealibrary.com on

tration procedures (20–22). Although the sensitivity ofthe ELISAs has been progressively increased, theseprocedures give no information on the viable biologicalactivity of the toxins.

The fact that microcystins inhibit PP1 and PP2Aenzymes (7–9) raised the possibility of a functionalassay for microcystin-LR and associated toxins basedon a protein phosphatase inhibition method. The ad-vantages over other techniques should be rapidity,economy, and sensitivity. Its ability to identify anytoxin that inhibits these enzymes excludes the need fortoxin identification and will therefore allow the detec-tion of previously unidentified toxins. A functional as-say combined with two physicochemical techniques(capillary electrophoresis coupled with a liquid chro-matography-linked protein phosphatase bioassay) wasdeveloped and detected both diarrhetic shellfish toxinsand hepatotoxic microcystins (23). The protein phos-phatase activity (PP1 and PP2A) was routinely mea-sured by the dephosphorylation of glycogen phosphor-ylase (24) with 32P, but the trend to nonradiactiveassays leads to develop a colorimetric protein phospha-tase inhibition assay for microcystins (25, 26) adaptedfrom (27), which found that the p-nitrophenyl phos-phate (pNPP) is a good substrate for PP1 and PP2A.

Based on previous work (28), in the present work westudied this substrate for the detection of microcystin-LR, the main toxin responsible of the hepatotoxicity inthe blooms of cyanobacteria in the Northern hemi-sphere. We found that the fluorogenic substrates ofprotein phosphatases provide a very good sensitivity toPP2A inhibition assays, even in 96-well microplates(28). The fluorogenic compounds are converted tohighly fluorescent products by the enzyme, the assayrequires only minimal processing, and the results canbe read on a standard multiwell scanning fluorimeter.We used both protein phosphatases, PP1 and PP2A,since the data on the literature are not uniform withrespect to which of the two is more sensitive to micro-cystin-LR (5, 7–9, 25, 29–31). In this assay a newprotein phosphatase fluorogenic substrate was tested,namely 6,8-difluoro-4-methylumbelliferyl phosphate(DiFMUP). It is a derivative of the 4-methylumbel-liferyl phosphate (4-MUP) that presents better spectralproperties and its hydrolysis product, 6,8-difluoro-4-methylumbelliferone (DiFMU), exhibits a lower pKa

than the hydrolysis product of 4-MUP, which is anadvantage when it works at pH 7.0, since the fluores-cence quantum yield is higher.

MATERIALS AND METHODS

Reagents

Microcystin-LR, from Biomol Res. Lab., Inc., wasdissolved in dimethyl sulfoxide to a final concentrationof 500 !g/mL and was maintained as a frozen stock at

!20°C. The two protein phosphatases tested, PP1 andPP2A, were provided by UBI (Upstate Biothecnology,Inc., New York) diluted in 50% glycerol, 20 mM 4-mor-pholinepropanesulfonic acid (pH 7.5), 60 mM 2-mer-captoethanol, 0.1 M NaCl, and 1 mg/mL serum albu-min. PP1 from rabbit skeletal muscle is a mixture ofisoforms, based on peptide sequencing. PP2A was iso-lated from human red blood cells as the heterodimer of60 kDa (A) and 36 kDa (C) subunits.

The fluorogenic compound 6,8-difluoro-4-methylum-belliferyl phosphate, ammonium salt (DiFMUP), andits fluorescent standard DiFMU were purchased fromMolecular Probes (Leiden, The Netherlands). TheDiFMUP stock solution was prepared in 50 mM HCl–Tris (pH 7.0) to a final concentration of 10 mM and wasstored at !20°C in aliquots of 50 !L. HCl–Tris, CaCl2,NiCl2, and bovine serum albumin (BSA) were providedby Sigma (Madrid, Spain).

Protein Phosphatase Inhibition Assay

Assays performed in a 96-well microplate. The pro-tein phosphatases 1 or 2A were diluted to 0.025 units/well with the phosphatase assay buffer (50 mM HCl–Tris, 0.1 mM CaCl2, pH 7.0). The microcystin-LR wasalso dissolved with the phosphatase assay buffer to theappropriate concentrations. The DiFMUP was dilutedin 50 mM HCl–Tris (pH 7.0) to a final concentration of100 !M for the assay with PP1 and to 50 !M when theassay is with PP2A. The final volume in each well was200 !L, distributed as follows: 5 !L of NiCl2 (40 mM),5 !L of BSA (5 mg/mL), 25 !L of protein phosphatase1 or 2A, 10 !L of microcystin-LR, and 35 !L of thephosphatase assay buffer to complete the 80-!L vol-ume. The enzyme substrate was added in a volume of120 !L. All assays were carried out at 37°C. The pro-tein phosphatase was incubated with the reagents 10min before the addition of microcystin-LR, and another10 min is necessary for the incubation of the enzymewith the toxin before the addition of the substrate. Thefluorescence was measured in the fluorescence platereader at 355–460 nm, at time intervals from 5 to 45min. The reading time for all assays was 30 min.

Assays performed in a quartz cuvette (1-cm lightpath). The protein phosphatase 2A was diluted to 0.1unit with the phosphatase assay buffer (50 mM HCl–Tris, 0.1 mM CaCl2, pH 7.0). The fluorogenic substrate(DiFMUP) and its reference standard (DiFMU) werediluted to the appropriate concentrations in 50 mMHCl–Tris (pH 7.0). The assay in a 1-mL reaction mix-ture contained 40 mM NiCl2 (25 !L), 5 mg/mL BSA (25!L), 0.1 units of PP2A (20 !L), phosphatase assaybuffer (330 !L), and the fluorogenic substrate (600 !L).The enzymatic reaction was initiated by adding thesubstrate to a cuvette where the other componentswere previously incubated for 2 min. Fluorescence was

290 FONTAL ET AL.

monitored for 5 min using a cuvette fluorescencereader (Perkin–Elmer Luminescence SpectrometerLS50B) at wavelengths of 354–452 nm. The slit widthswere set at 3 nm. The temperature was always 37°C.Specific activity was defined as nmol product/min/mgprotein.

Characterization of DiFMUP as PP1 and PP2ASubstrate

In the first step the ability of PP1 and PP2A todephosphorylate DiFMUP to DiFMU was determined.Both PP1 and PPA were tested in amounts ranging

FIG. 1. Arbitrary fluorescence units versus amount of protein phosphatase 1 (PP1) (A) and protein phosphatase 2A (PP2A) (B). The finalconcentration of DiFMUP was 50 !M in the assay with PP1 (A) and 100 !M in the assay with PP2A (B). The inset in (B) shows that above0.025 units of PP2A the relationship changes to nonlinear. The least-squares fit equation for the data is (fluorescence units) (y) " !12.448 #2561.3 $ PP1 (x), r2 " 0.99 (A) and (y) " !158,43 # 120441 $ PP2A (x), r2 " 0.99 (B).

291FLUORESCENT ASSAY FOR MICROCYSTIN-LR

from 0.0025 to 0.1 units per well. The fluorescenceenhancement derived from enzymatic hydrolysis wasmeasured at time intervals ranging from 5 to 45 min.The fluorescence was measured in a fluorescence platereader (Fluoroskan II; Labsystems, Finland) at 355–460 nm, since the product of the hydrolysis of DiFMUPis fluorescent at an excitation wavelength of 385 nmand an emission wavelength of 485 nm. We also deter-mined the substrate saturation concentration withDiFMUP concentrations ranging from 10 to 500 !M forPP1 and 5 to 100 !M for PP2A.A kinetic study for PP2A was performed by mixing

0.1 units of enzyme and different DiFMUP concen-trations (0.5 to 5 !M). Initial velocity was measuredby monitoring fluorescence versus time in the first 5min of the reaction. We assume that substrate con-centration during this time remains constant. Thefluorescence values were correlated with the amountof product using the standard curve. Fluorescencevalues were corrected to detect nonenzymatic hydro-lysis in the absence of PP2A. Km and Vmax values wereobtained from double reciprocal plots of the hydroly-sis curves.

IC50 Determination for Microcystin-LR for PP1 andPP2A

PP1 and PP2A were dissolved in the phosphataseassay buffer to give concentrations of 0.050, 0.025,and 0.0125 units per well. An estimation of the mi-crocystin-LR IC50 for these three different concentra-tions of protein phosphatase was obtained by usingtoxin concentrations ranging from 4 to 0.0004 nM.

RESULTS

DiFMUP as Substrate for PP1 and PP2A

This study was initiated by establishing whetherthe fluorogenic substrate DiFMUP would yield agood sensitivity to quantify the activity of PP1 andPP2A in an inhibition assay. As expected, the fluo-rescence readings increased with higher PP1 andPP2A concentrations. The increase in the fluores-cence showed a linear relationship between arbitraryfluorescence units and protein phosphatase concen-tration (Fig. 1). We selected 0.025 units of PP1 andPP2A for the assay. In both cases this enzyme con-centration was in the fluorescence linear range. Forthe same number of enzyme units the fluorescencelevels obtained were 24 times higher with PP2A thanwith PP1.

Since the relationship between fluorescence andtime is linear, we measured the fluorescence en-hancement at time intervals ranging from 5 to 45min (Fig. 2). For the microcystin assay we propose afixed-time mode, 30 min. At this time and for both

protein phosphatases the enhancement of fluores-cence was linear.

We studied next the saturation concentration ofthe substrate. As shown in the substrate saturationcurves (Fig. 3) this concentration is different for eachprotein phosphatase (0.025 units), with values of 100!M DiFMUP for PP1 (A) and 50 !M for PP2A (B).

Figure 4 shows the signal-to-noise ratio, lookingfor a suitable index to set the sensitivity of the assay.In order to obtain better results, the difference be-tween the control and the blank, in arbitrary fluo-rescence units, must to be as high as possible; in thisway the range of sensitivity will be wider and thedetected differences in the inhibition assays will bealso higher. Clearly, this ratio is better for PP2Athan for PP1.

The kinetic study for PP2A and DiFMUP was car-ried out in a cuvette fluorimeter with more sensitiv-

FIG. 2. Arbitrary fluorescence units versus time in minutes. Thefinal concentration of DiFMUP was 50 !M in the assay with PP1 (A)and 100 !M in the assay with PP2A (B). The PP1 and PP2A concen-tration was 0.025 units/well. The least-squares fit equation for thedata is (fluorescence units) (y) " 0.22 # 7.46 $ PP1 (x), r2 " 0.95 (A)and (y) " !30.46 # 584.2 $ PP2A (x), r2 " 0.99 (B).

292 FONTAL ET AL.

ity than the fluorescence plate reader. A lower rangeof DiFMUP concentrations (0.5 to 5 !M) was used.The cuvette fluorimeter allows a continous record offluorescence at the constant temperature of 37°C.

Figure 5 shows the double reciprocal plot of thehydrolysis curve, where the Km and Vmax values wereobtained as 9.3 !M and 3200 nmol/min/mg protein,respectively.

FIG. 3. Saturation curve of DiFMUP for PP1 (A) and PP2A (B). In two curves the concentration of protein phosphatase was 0.025 units/well.The DiFMUP saturation concentration for PP1 (100 !M) and for PP2A (50 !M) is in the linear portion of the curve. The inset shows the linearportion of the saturation curves of substrate. The least-squares fit equation for the data is (fluorescence units) (y) " 12.655 # 1.4535 $DiFMUP (x), r2 " 0.98 (A) and (y) " 103.76 # 33.613 $ DiFMUP (x), r2 " 0.99 (B). Means % SEM of three experiments.

293FLUORESCENT ASSAY FOR MICROCYSTIN-LR

IC50 of Microcystin-LR for PP1 and PP2A

The hydrolysis of the fluorescent substrate DiFMUPby the protein phosphatases 1 and 2A was potentlyinhibited by microcystin-LR, the 50% inhibition (IC50)taking place at 0.01 and 0.08 nM for PP1 for PP2A,respectively (Fig. 6). An accurate IC50 can only be es-timated by using the enzyme concentration belowwhich any change in IC50 value is observed with sub-sequent dilution of enzyme (8). The IC50 of PP1 formicrocystin-LR is two times higher using 0.05 unitsthan the IC50 obtained with 0.025 units of enzyme (theresults obtained with 0.0125 units are not differentthan those obtained with 0.025 units). When we usedPP2A, the IC50 value obtained with 0.0125 units ofenzyme was 0.08 nM; for 0.025 units this value wasonly slightly higher, 0.1 nM. From these results 0.025units of PP1 or PP2A was employed to create the stan-dard inhibition curve with different concentrations ofmicrocystin-LR.

Fluorescent Inhibition Protein Phosphatase Assay forMicrocystin-LR

Once the working conditions were established, wecarried out the standard inhibition curves for each oneof the protein phosphatases (0.025 units/well) witheight concentrations of microcystin-LR (4 to 0.0004nM) (Fig. 7). We used the linear portion of these stan-dard curves, which will allow an accurate determina-tion of the microcystin-LR concentration present in theassay. The linear response of PP1 inhibition with dif-ferent concentrations of microcystin-LR was estab-lished between 0.2 and 0.0004 nM. In the case of PP2Athis linear response ranged between 1 and 0.012 nM(Fig. 8).

DISCUSSION

This paper is the result of the investigations aboutthe application of fluorogenic substrates to functional

FIG. 4. Graphic representation of the signal-to-noise ratio for PP1and PP2A. Means % SEM of 10 experiments.

FIG. 5. The double reciprocal plot the of initial velocity versusDiFMUP concentration was linear with a regression coefficient of0.99.

FIG. 6. Inhibition of PP1 (A) and PP2A (B) activity by microcys-tin-LR in the presence of two concentrations of PP1 (0.05 and 0.025units/well) and three concentrations of PP2A (0.05, 0.025, and 0.0125units/well). The toxin amount ranged between 4 and 0.0004 nM. Thesubstrate concentration is 100 !M DiFMUP for PP1 (A) and 50 !Mfor PP2A (B). Means % SEM of three experiments.

294 FONTAL ET AL.

assays for the detection of serine/threonine proteinphosphatase inhibitors, since the sensitivity of fluori-metric methods is superior to conventional methodswith colorimetric substrates (32). Previously, a fluores-cent PP2A inhibition assay for the detection of okadaicacid and diarrheic analogs was described (28) and itsattractive possibilities were demonstrated. Now, afunctional assay is proposed for the detection of micro-cystin-LR, a potent hepatotoxin produced by cyanobac-teria that causes its toxic effect through the inhibitionof serine/threonine protein phosphatases 1 and 2A. Inthis case, we used a new fluorogenic substrate, namely,DiFMUP, which is a derivative of 4-MUP that showsmore fluorescence at pH 7.0 than the original molecule.The obtained results were better than those for othersubstrates tested before, mainly for PP2A, since thesignal-to-noise ratio was very high. Therefore the abil-ity to detect concentrations of inhibitor will be widerand more accurate.

The IC50 of microcystin-LR for PP1 and for PP2A is0.01 and 0.08 nM, respectively. Therefore PP1 fromrabbit skeletal muscle is more sensitive to this toxinthan PP2A from human red blood cells. The linearrange of the inhibition standard curve is moved tolower concentrations when the enzyme is PP1 (0.2 to0.00004 nM). In the case of the PP2A, the linear por-tion of the curve is shifted to the right, to slightlyhigher concentrations (1 to 0.012 nM). Although PP1 ismore sensitive than PP2A to the inhibitory effect ofmicrocystin-LR, the linearity range is broader forPP2A, which allows the detection of wider range oftoxin concentrations. In addition to this advantage, theuse of PP2A gives more accurate determinations since

the degree of dephosphorylation of DiFMUP is fivefoldfor PP2A. Even though both enzymes can be used todetect microcystin-LR, we are inclined to use PP2A dueto its wider range of sensitivity.

Overall, following the trend of fluorescent functionalassays, initiated for the determination of OA in shell-fish (28), we developed a sensitive method to detectvery low amounts of microcystin in solution, down to0.08 pg of microcystin-LR in the well, which representsa significant increase in the toxin detection limit infunctional assays (25). Also, we propose this fluores-cent assay as a useful tool in the study of different PP1and PP2A inhibitors, since its good resolution rendersa very sensitive method to detect low or very closeconcentrations.

ACKNOWLEDGMENT

This work was funded with a grant from CICYT (ALI95-1012-C05-03).

REFERENCES

1. Carmichael, W. W. (1988) in Marine Toxins and Venoms, Hand-book of Natural Toxins (Tu, A. T., Ed.), Vol. 3, pp. 121–147,Dekker, New York.

2. Falconer, I. R., Jackson, A. R. B., Langley, J., and Runnegar,M. T. C. (1981) Aust. J. Biol. Sci. 34, 179–187.

3. Eriksson, J. E., Toivola, D., Meriluoto, J. A. O., Haraki, H., Han,Y., and Hartshorne, D. (1990) Biochem. Biophys. Res. Commun.173, 1347–1353.

4. Carmichael, W. W. (1992) United States Environmental Protec-tion Agency/600/R-92/079, 32–33.

5. Matsushima, R., Yoshizawa, S., Watanabe, M. F., Harada, K.,Furusawa, M., Carmichael, W. W., and Fujiki, H. (1990) Bio-chem. Biophys. Res. Commun. 171, 867–874.

FIG. 7. Inhibitory effect of microcystin-LR on the activity of proteinphosphatase 1 and 2A (0.025 units/well). The activity of PP1 andPP2A was assayed in the presence of different concentrations ofmicrocystin-LR (4 to 0.0004 nM) using the fluorogenic substrateDiFMUP. PP1 and PP2A activity was expressed as a percentage ofthe activity in the absence of inhibitor. Means % SEM of five exper-iments.

FIG. 8. Linear response of PP1 and PP2A inhibition from differentconcentrations of microcystin-LR. The least-squares fit equation forthe data is (fluorescence units) (y) " 80.808 ! 226,87 (x), r2 " 0.99for PP1 and (y) " 75.634 ! 59.01 (x), r2 " 0.97 for PP2A. Means %SEM of five experiments.

295FLUORESCENT ASSAY FOR MICROCYSTIN-LR

6. Hooser, S. B., Kuhlenschmidt, M. S., Dahlem, A. M., Beasly,V. R., Carmichael, W. W., and Haschek, W. M. (1991) Toxicon 29,589–601.

7. Yoshizawa, S., Matsushima, R., Watanabe, M. F., Harada, K.,Ichihara, A., Carmichael, W. W., and Fujiki, H. (1990) J. CancerRes. Clin. Oncol. 116, 609–614.

8. Honkanen, R. E., Zwiller, J., Moore, R. E., Daily, S. L., Khatra,B. S., Dukelow, M., and Boynton, A. L. (1990) J. Biol. Chem. 265,19401–19404.

9. MacKintosh, C., Beattie, K. A., Klumpp, S., Cohen, P., and Codd,G. A. (1990) FEBS Lett. 264, 187–192.

10. Nishiwaki-Matsushima, R., Ohta, S., Nishiwaki, S., Suganuma,M., Kohyama, K., Ishiwaka, T., Carmichael, W. W., and Fujiki,H. (1992) J. Cancer Res. Clin. Oncol. 118, 420–424.

11. Falconer, I. R. (1991) Environ. Toxic Water Qual. 6, 177–184.12. Carmichael, W. W. (1994) Sci. Am. 1, 64–72.13. Yu, S.-Z. (1989) in Primary Liver Cancer (Tang, Z.-Y., Wu, M.-C.,

and Xia, S.-S., Eds.), pp. 1–37, China Academic Publishers/Springer, Berlin.

14. Ueno, Y., Nagata, S., Tsutsumi, T., Hasegawa, A., Watanabe,M. F., Park, H. D., Chen, G. C., Chen, G., and Yu, S. Z. (1996)Carcinogenesis 17, 1317–1321.

15. Watanabe, M. F., Oishi, S., Harda, K., Matsuura, K., Kawai, H.,and Suzuki, M. (1988) Toxicon 26, 1017–1025.

16. Meriluoto, J. A., Eriksson, J. E., Harada, K., Dahlem, A. M.,Sivonen, K., and Carmichael, W. W. (1990) J. Chromatogr. 509,390–395.

17. Kondo, F., Ikai, Y., Oka, H., Watanabe, M. F., Watanabe, M.,Harada, K., and Suzuki, M. (1992) Toxicon 30, 227–237.

18. Chu, F. S., Huang, X., and Wei, R. D. (1990) J. Assoc. Off. Anal.Chem. 73, 451–456.

19. McDermott, C. M., Feola, R., and Plude, J. (1995) Toxicon 33,1433–1442.

20. Nagata, S., Tsutsumi, T., Hasegawa, A., Watanabe, M. F., andUeno, Y. (1995) Jpn. J. Toxicol. Environ. Health 41, 10.

21. Nagata, S., Tsutsumi, T., Hasegawa, A., Yoshida, F., and Ueno,Y. (1997) J. Assoc. Off. Anal. Chem. Int. 80, 408–417.

22. Tsutsumi, T., Nagata, S., Yoshida, F., and Ueno, Y. (1998) Toxi-con 36-2, 235–245.

23. Boland, M. P., Smillie, M. A., Chen, D. Z. X., and Holmes, C. F. B.(1993) Toxicon 31, 1393–1406.

24. Cohen, P. (1989) Annu. Rev. Biochem. 58, 453–508.

25. An, J., and Carmichael, W. W. (1994) Toxicon 32, 1495–1508.

26. Ward, C. J., Beattie, K. A., Lee, E. Y., and Codd, G. A. (1997)FEMS Microbiol. Lett. 153, 465–473.

27. Takai, A. A. M., G. (1991) Biochem. J. 275, 233–239.

28. Vieytes, M. R., Fontal, O. I., Leira, F., Baptista de Sousa,J. M. V., and Botana, L. M. (1997) Anal. Biochem. 248, 258–264.

29. Ichinose, M., Endo, S., Critz, S. D., Shenolikar, S., and Byrne,J. H. (1990) Brain Res. 533, 137–140.

30. Suganuma, M., Fujiki, H., Okabe, S., Nishiwaki, S., Brautigan,D., Ingebritsen, T. S., and Rosner, M. R. (1992) Toxicon 30,873–878.

31. Lin, J. R., and Chu, F. S. (1994) Toxicon 32, 605–613.

32. Honkanen, R. E., Codispoti, B. A., Tse, K., and Boynton, A. L.(1994) Toxicon 32, 339–350.

33. Guilbault, G. G. (1990) in Practical Fluorescence (Guilbault,G. G., Ed.), Dekker, New York.

296 FONTAL ET AL.