Liquid chromatographic determination of microcystins in water samples following pre-column excimer fluorescence derivatization with 4-(1-pyrene)butanoic acid hydrazide

Analytica Chimica Acta 755 (2012) 93–99

Contents lists available at SciVerse ScienceDirect

Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca

Liquid chromatographic determination of microcystins in water samples following pre-column excimer fluorescence derivatization with 4-(1-pyrene)butanoic acid hydrazide Tadashi Hayama a , Kenji Katoh a , Takayoshi Aoki a , Miki Itoyama a , Kenichiro Todoroki b , Hideyuki Yoshida a , Masatoshi Yamaguchi a , Hitoshi Nohta a,∗ a

Faculty of Pharmaceutical Sciences, Fukuoka University, 8-19-1 Nanakuma, Johnan, Fukuoka 814-0180, Japan Laboratory of Analytical and Bio-Analytical Chemistry, Graduate School of Pharmaceutical Sciences, University of Shizuoka, 52-1 Yada, Suruga-ku, Shizuoka 422-8526, Japan b

h i g h l i g h t s

g r a p h i c a l

a b s t r a c t

 We have developed a novel method for LC determination of microcystins (MCs).  MCs were successfully derivatized with 4-(1-pyrene)butanoic acid hydrazide.  The derivatives could be specifically detected at excimer fluorescence wavelength.  The SPE method was used as pretreatment of MCs in environmental water samples.  This method can be used for evaluating MC contamination in water samples.

a r t i c l e

i n f o

Article history: Received 18 June 2012 Received in revised form 27 September 2012 Accepted 4 October 2012 Available online 12 October 2012 Keywords: Microcystins Excimer fluorescence derivatization Liquid chromatography Environmental water sample

a b s t r a c t A method to measure the concentrations of microcystins (MCs) in water samples has been developed by incorporating pre-column fluorescence derivatization and liquid chromatography (LC). A solid-phase extraction for pretreatment was used to extract the MCs in water samples. The MCs were derivatized with excimer-forming 4-(1-pyrene)butanoic acid hydrazide (PBH). The MCs could then be detected by fluorescence after separation with a pentafluorophenyl (PFP)-modified superficially porous (core shell) particle LC column. The derivatization reactions of MCs with PBH proceeded easily in the presence of 4,6-dimethoxy-1,3,5-triazin-2-yl-4-methylmorpholinium (DMT-MM) as a condensation reagent, and the resulting derivatives could be easily separated on the PFP column. The derivatives were selectively detected at excimer fluorescence wavelengths (440–540 nm). The instrument detection limit and the instrument quantification limit of the MCs standards were 0.4–1.2 ␮g L−1 and 1.4–3.9 ␮g L−1 , respectively. The method was validated at 0.1 and 1.0 ␮g L−1 levels in tap and pond water samples, and the recovery of MCs was between 67 and 101% with a relative standard deviation of 11%. The proposed method can be used to quantify trace amounts of MCs in water samples. © 2012 Elsevier B.V. All rights reserved.

1. Introduction

∗ Corresponding author. Tel.: +81 92 871 6631; fax: +81 92 863 0389. E-mail address: [email protected] (H. Nohta). 0003-2670/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.aca.2012.10.009

Cyanotoxins such as microcystins (MCs) are produced by cyanobacteria, which are known to reside widely in eutrophic environmental surface water. More than 80 MCs have been identified,

94

T. Hayama et al. / Analytica Chimica Acta 755 (2012) 93–99

and they consist of a cyclic heptapeptide with 3-amino-9methoxy-2,6,8-trimethyl-10-phenyl-4,6-decadienoic acid (Adda) as a common motif [1]. Because of the high toxicities of MCs [2–4], the World Health Organization (WHO) has established an allowable of MC level of 1 ␮g L−1 in drinking water [5]. Hence, in order to evaluate and control the quality of drinking water, it is necessary to determine MC content in environmental water sources, such as surface water and ground water used for drinking. Several analytical methods have been developed to quantify MCs such as enzymelinked immunosorbent assays (ELISA) [6], protein phosphatase inhibition assays (PPIA) [7], capillary electrophoresis (CE) [8], and liquid chromatography (LC) with UV [9,10] or mass spectrometry (MS) detection [11,12]. ELISA and PPIA have been used for fast screening analysis of MCs; they can determine total MC content but cannot identify individual MCs. By contrast, both identification and quantification of MCs can be performed by an LC method, although there are limited commercial standards. Unfortunately, UV detection is not sensitive enough to analyze the trace MCs, and therefore not only large sample volume but also tedious and time-consuming pretreatments are needed. On the other hand, the LC–MS methods have provided reasonable results when compared to other methods because of their high sensitivity and specificity. However, more affordable and convenient detection system is still required in regulatory environmental analysis. MCs contain the reactive functional groups, such as two carboxylic acids, in their structure, and they have potential for analysis by fluorescence detection following chemical derivatization with appropriate reagent. Such method would assist as a complement method for determining MCs by LC–MS and should also be helpful in regulatory monitoring in small laboratories. Previously, we developed the selective detection-oriented derivatization methods to detect poly-functional compounds such as polyamines [13–17], polycarboxylic acids [18–20], and polyphenols [21,22]. These methods worked by derivatization with pyrene followed by reversed-phase LC-separation and fluorescence detection. The polypyrene derivatives could be detected by intramolecular excimer fluorescence (440–540 nm), whereas monopyrene derivatives and reagent blanks, which emitted

COOH O

2. Materials and methods 2.1. Reagents and materials Standard solutions (10 mg L−1 each in 20% methanol) of MC-YR, MC-LR, and MC-RR were obtained from Kanto Chemical (Tokyo, Japan). Highly purified PBH and DMT-MM chloride were purchased from Sigma–Aldrich (St. Louis, MO, USA) and Wako Pure Chemicals (Osaka, Japan), respectively. All organic solvents were of LC grade; they were purchased from Wako Pure Chemicals and used as received. Ultrapure water, further purified using a Milli-Q

CH3 N CH2

HN OCH3

monomer fluorescence (360–420 nm), did not affect the analysis of the polypyrene derivatives. This derivatization strategy selectively detected different polyfunctional compounds, even in complex samples containing monofunctional compounds, in spite of only simple derivatization procedure without extensive samples pretreatment. In this study, excimer fluorescence derivatization combined with LC methods was applied to develop a simpler, universally feasible MC detection method. 4-(1-Pyrene)butanoic acid hydrazide (PBH) was attached to two carboxylic acids on the MC peptides in the presence of 4,6-dimethoxy-1,3,5-triazin-2yl-4-methylmorpholinium (DMT-MM) as a condensation reagent (Fig. 1). The derivatization conditions were designed to enhance the sensitivity of the PBH-derivatized MCs. The LC conditions using a pentafluorophenyl (PFP)-modified superficially porous (core shell) particle LC column have been optimized for good peak shape and sufficient separation. After these analytical conditions were optimized using standards of MC-YR, MC-LR, and MC-RR as the representative MCs, the proposed method was applied to the analysis of MCs in tap water and surface (pond) water samples by a combination of solid-phase extraction (SPE) and derivatization with PBH. To the best of our knowledge, the method presented here is the first report of the analysis of MCs by LC-fluorescence detection following excimer fluorescence derivatization in their intact forms.

O

O

O

NH

NH CH3 H N R2 O

H N N H

O

R1 O COOH

CH3 N CH2

HN OCH3

Microcystins

O

O

NH CH3 H N R2

+

O

O

O

O

NH

R1 O

NN O HH

Dipyrene-derivatized microcystins NHNH2

4-(1-Pyrene)butanoic acid hydrazide

Toxin

R1

R2

MC-RR

L-Arginine

L-Arginine

MC-YR

L-Tyrosine

L-Arginine

MC-LR

L-Leucine

L-Arginine

Fig. 1. Intramolecular excimer-forming fluorescence derivatization of MCs with PBH.

T. Hayama et al. / Analytica Chimica Acta 755 (2012) 93–99

Gradient system (Millipore, Billerica, MA, USA), was used for preparing all aqueous solutions. Working standards of the MCs were prepared by diluting the standard solutions with acetonitrile to the required concentrations before use. MCs and organic solvents are toxic if exposed to the eyes, lungs, or skin; therefore, they should be carefully handled in accordance with the latest material safety data sheets. An InertSep RP-1 cartridge (60 mg, 3 mL) purchased from GL Sciences (Tokyo, Japan) was rinsed with 1 mL each of acetonitrile and pure water before use. 2.2. Derivatization To sample solution (50 ␮L) in a 1.5-mL screw cap tube, 10 ␮L of 40 mM PBH in dimethylsulfoxide and 10 ␮L of 200 mM DMT-MM in 95% acetonitrile were added. After the tube was sealed tightly, the mixture was vortexed and then heated at 60 ◦ C for 30 min. Afterwards, the reaction mixture was cooled in an ice-water bath, and 130 ␮L of initial mobile phase was added to the tube. The resulting mixture was filtered and injected into the LC system. 2.3. LC systems and their conditions Chromatography was performed on a Shimadzu (Kyoto, Japan) liquid chromatography system, consisting of two LC-10ADVP pumps, a high-pressure gradient unit, a DGU-12A online degasser, a SIL-10ADVP autosampler, a CTO-10ACVP column oven, and an RF-20AXL fluorescence detector. The column was an Accucore PFP column (50 mm × 3.0 mm I.D., particle size 2.6 ␮m; ThermoFisher Scientific, San Jose, CA, USA). A 10-␮L aliquot of each sample was automatically injected. Solvent A was a mixture of water and acetonitrile (70:30, v/v) containing 0.15% trifluoroacetic acid (TFA) and 10 mM ammonium acetate. Solvent B was a mixture of methanol and acetonitrile (70:30, v/v) containing 0.15% TFA and 10 mM ammonium acetate. The gradient elution conditions were as follows: 0–15 min, linear change from 60% to 65% B; 15–15.01 min, linear change from 65% to 60% B; run time, 20 min. The flow rate and column oven temperature were set at 1.0 mL min−1 and 50 ◦ C, respectively. The fluorescence detector was operated at excitation and emission wavelengths of 345 nm and 475 nm, respectively. 2.4. Fluorescence spectrofluorometer Fluorescence spectra were measured on a Hitachi (Tokyo, Japan) F-2500 spectrofluorometer in 10 mm × 10 mm quartz cells; a spectral bandwidth of 5 nm was used for both the excitation and emission monochromators. Fluorescence emission spectra of PBHderivatized MCs were measured with an excitation wavelength of 345 nm. 2.5. LC–MS For mass spectra, an ABSciex API4000 Qtrap MS system with positive electrospray ionization (ESI) mode was used in place of the fluorescence detector. To confirm the derivative structures, the compounds were analyzed in quadrupole (Q1) scan and enhanced resolution (ER) mode after derivatization. The MS was configured with the following parameters: capillary voltage of 5500 V, a source temperature of 500 ◦ C, a curtain gas of 40 (arbitrary units), an ion source gas 1 pressure of 60 (arbitrary units), an ion source gas 2 setting of 80 (arbitrary units), and for Q1 scan and ER mode, a scan range of m/z 500–2000 Da and center masses of m/z 804, 1564, and 1614 Da, respectively. The elution conditions of LC were same as those in the fluorescence method.

95

2.6. Method validation Peak areas were integrated automatically and used for the quantification of MCs. Calibration solutions were prepared by diluting the stock solution of MCs. The concentration range of the calibration standards was 10–500 ␮g L−1 (10, 20, 50, 100, 200, and 500 ␮g L−1 ). The concentrations of MCs were calculated from the calibration curves by using the peak areas of PBH-derivatized MCs, and then, those were converted into the concentrations of MCs in samples. The inter-day precisions of the method were estimated using standard solutions (10, 50, and 100 ␮g L−1 ), by repeated analysis 6 times each day. The instrument detection limit (IDL) and the instrument quantification limit (IQL) were defined as the concentrations that gave signal-to-noise ratios (S/N) of 3 and 10, respectively.

2.7. Water sample preparation The tap water samples were taken from our laboratory (Fukuoka, Japan). The pond water samples were collected in Fujieda (Shizuoka, Japan). In this study, these tap and pond water samples were collected in prewashed glass bottles. For the tap water, 15 mg of ascorbic acid sodium salt was added to each 1 L of sample in order to neutralize residual chlorine. The samples were protected from light and stored at 4 ◦ C until analysis. To 10 mL of each water sample was added 2 mL of methanol. The tap and pond water samples were spiked with MC standards at levels of 0.1 and 1 ␮g L−1 . The entire sample was directly injected into the preconditioned InertSep RP-1 for SPE. The cartridge was washed with 1 mL of pure water, and then dried under a nitrogen stream for 10 min. After the MCs in the cartridge were desorbed with 1 mL of 0.05% (v/v) TFA in methanol, the eluate was concentrated under a stream of nitrogen. The residue was reconstituted with 50 ␮L of acetonitrile and then derivatized. The recovery rate was calculated as the ratio of peak areas of derivatives obtained with standard spiked water samples to those from equal standard amounts of MCs.

3. Results and discussion 3.1. Derivatization of MCs We chose PBH as a pre-column excimer derivatization reagent because it has been successfully applied to the measurement of a wide variety of carboxyl compounds [18,19]. In this study, the derivatization of MCs with PBH proceeded in the presence of DMT-MM as the condensation reagent. The derivatization reaction conditions were optimized by with the above-mentioned MC standards (each 100 ␮g L−1 ). Conditions such as the concentrations of PBH (0–50 mM) and DMT-MM (0–300 mM) and the reaction time (5–60 min) and temperature (rt – 90 ◦ C) were varied. By varying the concentration of PBH over 0–50 mM, we found that the maximum signal was seen with derivatization with over 40 mM PBH (Fig. 2A). The concentration of DMT-MM also affected the peak intensities of the derivatives. By increasing the concentration of DMT-MM, the peak areas of all derivatives increased. The DMT-MM concentration of over 100 mM was required for a constant, maximum peak area for examined MCs (Fig. 2B). For this study, the optimum concentration of DMTMM was found to be 200 mM. A reaction time of 30 min at 60 ◦ C was required to obtain maximum responses to MCs. Consequently, 40 mM PBH, 200 mM DMT-MM, and a 30 min reaction time of at 60 ◦ C were the optimal conditions for the derivatization of MCs. The derivatives were stable for at least 1 day in the dark at 4 ◦ C.

96

T. Hayama et al. / Analytica Chimica Acta 755 (2012) 93–99

Fig. 2. Effects of the concentrations of (A) PBH and (B) DMT-MM on the peak areas of PBH-derivatized MCs.

3.2. LC separation In the preliminary study, we used octadecylsilica (ODS) columns to analyze the derivatives of MCs. However, ODS columns were not amenable to the retention and separation of the PBH-derivatized MCs. Therefore, we chose an Accucore PFP column, which is a superficially porous (core shell) particle column that offers an alternative selectivity to ODS. This column not only reduced analytical run-time without decreasing resolution but also had very highly selective retention, which was most apparent when the functional groups were located on an aromatic or other rigid ring system. The derivatives were adequately separated with this column with the mixture of acetonitrile, methanol, and water in the mobile phase. Furthermore, good peak shapes for the derivatives were achieved with the addition of TFA (0.15%) and ammonium acetate (10 mM). Chromatograms were obtained from standard solutions of MCs and blank (deionized water) samples (Fig. 3). Under the LC conditions, all of the PBH-derivatized MCs could be detected and distinguished within 15 min. 3.3. Fluorescence spectra Fluorescence spectra were obtained from LC column eluates, as shown in Fig. 4. As expected, the derivatives in the eluate

emitted not only monomer fluorescence (360–420 nm) but also excimer fluorescence (440–550 nm), although the reagent blank solution showed only monomer fluorescence. Therefore, we chose a fluorescence emission wavelength of 475 nm as optimal for the highly sensitive and selective analysis of PBH-derivatized MCs.

3.4. Structure analysis by LC–MS The PBH-derivatized MC structures were analyzed by LCESI–MS in the positive mode. In this MS analysis, we first determined the protonated molecular ions of derivatives by Q1 scan mode (m/z 500–2000), and ions at m/z 1614, 1564, and 804 were found in the total ion chromatographic peaks corresponding retention times of PBH-derivatized MC-YR, MC-LR, and MC-RR, respectively. Because these were the expected ions of [M+H]+ for PBH-derivatized MC-YR (exact mass: 1612.79) and MC-LR (exact mass: 1562.81) and [M+2H]2+ for MC-RR (exact mass: 1605.83), we subsequently boosted the resolution by setting the MS to ER mode in the third quadrupole (Fig. 5). As a result, all MC derivatives in this study were indeed the di-pyrene derivatives.

1.5 uV 32500

Normalized fluorescence intensity

30000 27500 25000 22500 20000 17500 15000 12500 10000

2

1

7500

1

1.0

2 0.5

3

3

5000

4

(A) 0

2500

(B)

0 0.0

2.5

5.0

7.5

10.0

12.5

15.0 min

Fig. 3. Typical chromatograms obtained with (A) the standard solution of MCs (each 200 ␮g L−1 ) and (B) blank solution. Peaks: 1, MC-YR; 2, MC-LR; 3, MC-RR; others, reagent blanks.

350

400

450

500

550

600

Wavelength (nm) Fig. 4. Fluorescence emission spectra (excitation 345 nm) of the PBH-derivatized (1) MC-LR, (2) MC-YR, (3) MC-RR, and (4) reagent blank. Each spectrum was normalized to the peak at 375 nm.

T. Hayama et al. / Analytica Chimica Acta 755 (2012) 93–99

9.91 1.8e7

97

12.84

(A)

1.6e7

Intensity

1.4e7 1.2e7 1.0e7 8.0e6 6.0e6

6.97

4.0e6 2.0e6 2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0

18.0

Time (min)

(B)

1613.8

[M+H]+

1.8e5

5.5e5

1563.8

5.0e5

[M+H]+

2.4e4

4.0e5

1.4e5

1565.8

2.5e5

8.0e4

1.2e4

2.0e5

6.0e4

1616.8 8000.0

1617.7 1600 1604 1608 1612 1616 1620 1624

805.4

1.5e5

1566.8

4.0e4

4000.0

804.9

3.0e5

1.0e5

1.6e4

[M+2H]2+

3.5e5

1.2e5

2.0e4

804.0

4.5e5

1.6e5

1615.8

804.5

6.0e5

2.0e5

2.8e4

Intensity

(D)

1564.9

2.2e5

3.6e4 3.2e4

(C)

1614.8

1.0e5

2.0e4 1556

1560

1564

m/z

1568

805.9

5.0e4

1567.8 1572

1576

794

798

802

m/z

806

810

814

m/z

Fig. 5. ESI-ER-MS spectra of (A) total ion chromatographic peaks corresponding the retention times of (B) 6.97, (C) 9.91, and (D) 12.84 min.

3.5. Analysis of standards The relationships between the peak areas of derivatives and the concentrations of MCs were linear over the range 10–500 ␮g L−1 . The correlation coefficients of the calibration curves, the inter-day precision values, the IDL, and the IQL of MCs standards are shown in Table 1. The correlation coefficients of the calibration curves were greater than 0.9989. The relative standard deviation (RSD) of the inter-day precision values established by repeated determinations (n = 6) using standard solutions of MCs (10, 50, and 100 ␮g L−1 ) was within 7.4% RSD. The IDL values of MC-YR, MC-LR, and MC-RR were 1.2, 0.4, and 0.8 ␮g L−1 , respectively. The IQL values of MC-YR, MCLR, and MC-RR were 3.9, 1.4, and 2.8 ␮g L−1 , respectively.

The recovery rates from the SPE were determined by the addition of 0.1 and 1.0 ␮g L−1 of each MC to pure water. For all MCs examined, the recoveries were achieved in the range 89–103% within 2.8% (n = 3) as RSD (Table 2). The method we have introduced could satisfy enough the required detectable value for MC-LR in drinking water specified in WHO requirements with acceptable precision, but with only a small 10-mL sample [5]. 3.6. Application to actual water samples In order to demonstrate the feasibility of this method to actual water samples, tap and pond water samples were analyzed with this method. When these samples were spiked with MCs, they did

Table 1 Method validation of the present study. MCs

YR LR RR a b c d

Linearitya

0.9989 0.9996 0.9993

IDLb

IQLc

RSDd (%; n = 6)

(␮g L−1 )

(␮g L−1 )

10 ␮g L−1

50 ␮g L−1

100 ␮g L−1

1.2 0.4 0.8

3.9 1.4 2.8

7.4 3.2 4.7

2.9 2.4 1.9

0.9 1.2 0.6

Correlation coefficients of calibration curves of MC in the concentration range 10–500 ␮g L−1 . Instrument detection limit: defined as the concentration giving a signal-to-noise ratio of 3. Instrument quantification limit: defined as the concentration giving a signal-to-noise ratio of 10. Relative standard deviation of PBH-derivative peak areas.

98

T. Hayama et al. / Analytica Chimica Acta 755 (2012) 93–99

Table 2 Mean recovery (%; n = 3) of MCs from water samples. MCs

YR LR RR

Amount added (␮g L−1 )

Pure water

0.1 1 0.1 1 0.1 1

Tap water

Pond water

Recovery (%)

RSD (%)

Recovery (%)

RSD (%)

Recovery (%)

RSD (%)

103 95 91 95 103 89

2.1 2.7 2.8 2.3 1.6 2.0

82 92 85 84 67 68

11 4.0 2.2 2.4 5.6 4.1

84 97 101 100 84 83

4.1 3.7 3.9 0.9 3.6 1.5

(A)

(B)

uV

uV

22500

20000

20000

17500

17500 15000 15000 12500

1

12500

2

10000 10000

1

2

3 3

7500

5000

(a) (a)

5000

(b)

2500

(b)

2500

0 0.0

7500

(c) 2.5

5.0

7.5

10.0

12.5

15.0 min

(c) 0 0.0

2.5

5.0

7.5

10.0

12.5

15.0 min

Fig. 6. Chromatograms obtained with (a) 1 ␮g L−1 MCs standards spiked, (b) 0.1 ␮g L−1 MCs standards spiked, and (c) unspiked pond water samples. (A) Excimer and (B) monomer fluorescence intensities were monitored at emission wavelengths of 475 nm and 375 nm, respectively, with an excitation wavelength of 345 nm. Peaks: see Fig. 3.

register fluorescence signals (see Table 2 and Fig. 6). No interfering peaks were seen in the spiked pond water samples (Fig. 6A). Furthermore, when monomer fluorescence detection was monitored at 375 nm, the peaks for MC derivatives could hardly be identified (Fig. 6B). In excimer fluorescence detection, the limits of detection (LOD, S/N = 3) and the limits of quantification (LOQ, S/N = 10) of MC-YR, MC-LR, and MC-RR in pond water sample were 0.021, 0.010, and 0.025 ␮g L−1 and 0.070, 0.033, and 0.083 ␮g L−1 , respectively. Recovery from tap water with 0.1 and 1.0 ␮g L−1 of each MC added was 67–92% within 11% (n = 3) RSD. From pond water samples, recoveries ranged from 83 to 101% within 4.1% RSD. In tap water samples, although the recovery of MC-RR was significantly lower, the reason remains unknown. However, the lower recovery had no effect on the quantification with affordable reproducibility (RSD < 11%). This method may require appropriate pretreatment when it applied to another water samples. Nevertheless, it would offer a useful tool for regulatory evaluating the quality of drinking water according to WHO guideline of MCs.

Accucore PFP column. The derivatives could be specifically detected by excimer fluorescence and separated from other components by LC. This excimer fluorescence derivatization is sufficiently sensitive and provides a precise MC analysis. This method quantify a tenth of the maximum allowable value (0.1 ␮g L−1 ) of MCs in drinking water, as given by the WHO. Although the derivatization procedure may be laborious, the proposed method makes it useful for evaluating MC contamination in drinking water samples. Additional some environmental water sample analyses are in progress.

Acknowledgements We would like to extend our thanks to Mr. S. Inoue at Seikan Kensa Center Co., Ltd. in Shizuoka, Japan for providing pond water samples and to Mr. T. Inoue and Mses. S. Gotoh and Mses. T. Tao (Faculty of Pharmaceutical Sciences, Fukuoka University) for excellent technical assistance.

References 4. Conclusions An LC method for detecting MCs (MC-LR, MC-YR, and MCRR) in water samples has been developed. This method involves excimer fluorescence derivatization of MCs with PBH in the presence of DMT-MM and the separation of derivatives by LC on an

[1] F.F. del Campo, Y. Ouahid, Environ. Pollut. 158 (2010) 2906–2914. [2] S.G. Bell, G.A. Codd, Rev. Med. Microbiol. 5 (1994) 256–264. [3] W.W. Carmichael, S.M.F.O. Azevedo, J.S. An, R.J.R. Molica, E.M. Jochimsen, S. Lau, K.L. Rinehart, G.R. Shaw, G.K. Eaglesham, Environ. Health Perspect. 109 (2001) 663–668. [4] L. Pearson, T. Mihali, M. Moffitt, R. Kellmann, B. Neilan, Mar. Drugs 8 (2010) 1650–1680.

T. Hayama et al. / Analytica Chimica Acta 755 (2012) 93–99 [5] WHO (World Health Organization), Guidelines for Drinking-Water Quality, vol. 1: Recommendations, 2nd ed., WHO, Genova, 1998 (addendum). [6] P. Lindner, R. Molz, E. Yacoub-George, A. Durkop, H. Wolf, Anal. Chim. Acta 521 (2004) 37–44. [7] N. Bonacha, I. Maatouk, G. Vincent, Y. Levi, Food Chem. Toxicol. 40 (2002) 1677–1683. [8] P. Tong, L. Zhang, Y. He, S. Tang, J. Cheng, G. Chen, Talanta 82 (2010) 1101–1106. [9] ISO (International Organization for Standardization), ISO Standard 20179:2005. Water Quality – Determination of Microcystins – Method Using Solid Phase Extraction (SPE) and High Performance Liquid Chromatography (HPLC) with Ultraviolet (UV) Detection, ISO, Geneva, 2005. [10] L. Spoof, M.R. Neffing, J. Meriluoto, Toxicon 55 (2010) 954–964. [11] M.R. Neffing, J. Meriluoto, L. Spoof, Anal. Chim. Acta 653 (2009) 234–241. [12] A. Mekebri, G.J. Blondina, D.B. Crane, J. Chromatogr. A 1216 (2009) 3147–3155. [13] H. Nohta, H. Satozono, K. Koiso, H. Yoshida, J. Ishida, M. Yamaguchi, Anal. Chem. 72 (2000) 4199–4204.

99

[14] Y. Nakano, H. Nohta, H. Yoshida, T. Saita, H. Fujito, M. Mori, M. Yamaguchi, J. Chromatogr. B 774 (2002) 165–172. [15] T. Yoshitake, M. Yamaguchi, H. Nohta, F. Ichinose, H. Yoshida, S. Yoshitake, K. Fuxe, J. Kehr, J. Neurosci. Methods 127 (2003) 11–17. [16] H. Yoshida, H. Harada, Y. Nakano, H. Nohta, J. Ishida, M. Yamaguchi, Biomed. Chromatogr. 18 (2004) 687–693. [17] Y. Nakano, H. Nohta, H. Yoshida, K. Todoroki, T. Saita, H. Fujito, M. Mori, M. Yamaguchi, Anal. Sci. 20 (2004) 489–493. [18] H. Nohta, J. Sonoda, H. Yoshida, H. Satozono, J. Ishida, M. Yamaguchi, J. Chromatogr. A 1010 (2003) 37–44. [19] H. Yoshida, J. Sonoda, J. Araki, H. Nohta, J. Ishida, S. Hirose, M. Yamaguchi, Anal. Chim. Acta 534 (2005) 177–183. [20] K. Kato, T. Hayama, M. Itoyama, K. Todoroki, H. Yoshida, M. Yamaguchi, H. Nohta, Bunseki Kagaku 60 (2011) 39–44 (Japanese but English abstract). [21] H. Yoshida, H. Harada, H. Nohta, M. Yamaguchi, Anal. Chim. Acta 488 (2003) 211–221. [22] H. Yoshida, F. Kido, M. Yoshitake, K. Todoroki, H. Nohta, M. Yamaguchi, Anal. Sci. 23 (2007) 485–488.