Chromatographic separation for domoic acid using a fragment imprinted polymer

Analytica Chimica Acta 577 (2006) 1–7

Chromatographic separation for domoic acid using a fragment imprinted polymer Takuya Kubo a,∗ , Makoto Nomachi a , Koji Nemoto a , Tomoharu Sano b , Ken Hosoya c , Nobuo Tanaka c , Kunimitsu Kaya a a b

Graduate School of Environmental Studies, Tohoku University, Aoba 6-6-20, Aramaki, Aoba-ku, Sendai 980-8579, Japan Laboratory of Intellectual Fundamentals for Environmental Studies (LIFES), National Institute for Environmental Studies, Onogawa 16-2, Tsukuba 305-8506, Japan c Department of Polymer Science, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku 606-8585, Kyoto, Japan Received 14 February 2006; received in revised form 13 June 2006; accepted 14 June 2006 Available online 18 June 2006

Abstract We prepared molecularly imprinted polymers for an amnesic shellfish poison, domoic acid. To prepare the polymer, we tested several commercial aromatic dicarboxylic compounds such as isomers of phthalic acid for templates of molecularly imprinted polymers. The highest selective recognition ability of the polymer for domoic acid in the tested compounds was found when o-phthalic acid was used as the template. The ability was due to the acidity of the carboxylic acids in the domoic acid and the similarity of the shape around the carboxylic acids of domoic acid and the templates. The effective chromatographic separation of domoic acid in the extract from blue mussels was achieved with a LC column packed with the fragment imprinted polymer using o-phthalic acid as the template. This polymer can be utilized for a clean up procedure of domoic acid in toxic shellfish. © 2006 Elsevier B.V. All rights reserved. Keywords: Molecularly imprinted polymer; Domoic acid; Selective separation; LC

1. Introduction Amnesic shellfish poisoning (APS) is caused by consumption of shellfish that have accumulated domoic acid, a neurotoxin produced by some strains of phytoplankton. The most characteristic symptom is permanent loss of short-term memory. Ingestion of the intoxicated shellfish has resulted in death in both animal and human consumers in severe [1,2]. The neurotoxic properties of domoic acid result in neuronal degeneration and necrosis in specific regions of the hippocampus. Following the 1987 outbreak of ASP, the Canadian authorities imposed an action limit of 20 ␮g domoic acid/g wet tissues of shellfish. The principal toxin responsible for APS is domoic acid, an amino acid belonging to the kainoid class of compounds [3]. Domoic acid is produced by a number of marine algae, including microalgae of the genus Pseudonitzschia, and is accumulated by

∗

Corresponding author. Tel.: +81 22 795 7410; fax: +81 22 795 7410. E-mail address: [email protected] (T. Kubo).

0003-2670/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2006.06.028

shellfish filter feeding during blooms. Furthermore, 10 isomers of domoic acid have been identified in marine samples [3–5]. Several analytical methods have been developed for the quantitative determination of domoic acid [6–11]. The original method used for the detection of domoic acid was the mouse bioassay, involving intra-peritoneal injection of mice with an acidic extract from whole shellfishes [6]. As for biochemical methods, the applications of the assay to analyze domoic acid in sea water have been reported based on the antibodies [7]. However, the biochemical methods are costly and have poor reproducibility. Probably, fractionation of domoic acid from samples is necessary for the determination. Recently, some sensitive methods using liquid chromatographic-mass spectrometry (LC-MS) have also developed [8–11]. In the LC-MS methods, the ionization efficiency of domoic acid was interfered by some compounds which eluted around the domoic acid peak. In analytical methods of natural products, selective separation of target compounds is very important, since target compounds are extracted together with a lot of biological compounds which interfere with the analyses. In the case of domoic acid, effective

2

T. Kubo et al. / Analytica Chimica Acta 577 (2006) 1–7

selective pretreatment media for the quantitative analyses have not been developed. Lotierzo et al. [12] have reported a sensor for domoic acid prepared by a molecularly imprinted polymer film using domoic acid as the template. However, interfering compounds have not been examined in detail. Since molecular imprinting method is rather easy method, the molecular imprinted polymers (MIPs) have been used as various media such as an artificial antibody, a biosensor as well as stationary phases for high performance liquid chromatography (HPLC) [13–16]. However, the molecular imprinting method should require the template molecule directly to realize specific molecular recognition ability towards the real target molecule. But, it becomes a serious problem, if we have to obtain the selective molecular recognition ability toward some toxic compounds or very rare compound. Moreover, when we utilize the template molecule, it cannot be removed completely from the polymer even after tedious repeated washing process with some organic solvents, because the imprinted sites can be also formed not only on the surface but also deeply in the cross-linked polymer network structure, where organic solvent can hardly reach [17]. This can be the other serious problem for trace analyses of the environmental toxic compounds by gradually leaked template molecule from the polymer matrix. Therefore, some alternative molecular imprinting procedures have been reported [18–20] and we have also previously reported some alternative molecular imprinting procedures that the template molecule has been selected from analogues of the target compound or partially resemble compounds of the target compound, but not used the target compound itself [21–25]. In this study, we developed an effective molecularly imprinted polymer for domoic acid. To prepare the polymer, we tested several commercial aromatic dicarboxylic acid compounds such as isomers of phthalic acid for templates of MIPs. The selectivity of domoic acid adsorption of each tested polymer was evaluated using LC. The highest selective recognition polymer for domoic acid was observed when o-phthalic acid was used as the template of the MIP. The selectivity was due to the acidity of the carboxylic acids in the domoic acid and the similarity of the shape around the carboxylic acids of domoic acid and the templates. When the polymer was packed with a LC column for separation of domoic acid in the extract from blue mussels, the effective chromatographic separation was achieved. 2. Experimental section 2.1. Chemicals Ethylene glycol dimethacrylate (EDMA) and 4-vinylpyridine (4Vp) were obtained from Wako Chemicals (Osaka, Japan), and were distilled under vacuum for removal of polymerization inhibitors [26]. Acetic aid, o-, m-, p-, homo-phthalic acid, 2,2 -azobis-(2,4-dimethyl-valeronitrile) (ADVN) and acetonitrile (AN) in HPLC grade were also purchased from Wako Pure Chemicals. Domoic acid, dihydrokainic acid and trans-dl1,2-cyclopentanedicarboxylic acid (CDA) were purchased from Sigma Aldrich (USA).

Table 1 Composition and abbreviation of polymer particles Abbreviations

Crosslinker, porogen, functional monomer

Template (0.56 mmol)

Non-MIP

EDMA (26.5 mmol), toluene (5 mL), 4Vp (4.4 mmol) EDMA (26.5 mmol), toluene (5 mL), 4Vp (4.4 mmol) EDMA (26.5 mmol), toluene (5 mL), 4Vp (4.4 mmol) EDMA (26.5 mmol), toluene (5 mL), 4Vp (4.4 mmol) EDMA (26.5 mmol), toluene (5 mL), 4Vp (4.4 mmol) EDMA (26.5 mmol), toluene (5 mL), 4Vp (4.4 mmol)

–

p-ph-MIP m-ph-MIP o-ph-MIP H-ph-MIP CDA-MIP

p-Phthalic acid m-Phthalic acid o-Phthalic acid Homo-phthalic acid CDA

2.2. Preparation of polymer particles To prepare polymer-based separation media, we applied the two-step swelling and polymerization method [27,28]. The polystyrene seed particles were prepared through the emulsifier free emulsion polymerization [29]. The polymerization was carried out at 50 ◦ C for 24 h using 2.0 wt.% of ADVN as a radical initiator. After the polymerization procedure, the polymer particles were washed with MeOH, then tetrahydrofuran (THF). The washed particles were dried on a membrane filter at room temperature. The polymer particles had about 5.0 ␮m in diameter with excellent size uniformity as reported previously [30]. The compositions of prepared polymer particles are shown in Table 1, and the structure of template molecules, domoic acid and dihydrokainic acid are shown in Fig. 1. 2.3. Evaluation of the polymers using LC Polymer particles were packed into stainless steel columns for evaluation of the selective recognition effect. The two mobile phases were used for the evaluations; one was AN/10 mM phosphate buffer (pH 3.0, 7/3, v/v) for evaluation of the aromatic acidic solutes, another one was AN/0.05% aqueous acetic acid (7/3, v/v) for evaluation of domoic acid and dihydrokainic acid. Chromatographic data were acquired with a LC system (Shimadzu, Japan), consisting from a LC-10AD as a pump, a SPDM10A as a PDA detector, and a CTO-10AC as a column oven. The retention factor k and the separation factor α were defined as follows: k = (v(retention volume of solute) − v(void volume) )/v(void volume)

(a)

α(a/b) = k a/k b

(b)

(where k a = k of a, k b = k of b)

2.4. Separation of domoic acid in shellfish extract Blue mussels were obtained from a fish market. An extract was prepared from edible tissues (11.0 g) of blue mussels and 50 mL of 50% aqueous methanol (v/v) according to the method of Lopez-Rivera et al. [11]. Authentic domoic acid was added into the extract at the concentration of 25 ␮g mL−1 . The domoic

T. Kubo et al. / Analytica Chimica Acta 577 (2006) 1–7

3

Fig. 1. Structure of domoic acid, dihydrokainic acid and template molecules.

acid-containing extract was filtered using a membrane filter (0.2 ␮m). The domoic acid-containing extract of blue mussels was applied to the column packed with the o-phthalic acid imprinted polymers. The polymers were evaluated for the separation of domoic acid in blue mussels. 3. Results and discussion 3.1. Fundamental properties In order to clarify fundamental retention properties of the MIP particles, the polymers were evaluated using LC with the AN aqueous solution containing phosphate buffer (pH

3.0) as the mobile phase. Several acidic compounds were utilized for the evaluation solutes. The retention factors of the solutes are summarized in Fig. 2. As shown in the figure, the retentions of o-phthalic acid were higher on the all imprinted polymers. According to the pKa1 value of o-phthalic acid (Table 2), these facts suggested that the significant difference of the retention was due to the strength of acidity of the solutes. On the other hand, o-phthalic acid was not retained on nonMIP. These results indicated that all imprinted polymers formed recognition sites for dicarboxylic acid compounds as the template molecules. The separation factors of the acidic solutes toward benzoic acid were shown in Table 3. In each MIPs

Fig. 2. Retention factor of acidic solutes. LC conditions—mobile phase: AN/10 mM phosphate buffer (pH 3) = 7/3; column size: 150 mm × 4.6 mm i.d.; flow rate: 1.0 mL min−1 ; temperature: 30 ◦ C; detection: photo diode array.

4

T. Kubo et al. / Analytica Chimica Acta 577 (2006) 1–7

Table 2 pKa value of acidic compounds

Benzoic acid o-Phthalic acid m-Phthalic acid p-Phthalic acid Homo-phthalic acid CDA

pKa1

pKa2

4.20 2.94 3.70 3.54 3.58 4.18

– 5.43 4.60 4.34 – –

Table 3 Comparison of separation factor toward benzoic acid Polymers

p-Phthalic acid

m-Phthalic acid

o-Phthalic acid

Non-MIP p-ph-MIP m-ph-MIP o-ph-MIP H-ph-MIP CDA-MIP

1.06 1.98 1.54 1.99 1.36 1.47

0.97 1.39 1.45 1.71 1.20 1.26

1.86 12.5 11.3 19.4 9.37 9.80

LC conditions—mobile phase: AN/10 mM phosphate buffer (pH 3) = 7/3; column size: 150 mm × 4.6 mm i.d.; flow rate: 1.0 mL min−1 ; temperature: 30 ◦ C; detection: photo diode array; ␣: solute/benzoic acid.

prepared with phthalic acid as template molecule, the imprinting effect for template molecule was observed. Especially, the imprinting effect on o-ph-MIP was much higher.

for the selective separation of domoic acid, the o-ph-MIP column was able to separate domoic acid and phenol completely (Fig. 5). With a similar structure as domoic acid, dihydrokainic acid was also retained by the o-ph-MIP column. These findings suggest that the two carboxylic groups attached with the neighboring carbons on the ring are selectively recognized by the o-ph-MIP column. Additionally, the pKa values of domoic acid also correlate with the retention factors, since the pKa values of domoic acid are 2.10, 3.72 and 4.97 [31], and resemble closely those of o-phthalic acid. That is, domoic and dihydrokainic acids are the strong acids as o-phthalic acid. When homo-phthalic acid (H-ph) and cyclopentanedicarboxylic acid (CDA) were used as the template molecules of the imprinted polymers, the columns packed with the particles selectively retained domoic and dihydrokainic acid. The pKa values of H-ph and CDA are almost the same as those of m-, and p-phthalic acids. From the pKa values, domoic acid ought to be the low retention ability. The results of the retention of domoic acid by the columns prepared with H-ph and CDA were much different from the presumption by the pKa values. It is this difference that is due to the structural recognition of the moiety of the two carboxylic groups attached with the neighboring carbons on the ring. As the results, the selective recognition of domoic acid was achieved to utilize the o-phthalic acid as the template for the fragment imprinting. 3.3. Selective separation for domoic acid in an extract of blue mussel tissues

3.2. Selective separation of domoic acid The selective separation ability of the imprinted polymer particles against domoic acid and dihydrokainic acid was evaluated by LC. The retention factors and the separation factors of domoic and dihydrokainic acids toward benzoic acid were shown in Figs. 3 and 4, respectively. When the columns packed with the MIP particles prepared with the different templates were tested

When the blue mussel extract containing domoic acid was applied for the column packed with C18 and non-MIP, the peak of domoic acid was eluted within 5 min and overlapped with the peaks of blue mussel components (Fig. 6). On the other hands, when the o-ph-MIP column was used for the separation, the domoic acid peak was found at about 11 min and completely separated from the other peaks.

Fig. 3. Comparison of retention factor for domoic acid. LC conditions—mobile phase: AN/0.05% aqueous acetic acid = 7/3; column size: 150 mm × 4.6 mm i.d.; flow rate: 1.0 mL min−1 ; temperature: 30 ◦ C; detection: photo diode array; injection volume: 2 ␮L (1.0 mg mL−1 ).

T. Kubo et al. / Analytica Chimica Acta 577 (2006) 1–7

5

Fig. 4. Comparison of separation factor toward benzoic acid. LC conditions were same as Fig. 3.

This column will be able to use for analysis of domoic acid using the methods of the column-switching LC-MS and/or PDA. In this case, the novel o-ph-MIP column will be applied for the first column as a clean up column, and then domoic

acid will be introduced to the second column as an analytical column such as a C18 column. Moreover, the MIP may be used as a cartridge for the direct isolation of domoic acid from the biochemical samples.

Fig. 5. Chromatograms of domoic acid on each polymer columns. LC conditions were same as Fig. 3.

6

T. Kubo et al. / Analytica Chimica Acta 577 (2006) 1–7

Fig. 6. Chromatograms of the shellfish extract. LC conditions were same as Fig. 3. Injection: 20 ␮L of the extracted sample. (a) Extract containing domoic acid on the commonly used C18 column; (b) extract containing domoic acid on non-MIP packed column; (c) extract containing domoic acid on o-ph-MIP; (d) extract not containing domoic acid on o-ph-MIP.

4. Conclusion

References

We developed a novel separation medium for selective separation of domoic acid as an amnesic shellfish poisoning. The medium was prepared using molecularly imprinting technique with 4-vinylpyridine as the functional monomer and o-phthalic acid as the template. The selective recognition of domoic acid is due to the acidity of the carboxylic acids in the molecule and the structural recognition of the moiety of the two carboxylic groups attached with the neighboring carbons on the ring. The separation of domoic acid in the extract of blue mussels using the o-ph-MIP column was successfully achieved. Actually, the peak of domoic acid was completely isolated from other peaks. The novel o-ph-MIP column will be able to apply the domoic acid analyses for selective separation of domoic acid from shellfish components.

[1] J.L.C. Wright, R.K. Boyd, A.S. DeFreitas, M. Falk, R.A. Foxall, W.D. Jamieson, M.V. Laycock, A.W. McCulloch, A.G. McInnes, P. Odense, V.P. Pathak, M.A. Quilliam, M.A. Ragan, P.G. Sim, P. Thibault, J.A. Walter, M. Gilgan, D.J. Richard, D. Dwar, Can. J. Chem. 67 (1989) 481. [2] S.S. Bates, C.J. Bird, A.S. DeFreitas, R. Foxall, M. Gilgan, L.A. Hanic, G.R. Johnson, A.W. McCulloch, P. Odense, R. Pocklington, M.A. Quiliam, P.G. Sim, J.C. Smith, D.V. Subba Rao, E.C. Todd, J.A. Walter, J.L. Wright, Can. J. Fish Aquat. Sci. 46 (1989) 1203. [3] J.L.C. Wright, M. Falk, A.G. McInnes, J.A. Walter, Can. J. Chem. 68 (1990) 22. [4] J.L.C. Wright, M. Quiliam, Manual on Harmful Marine Microalgae, IOC Manuals and Guides, No. 33, 1995, 113 pp. [5] L. Zaman, O. Arakawa, A. Shimosu, Y. Onoue, S. Nishio, Y. Shida, T. Noguchi, Toxicon 35 (1997) 205. [6] F. Inverson, J. Truelove, E. Nera, L. Tryphonas, J. Campbell, E. Lok, Food Chem. Toxicol. 27 (1989) 377. [7] D.S. Smith, D.D. Kitts, Food Chem. Toxicol. 32 (1994) 1147. [8] J.F. Lawrence, B.P. Lau, C. Cleroux, D. Lewis, J. Chromatogr. A 659 (1994) 119. [9] T. Sun, W.H. Wong, J. Agric. Food Chem. 47 (1999) 4678. [10] A. Furey, M. Lehane, M. Gillman, P. Fernandez-Puente, K.J. James, J. Chromatogr. A 938 (2001) 167.

Acknowledgement This research was financial supported by Nanotechnology Project of the Ministry of Environment.

T. Kubo et al. / Analytica Chimica Acta 577 (2006) 1–7 [11] A. Lopez-Rivera, B.A. Suarez-Isla, P.P. Eilers, C.G. Beaudry, S. Hall, M. Fernandes Amandi, A. Furey, K.J. James, Anal. Bioanal. Chem. 381 (2005) 1540. [12] M. Lotierzo, O.Y.F. Henry, S. Piletsky, I. Tothill, D. Cullen, M. Kania, B. Hock, A.P.F. Turner, Biosens. Bioelectron. 20 (2004) 145. [13] G. Wulff, A. Sarhan, K. Zabrocki, Tetrahedron Lett. 44 (1973) 4329. [14] B. Sellergren, B. Ekberg, K. Mosbach, J. Chromatogr. 347 (1985) 1. [15] D. Spivak, K.J. Shea, J. Org. Chem. 64 (1999) 4627. [16] B. Sellergren, Molecular Imprinted Polymers, Elsevier Science, 2001. [17] J. Haginaka, H. Sanbe, Anal. Chem. 72 (1) (2000) 5206. [18] A. Rachkov, N. Minoura, J. Chromatogr. A 889 (2000) 111. [19] J. Matsui, K. Fujiwara, T. Takeuchi, Anal. Chem. 72 (2000) 1810. [20] M. Quaglia, K. Chenon, A.J. Hall, E.D. Lorenzi, B. Sellergren, J. Am. Chem. Soc. 123 (2001) 2146. [21] K. Hosoya, K. Yoshizako, H. Sasaki, K. Kimata, N. Tanaka, J. Chromatogr. A 828 (1998) 91. [22] T. Kubo, K. Hosoya, Y. Watabe, T. Ikegami, N. Tanaka, T. Sano, K. Kaya, J. Chromatogr. A 987 (2003) 389.

7

[23] T. Kubo, K. Hosoya, Y. Watabe, T. Ikegami, N. Tanaka, T. Sano, K. Kaya, J. Chromatogr. A 1029 (2004) 37. [24] T. Kubo, K. Hosoya, Y. Watabe, T. Ikegami, N. Tanaka, T. Sano, K. Kaya, J. Sep. Sci. 27 (2004) 316. [25] T. Kubo, K. Hosoya, T. Sano, M. Nomachi, N. Tanaka, K. Kaya, Anal. Chim. Acta 549 (2005) 45–50. [26] D.D. Perrin, W.L.F. Armarego, D.R. Perrin, Purification of Laboratory Chemicals, Pergamon Press, Oxford, 1980. [27] J. Ugelstad, K.H. Kaggerud, F.H. Hansen, A. Perge, Makromol. Chem. 180 (1979) 737. [28] J. Ugelstad, P.C. Mork, Advanced in Colloid and Interface, Science, vol. 13, Elsevier Scientific Publishing Company, Amsterdam, 1980, 101 pp. [29] V. Smigol, F. Svec, K. Hosoya, Q. Wang, J.M.J. Frechet, Angew. Makromol. Chem. 195 (1992) 151. [30] K. Hosoya, J.M.J. Frechet, J. Polym. Sci., Part A: Polym. Chem. 31 (1993) 2129. [31] N. Pieneiro, J.M. Leao, A.G. Martinez, J.A. Rodriquez Vazaquez, J. Chromatogr. A 847 (1999) 223.