Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Citric acid cycle wikipedia , lookup
Proteolysis wikipedia , lookup
Point mutation wikipedia , lookup
Fatty acid synthesis wikipedia , lookup
Fatty acid metabolism wikipedia , lookup
Nucleic acid analogue wikipedia , lookup
Peptide synthesis wikipedia , lookup
Size-exclusion chromatography wikipedia , lookup
Genetic code wikipedia , lookup
Amino acid synthesis wikipedia , lookup
JOURNAL OF MOLECULAR RECOGNITION J. Mol. Recognit. 2006; 19: 39–48 Published online 2 November 2005 in Wiley InterScience (www.interscience.wiley.com). DOI:10.1002/jmr.756 Application of cross-linked b-cyclodextrin polymer for adsorption of aromatic amino acids Shouwan Tang, Liang Kong, Junjie Ou, Yueqi Liu, Xin Li and Hanfa Zou* National Chromatographic Research and Analysis Center, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, People’s Republic of China b-Cyclodextrin (b-CyD) was cross-linked by hexamethylene diisocyanate and the polymer was investigated for adsorption of aromatic amino acids (AAA) from phosphate buffer. High adsorption rates were observed at the beginning and the adsorption equilibrium was then gradually achieved in about 45 min. The adsorption of AAA decreased with the increase of initial concentration and also temperature. Under the same conditions, the adsorption efficiencies of AAA were in the order of L-tryptophan (L-Trp) > Lphenylalanine (L-Phe) > L-tyrosine (L-Tyr). Much higher adsorption values, up to 52.4 and 43.0 mg/g for L-Trp and L-Phe, respectively, at 50 mmol/L and 3.2 mg/g for L-Tyr at 2 mmol/L, were obtained with the b-CyD polymer at 37 C. It was shown that the adsorption of AAA on the b-CyD polymer was consistent with the Freundlich isotherm equation. The adsorption of mixed aromatic amino acids and branched-chain amino acids (BCAA) showed that AAA were preferentially adsorbed with adsorption efficiencies 10–24%, while those of BCAA were lower than 2%. It seems that the structure and hydrophobicity of amino acid molecules are responsible for the difference in adsorption, by influencing the strength of interactions between amino acid molecule and the polymer. Copyright # 2005 John Wiley & Sons, Ltd. Keywords: b-cyclodextrin polymer; adsorption; aromatic amino acids; branched-chain amino acids Received 19 May 2005; revised 16 August 2005; accepted 23 August 2005 INTRODUCTION Molecular recognition plays an important role in nature, for instance, receptors, enzymes, transporters, and so forth. Developing artificial receptors, in other words constructing molecular recognition sites artificially, is one of the major goals in chemistry. To date, various synthetic molecular clefts and cavities (e.g., crown ethers, cyclodextrins, and cyclophanes) have been developed by utilizing interactions such as hydrogen bonding, ionic interaction, van der Waals interaction and hydrophobic effect. The cyclodextrins (CyDs) are a series of cyclic oligosaccharides with a hydrophilic exterior and a hydrophobic interior capable of binding small hydrophobic structures. -CyD and their *Correspondence to: Prof. Dr Hanfa Zou, National Chromatographic Research and Analysis Center, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, People’s Republic of China. E-mail: [email protected] Contract/grant sponsor: Natural Science Foundation of China; contract/grant number: 20475054. Contract/grant sponsor: Knowledge Innovation Program of DICP. Abbreviations used: AAA, aromatic amino acids; BCAA, branched-chain amino acids; CyD, cyclodextrin; DMF, dimethylformamide; DNFB, 2,4-dinitrofluorobenzene; DNP-OH, 2,4-dinitrophenol; (DNP)2O, diDNP ether; FT-IR, Fourier transform infrared; Gly, glycine; HMDI, hexamethylene diisocyanate; HPLC, high performance liquid chromatography; Ile, isoleucine; Leu, leucine; Phe, phenylalanine; MIPs, molecularly imprinted polymers; SEM, scanning electron microscopic; Trp, tryptophan; Tyr, tyrosine; Val, valine. Copyright # 2005 John Wiley & Sons, Ltd. derivatives are well known for their ability to form inclusion complexes with a variety of compounds. -CyD dimers and trimers have been prepared and investigated as binders for a series of compounds containing t-butylphenyl groups, dimeric and trimeric amino acid amides, respectively (Breslow et al., 1995; Leung et al., 2001). Recently, molecular imprinting as a technique to generate recognition sites of predetermined selectivity has received considerable attention (Andersson and Nicholls, 1997). Molecularly imprinted polymers (MIPs) based on methacrylic acid and 4vinylpyridine, Cu(II)-N-(4-vinylbenzyl) iminodiacetic acid, etc. have been used for chiral recognition of aromatic amino acids (AAA) (Huang et al., 2003, 2004; Vidyasankar et al., 1997). MIPs based on -CyD have been extensively examined for efficient recognition of steroids, peptides, Dphenylalanine (D-Phe), amino acid derivatives, antibiotics, etc. (Akiyama et al., 2001; Asanuma et al., 1997, 1998, 1999, 2001, 2004; Hishiya et al., 1999, 2002; Piletsky et al., 1998, 1999; Sreenivasan, 1998; Sreenivasan and Sivakumar, 1998; Zhong et al., 2001). Highly enantioselective polymers for L-Phe, D- or L-tryptophan methyl ester were prepared by the so-called ‘surface molecular imprinting technique’ (Nakamura et al., 1999; Yoshida et al., 2000a,b). Pilestsky et al. (1998, 1999), demonstrated the rational use of hydrophobic and electrostatic ligand– polymer interactions for recognition of D-Phe in a novel MIP system. MIPs with element of CyDs were also found to have enhanced adsorption capacity in addition to the 40 S. TANG ET AL. selectivity acquired by the imprinting, which was attributed to the presence of -CyD moieties in the copolymer (Sreenivasan, 1998; Sreenivasan and Sivakumar, 1998). In this study, we developed an alternative sorbent— cross-linked -CyD polymer for adsorption of AAA based on cavity inclusion effect. Polymeric sorbents have been extensively tested for amino acid adsorption (Doulia et al., 2001; Dı́ez et al., 1998; Grzegorczyck and Carta, 1996a,b). The -CyD polymer has been shown to be a good sorbent for bilirubin in our previous study (Zheng et al., 2004). In the present article, we report our preliminary results related to the AAA adsorption behavior of this polymer in phosphate buffer. structures of amino acids used in this paper are shown in Figure 1. Acetonitrile of HPLC grade was purchased from Merck (Darmstadt, Germany). Double distilled water purified by Milli-Q (Millipore Co., Miford, MA, USA) was utilized throughout the experiments. Other reagents were analytical grade. Instruments V-550 UV/VIS spectrophotometer was from Jasco (Tokyo, Japan). HPLC system consisted of two LC-10ADvp pumps (Shimadzu, Kyoto, Japan), a SPD-10Avp UV-Vis detector (Shimadzu, Kyoto, Japan), a column (250 4.6 mm I.D.) packed with 5 mm Kromasil C18 (Akzo Noble AB, Nacka, Sweden), a TC-100 temperature controller (Autoscience, Tianjin, China) and a WDL-95 chromatographic workstation (National Chromatographic R&A Center, Dalian, China). EXPERIMENTAL Materials -Cyclodextrin (-CyD) of chemical grade (Shanghai Chemical Reagent Co., Ltd, Shanghai, China) was recrystallized three times and dried under vacuum at 110 C for 24 h before use. Pyridine of analytical grade (Tianjin Kermel Chemical Development Center, Tianjin, China) was dried with molecular sieve 4A, refluxed with CaH2 for 2 h, and then distilled for use. L-Tryptophan (L-Trp), Ltyrosine (L-Tyr), L-phenylalanine (L-Phe), 2,4-dinitrofluorobenzene (DNFB), and hexamethylene diisocyanate (HMDI, 99% pure) were obtained from Acros (New Jersey, USA). L-Valine (L-Val), L-isoleucine (L-Ile), and L-leucine (L-Leu) were obtained from TCI (Tokyo, Japan). The Preparation of the b-CyD polymer The -CyD polymer was prepared according to the procedures described elsewhere (Hishiya et al., 1999) with a few modifications. -CyD (5 g) was dissolved in dry pyridine (50 ml) and then HMDI (4.5 ml) was added. After being magnetically stirred at 65 C for 2 h, the mixture was poured into acetone to obtain the cross-linked polymer. Then the polymer was washed with hot water and hot ethanol to remove completely unreacted cross-linker and free -CyD. O H2N CH O C OH H2N CH2 CH C O H2N OH CH C OH CH2 CH2 HN OH (a) (b) O H2N CH C CH2 CH CH3 (c) O OH H2N CH C CH CH3 O OH H2N CH C CH CH3 OH CH2 CH3 CH3 (d ) (e) CH3 (f ) Figure 1. Structures of amino acid molecules. (a) Tryptophan; (b) phenylalanine; (c) tyrosine; (d) leucine; (e) isoleucine; (f) valine. Copyright # 2005 John Wiley & Sons, Ltd. J. Mol. Recognit. 2006; 19: 39–48 ADSORPTION OF AROMATIC AMINO ACIDS Finally, the polymer was dried under vacuum at 70 C for 24 h before use. Sample preparation Preparation of single amino acid samples. Stock solutions (2.0 mmol/L) of L-Trp, L-Phe or L-Tyr were prepared by dissolving them in 10 mmol/L phosphate buffer (pH 7.4) separately. AAA solutions (0.01–0.40 mmol/L) were obtained by dilution of the stock solutions with 10 mmol/L phosphate buffer (pH 7.4). Preparation of mixed amino acid samples. Stock solutions of mixed AAA (2.0 mmol/L for each amino acid) and aromatic amino acids with branched-chain amino acids (BCAA) solution (2.0 mmol/L for each AAA and 1.0 mmol/L for each BCAA) were prepared by dissolving L-Trp, L-Phe and L-Tyr or L-Trp, L-Phe, L-Tyr, L-Val, L-Leu and L-Ile in 10 mmol/L phosphate buffer (pH 7.4), respectively. Mixed AAA solutions (0.02–0.40 mmol/L for each AAA) or AAA-BCAA solution (0.2 mmol/L for each AAA and 0.1 mmol/L for each BCAA) were obtained by dilution of the corresponding stock solution above with 10 mmol/L phosphate buffer (pH 7.4). Determination of the concentration of amino acids Adsorption of AAA from phosphate buffer with the -CyD polymer was studied batchwise. Five milliliters of amino acid solution was incubated with 100 mg of the -CyD polymer and shaken for a certain time. For single amino acid samples, UV-Vis spectrophotometer was used to determine the concentration of amino acids after adsorption. The detection wavelengths of L-Trp, L-Phe, L-Tyr were 280, 257, 275 nm, respectively. For mixed amino acid samples, they were first derivatized with DNFB (Robert and Gerhard, 1988) and then separated and detected by HPLC. 41 0.5 mol/L NaHCO3 (pH 9.0) and 50 ml of acetonitrile containing 1% DNFB (v/v) were added and mixed. Derivatization was carried out at 60 C for 1 h while shielded from light. Finally, 600 ml of 50 mmol/L KH2PO4 (pH 7.0) was added after the mixture cooled down to room temperature. Chromatographic analysis of mixed amino acids A 250 4.6 mm I.D. column packed with 5 mm Kromasil C18 was used for analysis of the derivatized amino acids by keeping column at 33 C. Mobile phase A was acetonitrile-water (50:50, v/v) and B was 100 mmol/L sodium acetate buffer (pH 6.4), containing 1% dimethylformamide (DMF) (v/v). Gradient elution was carried out as follows: 0–0.3 min, 84%B; 0.3–4 min, 84%B–70%B; 4–7 min, 70% B–66%B; 7–12 min, 66%B–57%B; 12–22 min, 57%B–45% B; 22–25 min, 45%B; 25–40 min, 45%B–0%B, at a flow rate of 1.0 ml/min. Detection was by UV at 360 nm. Characterizing the b-CyD polymer The polymer dried under vacuum at 70 C for 24 h was placed on a sticky carbon foil which was attached to a standard aluminum specimen stubs. The carbon foil was to increase conductivity. The sample was vapor-deposited with gold (Eiko IB3 ion coater). Scanning electron microscopic (SEM) images of the -CyD polymer were obtained with a JSM-6360 LV scanning electron microscope (Jeol, Tokyo, Japan). -CyD was dried under vacuum at 110 C for 24 h. The IR spectra for -CyD and the -CyD polymer were obtained with an FT-IR spectrophotometer (Perkin-Elmer, Boston, MA, USA) as KBr tablets in the 4000–400 cm1 region. RESULTS AND DISCUSSION Characterization of the b-CyD polymer Derivatization procedure The dinitrophenylation of mixed amino acids was according to the following procedure: to 100 ml of sample, 250 ml of The morphology of the polymer was closely related to the polymerization conditions. The scanning electron micrographs for the internal structure of the -CyD polymer are shown in Figure 2. Figure 2. Scanning electron micrograph of the -CyD polymer. (a) 1,000 ; (b) 10,000 . Copyright # 2005 John Wiley & Sons, Ltd. J. Mol. Recognit. 2006; 19: 39–48 42 S. TANG ET AL. From the FT-IR (KBr) of -CyD and the -CyD polymer as shown in Figure 3, it can be seen that: for -CyD, 3600– 3000 ((O-H)), 2928 cm1 ((C-H)); while for the -CyD polymer, 3600–3000 ((O-H)), 2933 ((C-H)), 1701 ((C ¼ O)) and 1541 cm1 ((N-H)). The existence of 1701 ((C ¼ O)) and 1541 cm1 ((N-H)) is the evidence of the polymer being cross-linked by HMDI. Adsorption of single AAA Figure 4 gives the adsorption rate of different AAA. The -CyD polymer was incubated with the samples for 5, 15, 30, 45, and 60 min at 37 C. As shown in Figure 4, the adsorption process was completed within about 45 min and this value can be considered as the equilibrium time for AAA adsorption. Relatively high adsorption rates of AAA were observed at the beginning and more than 82% of the equilibrium amount was adsorbed within 5 min for AAA. Under the same conditions, the adsorption efficiencies of AAA were in the order of L-Trp > L-Phe > L-Tyr. This is due to the complementarity of the structure and hydrophobicity of AAA molecules to the cavity of the -CyD polymer. Since the polymer was obtained by cross-linking of hydro- xyl groups on the primary faces of two or more -CyDs, the hydrophobic cavity was little affected (the cavity volume of -CyD [Å3], 262). As shown in Table 1, Trp is the biggest (volume [Å3], 227.8) and most hydrophobic among AAA molecules. So, the interactions between L-Trp and the -CyD polymer are the strongest among AAA, resulting in the highest adsorption efficiency. As shown in Figure 1, one hydroxyl group distinguishes L-Tyr from L-Phe, but the adsorption efficiency of L-Phe was 3–4% higher than L-Tyr in the adsorption process studied. This is probably due to the hydroxyl group decreasing considerably the hydrophobicity of the molecule while increasing the volume of the molecule slightly. For the reason above, the volume of Tyr is slightly bigger than Phe, but the hydrophobicity of Phe is much higher than Tyr as shown in Table 1. From these two parameters—molecular volume and hydrophobicity, the interactions between L-Phe and the -CyD polymer are stronger than those of L-Tyr. Thus the adsorption efficiency of L-Phe is higher than L-Tyr. Figure 5 gives adsorption efficiency versus initial concentration of AAA. The -CyD polymer was incubated with AAA samples containing different amounts of AAA (0.02– 0.40 mmol/L) at 37 C for 1 h. The adsorption efficiency of AAA decreased with increasing AAA initial concentration. Figure 3. FT-IR spectra of -CyD and the -CyD polymer. (a) -CyD; (b) the -CyD polymer. Copyright # 2005 John Wiley & Sons, Ltd. J. Mol. Recognit. 2006; 19: 39–48 ADSORPTION OF AROMATIC AMINO ACIDS obtained. But from the trend of adsorption capacity with respect to concentrations as shown in Figure 6(c), the adsorption was far from saturation. Adsorption of AAA from phosphate buffer onto the surface of the -CyD polymer can be explained by the Freundlich isotherm using the following equation 30 Adsorption efficiency (%) 43 25 q ¼ kC 1=n 20 ð1Þ where q is the amount of substance adsorbed per unit mass of sorbent at equilibrium (mg/g), C is the equilibrium concentration of the solute being adsorbed (mmol/L), while k and n are empirical constants dependent on the nature of sorbent and solute and the temperature. By taking logarithms of both sides of equation (1), we obtain 15 1 lg q ¼ lg k þ lg C n 10 0 10 20 30 40 50 60 Adsorption time (min) Figure 4. Adsorption rates of L-Trp (~), L-Phe (*) and L-Tyr (&). Experimental conditions: 5 ml of 0.2 mmol/L L-Trp, L-Phe or L-Tyr incubated with 100 mg of the -CyD polymer for 5, 15, 30, 45, 60 min, respectively; temperature, 37 C. When the concentration of AAA increased from 0.02 to 0.40 mmol/L, the adsorption efficiencies of L-Trp, L-Phe, and L-Tyr decreased from 35.55, 34.31, and 32.05% to 24.27, 18.81, and 12.17%, respectively. Also the adsorption efficiencies remained in the same order, L-Trp > L-Phe > L-Tyr. The effect of temperature was also studied. The AAA adsorption at 25, 37, and 50 C, representing the amount of AAA adsorbed with respect to the AAA initial concentrations, is shown in Figure 6. The amount of adsorbed AAA per unit amount of the polymer decreased with increasing temperature. Note that the adsorption values of L-Trp and L-Phe reached 52.4, and 43.0 mg/g polymer, respectively, at 50 mmol/L, while 3.2 mg/g polymer for L-Tyr at 2 mmol/L especially at 37 C. Because of low solubility of L-Tyr in water (solubility at 25 C, 0.46 g/kg H2O) (Lide, 2003), the adsorption values at higher concentrations were not ð2Þ and hence a plot of lg q versus lg C should be linear with slope equal to 1/n and intercept equal to lg k. The adsorption isotherms of L-Trp, L-Phe, and L-Tyr adsorbed on the -CyD polymer are presented in Figure 6, and the corresponding Freundlich adsorption isotherms are shown in Figure 7. The empirical constants k, 1/n and correlation coefficients obtained from Figure 7 are listed in Table 2. Plots in Figure 7 with correlation coefficients varying from 0.9901 to 0.9985 indicate that the adsorption of AAA on the -CyD polymer was consistent with the Freundlich isotherm equation, which indicates that the adsorption mechanism is a monolayer adsorption. Adsorption of mixed amino acids In order to study the adsorption of mixed amino acids on the -CyD polymer, HPLC method was established for quantitative determination of amino acids as described in experimental section. Amino acid analysis is usually carried out by ion exchange chromatography followed by postcolumn derivatization with ninhydrin or fluorescent reagents (Benson and Hare, 1975). These methods are limited by the fact that the equipment is elaborate and dedicated to the sole purpose of amino acid analysis. An alternate approach involves the general method of precolumn derivatization Table 1. The physical properties of amino acids Amino acids Volume (Å3)a Residue non-polar surface area (Å2)bc Trp Phe Tyr Val Ile Leu 227.8 189.9 193.6 140.0 166.7 166.7 37 þ 199 39 þ 155 38 þ 116 135 155 164 Estimated hydrophobic effect for residue burial (kcal/mol)b 4.11 3.46 2.81 3.38 3.88 4.10 Estimated hydrophobic effect for side chain burial (kcal/mol)bd 2.9 2.3 1.6 2.2 2.7 2.9 a Values from Zamyatin, 1972. Values from Karplus, 1997. All surfaces associated with main- and side-chain carbon atoms were included except for amide, carboxylate and guanidino carbons. For aromatic side chains, the aliphatic and aromatic surface areas are reported separately. d The values are obtained from the previous column by substracting the value for Gly (1.18 kcal/mol) from each residue. b c Copyright # 2005 John Wiley & Sons, Ltd. J. Mol. Recognit. 2006; 19: 39–48 44 S. TANG ET AL. (a) 60 Adsorption capacity (mg/g) Adsorption efficiency (%) 35 30 25 20 40 20 0 0 10 15 20 30 40 50 40 50 Concentration (mmol/L) 0.0 0.1 0.2 0.3 0.4 Concentration (mmol/L) Figure 5. Effect of concentration on adsorption of L-Trp (~), LPhe (*) and L-Tyr (&). Experimental conditions: L-Trp, L-Phe or LTyr at a concentration of 0.02–0.40 mmol/L incubated with 100 mg of the -CyD polymer for 1 h. Other conditions are the same as in Figure 4. with chromophores followed by separation of the amino acid derivatives. The derivatization procedures in current usage include o-phthalaldehyde (Jones et al., 1981), dansylation (De Jong et al., 1982) and dabsylation (Chang et al., 1981). However, these types of amino acid modification procedures may lead to problems, such as the inability of the reagent to react with all amino acids, fluorescence quenching by the buffers used for amino acid elution, poor detection of some amino acids and instability of certain amino acid derivatives. A dinitrophenyl derivative, which is frequently overlooked, is formed by reaction with DNFB (Robert et al., 1988). These derivatives have extremely high extinction coefficients and are also very stable. Dinitrophenylation thus seems to be well suited for precolumn derivatization of amino acids. The representative chromatogram of amino acid derivatives with DNFB as shown in Figure 8 clearly indicates that baseline resolution of 2,4-DNP-amino acids can be achieved in just over 45 min. Dinitrophenylation is quantitative as demonstrated by the calibration equation with correlation coefficient in the range of 0.9991–0.9998 as shown in Table 3. This establishes that for a given sample of derivatized amino acid the precise number of moles can be calculated from the integrated peak areas. Table 3 gives the analytical parameters of this method. In the range of 0.01–0.40 mmol/L good linearities were found for L- Trp, L-Phe and L-Tyr, while 0.005–0.20 mmol/L for L-Val, L-Ile and L-Leu. The RSDs of the peak areas for the calibration curves were between 1.53 and 4.71% for the tested amino acids. Figure 9 gives the adsorption rate of mixed AAA at 37 C. Solutions at a concentration of 0.20 mmol/L for each AAA were used. The characteristics of adsorption rates of mixed AAA were almost the same as single AAA as shown in Figure 4, that is, the adsorption efficiencies of AAA were in the order of L-Trp > L-Phe > L-Tyr and the adsorption equilibrium was gradually achieved in about 45 min. The -CyD polymer was incubated with mixed AAA samples containing different amounts of AAA (0.02– Copyright # 2005 John Wiley & Sons, Ltd. 40 20 0 0 10 20 30 Concentration (mmol/L) (c) 4 Adsorption capacity (mg/g) 10 Adsorption capacity (mg/g) (b) 3 2 1 0 0.0 0.5 1.0 1.5 2.0 Concentration (mmol/L) Figure 6. Adsorption isotherms of AAA at 25 C (^), 37 C (*) and 50 C (!). (a) L-Trp; (b) L-Phe; (c) L-Tyr. Experimental conditions: L-Trp, L-Phe or L-Tyr at various concentrations incubated with the -CyD polymer for 1 h. Other conditions are the same as in Figure 5. 0.40 mmol/L each) at 37 C for 1 h. Figure 10 shows the adsorption efficiencies with respect to initial concentrations of mixed AAA. Compared with single amino acid as shown in Figure 5, there are similarities and also differences. The adsorption efficiencies of AAA remained in the same order as L-Trp > L-Phe > L-Tyr, but were lower than when loaded individually under the same conditions. For L-Trp, the adsorption behaviors were almost the same in mixed and single AAA solutions. For L-Phe, the adsorption efficiency declined slightly (3.33%) in the concentration range studied in mixed AAA solutions, while dramatically (14.27%) in the range of 0.02–0.10 mmol/L and slightly (2.02%) in the range of 0.10–0.40 mmol/L in single AAA solutions. For L-Tyr, the adsorption efficiency decreased slightly (3.91%, including 3.05% from 0.10 to 0.20 mmol/L) when concentration increased from 0.02 to 0.40 mmol/L in mixed AAA solutions, and more importantly (15.29%) before and slightly (4.49%) after 0.10 mmol/L in single AAA solutions. The structure of AAA is one of the key factors that affects J. Mol. Recognit. 2006; 19: 39–48 ADSORPTION OF AROMATIC AMINO ACIDS 45 Table 2. The values of different parameters of the Freundlich adsorption isotherms for L-Trp, L-Phe and L-Tyr at 25, 37, and 50 C Amino acids L-Trp L-Phe L-Tyr Temperature ( C) k 1/n Correlation coefficient (r) 25 37 50 25 37 50 25 37 50 4.00 3.29 3.16 3.92 3.78 3.88 1.51 1.47 1.32 0.69 0.72 0.64 0.67 0.65 0.62 1.07 1.04 0.94 0.9952 0.9970 0.9961 0.9946 0.9970 0.9901 0.9938 0.9979 0.9985 Figure 8. Chromatogram for HPLC separation of DNFB derivatives of amino acids. Column, Kromasil C18, 5 mm, 250 4.6 mm I.D., maintained at 33 C. Mobile phase A, acetonitrile–water (50:50, v/v) and B, 100 mmol/L sodium acetate buffer (pH 6.4), containing 1% DMF (v/v). Gradient elution was carried out as described in experimental section at a flow rate of 1 ml/min. Detection: UV at 360 nm. Peak identification: (1) DNP-OH; (2) Val; (3) Ile; (4) Leu; (5) Trp; (6) Phe; (7) (DNP)2O; (8) Tyr. competed for the interaction sites of the -CyD polymer, resulting in lower adsorption efficiencies than when loaded individually. With the strongest interactions with the -CyD polymer, the adsorption behavior of L-Trp was slightly affected by the presence of the other AAA, whose interactions with the -CyD polymer were weaker than L-Trp, showing almost the same behavior as in single AAA solutions. But for L-Phe and L-Tyr, which interact weakly with the -CyD polymer, their adsorption behaviors were greatly influenced by the presence of L-Trp in mixed AAA solutions. In the case of L-Tyr, the presence of L-Phe may also affect the adsorption behavior. Certainly, the existence of AAA with weaker interactions with the CyD polymer also has some limited influence on the L-Tyr Figure 7. The Freundlich isotherms of AAA adsorbed on the CyD polymer at 25 C (^), 37 C (*) and 50 C (!). (a) L-Trp; (b) LPhe; (c) L-Tyr. the stability of the complex, which is decisive for the strength and positioning of the AAA–polymer interactions. As described above, the interaction strength between AAA and the -CyD polymer was in the order of L-Trp > L-Phe > L-Tyr. When loaded together, L-Trp, L-Phe and Copyright # 2005 John Wiley & Sons, Ltd. J. Mol. Recognit. 2006; 19: 39–48 46 S. TANG ET AL. Table 3. Analytical parameters for the determination of derivatized amino acids by HPLC Amino acids Calibration range (mmol/L) Calibration equationa Correlation coefficient(r) 0.01–0.40 0.01–0.40 0.01–0.40 0.005–0.20 0.005–0.20 0.005–0.20 Y ¼ (1.90E-6 2.41E-8) X þ (0.00235 0.00166) Y ¼ (1.67E-6 2.38E-8) X þ (0.00306 0.00185) Y ¼ (1.69E-6 3.34E-8) X þ (0.00290 0.00257) Y ¼ (1.61E-6 3.17E-8) X þ (-0.0137 0.00155) Y ¼ (1.56E-6 2.72E-8) X þ (0.00104 0.00114) Y ¼ (1.60E-6 3.90E-8) X þ (0.00185 0.00158) 0.9998 0.9997 0.9991 0.9994 0.9995 0.9991 L-Trp L-Phe L-Tyr L-Val L-Ile L-leu a b Repeatability (RSD,%)b 2.63 1.53 1.67 2.11 2.14 4.71 Y, the concentration of amino acid before derivatization (mmol/L); X, the peak areas of amino acids. n ¼ 5. n ¼ 3. Adsorption efficiency (%) 25 20 15 10 0 10 20 30 40 50 60 Adsorption time (min) Figure 9. Adsorption rates of L-Trp (~), L-Phe (*) and L-Tyr (&) in mixed AAA solutions. Experimental conditions: 5 ml of 0.20 mmol/L L-Trp, L-Phe and L-Tyr incubated with the -CyD polymer. Other conditions are the same as in Figure 4. Adsorption efficiency (%) 35 30 25 adsorption behaviors of AAA with stronger interactions with the -CyD polymer. So, the different AAA adsorption behaviors in mixed and single AAA solutions are the consequence of different interaction strength between AAA and the -CyD polymer, which is determined by the complementarity of the structure and hydrophobicity of AAA molecules to the cavity of the -CyD polymer. To clarify if the -CyD polymer was a specific sorbent for AAA adsorption, the effect of BCAA on AAA adsorption was investigated. The -CyD polymer was incubated with mixed AAA-BCAA solution at 37 C for 1 h. The pre- and post-samples were collected, derivatized with DNFB and then analyzed by HPLC as described in experimental section. Table 4 gives the results for adsorption of mixed amino acids. It can be seen that AAA were preferentially adsorbed on the -CyD polymer with adsorption efficiencies in the range of 10.40–23.88%, while those of BCAA were lower than 2%. The adsorption efficiencies of BCAA were in the order of L-Leu > L-Ile > L-Val, which is also consistent with the volume and hydrophobicity of BCAA molecules shown in Table 1. The structure of molecules is the key factor for the big difference between AAA and BCAA in adsorption efficiencies, by influencing the formation and strength of interactions between amino acid molecules and the polymer. Compared with BCAA molecules, AAA molecules contain a benzene ring, making them more suitable to the cavity of the -CyD polymer. Since the BCAA molecule is too small to allow it to fit well into the cavity of the polymer, it leads to low adsorption efficiency. The ratio of the number (as mmol) of AAA molecules to BCAA molecules adsorbed 20 Table 4. Adsorption of AAA and BCAA on the b-CyD polymer in the mixed solution 15 Amino Before adsorption acids (mmol/L) 10 L-Val 0.0 0.1 0.2 0.3 0.4 Concentration (mmol/L) Figure 10. Effect of concentration on adsorption of L-Trp (~), LPhe (*) and L-Tyr (&) in mixed AAA solutions. Experimental conditions: the concentration of mixed AAA varied from 0.02 to 0.40 mmol/L for each AAA. Other conditions are the same as in Figure 5. Copyright # 2005 John Wiley & Sons, Ltd. L-Ile L-Leu L-Trp L-Phe L-Tyr 0.103 0.096 0.101 0.201 0.206 0.202 After adsorption (mmol/L) 0.102 0.095 0.099 0.153 0.164 0.181 Adsorption efficiency (%) 0.97 1.04 1.98 23.88 20.39 10.40 100 mg of the -CyD polymer was incubated with 5 ml of mixed AAABCAA at 37 C for 1 h. J. Mol. Recognit. 2006; 19: 39–48 ADSORPTION OF AROMATIC AMINO ACIDS on the sorbent was about 27 in this case. It indicates that the -CyD polymer was a specific sorbent for AAA adsorption, while being non-specific for BCAA. CONCLUSION In this article, we studied the adsorption behavior of AAA on the -CyD polymer in phosphate buffer. High adsorption rates were observed at the beginning and the adsorption equilibrium was then gradually achieved in about 45 min. The adsorption of AAA decreased with the increase of initial concentration and also temperature. Under the same conditions, the adsorption efficiencies of AAA were in the order of L-Trp > L-Phe > L-Tyr. Much higher values up to 52.4, 43.0 (at 50 mmol/L), and 3.2 (at 2 mmol/L) mg/g polymer for L-Trp, L-Phe and L-Tyr, respectively, were obtained at 37 C. The ratio (as mmol) of AAA molecules 47 to BCAA molecules adsorbed on the sorbent was about 27, indicating that AAA molecules were preferentially adsorbed on the -CyD polymer. It was found that the structure and hydrophobicity of amino acid molecules were responsible for the difference in adsorption, by influencing the strength of interactions between amino acid molecule and the polymer. These results together with those obtained previously for adsorption of bilirubin (Zheng et al., 2004) indicate that this inexpensive sorbent may be an important alternative to existing sorbents used in the therapy of liver patients showing elevated bilirubin and AAA. Acknowledgments Financial supports from the Natural Science Foundation of China (No. 20475054) and the Knowledge Innovation Program of DICP to Prof. Hanfa Zou are gratefully acknowledged. REFERENCES Akiyama T, Hishiya T, Asanuma H, Komiyama M. 2001. Molecular imprinting of cyclodextrin on silica-gel support for the stationary phase of high-performanceliquid-chromatography. J. Incl. Phenom. Macro. 41: 149–153. Andersson HS, Nicholls IA. 1997. Molecular imprinting: recent innovations in synthetic polymer receptor and enzyme mimics. Recent Res. Develop. Pure Appl. Chem. 1: 133–157. Asanuma H, Akiyama T, Kajiya K, Hishiya T, Komiyama M. 2001. Molecular imprinting of cyclodextrin in water for the recognition of nanometer-scaled guests. Anal. Chim. Acta. 435: 25–33. Asanuma H, Hishiya T, Komiyama M. 2004. Efficient separation of hydrophobic molecules by molecularly imprinted cyclodextrin polymers. J. Incl. Phenom. Macro. 50: 51–55. Asanuma H, Kajiya K, Hishiya T, Komiyama M. 1999. Molecular imprinting of cyclodextrin in water for the recognition of peptides. Chem. Lett. 665–666. Asanuma H, Kakazu M, Shibata M, Hishiya T, Komiyama M. 1997. Molecularly imprinted polymer of -cyclodextrin for the efficient recognition of cholesterol. Chem. Commun. 1971–1972. Asanuma H, Kakazu M, Shibata M, Hishiya T, Komiyama M. 1998. Synthesis of molecularly imprinted polymer of -cyclodextrin for the efficient recognition of cholesterol. Supramol. Sci. 5: 417–421. Benson JR, Hare PE. 1975. Ortho-phthalaldehyde—fluorogenic detection of primary amines in picomole range—comparison with fluorescamine and ninhydrin. Proc. Nat. Acad. Sci. USA 72: 619–622. Breslow R, Halfon S, Zhang BL. 1995. Molecular recognition by cyclodextrin dimers. Tetrahedron. 51: 377–388. Chang JY, Knecht R, Braun DG. 1981. Amino-acid-analysis at the picomole level—application to the c-terminal sequenceanalysis of polypeptides. Biochem. J. 199: 547–555. De Jong C, Hughes GJ, Van Wieringen E, Wilson KJ. 1982. Amino acid analyses by high-performance liquid chromatography: an evaluation of the usefulness of precolumn Dns derivatization. J. Chromatogr. 241: 345–359. Dı́ez S, Leitão A, Ferreira L, Rodrigues A. 1998. Adsorption of phenylalanine onto polymeric resins: equilibrium, kinetics and operation of a parametric pumping unit. Sep. Purif. Technol. 13: 25–35. Copyright # 2005 John Wiley & Sons, Ltd. Doulia D, Rigas F, Gimouhopoulos C. 2001. Removal of amino acids from water by adsorption on polystyrene resins. J. Chem. Technol. Biotechnol. 76: 83–89. Grzegorczyck DS, Carta G. 1996a. Adsorption of aminoacids on porous polymeric adsorbents I-equilibrium. Chem. Eng. Sci. 51: 807–817. Grzegorczyck DS, Carta G. 1996b. Adsorption of aminoacids on porous polymeric adsorbents II. Intraparticle mass transport. Chem. Eng. Sci. 51: 819–826. Hishiya T, Akiyama T, Asanuma H, Komiyama M. 2002. Molecular imprinting of cyclodextrins leading to synthetic antibodies. J. Incl. Phenom. Macro. 44: 365–367. Hishiya T, Shibata M, Kakazu M, Asanuma H, Komiyama M. 1999. Molecularly imprinted cyclodextrins as selective receptors for steroids. Macromolecules 32: 2265–2269. Huang XD, Zou HF, Chen XM, Luo QZ, Kong L. 2003. Molecularly imprinted monolithic stationary phases for liquid chromatographic separation of enantiomers and diastereomers. J. Chromatogr. A 984: 273–282. Huang XD, Qin F, Chen XM, Liu YQ, Zou HF. 2004. Short columns with molecularly imprinted monolithic stationary phases for rapid separation of diastereomers and enantiomers. J. Chromatogr. B 804: 13–18. Jones BN, Paabo S, Stein S. 1981. Amino-acid-analysis and enzymatic sequence determination of peptides by an improved ortho-phthaldialdehyde pre-column labeling procedure. J. Liq. Chromatogr. 4: 565–586. Karplus PA. 1997. Hydrophobicity regained. Protein Sci. 6: 1302– 1307. Leung DK, Atkins JH, Breslow R. 2001. Synthesis and binding properties of cyclodextrin trimers. Tetrahedron Lett. 42: 6255–6258. Lide DR (ed.). 2003. CRC Handbook of Chemistry and Physics (84th edn.) CRC Press LLC: Boca Raton, London, New York, Washington, D.C.; 7–1. Nakamura S, Takeuchi H, Nakamura H, Maeda M, Takagi M. 1999. Molecularly imprinted resin for recognition of amino acids using self-assembly at o/w interface. Trans. Mater. Res. Soc. Jpn. 24: 453–456. Piletsky SA, Andersson HS, Nicholls IA. 1998. The rational use of hydrophobic effect-based recognition in molecularly imprinted polymers. J. Mol Recogn. 11: 94–97. Piletsky SA, Andersson HS, Nicholls IA. 1999. Combined hydrophobic and electrostatic interaction-based recognition in J. Mol. Recognit. 2006; 19: 39–48 48 S. TANG ET AL. molecularly imprinted polymers. Macromolecules. 32: 633–636. Robert CM, Gerhard EG. 1988. Amino acid analysis by dinitrophenylation and reverse-phase high-pressure liquid chromatography. Anal. Biochem. 170: 220–227. Sreenivasan K. 1998. Synthesis and evaluation of a beta cyclodextrin-based molecularly imprinted copolymer. J. Appl. Polym. Sci. 70: 15–18. Sreenivasan K, Sivakumar R. 1998. Evaluation of a methyl betacyclodextrin based molecularly imprinted copolymer as novel absorbent for steroids. Macromol.—New Front., Proc. IUPAC Int. Symp. Adv. Polym. Sci. Technol. 2: 594–596. Vidyasankar S, Ru M, Arnold FH. 1997. Molecularly imprinted ligand-exhange adsorbents for the chiral separation of underivatized amino acids. J. Chromatogr. A 775: 51–63. Copyright # 2005 John Wiley & Sons, Ltd. Yoshida M, Uezu K, Goto M, Furusaki S. 2000a. Surface imprinted polymers recognizing amino acid chirality. J. Appl. Polym. Sci. 78: 695–703. Yoshida M, Hatate Y, Uezu K, Goto M, Furusaki S. 2000b. Chiral-recognition polymer prepared by surface molecular imprinting technique. Collids Surf. A 169: 259–269. Zamyatin AA. 1972. Protein volume in solution. Prog. Biophys. Mol. Biol. 24: 107–123. Zheng CJ, Huang XD, Kong L, Li X, Zou HF. 2004. Cross-linked -cyclodextrin polymer used for bilirubin removal. Chin. J. Chromatogr. 22: 128–130. Zhong N, Byun H-S, Bittman R. 2001. Hydrophilic cholesterolbinding molecular imprinted polymers. Tetrahedron Lett. 42: 1839–1841. J. Mol. Recognit. 2006; 19: 39–48