- Email: [email protected]
Temperature-responsive stationary phase utilizing a polymer of proline derivative for hydrophobic interaction chromatography using an aqueous mobile phase
Journal of Chromatography A, 1106 (2006) 152–158
Temperature-responsive stationary phase utilizing a polymer of proline derivative for hydrophobic interaction chromatography using an aqueous mobile phase Hideko Kanazawa a,∗ , Eri Ayano a , Chikako Sakamoto a , Reiko Yoda a , Akihiko Kikuchi b , Teruo Okano b b
a Department of Physical Pharmaceutical Chemistry, Kyoritsu University of Pharmacy, 1-5-30 Shibakoen, Minato-ku, Tokyo 105-8512, Japan Institute of Advanced Biomedical Engineering and Science, Tokyo Women’s Medical University, 8-1 Kawadacho, Shinjyuku-ku, Tokyo 162-8666, Japan
Available online 10 October 2005
Abstract A new method of chromatography is proposed, utilizing a thermo-responsive polymer carrying an amino acid ester residue for the stationary phase of high-performance liquid chromatography (HPLC). We have been investigating the new concept of chromatography, a temperature-responsive chromatography, using temperature-responsive poly(N-isopropylacrylamide) (PNIPAAm)-modified surface for HPLC with a constant aqueous media as the mobile phase. In this study, we designed and synthesized thermo-responsive poly(acryloyl-l-proline methyl ester) and its copolymer with N-isopropylacrylamide (NIPAAm). Homopolymers of acryloyl-l-proline methyl ester and copolymer were prepared by the reaction of radical telomerization. These polymers underwent a reversible phase transition from water-soluble forms into aggregates by changing the temperature, similar to PNIPAAm. The surface properties and functions of stationary phases modified with poly(acryloyl-l-proline methyl ester) were controlled by the external temperature. In the chromatographic system, we separated steroids and amino acids with a variety of hydrophobicities using a sole aqueous mobile phase. In contrast to a PNIPAAm-modified surface, a poly(acryloyl-l-proline methyl ester)-modified surface showed a greater affinity for hydrophobic amino acids. © 2005 Elsevier B.V. All rights reserved. Keywords: Temperature-responsive polymer; HPLC; Hydrophobic interaction chromatography; Separation; Amino acids; Steroids; Poly(acryloyl-l-proline methyl ester); Poly(N-isopropylacrylamide)
1. Introduction Recently, HPLC techniques have become increasingly important for the separation of proteins without denaturation, which is often observed with the use of the reversed-phase mode. Many reversed-phase methods for proteins have employed C18 columns with mobile phases containing acetonitrile in lowpH buffers. Little attention has been paid to recovering biological activity. Acetonitrile almost always leads to the total denaturation of proteins. Similarly, acidic environments, including the standard trifluoroacetic acid-based buffers, are sometimes destructive to the activities of many enzymes. Thus, these conditions should be avoided in the separation of most proteins.
∗
Corresponding author. Tel.: +81 3 5400 2657; fax: +81 3 5400 1378. E-mail address: [email protected] (H. Kanazawa).
0021-9673/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2005.09.052
In recent years, polymers that demonstrate good solubility in aqueous solutions at low temperature, but separate from solution when the temperature is raised above the lower critical solution temperature (LCST), have received increased attention [1–3]. Polymers exhibiting LCST behavior have been investigated for such applications as drug delivery system [4–6] and cell culture substrates [7–11]. However, there have been few reports concerning the use of these temperature-responsive polymers in chromatographic separation [12,13]. In previous studies, we developed a new method of HPLC using packing materials modified with a temperature-responsive polymer, poly(N-isopropylacrylamide) (PNIPAAm) [14–21]. PNIPAAm is one of the extensively studied thermosensitive polymers. PNIPAAm exhibits a thermally reversible soluble/insoluble changes phase transition in aqueous solution at 32 ◦ C. The surface properties and functions of the stationary phases are controlled by the external temperature. In the chromatographic system utilizing the PNIPAAm-modified stationary
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
phase, elution of target substances is controlled only by a small change in the column temperatures. We also observed in these studies that the LCST is modulated by varying the monomer composition [22]. In general, the incorporation of hydrophobic comonomers leads to a lower LCST, and hydrophilic comonomers lead to a higher LCST [23]. It has been reported that copolymerizing NIPAAm with butyl methacrylate (BMA) as hydrophobic comonomers leads to a lower LCST [15]. These features also showed that certain polypeptide solutions demonstrate temperature-induced phase separation [24,25]. It was found that an increase in the hydrophobicity of polypeptides lowered the temperature of the phase transition and increased the heat of the transition. The interpretation is that the increase in endothermic heat required to drive the transition reflects the energy required to destructive the greater numbers of waters of hydrophobic hydration [26]. In this study, we designed and synthesized thermo-responsive polymer carrying amino acid ester residue as a side chain. These polymers undergo a reversible phase transition from watersoluble forms into aggregates similar to PNIPAAm. Here, we demonstrated how to separate steroids and amino acids with a variety of hydrophobicities using a temperature-responsive silicagel surface modified with polymers of acryloyl-l-proline methyl ester as packing materials. 2. Experimental 2.1. Materials N-Isopropylacrylamide (NIPAAm) was kindly provided by KOHJIN (Tokyo, Japan) and was purified by recrystallization from n-hexane and dried at 25 ◦ C in vacuo. land d-proline methyl ester hydrochlorides were purchased from Kokusan Chemical (Tokyo, Japan). 3-Mercaptopropionic acid (MPA; Wako Pure Chemicals, Tokyo, Japan) was distilled under reduced pressure and the fraction boiling at 95 ◦ C (5 mmHg) was used. 2,2 -Azobisisobutyronitrile
153
(AIBN), N,N-dimethylformamide (DMF), acryloyl chloride, dichloromethane, and ethyl acetate (EtOAc) were obtained from Wako Pure Chemicals (Tokyo, Japan) and purified by conventional methods. N,N -Dicyclohexylcarbodiimide (DCC) and Nhydroxysuccinimide (SuOH) were purchased from Wako Pure Chemicals. Aminopropyl silica (average diameter of 5 m, pore size 12 nm) was purchased from Nishio Kogyo, (Tokyo, Japan). Sulfo-succinimidyl-4-O-(4,4 -dimethoxytrityl) butyrate (s-SDTB) was obtained from Pierce (Rockford, IL, USA). Cortisone acetate, hydrocortisone, hydrocortisone acetate, amino acids phenylthiohydantoins (PTH-amino acids) were from Wako Pure Chemicals (Tokyo, Japan) and other steroids were from Sigma Chemicals (St. Louis, MO, USA). Milli-Q grade water was used for preparation of sample solutions. Other reagents and solvents were commercially obtained and used without further purification. 2.2. Synthesis of acryloyl-l-proline (or d-proline) methyl ester monomer and polymerization procedure N-Acryloyl derivative of proline methyl ester was synthesized by the reaction of proline methyl ester with acryloyl chloride in dichloromethane. l- or d-proline methyl ester hydrochlorides were neutralized by NaOH aqueous solution. Proline methyl ester (0.22 mol) was dissolved in 270 mL of dichloromethane. Then acryloyl chloride (0.242 mol) and triethylamine (0.264 mol) were added to the reaction mixture with stirring below −2 ◦ C. After the reaction proceeded for 2 h with stirring, the reaction mixture was extracted by 200 mL of HC1 aqueous solution (HCl:water = 1:4) and 50 mL of water. Then the dichloromethane layer was dried by 15 g of sodium sulfate and under reduced pressure to remove the dichloromethane. The resulting product was evaporated in vacuo by a rotary evaporator. The yield was 68.4%. Poly(acryloyl-l-proline methyl ester) and copolymers with NIPAAm were prepared by radical polymerization in DMF, as shown in the reaction scheme (Fig. 1). Typical procedures
Fig. 1. Synthesis of poly(NIPAAm-co-acryloyl-l-proline methyl ester) copolymer.
154
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
were as follows: acryloyl-l-proline methyl ester (40 mmol) and NIPAAm (40 mmol) were dissolved in 50 mL of DMF. AIBN (0.88 mmol), and MPA (6.25 mmol) were used as an initiator, and chain transfer agent, respectively. The reaction mixture was degassed by subjecting to freeze-thaw cycles and the ampoules containing the mixture were sealed under reduced pressure. The reaction proceeded at 70 ◦ C for 2 h. After the reaction solution was concentrated by evaporation, the reaction mixture was poured into 50 mLof diethyl ether to precipitate the polymers. The polymer was further purified by repeated precipitation from 10 mL of acetone into 500 mL of diethyl ether. The molecular weight of the polymers was determined by end group titration with 0.01 M NaOH using phenolphthalein as an indicator. Specific rotations at the sodium D line, [α]D , and near-UV (220–420 nm) circular dichroism (CD) spectrum were obtained by a DIP-1000 polarimeter and a CD-1595 circular dichroism detector (JASCO Corporation, Tokyo, Japan), respectively. 2.3. Transmittance measurements The LCSTs of the polymers were determined by measuring the optical transmittance of polymer aqueous solutions. The optical transmittance of polymer solutions (5 mg/mL) was measured at 500 nm at various temperatures using a UV–vis spectrophotometer (U-3000, Hitachi, and Tokyo, Japan). The temperature of observation cell was controlled with a LAUDARCS20D water bath with a deviation of 0.02 ◦ C. The LCST values for each polymer were defined as the temperature where 50% optical transmittance of polymer aqueous solutions was observed.
(0.077 mg/mL). Standard solutions of steroids were prepared with dexamethasone (DX; 0.061 mg/mL), hydrocortisone (HC; 0.101 mg/mL), hydrocortisone acetate (HCA; 0.007 mg/mL), prednisolone (PRE; 0.166 mg/mL), and testosterone (TES; 0.027 mg/mL). The log P values of steroids were calculated by the CAChe system (FUJITSU, Japan). 3. Results and discussion 3.1. Characterization of thermo-responsive polymer carrying amino acids residue Fig. 2 shows the temperature dependence for the optical transmittance of polymer solutions. PNIPAAm and poly(acryloyll-proline methyl ester) exhibit an LCST in water. Both of poly(acryloyl-l-proline methyl ester) and its copolymers with NIPAAm also exhibit thermally reversible soluble-insoluble changes in response to temperature changes. The physicochemical data for these polymers are given in Table 1. The polymer of acryloyl-d-proline methyl ester was also synthesized by the same way as that of the l-proline derivative to compare with the property of the polymer of the l-proline derivative. The chiroptical properties of these polymers were evaluated using polarimeter (optical rotation) and CD spectroscopy. Specific
2.4. Modification of aminopropyl silica with NIPAAm polymer Modification of aminopropyl silica with NIPAAm copolymer by activated ester-amine coupling was the same procedure as described in our previous report [27]. The surface coverage degrees of three columns were calculated by quantification of residual amount of amino group on the supports. The amount of amino group on silica supports was determined by spectrophotometric method using s-SDTB [27]. Introduction rate of polymers on the supports were 77–81%. 2.5. Temperature-responsive chromatography A polymer-grafted silica support was packed into a stainlesssteel column (length: 150 mm × 4.6 mm I.D.). The column was connected to an HPLC system (HITACHI Model L-6200 intelligent pump; L-4000 UV-monitor, D-2500 data processor). The column oven was a Shodex AO-30 (Showa Denko, Tokyo, Japan). Milli-Q grade water was used as the mobile phase. The elution behaviors of the samples were recorded at a flowrate of 1 mL/min at various temperatures. Standard solutions of PTH-amino acids were prepared with Arg (initial concentration 1.000 mg/mL); Asn (0.404 mg/mL); Asp (1.084 mg/mL); Gin (0.124 mg/mL); Glu (0.471 mg/mL); His (1.000 mg/mL); Leu (0.014 mg/mL); Met (0.063 mg/mL); Phe (0.038 mg/mL); Pro (0.038 mg/mL); Thr (0.019 mg/mL); Trp (0.019 mg/mL); Tyr
Fig. 2. Temperature dependence of the optical transmittance of polymer solutions at 500 nm: () poly(acryloyl-l-proline methyl ester); () poly(NIPAAmco-acryloyl-l-proline methyl ester) (1:1); () PNIPAAm.
Table 1 Characterization of poly(NIPAAm-co-acryloyl-proline methyl ester) Polymer unit
LCST (◦ C)a
Acryloyl-l-proline methyl ester NIPAAm-co-acryloyl-l-pro line methyl ester NIPAAm-co-acryloyl-d-proline methyl ester NIPAAm
20.7 24.0
−107 −41
3200 7500
23.5
43
4300
a b c
31.5
b [α]25 D
–
M.W.c
4500
Determined by transmittance measurement at 500 nm. Mesured by polarimeter at 25 ◦ C (c = 0.5% (w/v) in methanol). Number-averaged molecular weight determined by end group titration.
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
155
Fig. 4. Chromatograms of a mixture of benzene and steroids on a poly(NIPAAmco-acryloyl-l-proline methyl ester)-modified silica column at (a) 10 ◦ C and (b) 40 ◦ C. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm; injection volume, 20 L. The peaks are: (1) BEN; (2) HC; (3) COR; (4) PRE; (5) DX; (6) HCA; (7) TES.
Fig. 3. CD and UV spectra of (1) poly(NIPAAm-co-acryloyl-l-proline methyl ester) and (2) poly(NIPAAm-co-acryloyl-d-proline methyl ester). The mobile phase is a 50% methanol water solution; injection volume, 10 L.
rotations at 25 ◦ C, [α]25 D , and CD spectra for these two polymers were obtained. The [α]D value was positive and negative for a polymer of l-proline and that of d-proline, respectively. The circular dichroism spectrum can be used to characterize the conformation of the chiral polymer in solution [28,29]. The signs of circular dichroism at all wavelengths for these two polymers are opposite, as shown in Fig. 3. The results seemed to indicate that these polymers maintained the optical active properties of its monomer used. 3.2. Temperature-responsive chromatographic separation of steroids Using a column packed with poly(acryloyl-l-proline methyl ester)-modified silica in a HPLC system, the separation of steroids was carried out by changing the column temperature. Milli-Q water was used as the sole mobile phase. Fig. 4 shows typical chromatograms of steroids on a copolymer of NIPAAm and acryloyl-l-proline methyl ester grafted silica column. The temperature-dependent resolution of steroids was achieved using only water as a mobile phase. Although the peaks of steroids are not properly resolved at 10 ◦ C, they were well resolved at 40 ◦ C. With increasing the temperature, increased interactions between the solutes and the poly(NIPAAm-coacryloyl-l-proline methyl ester)-grafted surfaces of the stationary phases were observed. We previously demonstrated that a hydrophobic interaction between steroids and PNIPAAmgrafted surfaces was readily modulated by the temperature [14–17,30]. A PNIPAAm-modified surface showed remarkable surface wettability changes by changing the temperature [30].
Therefore, a hydrophobic interaction with hydrophobic steroids is greatly affected by the temperature on this surface. The temperature-dependent elution behavior of steroids was examined on a PNIPAAm-modified silica column, a poly(NIPAAmco-acryloyl-l-proline methyl ester)-modified silica column, and a poly(acryloyl-l-proline methyl ester)-modified silica column. Fig. 5 shows the effect of temperature on the retention of steroids on three types of thermo-responsive polymer-modified silica columns. The elution profiles of steroids on these columns are opposite from those observed using conventional C18 column, in which the elution time was shortened with the temperature. The results indicated that the copolymer-modified surface of the stationary phase exhibited temperature-controlled hydrophilic–hydrophobic changes. On those three thermoresponsive polymer-modified stationary phases, the hydrophobicity of the surfaces reduced at a lower temperature and increased at an elevated temperature. A substance’s hydrophobicity is usually defined as the logarithm of its partition coefficients in the n-octanol/water system, log P values. The log P values for each steroid are 1.61 for hydrocortisone, 1.62 for prednisolone, 1.83 for dexamethasone, 2.30 for hydrocortisone acetate, and 3.32 for testosterone. The more hydrophobic steroids showed longer retention times on the poly(acryloyll-proline methyl ester)-modified column. These results indicate that a stronger hydrophobic interaction is the primary driving force for the retention of steroids on poly(acryloyl-lproline methyl ester)-modified matrices, similar to the case of PNIPAAm-modified matrices. The elution orders of steroids were the same on these three columns at a higher temperature than LCST. Interestingly, there was a change of elution order between dexamethasone and hydrocortisone acetate around at the LCST on the poly(acryloyl-l-proline methyl ester)-modified column. This phenomenon was also observed on copolymer of NIPAAm with acryloyl-l-proline methyl ester-modified col-
156
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
Fig. 5. Effect of acryloyl-l-proline methyl ester introduction on the retention times of steroids on (a) a PNIPAAm-modified silica column; (b) a poly(NIPAAmco-acryloyl-l-proline methyl ester)-modified silica column and (c) a poly(acryloyl-l-proline methyl ester)-modified silica column: () HC; (䊉) PRE; (♦) DX; () HCA; () TES. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm; injection volume, 5 L.
umn. Therefore, it may be due to the difference in side chain structure that attributed to the interaction with two steroids. 3.3. Temperature-responsive chromatographic separation of PTH-amino acids We previously demonstrated the temperature-responsive chromatographic separation of PTH-amino acids using aqueous media as the mobile phase on PNIPAAm grafted silica columns [21]. Fig. 6 shows typical chromatograms of a mixture of PTHamino acids at below and above the LCST on poly(acryloyll-proline methyl ester)-modified silica columns. Milli-Q water was used as the sole mobile phase. PTH-amino acids were dissolved in Milli-Q water, and a sample mixture solution was injected into a poly(acryloyl-l-proline methyl ester)-modified column. As shown in Fig. 6, the peaks were not properly resolved at 5 ◦ C, which was lower than the LCST. As the column temperature was raised to 40 ◦ C, the nine PTH-amino acids were well
resolved. Fig. 7 shows the effect of temperature on the retention of PTH-amino acids on three types of thermo-responsive polymer-modified silica columns. The interaction between PTHamino acids and these temperature-responsive surfaces becomes stronger at elevated temperature. In a non-polar hydrophobic amino acid, such as Tyr, Phe and Trp, the retention times were retarded with increasing the temperature on the poly(acryloyl-lproline methyl ester)-modified column. However, the retention times of the amino acids, such as Cys, Gly and Ala, were slightly changed. In the separation of steroids, the retention times have a linear relationship with the log P values of steroids having different hydrophobicities. As can also be seen in Fig. 7, PTHamino acids eluted with the order of their hydrophobicity, and the more hydrophobic one showed a longer retention time at higher temperature. This suggests that the hydrophobic interaction is the primary driving force for the separation of PTH-amino acids with poly(acryloyl-l-proline methyl ester)-modified matrices.
Fig. 6. Chromatograms of PTH-amino acids on a poly(acryloyl-l-proline methyl ester)-modified silica column at (a) 5 ◦ C and (b) 40 ◦ C. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm. The peaks are: (1) Gly; (2) Ala; (3) Pro; (4) Val; (5) Met; (6) Leu; (7) Phe; (8) Tyr; (9) Trp.
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
157
Fig. 7. Effect of l-proline derivative introduction on the retention times of PTH amino acids on (a) a PNIPAAm-modified silica column; (b) a poly(NIPAAm-coacryloyl-l-proline methyl ester)-modified silica column and (c) a poly(acryloyl-l-proline methyl ester)-modified silica column: (×) Cys; () Gly; +, Ala; () Pro; () Met; () Thr; (♦) Leu; (䊉) Phe; () Tyr; () Tip. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm; injection volume, 5 L.
As shown in Fig. 5, an alteration of the retention times of steroids in response to the temperature on a poly(acryloyll-proline methyl ester)-modified column is less than that on a PNIPAAm-modified and a poly(NIPAAm-co-acryloyl-lproline methyl ester)-modified column. However, as shown in Fig. 7, the retention of PTH-amino acids was dramatically increased on a poly(acryloyl-l-proline methyl ester)-modified column in contrast to that on a homogeneous PNIPAAmmodified silica column. Fig. 8 shows a comparison of the separation selectivity in the aqueous mobile phase on PNIPAAmmodified and poly(acryloyl-l-proline methyl ester)-modified packing materials. The poly(acryloyl-l-proline methyl ester)modified stationary phase showed greater affinity for hydrophobic aromatic amino acids, such as Tyr, Trp and Phe, than the PNIPAAm-modified surface. These observations would be
caused by the interaction between the hydrophobic PTH-amino acids and proline moiety of the polymer chain on a grafted silica surface. Consequently, it is possible to determine the optimum separation conditions by setting an appropriate column temperature without modifying the eluent composition. In such separations, retention on the supports may also involve hydrophobic interaction between protein and polymer chains. The polypeptide and protein separations on temperature-responsive chromatography are currently in progress in our laboratory. 4. Conclusion The use of poly(acryloyl-l-proline methyl ester) as a silica surface modifier for temperature-responsive chromatography increased the selectivity and retention of PTH-amino acids. This study is the first report of the application of the polymer of the amino acid derivative for separation. By altering the copolymer composition we could control the LCST and the hydrophobicity of the surface of the stationary phases, and thus the elution time of the temperature-responsive chromatography. We succeeded to demonstrate hydrophobic interaction chromatography-like separation selectivity of the PTH-amino acid by simply changing the column temperature with pure water as a mobile phase. The ability of the proposed thermo-responsive polymer-modified stationary phase to separate the solutes without the use of an organic solvent is advantageous from the point of view of keeping the biological activity, environmental reasons, and mobile-phase reagent cost. We are currently performing studies to utilize a thermo-responsive polymer carrying amino acid ester residue for the chiral stationary phase. Acknowledgment
Fig. 8. Comparison of the retention times of PTH-amino acids on a poly(acryloyl-l-proline methyl ester)-modified silica column and a PNIPAAmmodified silica column at 40 ◦ C. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm.
This work was supported in part by a Grant-in-Aid for Scientific Research No. 15590049 from the Ministry of Education, Science, Sports, and Culture of Japan.
158
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
References [1] M. Heskins, J.E. Guillet, E. James, J. Macromol. Sci. Chem. A 2 (1968) 1441. [2] Y.H. Bae, T. Okano, S.W. Kim, J. Polym. Sci. Polym. Phys. 28 (1990) 923. [3] T. Tanaka, I. Nishio, S. Sun, S. Ueno-Nishio, Science 218 (1981) 467. [4] A. Chilkoti, M.R. Dreher, D.E. Meyer, D. Raucher, Adv. Drug Deliv. Rev. 54 (2002) 613. [5] R. Yoshida, K. Sakai, T. Okano, Y. Sakurai, Y.H. Bae, S.W. Kim, J. Biomater. Sci. Polym. Ed. 3 (1991) 155. [6] R. Yoshida, K. Sakai, T. Okano, Y. Sakurai, J. Biomater. Sci. Polym. Ed. 3 (1991) 243. [7] N. Yamada, T. Okano, H. Sakai, F. Karikusa, Y. Sawasaki, Y. Sakurai, Makromol. Chem. 11 (1990) 571. [8] T. Okano, N. Yamada, H. Sakai, Y. Sakurai, J. Biomed. Mater. Res. 27 (1993) 1243. [9] T. Okano, N. Yamada, M. Okumura, H. Sakai, Y. Sakurai, Biomaterials 16 (1995) 297. [10] M. Yamato, M. Utsumi, A. Kushida, C. Konno, A. Kikuchi, T. Okano, Tissue Eng. 7 (2001) 473. [11] T. Shimizu, M. Yamato, A. Kikuchi, T. Okano, Tissue Eng. 7 (2001) 141. [12] M. Gewehr, K. Nakamura, N. Ise, H. Kitano, Macromol. Chem. 193 (1992) 249. [13] K. Hosoya, K. Kimata, T. Araki, N. Tanaka, J.M.J. Frechet, Anal. Chem. 67 (1995) 1907. [14] H. Kanazawa, K. Yamamoto, Y. Matsushima, N. Takai, A. Kikuchi, Y. Sakurai, T. Okano, Anal. Chem. 68 (1996) 100.
[15] H. Kanazawa, Y. Kashiwase, K. Yamamoto, Y. Matsushima, A. Kikuchi, Y. Sakurai, T. Okano, Anal. Chem. 69 (1997) 823. [16] H. Kanazawa, Y. Matsushima, T. Okano, Trends Anal. Chem. 17 (1998) 435. [17] H. Kanazawa, T. Sunamoto, E. Ayano, Y. Matsushima, A. Kikuchi, T. Okano, Anal. Sci. 18 (2002) 45. [18] K. Yamamoto, H. Kanazawa, Y. Matsushima, K. Oikawa, A. Kikuchi, T. Okano, Environ. Sci. 7 (2000) 47. [19] J. Kobayashi, A. Kikuchi, K. Sakai, T. Okano, Anal. Chem. 73 (2001) 2027. [20] J. Kobayashi, A. Kikuchi, K. Sakai, T. Okano, J. Chromatogr. A 958 (2002) 109. [21] H. Kanazawa, T. Sunamoto, Y. Matsushima, A. Kikuchi, T. Okano, Anal. Chem. 72 (2000) 5961. [22] Y.G. Takei, T. Aoki, K. Sanui, N. Ogata, T. Okano, Y. Sakurai, Bioconjug. Chem. 4 (1993) 341. [23] L.D. Taylor, L.D. Cerankowski, J. Polym. Sci. Polym. Chem. 13 (1975) 2551. [24] D.W. Urry, Prog. Biophys. Mol. Biol. 57 (1992) 23. [25] D.W. Urry, C.H. Luan, T.M. Parker, D.C. Gowda, K.U. Prasad, M.C. Reid, A. Safavy, J. Am. Chem. Soc. 113 (1991) 4346. [26] C.H. Luan, R.D. Harris, K.U. Prasad, D.W. Urry, Biopolymers 29 (1990) 1699. [27] K. Yamamoto, H. Kanazawa, Y. Matsushima, N. Takai, A. Kikuchi, T. Okano, Chromatography 21 (2000) 209. [28] K. Morino, K. Maeda, Y. Okamoto, E. Yashima, T. Sato, Chem. Eur. J. 8 (22) (2002) 5112. [29] T. Uno, S. Habaue, Y. Okamoto, Enantiomer 5 (1) (2000) 29–36. [30] T. Yakushiji, K. Sakai, A. Kikuchi, T. Aoyagi, Y. Sakurai, T. Okano, Anal. Chem. 71 (1999) 1125.
Temperature-responsive stationary phase utilizing a polymer of proline derivative for hydrophobic interaction chromatography using an aqueous mobile phase Hideko Kanazawa a,∗ , Eri Ayano a , Chikako Sakamoto a , Reiko Yoda a , Akihiko Kikuchi b , Teruo Okano b b
a Department of Physical Pharmaceutical Chemistry, Kyoritsu University of Pharmacy, 1-5-30 Shibakoen, Minato-ku, Tokyo 105-8512, Japan Institute of Advanced Biomedical Engineering and Science, Tokyo Women’s Medical University, 8-1 Kawadacho, Shinjyuku-ku, Tokyo 162-8666, Japan
Available online 10 October 2005
Abstract A new method of chromatography is proposed, utilizing a thermo-responsive polymer carrying an amino acid ester residue for the stationary phase of high-performance liquid chromatography (HPLC). We have been investigating the new concept of chromatography, a temperature-responsive chromatography, using temperature-responsive poly(N-isopropylacrylamide) (PNIPAAm)-modified surface for HPLC with a constant aqueous media as the mobile phase. In this study, we designed and synthesized thermo-responsive poly(acryloyl-l-proline methyl ester) and its copolymer with N-isopropylacrylamide (NIPAAm). Homopolymers of acryloyl-l-proline methyl ester and copolymer were prepared by the reaction of radical telomerization. These polymers underwent a reversible phase transition from water-soluble forms into aggregates by changing the temperature, similar to PNIPAAm. The surface properties and functions of stationary phases modified with poly(acryloyl-l-proline methyl ester) were controlled by the external temperature. In the chromatographic system, we separated steroids and amino acids with a variety of hydrophobicities using a sole aqueous mobile phase. In contrast to a PNIPAAm-modified surface, a poly(acryloyl-l-proline methyl ester)-modified surface showed a greater affinity for hydrophobic amino acids. © 2005 Elsevier B.V. All rights reserved. Keywords: Temperature-responsive polymer; HPLC; Hydrophobic interaction chromatography; Separation; Amino acids; Steroids; Poly(acryloyl-l-proline methyl ester); Poly(N-isopropylacrylamide)
1. Introduction Recently, HPLC techniques have become increasingly important for the separation of proteins without denaturation, which is often observed with the use of the reversed-phase mode. Many reversed-phase methods for proteins have employed C18 columns with mobile phases containing acetonitrile in lowpH buffers. Little attention has been paid to recovering biological activity. Acetonitrile almost always leads to the total denaturation of proteins. Similarly, acidic environments, including the standard trifluoroacetic acid-based buffers, are sometimes destructive to the activities of many enzymes. Thus, these conditions should be avoided in the separation of most proteins.
∗
Corresponding author. Tel.: +81 3 5400 2657; fax: +81 3 5400 1378. E-mail address: [email protected] (H. Kanazawa).
0021-9673/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2005.09.052
In recent years, polymers that demonstrate good solubility in aqueous solutions at low temperature, but separate from solution when the temperature is raised above the lower critical solution temperature (LCST), have received increased attention [1–3]. Polymers exhibiting LCST behavior have been investigated for such applications as drug delivery system [4–6] and cell culture substrates [7–11]. However, there have been few reports concerning the use of these temperature-responsive polymers in chromatographic separation [12,13]. In previous studies, we developed a new method of HPLC using packing materials modified with a temperature-responsive polymer, poly(N-isopropylacrylamide) (PNIPAAm) [14–21]. PNIPAAm is one of the extensively studied thermosensitive polymers. PNIPAAm exhibits a thermally reversible soluble/insoluble changes phase transition in aqueous solution at 32 ◦ C. The surface properties and functions of the stationary phases are controlled by the external temperature. In the chromatographic system utilizing the PNIPAAm-modified stationary
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
phase, elution of target substances is controlled only by a small change in the column temperatures. We also observed in these studies that the LCST is modulated by varying the monomer composition [22]. In general, the incorporation of hydrophobic comonomers leads to a lower LCST, and hydrophilic comonomers lead to a higher LCST [23]. It has been reported that copolymerizing NIPAAm with butyl methacrylate (BMA) as hydrophobic comonomers leads to a lower LCST [15]. These features also showed that certain polypeptide solutions demonstrate temperature-induced phase separation [24,25]. It was found that an increase in the hydrophobicity of polypeptides lowered the temperature of the phase transition and increased the heat of the transition. The interpretation is that the increase in endothermic heat required to drive the transition reflects the energy required to destructive the greater numbers of waters of hydrophobic hydration [26]. In this study, we designed and synthesized thermo-responsive polymer carrying amino acid ester residue as a side chain. These polymers undergo a reversible phase transition from watersoluble forms into aggregates similar to PNIPAAm. Here, we demonstrated how to separate steroids and amino acids with a variety of hydrophobicities using a temperature-responsive silicagel surface modified with polymers of acryloyl-l-proline methyl ester as packing materials. 2. Experimental 2.1. Materials N-Isopropylacrylamide (NIPAAm) was kindly provided by KOHJIN (Tokyo, Japan) and was purified by recrystallization from n-hexane and dried at 25 ◦ C in vacuo. land d-proline methyl ester hydrochlorides were purchased from Kokusan Chemical (Tokyo, Japan). 3-Mercaptopropionic acid (MPA; Wako Pure Chemicals, Tokyo, Japan) was distilled under reduced pressure and the fraction boiling at 95 ◦ C (5 mmHg) was used. 2,2 -Azobisisobutyronitrile
153
(AIBN), N,N-dimethylformamide (DMF), acryloyl chloride, dichloromethane, and ethyl acetate (EtOAc) were obtained from Wako Pure Chemicals (Tokyo, Japan) and purified by conventional methods. N,N -Dicyclohexylcarbodiimide (DCC) and Nhydroxysuccinimide (SuOH) were purchased from Wako Pure Chemicals. Aminopropyl silica (average diameter of 5 m, pore size 12 nm) was purchased from Nishio Kogyo, (Tokyo, Japan). Sulfo-succinimidyl-4-O-(4,4 -dimethoxytrityl) butyrate (s-SDTB) was obtained from Pierce (Rockford, IL, USA). Cortisone acetate, hydrocortisone, hydrocortisone acetate, amino acids phenylthiohydantoins (PTH-amino acids) were from Wako Pure Chemicals (Tokyo, Japan) and other steroids were from Sigma Chemicals (St. Louis, MO, USA). Milli-Q grade water was used for preparation of sample solutions. Other reagents and solvents were commercially obtained and used without further purification. 2.2. Synthesis of acryloyl-l-proline (or d-proline) methyl ester monomer and polymerization procedure N-Acryloyl derivative of proline methyl ester was synthesized by the reaction of proline methyl ester with acryloyl chloride in dichloromethane. l- or d-proline methyl ester hydrochlorides were neutralized by NaOH aqueous solution. Proline methyl ester (0.22 mol) was dissolved in 270 mL of dichloromethane. Then acryloyl chloride (0.242 mol) and triethylamine (0.264 mol) were added to the reaction mixture with stirring below −2 ◦ C. After the reaction proceeded for 2 h with stirring, the reaction mixture was extracted by 200 mL of HC1 aqueous solution (HCl:water = 1:4) and 50 mL of water. Then the dichloromethane layer was dried by 15 g of sodium sulfate and under reduced pressure to remove the dichloromethane. The resulting product was evaporated in vacuo by a rotary evaporator. The yield was 68.4%. Poly(acryloyl-l-proline methyl ester) and copolymers with NIPAAm were prepared by radical polymerization in DMF, as shown in the reaction scheme (Fig. 1). Typical procedures
Fig. 1. Synthesis of poly(NIPAAm-co-acryloyl-l-proline methyl ester) copolymer.
154
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
were as follows: acryloyl-l-proline methyl ester (40 mmol) and NIPAAm (40 mmol) were dissolved in 50 mL of DMF. AIBN (0.88 mmol), and MPA (6.25 mmol) were used as an initiator, and chain transfer agent, respectively. The reaction mixture was degassed by subjecting to freeze-thaw cycles and the ampoules containing the mixture were sealed under reduced pressure. The reaction proceeded at 70 ◦ C for 2 h. After the reaction solution was concentrated by evaporation, the reaction mixture was poured into 50 mLof diethyl ether to precipitate the polymers. The polymer was further purified by repeated precipitation from 10 mL of acetone into 500 mL of diethyl ether. The molecular weight of the polymers was determined by end group titration with 0.01 M NaOH using phenolphthalein as an indicator. Specific rotations at the sodium D line, [α]D , and near-UV (220–420 nm) circular dichroism (CD) spectrum were obtained by a DIP-1000 polarimeter and a CD-1595 circular dichroism detector (JASCO Corporation, Tokyo, Japan), respectively. 2.3. Transmittance measurements The LCSTs of the polymers were determined by measuring the optical transmittance of polymer aqueous solutions. The optical transmittance of polymer solutions (5 mg/mL) was measured at 500 nm at various temperatures using a UV–vis spectrophotometer (U-3000, Hitachi, and Tokyo, Japan). The temperature of observation cell was controlled with a LAUDARCS20D water bath with a deviation of 0.02 ◦ C. The LCST values for each polymer were defined as the temperature where 50% optical transmittance of polymer aqueous solutions was observed.
(0.077 mg/mL). Standard solutions of steroids were prepared with dexamethasone (DX; 0.061 mg/mL), hydrocortisone (HC; 0.101 mg/mL), hydrocortisone acetate (HCA; 0.007 mg/mL), prednisolone (PRE; 0.166 mg/mL), and testosterone (TES; 0.027 mg/mL). The log P values of steroids were calculated by the CAChe system (FUJITSU, Japan). 3. Results and discussion 3.1. Characterization of thermo-responsive polymer carrying amino acids residue Fig. 2 shows the temperature dependence for the optical transmittance of polymer solutions. PNIPAAm and poly(acryloyll-proline methyl ester) exhibit an LCST in water. Both of poly(acryloyl-l-proline methyl ester) and its copolymers with NIPAAm also exhibit thermally reversible soluble-insoluble changes in response to temperature changes. The physicochemical data for these polymers are given in Table 1. The polymer of acryloyl-d-proline methyl ester was also synthesized by the same way as that of the l-proline derivative to compare with the property of the polymer of the l-proline derivative. The chiroptical properties of these polymers were evaluated using polarimeter (optical rotation) and CD spectroscopy. Specific
2.4. Modification of aminopropyl silica with NIPAAm polymer Modification of aminopropyl silica with NIPAAm copolymer by activated ester-amine coupling was the same procedure as described in our previous report [27]. The surface coverage degrees of three columns were calculated by quantification of residual amount of amino group on the supports. The amount of amino group on silica supports was determined by spectrophotometric method using s-SDTB [27]. Introduction rate of polymers on the supports were 77–81%. 2.5. Temperature-responsive chromatography A polymer-grafted silica support was packed into a stainlesssteel column (length: 150 mm × 4.6 mm I.D.). The column was connected to an HPLC system (HITACHI Model L-6200 intelligent pump; L-4000 UV-monitor, D-2500 data processor). The column oven was a Shodex AO-30 (Showa Denko, Tokyo, Japan). Milli-Q grade water was used as the mobile phase. The elution behaviors of the samples were recorded at a flowrate of 1 mL/min at various temperatures. Standard solutions of PTH-amino acids were prepared with Arg (initial concentration 1.000 mg/mL); Asn (0.404 mg/mL); Asp (1.084 mg/mL); Gin (0.124 mg/mL); Glu (0.471 mg/mL); His (1.000 mg/mL); Leu (0.014 mg/mL); Met (0.063 mg/mL); Phe (0.038 mg/mL); Pro (0.038 mg/mL); Thr (0.019 mg/mL); Trp (0.019 mg/mL); Tyr
Fig. 2. Temperature dependence of the optical transmittance of polymer solutions at 500 nm: () poly(acryloyl-l-proline methyl ester); () poly(NIPAAmco-acryloyl-l-proline methyl ester) (1:1); () PNIPAAm.
Table 1 Characterization of poly(NIPAAm-co-acryloyl-proline methyl ester) Polymer unit
LCST (◦ C)a
Acryloyl-l-proline methyl ester NIPAAm-co-acryloyl-l-pro line methyl ester NIPAAm-co-acryloyl-d-proline methyl ester NIPAAm
20.7 24.0
−107 −41
3200 7500
23.5
43
4300
a b c
31.5
b [α]25 D
–
M.W.c
4500
Determined by transmittance measurement at 500 nm. Mesured by polarimeter at 25 ◦ C (c = 0.5% (w/v) in methanol). Number-averaged molecular weight determined by end group titration.
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
155
Fig. 4. Chromatograms of a mixture of benzene and steroids on a poly(NIPAAmco-acryloyl-l-proline methyl ester)-modified silica column at (a) 10 ◦ C and (b) 40 ◦ C. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm; injection volume, 20 L. The peaks are: (1) BEN; (2) HC; (3) COR; (4) PRE; (5) DX; (6) HCA; (7) TES.
Fig. 3. CD and UV spectra of (1) poly(NIPAAm-co-acryloyl-l-proline methyl ester) and (2) poly(NIPAAm-co-acryloyl-d-proline methyl ester). The mobile phase is a 50% methanol water solution; injection volume, 10 L.
rotations at 25 ◦ C, [α]25 D , and CD spectra for these two polymers were obtained. The [α]D value was positive and negative for a polymer of l-proline and that of d-proline, respectively. The circular dichroism spectrum can be used to characterize the conformation of the chiral polymer in solution [28,29]. The signs of circular dichroism at all wavelengths for these two polymers are opposite, as shown in Fig. 3. The results seemed to indicate that these polymers maintained the optical active properties of its monomer used. 3.2. Temperature-responsive chromatographic separation of steroids Using a column packed with poly(acryloyl-l-proline methyl ester)-modified silica in a HPLC system, the separation of steroids was carried out by changing the column temperature. Milli-Q water was used as the sole mobile phase. Fig. 4 shows typical chromatograms of steroids on a copolymer of NIPAAm and acryloyl-l-proline methyl ester grafted silica column. The temperature-dependent resolution of steroids was achieved using only water as a mobile phase. Although the peaks of steroids are not properly resolved at 10 ◦ C, they were well resolved at 40 ◦ C. With increasing the temperature, increased interactions between the solutes and the poly(NIPAAm-coacryloyl-l-proline methyl ester)-grafted surfaces of the stationary phases were observed. We previously demonstrated that a hydrophobic interaction between steroids and PNIPAAmgrafted surfaces was readily modulated by the temperature [14–17,30]. A PNIPAAm-modified surface showed remarkable surface wettability changes by changing the temperature [30].
Therefore, a hydrophobic interaction with hydrophobic steroids is greatly affected by the temperature on this surface. The temperature-dependent elution behavior of steroids was examined on a PNIPAAm-modified silica column, a poly(NIPAAmco-acryloyl-l-proline methyl ester)-modified silica column, and a poly(acryloyl-l-proline methyl ester)-modified silica column. Fig. 5 shows the effect of temperature on the retention of steroids on three types of thermo-responsive polymer-modified silica columns. The elution profiles of steroids on these columns are opposite from those observed using conventional C18 column, in which the elution time was shortened with the temperature. The results indicated that the copolymer-modified surface of the stationary phase exhibited temperature-controlled hydrophilic–hydrophobic changes. On those three thermoresponsive polymer-modified stationary phases, the hydrophobicity of the surfaces reduced at a lower temperature and increased at an elevated temperature. A substance’s hydrophobicity is usually defined as the logarithm of its partition coefficients in the n-octanol/water system, log P values. The log P values for each steroid are 1.61 for hydrocortisone, 1.62 for prednisolone, 1.83 for dexamethasone, 2.30 for hydrocortisone acetate, and 3.32 for testosterone. The more hydrophobic steroids showed longer retention times on the poly(acryloyll-proline methyl ester)-modified column. These results indicate that a stronger hydrophobic interaction is the primary driving force for the retention of steroids on poly(acryloyl-lproline methyl ester)-modified matrices, similar to the case of PNIPAAm-modified matrices. The elution orders of steroids were the same on these three columns at a higher temperature than LCST. Interestingly, there was a change of elution order between dexamethasone and hydrocortisone acetate around at the LCST on the poly(acryloyl-l-proline methyl ester)-modified column. This phenomenon was also observed on copolymer of NIPAAm with acryloyl-l-proline methyl ester-modified col-
156
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
Fig. 5. Effect of acryloyl-l-proline methyl ester introduction on the retention times of steroids on (a) a PNIPAAm-modified silica column; (b) a poly(NIPAAmco-acryloyl-l-proline methyl ester)-modified silica column and (c) a poly(acryloyl-l-proline methyl ester)-modified silica column: () HC; (䊉) PRE; (♦) DX; () HCA; () TES. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm; injection volume, 5 L.
umn. Therefore, it may be due to the difference in side chain structure that attributed to the interaction with two steroids. 3.3. Temperature-responsive chromatographic separation of PTH-amino acids We previously demonstrated the temperature-responsive chromatographic separation of PTH-amino acids using aqueous media as the mobile phase on PNIPAAm grafted silica columns [21]. Fig. 6 shows typical chromatograms of a mixture of PTHamino acids at below and above the LCST on poly(acryloyll-proline methyl ester)-modified silica columns. Milli-Q water was used as the sole mobile phase. PTH-amino acids were dissolved in Milli-Q water, and a sample mixture solution was injected into a poly(acryloyl-l-proline methyl ester)-modified column. As shown in Fig. 6, the peaks were not properly resolved at 5 ◦ C, which was lower than the LCST. As the column temperature was raised to 40 ◦ C, the nine PTH-amino acids were well
resolved. Fig. 7 shows the effect of temperature on the retention of PTH-amino acids on three types of thermo-responsive polymer-modified silica columns. The interaction between PTHamino acids and these temperature-responsive surfaces becomes stronger at elevated temperature. In a non-polar hydrophobic amino acid, such as Tyr, Phe and Trp, the retention times were retarded with increasing the temperature on the poly(acryloyl-lproline methyl ester)-modified column. However, the retention times of the amino acids, such as Cys, Gly and Ala, were slightly changed. In the separation of steroids, the retention times have a linear relationship with the log P values of steroids having different hydrophobicities. As can also be seen in Fig. 7, PTHamino acids eluted with the order of their hydrophobicity, and the more hydrophobic one showed a longer retention time at higher temperature. This suggests that the hydrophobic interaction is the primary driving force for the separation of PTH-amino acids with poly(acryloyl-l-proline methyl ester)-modified matrices.
Fig. 6. Chromatograms of PTH-amino acids on a poly(acryloyl-l-proline methyl ester)-modified silica column at (a) 5 ◦ C and (b) 40 ◦ C. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm. The peaks are: (1) Gly; (2) Ala; (3) Pro; (4) Val; (5) Met; (6) Leu; (7) Phe; (8) Tyr; (9) Trp.
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
157
Fig. 7. Effect of l-proline derivative introduction on the retention times of PTH amino acids on (a) a PNIPAAm-modified silica column; (b) a poly(NIPAAm-coacryloyl-l-proline methyl ester)-modified silica column and (c) a poly(acryloyl-l-proline methyl ester)-modified silica column: (×) Cys; () Gly; +, Ala; () Pro; () Met; () Thr; (♦) Leu; (䊉) Phe; () Tyr; () Tip. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm; injection volume, 5 L.
As shown in Fig. 5, an alteration of the retention times of steroids in response to the temperature on a poly(acryloyll-proline methyl ester)-modified column is less than that on a PNIPAAm-modified and a poly(NIPAAm-co-acryloyl-lproline methyl ester)-modified column. However, as shown in Fig. 7, the retention of PTH-amino acids was dramatically increased on a poly(acryloyl-l-proline methyl ester)-modified column in contrast to that on a homogeneous PNIPAAmmodified silica column. Fig. 8 shows a comparison of the separation selectivity in the aqueous mobile phase on PNIPAAmmodified and poly(acryloyl-l-proline methyl ester)-modified packing materials. The poly(acryloyl-l-proline methyl ester)modified stationary phase showed greater affinity for hydrophobic aromatic amino acids, such as Tyr, Trp and Phe, than the PNIPAAm-modified surface. These observations would be
caused by the interaction between the hydrophobic PTH-amino acids and proline moiety of the polymer chain on a grafted silica surface. Consequently, it is possible to determine the optimum separation conditions by setting an appropriate column temperature without modifying the eluent composition. In such separations, retention on the supports may also involve hydrophobic interaction between protein and polymer chains. The polypeptide and protein separations on temperature-responsive chromatography are currently in progress in our laboratory. 4. Conclusion The use of poly(acryloyl-l-proline methyl ester) as a silica surface modifier for temperature-responsive chromatography increased the selectivity and retention of PTH-amino acids. This study is the first report of the application of the polymer of the amino acid derivative for separation. By altering the copolymer composition we could control the LCST and the hydrophobicity of the surface of the stationary phases, and thus the elution time of the temperature-responsive chromatography. We succeeded to demonstrate hydrophobic interaction chromatography-like separation selectivity of the PTH-amino acid by simply changing the column temperature with pure water as a mobile phase. The ability of the proposed thermo-responsive polymer-modified stationary phase to separate the solutes without the use of an organic solvent is advantageous from the point of view of keeping the biological activity, environmental reasons, and mobile-phase reagent cost. We are currently performing studies to utilize a thermo-responsive polymer carrying amino acid ester residue for the chiral stationary phase. Acknowledgment
Fig. 8. Comparison of the retention times of PTH-amino acids on a poly(acryloyl-l-proline methyl ester)-modified silica column and a PNIPAAmmodified silica column at 40 ◦ C. The mobile phase is water; flow rate, 1.0 mL/min; absorption at 254 nm.
This work was supported in part by a Grant-in-Aid for Scientific Research No. 15590049 from the Ministry of Education, Science, Sports, and Culture of Japan.
158
H. Kanazawa et al. / J. Chromatogr. A 1106 (2006) 152–158
References [1] M. Heskins, J.E. Guillet, E. James, J. Macromol. Sci. Chem. A 2 (1968) 1441. [2] Y.H. Bae, T. Okano, S.W. Kim, J. Polym. Sci. Polym. Phys. 28 (1990) 923. [3] T. Tanaka, I. Nishio, S. Sun, S. Ueno-Nishio, Science 218 (1981) 467. [4] A. Chilkoti, M.R. Dreher, D.E. Meyer, D. Raucher, Adv. Drug Deliv. Rev. 54 (2002) 613. [5] R. Yoshida, K. Sakai, T. Okano, Y. Sakurai, Y.H. Bae, S.W. Kim, J. Biomater. Sci. Polym. Ed. 3 (1991) 155. [6] R. Yoshida, K. Sakai, T. Okano, Y. Sakurai, J. Biomater. Sci. Polym. Ed. 3 (1991) 243. [7] N. Yamada, T. Okano, H. Sakai, F. Karikusa, Y. Sawasaki, Y. Sakurai, Makromol. Chem. 11 (1990) 571. [8] T. Okano, N. Yamada, H. Sakai, Y. Sakurai, J. Biomed. Mater. Res. 27 (1993) 1243. [9] T. Okano, N. Yamada, M. Okumura, H. Sakai, Y. Sakurai, Biomaterials 16 (1995) 297. [10] M. Yamato, M. Utsumi, A. Kushida, C. Konno, A. Kikuchi, T. Okano, Tissue Eng. 7 (2001) 473. [11] T. Shimizu, M. Yamato, A. Kikuchi, T. Okano, Tissue Eng. 7 (2001) 141. [12] M. Gewehr, K. Nakamura, N. Ise, H. Kitano, Macromol. Chem. 193 (1992) 249. [13] K. Hosoya, K. Kimata, T. Araki, N. Tanaka, J.M.J. Frechet, Anal. Chem. 67 (1995) 1907. [14] H. Kanazawa, K. Yamamoto, Y. Matsushima, N. Takai, A. Kikuchi, Y. Sakurai, T. Okano, Anal. Chem. 68 (1996) 100.
[15] H. Kanazawa, Y. Kashiwase, K. Yamamoto, Y. Matsushima, A. Kikuchi, Y. Sakurai, T. Okano, Anal. Chem. 69 (1997) 823. [16] H. Kanazawa, Y. Matsushima, T. Okano, Trends Anal. Chem. 17 (1998) 435. [17] H. Kanazawa, T. Sunamoto, E. Ayano, Y. Matsushima, A. Kikuchi, T. Okano, Anal. Sci. 18 (2002) 45. [18] K. Yamamoto, H. Kanazawa, Y. Matsushima, K. Oikawa, A. Kikuchi, T. Okano, Environ. Sci. 7 (2000) 47. [19] J. Kobayashi, A. Kikuchi, K. Sakai, T. Okano, Anal. Chem. 73 (2001) 2027. [20] J. Kobayashi, A. Kikuchi, K. Sakai, T. Okano, J. Chromatogr. A 958 (2002) 109. [21] H. Kanazawa, T. Sunamoto, Y. Matsushima, A. Kikuchi, T. Okano, Anal. Chem. 72 (2000) 5961. [22] Y.G. Takei, T. Aoki, K. Sanui, N. Ogata, T. Okano, Y. Sakurai, Bioconjug. Chem. 4 (1993) 341. [23] L.D. Taylor, L.D. Cerankowski, J. Polym. Sci. Polym. Chem. 13 (1975) 2551. [24] D.W. Urry, Prog. Biophys. Mol. Biol. 57 (1992) 23. [25] D.W. Urry, C.H. Luan, T.M. Parker, D.C. Gowda, K.U. Prasad, M.C. Reid, A. Safavy, J. Am. Chem. Soc. 113 (1991) 4346. [26] C.H. Luan, R.D. Harris, K.U. Prasad, D.W. Urry, Biopolymers 29 (1990) 1699. [27] K. Yamamoto, H. Kanazawa, Y. Matsushima, N. Takai, A. Kikuchi, T. Okano, Chromatography 21 (2000) 209. [28] K. Morino, K. Maeda, Y. Okamoto, E. Yashima, T. Sato, Chem. Eur. J. 8 (22) (2002) 5112. [29] T. Uno, S. Habaue, Y. Okamoto, Enantiomer 5 (1) (2000) 29–36. [30] T. Yakushiji, K. Sakai, A. Kikuchi, T. Aoyagi, Y. Sakurai, T. Okano, Anal. Chem. 71 (1999) 1125.