- Email: [email protected]
Effect of amino acid sequence on the hydrophobicity of small peptides
Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
Effect of amino acid sequence on the hydrophobicity of small peptides Chih-I. Liu a,c , Ying-Chih Chan a , Wen-Yih Chen b , Ruoh-Chyu Ruaan b,∗ b
a Department of Chemical Engineering, Chung-Yuan University, Chung-Li 320, Taiwan Department of Chemical Engineering, National Central University, Chung-Li 320, Taiwan c Department of Nursing, Mei-Ho Institute of Technology, Pintung 912, Taiwan
Received 5 September 2004; received in revised form 7 June 2005; accepted 14 June 2005
Abstract Hydrophobic adsorbents, which contain alkyl ligands of various chain lengths, were used to study the hydrophobic interactions between alkyl chains and small peptides. We investigated the adsorption of four similar peptides: GWWG, GWGW, WGWG, WGGW. All of them contained two glycines (G) and two tryptophans (W) but the amino acids were arranged in different orders. The capacity factors of peptides between 10 and 35 ◦ C were measured and then the thermodynamic parameters, such as enthalpy and entropy changes (H◦ and S◦ ) of adsorption, were estimated. It implied that enthalpy was the major driving force in all the adsorption processes. Furthermore, H◦ and S◦ became more negative as the alkyl chain length was increased. It revealed that the van der Waal’s interaction had greater influence on the adsorption as the chain length increased. It was also found that, the contribution of each amino acid to peptide’s hydrophobicity was affected by the position of the amino acid. When a hydrophobic amino acid was positioned in the middle of a peptide chain, it exhibited the highest hydrophobicity. Interestingly, tryptophan at the carboxyl end was found more hydrophobic than it at the amino end of the peptide. © 2005 Elsevier B.V. All rights reserved. Keywords: Hydrophobic interaction; Gibbs free energy; Enthalpy; Entropy; Hydrophobic amino acids; Octyl; Decyl-CM-sepharose
1. Introduction Hydrophobic interaction played an important role in various biological systems and biochemical related applications. For example, the interaction between cells and artificial surface is strongly related to the adsorbed proteins on the surface [1–3]. The adsorption of proteins from liquid medium is often driven by the so-called hydrophobic interaction [4,5]. It was also a dominant interaction between the antigen–antibody binding [6,7] and the association between apolar hormones and their corresponding cell surface receptors [8]. Furthermore, the hydrophobic interaction chromatography is a direct application to use this type of interaction to separate biomolecules [9–13]. The study on how various peptides and ∗ Corresponding author. Tel.: +886 3 422 7151x34232; fax: +886 3 280 4341. E-mail address: [email protected] (R.-C. Ruaan).
0927-7757/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2005.06.023
proteins interact with hydrophobic surfaces has long been a major research focus [5,14,15]. Hydrophobic interaction is contributed not by a single force but by the joined efforts of various inter-molecular forces, such as van der Waal’s forces, electrostatic interactions and hydrogen bonds. Unlike electrostatic interaction that has been well characterized by Poisson equation, hydrophobic interaction can only be characterized experimentally because of its complexity. To find a way to quantify the hydrophobic interaction is an important matter. Hydrophobic interaction chromatography (HIC) has recently been adopted to study the hydrophobic interaction of proteins. Thermodynamics properties were used to characterize the retention behaviors. Haidacher et al. [16] investigated the effects of temperature on the retention of dansyl amino acids. They found the adsorption of dansyl amino acids on HIC columns (Spherogel, Synchropak propyl and butyl-NPR) was entropy-driven at low temperatures and
176
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
enthalpy-driven at high temperatures. Furthermore, Vailaya and Horvath [17] studied the effect of molecular structure and types of stationary phase on the adsorption enthalpy and entropy. They found that both the adsorption enthalpy and entropy of dansyl amino acids on HIC columns became more positive as the size of amino acids. The hydrophobicity of amino acids and proteins can be scaled by their retention factors on the column [18]. However, the hydrophobicity of a protein is not only affected by its amino acid composition but also related to the size, the secondary structure and the 3D structure of the protein. Therefore, the hydrophobicity of each amino acid has no direct correlation to the hydrophobicity to a protein, since the arrangement of amino acid in a peptide may also affect the overall hydrophobicity. Understanding the sequencedependent hydrophobic behavior for small molecules can be helpful for estimating the affect of sequence in large proteins. We attempt to estimate the hydrophobicity of small peptides of the same amino acid composition but in different sequences. Four peptides GWWG, GWGW, WGWG and WGGW were under investigation. The capacity factors of these peptides on various hydrophobic interaction columns are measured between 10 and 35 ◦ C. The Gibbs energy, enthalpy, and entropy of adsorption are calculated. The effect of amino acid sequence and the effect of solid surface are discussed.
alkylamine, EDC and NHS. The mixed was reacted for 20 h at 25 ◦ C in a shaker incubator. After reaction, the CM-Sepharose gel was washed continuously by 20% ethanol solution containing 0.2 M NaCl until no alkylamine can be detected. The eluted alkylamine was tested by ninhydrin reaction. The resulting alkyl-CM-Sepharose may be subject to repeated reaction to obtain higher alkyl substitution. The alkyl content was estimated by titrating the remaining carboxyl groups.
2. Experimental
2.4. Calculation of the thermodynamics parameters
2.1. Material
We may assume that the adsorption reaction takes place as follows.
Tetrahydrofuran (THF), pyridine, hydrochloride and sodium chloride were purchased from Riedel-deHa¨en (Chinosol, Seelze, Germany). Ammonium sulfate and tris(hydroxymethyl) aminomethane were purchased from Merck (Germany). Butyl-, hexyl-, octyl-, decylamine, 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide methiodode (EDC) were obtained from Aldrich (Milwaukee, WI, USA), Nhydroxysuccinimide (NHS), glycyl–glycyl–glycyl–glycine (GGGG), were purchased from Sigma (St. Louis, U.S.A). CM-Sepharose gel was purchased from Pharmacia Biotech. (Uppsala, Sweden). The peptides, glycyl–tryptophanyl–tryptophanyl–glycine (GWWG), tryptophanyl– glycyl–tryptophanyl–glycine (WGWG), glycyl–tryptophanyl–glycyl–tryptophan (GWGW) and tryptophanyl– glycyl–glycyl–tryptophanyl (WGGW) were obtained from Digital GENE Biosciences (Taipei, Taiwan). 2.2. Synthesis of the hydrophobic adsorbents The HIC adsorbent was synthesized by coupling alkylamine to CM-Sepharose particles through the carbodiimide mediated coupling reaction. Firstly CM-Sepharose was washed extensively with deionized water to remove ethanol. The resins were then washed with THF before the addition of
2.3. Chromatographic operation Hydrophobic interaction chromatography was carried out through a HPLC system which include a GL Sciences model PU-610 pump (Tokyo, Japan), a Lab Alliance model 500 UV monitor (PA, U.S.A.), a Spark model 816 autosampler (Emmen, Holland). The signals were processed through a Chem-Lab data station (SISC, Taipei, Taiwan). The volume of the sample loop is 100 L. Five milliliters of hydrophobic beads were packed in a 10 cm × 1 cm Sigma jacketed column at a flow rate of 1 mL/min. The final bed height was kept at 5.5 cm. The column temperature was controlled by Firstek B402-D (Shinjuang, Taiwan) circulation water bath. The column was first equilibrated with 20 mM Tris buffer (pH 6) containing (NH4 )2 SO4 (0.25 M) and all the samples were eluted isocratically in the same elution buffer. Eluent was monitored at UV 220 nm through a Lab Alliance UV detector (model 500).
M + L → ML
(1)
where M represents the adsorbed molecule and L represents the ligand on the adsorbent matrix. Then, the equilibrium constant could be calculated as follows. Keq =
CML CM CL
(2)
where CM , CL and CML are the concentrations of M, L and ML in the liquid phase. Keq =
ns 1 k = nm CL CL
(3)
where ns and nm represent the amount of solute in the stationary phase and mobile phase, respectively. The retention factor (k ) can be measured by chromatography and can be related to the equibrium constant (Keq ) as: k ≡
tR − t0 CML = = Keq CL t0 CM
where tR is the retention time of the sample and t0 is the retention time of non-retained molecules (GGGG). Therefore, k = Keq CL ≈ Keq Lm
(4)
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
177
Since CM is usually small, CL is close to Lm , the maximum concentration of ligand in the matrix. Lm can be calculated by the ligand density based on total volume (solid phase + mobile phase). Since G◦ = −RT ln Keq
(5)
and G◦ = H ◦ − T S ◦ ln Keq = −
H ◦ RT
+
S ◦ R
(6) (7)
so that S ◦ H ◦ + + ln Lm (8) RT R where G◦ denotes the standard Gibbs free change, H◦ and S◦ are the standard enthalpy and entropy changes. ln k = −
3. Results and discussions 3.1. The retention on octyl-Sepharose columns (k )
The capacity factor of the four peptides on a column packed with octyl-Sepharose resins were measured at various temperatures from 10 to 25 ◦ C. The natural logarithm of k was plotted against the inverse temperature and was shown in Fig. 1. It was found that the peptide GWWG was the most retentive for all the measured temperatures, followed by GWGW and WGWG. The peptide WGGW was the least retentive. Since there is no rigid secondary structure in such a small peptide, we might conclude from this result that the amino acid, tryptophan, exhibited the highest hydrophobicity when it was located in the middle of the peptide chain, but it’s hydrophobicity was reduced when it was located at the ends of the peptide, especially at the amino end.
Fig. 2. Van’t Hoff plot for the adsorption of peptides (GWWG, WGWG, GWGW, WGGW) onto the octyl-CM-Sepharose column. The ligand density is 33.3 mol/mL.
3.2. The retention on alkyl-CM-Sepharose columns We attached butyl-, hexyl-, octyl- and decyl functional groups on commercial CM-Sepharose particles via carbodiimide mediated method. The density of alkyl groups was controlled within 33.3–47.9 mol/mL gel. The capacity factors (k ) of the four peptides were measured. It was found in Fig. 2 that the elution of the four peptides from all the alkylCM-Sepharose columns followed the same order WGGW, GWGW, WGWG and then GWWG, which was different from that on octyl-Sepharose. It was an interesting phenomenon, which required further investigation. The retention time was measured from T = 283 to 318 K. It was found that the van’t Hoff plot was no longer linear at temperature higher than 303 K. We try to limit our estimation of H◦ within linear region to avoid nonlinear fitting. Therefore, the fitted data was restricted to the temperature lower than 303 K in enthalpy estimation by van’t Hoff plot. 3.3. The Gibbs energy, enthalpy and entropy of peptide adsorption
Fig. 1. Van’t Hoff plot for the adsorption of peptides (GWWG, WGWG, GWGW, WGGW) onto an octyl-Sepharose column (ligand density is 40 mol/mL).
The Gibbs energy adsorption could be calculated from Eq. (5) and the adsorption enthalpy could be estimated from the van’t Hoff’s plot. The entropy of adsorption could therefore be calculated. Fig. 3a showed the Gibbs energy, enthalpy, and entropy of adsorption of each peptide on octyl-Sepharose and Fig. 3b showed the thermodynamic parameters of peptides on octyl-CM-Sepharose. It was found that the entropies and enthalpies were all negative, which indicated that the adsorption was enthalpy driven in this lower temperature range 10–25 ◦ C. According to Lin et al. [19], the hydrophobic interaction was driven either by the entropy increase because of the repelling of ordered water from the hydrophobic surface, or by the negative enthalpy caused by the direct interaction between peptides and ligands. This result showed that the
178
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
Fig. 4. Effect of ligand length on the retention of peptides (GWWG, WGWG, GWGW, WGGW) onto butyl-, hexyl-, octyl-CM-Sepharose, and decyl-CMsepharose ligand density is 35, 35.3, 33.3, 39 mol/mL, respectively, at 298 K.
G◦ , H◦ or S◦ when the alkyl chain length increased from hexyl to octyl. This result is similar to Gao’s [20] observation that the alkyl side chains of adsorbents shifted from random coil to hypercoil as the carbon number exceeded six. However, Gao et al. found that the hypercoil conformation exhibited weaker hydrophobicity than the random coil state, which was different from our observations. 3.5. Effect of the carboxymethyl groups Fig. 3. G◦ (KJ/g mol),H◦ (KJ/g mol) and T S◦ (KJ/g mol) of peptides (GWWG, WGWG, GWGW, WGGW) onto (a) an octyl-Sepharose column (ligand density is 40 mol/mL) and (b) an octyl-CM-Sepharose column (ligand density is 33.3 mol/mL) at 298 K.
direct peptide–ligand interaction was the major driving force of the adsorption process. 3.4. Effect of alkyl chain length The capacity factors of peptides on various alkyl-CMSepharose columns were measured at 10–25 ◦ C. As shown in Fig. 4, the elution order of peptides on alkyl-CM-Sepharose columns were all the same but were different from that on octyl-Sepharose. The thermodynamic parameters were also calculated. Fig. 5a–d plotted the Gibbs free energy, enthalpy and entropy of adsorption of each peptide on resins of different alkyl chain length. It was found that the Gibbs free energy of adsorption increased as the increase of the alkyl chain length. The adsorption enthalpy and entropy also became more negative as the alkyl chain length increased. It seemed that the negativity of Gibbs free energy, enthalpy and entropy of adsorption increased as the increase in ligand’s hydrophobicity. It was also found that none of G◦ , H◦ and S◦ increased or decreased linearly with the increase in the number of carbon atoms in alkyl chains. There was a jump in
The elution order of peptides on alkyl-CM-Sepharose was different from that on octyl-Sepharose. We considered that the carboxymethyl groups on alkyl-CM-Sepharose were responsible for the alteration of the elution order. The capacity factors of these peptides on octyl-, decyl-CM-Sepharose of various octyl, decyl group densities were measured and the thermodynamic parameters were calculated. It was found from Fig. 4 that the elution order of peptides on all the resins was the same. However, the order of adsorption enthalpy varied as the ligand density increased. As shown in Fig. 6, the adsorption heat released from the peptide WGWG was lower than that from GWGW at a ligand density of 48 mol/mL. As the ligand density decreased, that is, as the carboxymethyl groups increased, the adsorption heat released from WGWG became higher than that from GWGW. We suspected that it was due to the electrostatic interaction between the carboxymethyl groups on the resin and the amino ends of the peptides. Although the column was operated at high ionic strength, the adsorption of peptides on the hydrophobic ligands might still bring the positively charged amino end of the peptide close enough to the negatively charged carboxymethyl group. The electrostatic interaction might cause the orientation of the peptides. Most peptides oriented their amino ends toward the solid surface because of the electrostatic interaction. Consequently, the carboxymethyl groups kept the peptides a small
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
179
Fig. 5. G◦ (KJ/g mol), H◦ (KJ/g mol) and T S◦ (KJ/g mol) of peptide (a) GWWG, (b) WGWG, (c) GWGW, and (d) WGGW onto butyl-, hexyl-, octyl, and decyl-CM-Sepharose columns at 298 K. The ligand density is 35, 35.3, 33.3, 39 mol/mL, respectively.
distance from the solid surface. As the distance between peptides and the solid surface increased, the interaction between peptides and the ligands decreased. As a result the adsorption enthalpies and entropies of peptides on octyl-CM-Sepharose
were all higher than those on octyl-Sepharose, which could be observed in Fig. 3a and b. Furthermore, as most peptides orient their amino ends toward the solid surface, the amino acid on the carboxyl end might not be reached by the octyl ligand. As a result, the carboxyl end tryptophans on the peptides GWGW and WGGW may be inaccessible to the octyl ligand, resulting in reduction of the adsorption enthalpy and entropy of GWGW and WGGW. This might be the reason why the retention of GWGW on alkyl-CM-Sepharose was less than that of WGWG.
4. Conclusion
Fig. 6. Effect of ligand density on the adsorption enthalpy of peptides on Octyl-CM-Sepharose ligand density is 18, 33.3, 47.9 mol/mL, respectively.
According to the elution order of these four peptides on the column of octyl-Sepharose, it was found that the amino acid tryptophan was more hydrophobic when it was located in the middle of a peptide chain than at the ends and the amino end tryptophan seemed less hydrophobic than the carboxyl one. The adsorption of these tryptophan containing peptides was driven by the adsorption enthalpy since both the enthalpy and entropy of adsorption were negative. As we increased the length of the alkyl ligand, it was found that the negativity of
180
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
the Gibbs free energy, enthalpy, and entropy of adsorption were all increased. It indicated that the increase in ligand chain length promoted the direct interactions between ligands and peptides. The elution order of peptides on negative charges containing hydrophobic resins was found different from that on octyl-Sepharose. We suspected that it was due to the electrostatic interaction between the surface charges on the resin and the positively charged amino end of the peptide. Acknowledgement The authors would like to thank the National Science Council of Taiwan for the financial support of this project (NCS 90-2214-E-033-008). References [1] G. Halperin, M. Tauber-Finkelstein, S. Shaltiel, J. Chromatogr. 317 (1984) 103. [2] M. Uchida, T. Kunitake, T. Kajiama, Int. J. Polym. Mater. 4 (1994) 199. [3] F. Macritchie, J. Colloid Interface Sci. 38 (1972) 484. [4] J.L. Piehler, C. Roisman, G. Schreiber, J. Biol. Chem. 275 (2000) 40425.
[5] R.J. Marsh, R.A.L. Jones, M. Sferrazza, Colloids Surf. B: Biointerfaces 23 (2002) 31. [6] E.A. Worobec, W. Paranchych, J.M.R. Parker, A.K. Taneja, R.S. Hodges, J. Biol Chem. 260 (1985) 938. [7] K.P. Murphy, E. Freire, Y. Paterson, Proteins 21 (1995) 83. [8] M. Pellegrini, A. Bisello, M. Rosenblatt, M. Chorev, F. Mierke, Anal. Biochem. 37 (1998) 12737. [9] M. Belew, Y. Zhou, S. Wang, L.E. Nystroem, J.C. Janson, J. Chromatogr. A 679 (1994) 67. [10] D. Riendeau, D.G. Harnish, R.C. Bleackley, V. Paetkau, J. Biochem. 258 (1983) 12114. [11] Y. Kato, T. Kitamura, K. Nakamura, A. Mitsui, Y. Yamasaki, T. Hashimoto, J. Chromatogr. 391 (1987) 395. [12] V. Stocchi, P. Cardoni, P. Ceccaroli, G. Piccoli, L. Cucchiarini, R. De Bellis, M. Dacha, J. Chromatogr. A 676 (1994) 51. [13] P. Ceccaroli, P. Cardoni, M. Buffalini, R. De Bellis, G. Piccoli, V. Stocchi, J. Chromatogr. B 702 (1997) 41. [14] J. Chen, J.D. Andrade, R.A. VanWagenen, Biomaterials 6 (1985) 231. [15] B.R. Young, W.G. Pitt, S.L. Cooper, J. Colloid Interface Sci. 125 (1988) 246. [16] D. Haidacher, A. Vailaya, Cs. Horvath, Proc. Natl. Acad. Sci. 93 (1996) 2290. [17] A. Vailaya, Cs. Horvath, Ind. Eng. Chem. Res. 35 (1996) 2964. [18] V.A. Basiuk, T.Y. Gromovoy, Colloids Surf. A: Physicochem. Eng. Aspects 118 (1996) 127. [19] F.Y. Lin, C.S. Chen, W.Y. Chen, S. Yamamoto, J. Chromatogr. A 912 (2001) 281. [20] J.Y. Gao, P.L. Dubin, Biopolymers 49 (1999) 185.
Effect of amino acid sequence on the hydrophobicity of small peptides Chih-I. Liu a,c , Ying-Chih Chan a , Wen-Yih Chen b , Ruoh-Chyu Ruaan b,∗ b
a Department of Chemical Engineering, Chung-Yuan University, Chung-Li 320, Taiwan Department of Chemical Engineering, National Central University, Chung-Li 320, Taiwan c Department of Nursing, Mei-Ho Institute of Technology, Pintung 912, Taiwan
Received 5 September 2004; received in revised form 7 June 2005; accepted 14 June 2005
Abstract Hydrophobic adsorbents, which contain alkyl ligands of various chain lengths, were used to study the hydrophobic interactions between alkyl chains and small peptides. We investigated the adsorption of four similar peptides: GWWG, GWGW, WGWG, WGGW. All of them contained two glycines (G) and two tryptophans (W) but the amino acids were arranged in different orders. The capacity factors of peptides between 10 and 35 ◦ C were measured and then the thermodynamic parameters, such as enthalpy and entropy changes (H◦ and S◦ ) of adsorption, were estimated. It implied that enthalpy was the major driving force in all the adsorption processes. Furthermore, H◦ and S◦ became more negative as the alkyl chain length was increased. It revealed that the van der Waal’s interaction had greater influence on the adsorption as the chain length increased. It was also found that, the contribution of each amino acid to peptide’s hydrophobicity was affected by the position of the amino acid. When a hydrophobic amino acid was positioned in the middle of a peptide chain, it exhibited the highest hydrophobicity. Interestingly, tryptophan at the carboxyl end was found more hydrophobic than it at the amino end of the peptide. © 2005 Elsevier B.V. All rights reserved. Keywords: Hydrophobic interaction; Gibbs free energy; Enthalpy; Entropy; Hydrophobic amino acids; Octyl; Decyl-CM-sepharose
1. Introduction Hydrophobic interaction played an important role in various biological systems and biochemical related applications. For example, the interaction between cells and artificial surface is strongly related to the adsorbed proteins on the surface [1–3]. The adsorption of proteins from liquid medium is often driven by the so-called hydrophobic interaction [4,5]. It was also a dominant interaction between the antigen–antibody binding [6,7] and the association between apolar hormones and their corresponding cell surface receptors [8]. Furthermore, the hydrophobic interaction chromatography is a direct application to use this type of interaction to separate biomolecules [9–13]. The study on how various peptides and ∗ Corresponding author. Tel.: +886 3 422 7151x34232; fax: +886 3 280 4341. E-mail address: [email protected] (R.-C. Ruaan).
0927-7757/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2005.06.023
proteins interact with hydrophobic surfaces has long been a major research focus [5,14,15]. Hydrophobic interaction is contributed not by a single force but by the joined efforts of various inter-molecular forces, such as van der Waal’s forces, electrostatic interactions and hydrogen bonds. Unlike electrostatic interaction that has been well characterized by Poisson equation, hydrophobic interaction can only be characterized experimentally because of its complexity. To find a way to quantify the hydrophobic interaction is an important matter. Hydrophobic interaction chromatography (HIC) has recently been adopted to study the hydrophobic interaction of proteins. Thermodynamics properties were used to characterize the retention behaviors. Haidacher et al. [16] investigated the effects of temperature on the retention of dansyl amino acids. They found the adsorption of dansyl amino acids on HIC columns (Spherogel, Synchropak propyl and butyl-NPR) was entropy-driven at low temperatures and
176
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
enthalpy-driven at high temperatures. Furthermore, Vailaya and Horvath [17] studied the effect of molecular structure and types of stationary phase on the adsorption enthalpy and entropy. They found that both the adsorption enthalpy and entropy of dansyl amino acids on HIC columns became more positive as the size of amino acids. The hydrophobicity of amino acids and proteins can be scaled by their retention factors on the column [18]. However, the hydrophobicity of a protein is not only affected by its amino acid composition but also related to the size, the secondary structure and the 3D structure of the protein. Therefore, the hydrophobicity of each amino acid has no direct correlation to the hydrophobicity to a protein, since the arrangement of amino acid in a peptide may also affect the overall hydrophobicity. Understanding the sequencedependent hydrophobic behavior for small molecules can be helpful for estimating the affect of sequence in large proteins. We attempt to estimate the hydrophobicity of small peptides of the same amino acid composition but in different sequences. Four peptides GWWG, GWGW, WGWG and WGGW were under investigation. The capacity factors of these peptides on various hydrophobic interaction columns are measured between 10 and 35 ◦ C. The Gibbs energy, enthalpy, and entropy of adsorption are calculated. The effect of amino acid sequence and the effect of solid surface are discussed.
alkylamine, EDC and NHS. The mixed was reacted for 20 h at 25 ◦ C in a shaker incubator. After reaction, the CM-Sepharose gel was washed continuously by 20% ethanol solution containing 0.2 M NaCl until no alkylamine can be detected. The eluted alkylamine was tested by ninhydrin reaction. The resulting alkyl-CM-Sepharose may be subject to repeated reaction to obtain higher alkyl substitution. The alkyl content was estimated by titrating the remaining carboxyl groups.
2. Experimental
2.4. Calculation of the thermodynamics parameters
2.1. Material
We may assume that the adsorption reaction takes place as follows.
Tetrahydrofuran (THF), pyridine, hydrochloride and sodium chloride were purchased from Riedel-deHa¨en (Chinosol, Seelze, Germany). Ammonium sulfate and tris(hydroxymethyl) aminomethane were purchased from Merck (Germany). Butyl-, hexyl-, octyl-, decylamine, 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide methiodode (EDC) were obtained from Aldrich (Milwaukee, WI, USA), Nhydroxysuccinimide (NHS), glycyl–glycyl–glycyl–glycine (GGGG), were purchased from Sigma (St. Louis, U.S.A). CM-Sepharose gel was purchased from Pharmacia Biotech. (Uppsala, Sweden). The peptides, glycyl–tryptophanyl–tryptophanyl–glycine (GWWG), tryptophanyl– glycyl–tryptophanyl–glycine (WGWG), glycyl–tryptophanyl–glycyl–tryptophan (GWGW) and tryptophanyl– glycyl–glycyl–tryptophanyl (WGGW) were obtained from Digital GENE Biosciences (Taipei, Taiwan). 2.2. Synthesis of the hydrophobic adsorbents The HIC adsorbent was synthesized by coupling alkylamine to CM-Sepharose particles through the carbodiimide mediated coupling reaction. Firstly CM-Sepharose was washed extensively with deionized water to remove ethanol. The resins were then washed with THF before the addition of
2.3. Chromatographic operation Hydrophobic interaction chromatography was carried out through a HPLC system which include a GL Sciences model PU-610 pump (Tokyo, Japan), a Lab Alliance model 500 UV monitor (PA, U.S.A.), a Spark model 816 autosampler (Emmen, Holland). The signals were processed through a Chem-Lab data station (SISC, Taipei, Taiwan). The volume of the sample loop is 100 L. Five milliliters of hydrophobic beads were packed in a 10 cm × 1 cm Sigma jacketed column at a flow rate of 1 mL/min. The final bed height was kept at 5.5 cm. The column temperature was controlled by Firstek B402-D (Shinjuang, Taiwan) circulation water bath. The column was first equilibrated with 20 mM Tris buffer (pH 6) containing (NH4 )2 SO4 (0.25 M) and all the samples were eluted isocratically in the same elution buffer. Eluent was monitored at UV 220 nm through a Lab Alliance UV detector (model 500).
M + L → ML
(1)
where M represents the adsorbed molecule and L represents the ligand on the adsorbent matrix. Then, the equilibrium constant could be calculated as follows. Keq =
CML CM CL
(2)
where CM , CL and CML are the concentrations of M, L and ML in the liquid phase. Keq =
ns 1 k = nm CL CL
(3)
where ns and nm represent the amount of solute in the stationary phase and mobile phase, respectively. The retention factor (k ) can be measured by chromatography and can be related to the equibrium constant (Keq ) as: k ≡
tR − t0 CML = = Keq CL t0 CM
where tR is the retention time of the sample and t0 is the retention time of non-retained molecules (GGGG). Therefore, k = Keq CL ≈ Keq Lm
(4)
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
177
Since CM is usually small, CL is close to Lm , the maximum concentration of ligand in the matrix. Lm can be calculated by the ligand density based on total volume (solid phase + mobile phase). Since G◦ = −RT ln Keq
(5)
and G◦ = H ◦ − T S ◦ ln Keq = −
H ◦ RT
+
S ◦ R
(6) (7)
so that S ◦ H ◦ + + ln Lm (8) RT R where G◦ denotes the standard Gibbs free change, H◦ and S◦ are the standard enthalpy and entropy changes. ln k = −
3. Results and discussions 3.1. The retention on octyl-Sepharose columns (k )
The capacity factor of the four peptides on a column packed with octyl-Sepharose resins were measured at various temperatures from 10 to 25 ◦ C. The natural logarithm of k was plotted against the inverse temperature and was shown in Fig. 1. It was found that the peptide GWWG was the most retentive for all the measured temperatures, followed by GWGW and WGWG. The peptide WGGW was the least retentive. Since there is no rigid secondary structure in such a small peptide, we might conclude from this result that the amino acid, tryptophan, exhibited the highest hydrophobicity when it was located in the middle of the peptide chain, but it’s hydrophobicity was reduced when it was located at the ends of the peptide, especially at the amino end.
Fig. 2. Van’t Hoff plot for the adsorption of peptides (GWWG, WGWG, GWGW, WGGW) onto the octyl-CM-Sepharose column. The ligand density is 33.3 mol/mL.
3.2. The retention on alkyl-CM-Sepharose columns We attached butyl-, hexyl-, octyl- and decyl functional groups on commercial CM-Sepharose particles via carbodiimide mediated method. The density of alkyl groups was controlled within 33.3–47.9 mol/mL gel. The capacity factors (k ) of the four peptides were measured. It was found in Fig. 2 that the elution of the four peptides from all the alkylCM-Sepharose columns followed the same order WGGW, GWGW, WGWG and then GWWG, which was different from that on octyl-Sepharose. It was an interesting phenomenon, which required further investigation. The retention time was measured from T = 283 to 318 K. It was found that the van’t Hoff plot was no longer linear at temperature higher than 303 K. We try to limit our estimation of H◦ within linear region to avoid nonlinear fitting. Therefore, the fitted data was restricted to the temperature lower than 303 K in enthalpy estimation by van’t Hoff plot. 3.3. The Gibbs energy, enthalpy and entropy of peptide adsorption
Fig. 1. Van’t Hoff plot for the adsorption of peptides (GWWG, WGWG, GWGW, WGGW) onto an octyl-Sepharose column (ligand density is 40 mol/mL).
The Gibbs energy adsorption could be calculated from Eq. (5) and the adsorption enthalpy could be estimated from the van’t Hoff’s plot. The entropy of adsorption could therefore be calculated. Fig. 3a showed the Gibbs energy, enthalpy, and entropy of adsorption of each peptide on octyl-Sepharose and Fig. 3b showed the thermodynamic parameters of peptides on octyl-CM-Sepharose. It was found that the entropies and enthalpies were all negative, which indicated that the adsorption was enthalpy driven in this lower temperature range 10–25 ◦ C. According to Lin et al. [19], the hydrophobic interaction was driven either by the entropy increase because of the repelling of ordered water from the hydrophobic surface, or by the negative enthalpy caused by the direct interaction between peptides and ligands. This result showed that the
178
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
Fig. 4. Effect of ligand length on the retention of peptides (GWWG, WGWG, GWGW, WGGW) onto butyl-, hexyl-, octyl-CM-Sepharose, and decyl-CMsepharose ligand density is 35, 35.3, 33.3, 39 mol/mL, respectively, at 298 K.
G◦ , H◦ or S◦ when the alkyl chain length increased from hexyl to octyl. This result is similar to Gao’s [20] observation that the alkyl side chains of adsorbents shifted from random coil to hypercoil as the carbon number exceeded six. However, Gao et al. found that the hypercoil conformation exhibited weaker hydrophobicity than the random coil state, which was different from our observations. 3.5. Effect of the carboxymethyl groups Fig. 3. G◦ (KJ/g mol),H◦ (KJ/g mol) and T S◦ (KJ/g mol) of peptides (GWWG, WGWG, GWGW, WGGW) onto (a) an octyl-Sepharose column (ligand density is 40 mol/mL) and (b) an octyl-CM-Sepharose column (ligand density is 33.3 mol/mL) at 298 K.
direct peptide–ligand interaction was the major driving force of the adsorption process. 3.4. Effect of alkyl chain length The capacity factors of peptides on various alkyl-CMSepharose columns were measured at 10–25 ◦ C. As shown in Fig. 4, the elution order of peptides on alkyl-CM-Sepharose columns were all the same but were different from that on octyl-Sepharose. The thermodynamic parameters were also calculated. Fig. 5a–d plotted the Gibbs free energy, enthalpy and entropy of adsorption of each peptide on resins of different alkyl chain length. It was found that the Gibbs free energy of adsorption increased as the increase of the alkyl chain length. The adsorption enthalpy and entropy also became more negative as the alkyl chain length increased. It seemed that the negativity of Gibbs free energy, enthalpy and entropy of adsorption increased as the increase in ligand’s hydrophobicity. It was also found that none of G◦ , H◦ and S◦ increased or decreased linearly with the increase in the number of carbon atoms in alkyl chains. There was a jump in
The elution order of peptides on alkyl-CM-Sepharose was different from that on octyl-Sepharose. We considered that the carboxymethyl groups on alkyl-CM-Sepharose were responsible for the alteration of the elution order. The capacity factors of these peptides on octyl-, decyl-CM-Sepharose of various octyl, decyl group densities were measured and the thermodynamic parameters were calculated. It was found from Fig. 4 that the elution order of peptides on all the resins was the same. However, the order of adsorption enthalpy varied as the ligand density increased. As shown in Fig. 6, the adsorption heat released from the peptide WGWG was lower than that from GWGW at a ligand density of 48 mol/mL. As the ligand density decreased, that is, as the carboxymethyl groups increased, the adsorption heat released from WGWG became higher than that from GWGW. We suspected that it was due to the electrostatic interaction between the carboxymethyl groups on the resin and the amino ends of the peptides. Although the column was operated at high ionic strength, the adsorption of peptides on the hydrophobic ligands might still bring the positively charged amino end of the peptide close enough to the negatively charged carboxymethyl group. The electrostatic interaction might cause the orientation of the peptides. Most peptides oriented their amino ends toward the solid surface because of the electrostatic interaction. Consequently, the carboxymethyl groups kept the peptides a small
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
179
Fig. 5. G◦ (KJ/g mol), H◦ (KJ/g mol) and T S◦ (KJ/g mol) of peptide (a) GWWG, (b) WGWG, (c) GWGW, and (d) WGGW onto butyl-, hexyl-, octyl, and decyl-CM-Sepharose columns at 298 K. The ligand density is 35, 35.3, 33.3, 39 mol/mL, respectively.
distance from the solid surface. As the distance between peptides and the solid surface increased, the interaction between peptides and the ligands decreased. As a result the adsorption enthalpies and entropies of peptides on octyl-CM-Sepharose
were all higher than those on octyl-Sepharose, which could be observed in Fig. 3a and b. Furthermore, as most peptides orient their amino ends toward the solid surface, the amino acid on the carboxyl end might not be reached by the octyl ligand. As a result, the carboxyl end tryptophans on the peptides GWGW and WGGW may be inaccessible to the octyl ligand, resulting in reduction of the adsorption enthalpy and entropy of GWGW and WGGW. This might be the reason why the retention of GWGW on alkyl-CM-Sepharose was less than that of WGWG.
4. Conclusion
Fig. 6. Effect of ligand density on the adsorption enthalpy of peptides on Octyl-CM-Sepharose ligand density is 18, 33.3, 47.9 mol/mL, respectively.
According to the elution order of these four peptides on the column of octyl-Sepharose, it was found that the amino acid tryptophan was more hydrophobic when it was located in the middle of a peptide chain than at the ends and the amino end tryptophan seemed less hydrophobic than the carboxyl one. The adsorption of these tryptophan containing peptides was driven by the adsorption enthalpy since both the enthalpy and entropy of adsorption were negative. As we increased the length of the alkyl ligand, it was found that the negativity of
180
Chih-I. Liu et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 266 (2005) 175–180
the Gibbs free energy, enthalpy, and entropy of adsorption were all increased. It indicated that the increase in ligand chain length promoted the direct interactions between ligands and peptides. The elution order of peptides on negative charges containing hydrophobic resins was found different from that on octyl-Sepharose. We suspected that it was due to the electrostatic interaction between the surface charges on the resin and the positively charged amino end of the peptide. Acknowledgement The authors would like to thank the National Science Council of Taiwan for the financial support of this project (NCS 90-2214-E-033-008). References [1] G. Halperin, M. Tauber-Finkelstein, S. Shaltiel, J. Chromatogr. 317 (1984) 103. [2] M. Uchida, T. Kunitake, T. Kajiama, Int. J. Polym. Mater. 4 (1994) 199. [3] F. Macritchie, J. Colloid Interface Sci. 38 (1972) 484. [4] J.L. Piehler, C. Roisman, G. Schreiber, J. Biol. Chem. 275 (2000) 40425.
[5] R.J. Marsh, R.A.L. Jones, M. Sferrazza, Colloids Surf. B: Biointerfaces 23 (2002) 31. [6] E.A. Worobec, W. Paranchych, J.M.R. Parker, A.K. Taneja, R.S. Hodges, J. Biol Chem. 260 (1985) 938. [7] K.P. Murphy, E. Freire, Y. Paterson, Proteins 21 (1995) 83. [8] M. Pellegrini, A. Bisello, M. Rosenblatt, M. Chorev, F. Mierke, Anal. Biochem. 37 (1998) 12737. [9] M. Belew, Y. Zhou, S. Wang, L.E. Nystroem, J.C. Janson, J. Chromatogr. A 679 (1994) 67. [10] D. Riendeau, D.G. Harnish, R.C. Bleackley, V. Paetkau, J. Biochem. 258 (1983) 12114. [11] Y. Kato, T. Kitamura, K. Nakamura, A. Mitsui, Y. Yamasaki, T. Hashimoto, J. Chromatogr. 391 (1987) 395. [12] V. Stocchi, P. Cardoni, P. Ceccaroli, G. Piccoli, L. Cucchiarini, R. De Bellis, M. Dacha, J. Chromatogr. A 676 (1994) 51. [13] P. Ceccaroli, P. Cardoni, M. Buffalini, R. De Bellis, G. Piccoli, V. Stocchi, J. Chromatogr. B 702 (1997) 41. [14] J. Chen, J.D. Andrade, R.A. VanWagenen, Biomaterials 6 (1985) 231. [15] B.R. Young, W.G. Pitt, S.L. Cooper, J. Colloid Interface Sci. 125 (1988) 246. [16] D. Haidacher, A. Vailaya, Cs. Horvath, Proc. Natl. Acad. Sci. 93 (1996) 2290. [17] A. Vailaya, Cs. Horvath, Ind. Eng. Chem. Res. 35 (1996) 2964. [18] V.A. Basiuk, T.Y. Gromovoy, Colloids Surf. A: Physicochem. Eng. Aspects 118 (1996) 127. [19] F.Y. Lin, C.S. Chen, W.Y. Chen, S. Yamamoto, J. Chromatogr. A 912 (2001) 281. [20] J.Y. Gao, P.L. Dubin, Biopolymers 49 (1999) 185.