Molecular Insight into Affinity Interaction between Cibacron Blue and

Aug 3, 2017 - ... Temperatures on the Flow-Induced Crystallization of Isotactic Polypropylene/Poly(ethylene terephthalate) Blends in Microinjection Mo...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/IECR

Molecular Insight into Affinity Interaction between Cibacron Blue and Proteins Juan Liang-Schenkelberg,† Georg Fieg, and Thomas Waluga* Institute of Process and Plant Engineering, Hamburg University of Technology, D-21073 Hamburg, Germany

ABSTRACT: Molecular dynamics (MD) simulations with all-atom models combining adsorption experiments are used to gain a better understanding of affinity interaction between proteins and Cibacron Blue (CB). These interactions with three different proteins are studied. The interactions with CB of different proteins are varying, although HSA and BSA have very similar structures. The influence of pH and salt concentration on these protein−CB bindings is investigated. Different parts of the CB molecule play miscellaneous roles in these interactions. The triazine part plays an important role in the hydrogen bonding between CB and bHb. The binding sites on protein surface differ by changing pH, salt concentration, and protein types. However, two parts seem to be very important to the binding of serum albumins.

1. INTRODUCTION

Despite the extensive application, the binding mechanisms between the dye ligand and proteins remain poorly understood. CB has a special chemical structure, so that its binding with the proteins is complicated (see Figure 1a). The anthraquinone chromophore has a structure of multiple aromatic, which makes it possible for CB to bind with the hydrophobic parts of the proteins. Several aromatics of CB have been substituted by sulfonic groups, so that it has also strong hydrophilic character to interact with proteins through electrostatic force for proper pH. Therefore, CB can bind many proteins by complex combination of electrostatic, hydrophobic, hydrogen bonding, and charge-transfer interactions.17,18 There are several factors that affect the affinity interaction of dye ligands, such as pH,9,19 ionic strength,9,20,21 and the density of immobilized ligands.19,22,23 The common methods used for the study of the interaction mechanisms between proteins and dye ligands are spectrometry,24−26 enzyme inhibition kinetics,27 affinity labeling,28 and X-ray diffraction (XRD).29 However, the binding mechanism between dye ligands and nonenzyme proteins (such as serum albumin and hemoglobin) is still not clear, since it could not be studied using enzyme inhibition kinetics methods. Moreover, most of the previous research works focus on the binding mechanisms between unbound dye ligands and the proteins. The dye ligands are mostly immobilized onto matrices in practical applications,

Affinity chromatography is a well-established method for the identification, purification, and separation of macromolecules based on highly specific molecular recognition. In this method, the target macromolecules are selectively adsorbed onto the chromatographic media through its affinity interaction with the ligands under favorable conditions. These adsorbed target macromolecules are then eluted by proper eluent under conditions favoring desorption. Dye−ligand affinity chromatography has been widely utilized for protein purification.1−3 Immobilized textile triazine dyesparticularly, Cibacron Blue 3GA (CB)has been used as an affinity chromatography tool for protein purification for a long time, because of its low cost, ease of immobilization, resistance to biological and chemical degradation, and the high protein-binding capacity of the corresponding adsorbents. CB is able to bind various proteins, such as kinases and dehydrogenases,4−6 phospholipase A2,7 lysozyme,8 γ-globulin,9 alkaline phosphatase,10 human interferon α-2b,11 serum albumin,12−14 reductases,15 and hemoglobin.16 However, its interaction with a large number of seemingly unrelated proteins inevitably compromises its protein binding specificity and endows their molecules with a serious drawback. One way to cope with the lack of specificity of chromatography is to use specific eluents that elute target proteins specifically with minimal contamination. Another strategy is to design new dye ligands for improved affinity and specificity of target protein. Both of them require a precise understanding of the binding and specific mechanisms between dye ligand and proteins. © XXXX American Chemical Society

Received: Revised: Accepted: Published: A

April 13, 2017 August 3, 2017 August 3, 2017 August 3, 2017 DOI: 10.1021/acs.iecr.7b01556 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Figure 1. (a) Structure of Cibacron Blue, being divided into five main parts. (b) Snapshot of the initial configuration of HSA simulation with 12 CB molecules. For a better view, water and ion particles are not shown here.

(CB) affinity chromatographic media. A combination of MD simulation and adsorption experiments will give a comprehensive understanding of the binding and specific mechanisms of the affinity adsorption from both microscopic and macroscopic point of view.

where the influence of factors such as steric hindrance and dyeligand density must be considered. Zhang performed adsorption equilibria experiments under different conditions to study the macroscopic adsorption mechanism of proteins onto dye-ligand adsorbents.9,30 The same experimental methods have also been well used in our former work for the investigation of ion-exchange adsorption of proteins.30 Besides, Zhang investigated also the adsorption of binary proteins onto dye-ligand affinity chromatography.16 However, Zhang16 focused mainly on the applicability of a mathematical model, studied only one initial concentration ratio of the proteins (1:10), without much exploration of the competition and specificity of the adsorption. Because of the complexity of the adsorption and the lack of effective experimental technique for a comprehensive investigation, the binding and specific mechanisms between immobilized dye ligand and nonenzyme proteins remain poorly understood. Molecular dynamics (MD) simulation is a research tool with sufficiently small scales in both time and space, to offer clear microscopic information in a direct manner. MD simulation computes the equilibrium and transport properties of a classical many-body system based on statistical mechanics (controlled by the laws of Newtonian dynamics), and provides a suitable method to analyze the interaction mechanism between molecules. It describes the time-dependent behavior of a molecular system. In MD simulation, the system can be visualized using all-atom (atomistic) models or coarse-grained (simplified) models, depending on the description precision required and the computational power provided. MD simulations have been successfully used to examine adsorption processes at different levels, including adsorbent modeling and visualization, adsorption of small molecules and protein adsorption on solid surfaces.32,33 It was also used to predict the binding properties of CB to cytosolic quinone reductase.15 In our former work, MD simulations with coarse-grained models have been successfully used to investigate the adsorption of serum albumin and hemoglobin onto ionexchange chromatographic media.34 The interaction between ion-exchange ligands and proteins as well as the binding sites of the proteins were studied. The results of the experiments and those of the MD simulations are qualitatively consistent.34,35 In this work, MD simulations are used innovatively to investigate the detail of dye-ligand affinity chromatographic adsorption of proteins. All-atom models are applied for a higher precision. Two serum albumins and hemoglobin are studied as model proteins to investigate their adsorption onto dye-ligand

2. METHODS 2.1. Molecular Details. The simulations were performed using the GROMACS 4.6 program.36 Since higher precision is required to investigate these complicated affinity interactions between protein and CB molecules, all-atom models were applied with Gromos 43a1 force field. The model of CB was created according to its all-atom structure (see Figure 1a). The all-atom models of proteins were obtained directly from the Protein Data Bank (HSA, PDB ID 1QRD; BSA, PDB ID 1AO6; bHb, PDB ID 1G0A). In the all-atom structure of both serum albumins, chain B, which is identical to chain A, was deleted. Only the protein atomistic structures were used. All other atoms in the crystal structures were removed. Simulation box (14 nm × 14 nm × 14 nm) with one protein and 12 CB molecules was set up (see Figure 1b). The distance between each CB and protein surface was kept to be 3.0 nm. The distances between every two CBs were set to be equal. The simulation box was further filled with water particles (SPC) to a density of 1 g/cm3. A certain number of counterions (sodium or chloride) were added in each simulation box to keep the system electrically neutral. To study the influence of pH value and salt concentration, simulations of each protein were run under three different conditions (pH 7 with 0.01 M salts, pH 7 with 0.1 M salts, and pH 12 with 0.02 M salts). In the simulations at pH 7, aspartic, glutamic acids, and C-terminus are negatively charged, while lysine, arginine, and N-terminus are positively charged. HSA and BSA have a charge of −15 and bHb has a charge of +2 in simulations at pH 7. At pH 12, lysine, arginine, and N-terminus do not charge anymore. HSA and BSA have a charge of −97, and bHb has a charge of −60. To keep the simulation box neutral, pH 7 and pH 12 was simulated with 0.01 and 0.02 M salt concentrations, respectively. To simulate the interactions under higher salt concentration, another condition was set at pH 7 with 0.1 M salt concentration. Periodic boundary conditions in the x-, y-, and z-directions were used. The reference temperature was coupled (coupling constant = 0.1 ps) to T = 298.15 K using a v-rescale thermostat. The pressure was coupled (coupling constant = 0.5 ps) to 1 bar using a B

DOI: 10.1021/acs.iecr.7b01556 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

former experimental research.37 For all three proteins, fewer CBs bind onto them at higher pH value. This is mainly due to the higher negative charge on protein surface at high pH, which results in stronger repulsion between proteins and negatively charged CB. For HSA, a stronger interaction is seen at pH 12, indicated by a lower value of potential energy. A possible reason for this contradiction may be more hydrogen bonding formed between HSA and CB under this condition (see Table 1). A lower amount of CB is bound, but they are bound more strongly, compared to the average interaction at pH 7. Comparing the results with salt concentrations at 0.01 and 0.1 M, it shows that more CBs bind onto HSA when salt concentration is higher. Higher salt concentration leads to a decrease in electrostatic repulsion between HSA and CB. It indicates the importance of HSA-CB electrostatic interaction under this condition. Different from HSA, much less hydrogen bonding (0.66) is seen between BSA and CB at pH 12, although the interaction between them is not obviously influenced. Thus, BSA-CB hydrophobic interaction is the main binding force under this condition. The strength and hydrogen bonding of bHb interactions with CB do not change much with pH or salt concentration. The CB−protein interaction is a complex combination of electrostatic, hydrophobic interactions and hydrogen bonding. For a deeper understanding of these interactions, further analysis is performed by comparing the contributions of different CB parts to protein−CB interaction potential energies. According to different properties, CB molecule is divided into five parts (see Figure 1a), including the multiple aromatic part (Part I), three sulfonic groups (Part II, III, and IV), and the rest of CB with triazine (Part V). The contributions of these five CB parts to their interaction with HSA, BSA, and bHb at the end of simulations, when the interactions reach equilibrations, are presented in Figure 2. As can be seen in Figure 2, all parts of the CB molecule interact with these three proteins under all conditions studied. It must be mentioned that Part V is usually used for CB immobilization.17 Therefore, it will not be considered in the following discussion. The interactions of different parts differ according to the protein type, pH, and salt concentration. It again indicates that the affinity interaction between protein and CB is a complex combination. Figure 2 shows that the interactions of CB parts with HSA and BSA are similar, especially at pH 7 with 0.1 M salt concentration. This is because of the similar structure of HSA and BSA. At higher salt concentration, which leads to reduce of electrostatic repulsion, a steep increase of Part IV interaction is seen for HSA and BSA. However, difference is obviously seen for Part I and Part II. Comparing with at pH 12, at pH 7 with 0.01 M ions HSA interacts stronger with CB through Part I and BSA bonds more with Part II of the CB molecule. This again indicates the difference in interactions between CB and different serum albumins. For bHb, change of pH has an obvious influence on the bindings of CB Part II and CB Part IV. Increase of pH results in stronger interaction of Part II and weaker interaction of Part IV. Combining the results shown in Table 1, showing that less hydrogen bonding is formed at pH 12, which leads to less binding of CB on bHb, it can be concluded that Part IV is the key part for hydrogen bonding between CB and bHb. In order to understand these affinity interactions more in detail and to provide favorable initial orientations for the further simulations of adsorption processes, the binding sites of proteins are particularly interesting. To obtain these binding

Berendsen thermostat. All systems were energy-minimized before starting a MD simulation. MD simulations performed in this study contain two main parts: (1) 500 ps pre-equilibration of system with protein atoms restrained under the isothermal−isobaric (NPT) ensemble and under canonical (NVT) ensemble, respectively; and (2) 40 ns MD simulation with an unrestrained protein. The second part of simulation was repeated three times for each protein under each situation. Both parts, as well as the energy minimization, have been calculated and analyzed. Only the second part is going to be discussed here. The results were calculated and analyzed by the GROMACS suite of programs, Excel and Matlab. 2.2. Experiments. BSA and bHb were purchased from Sigma−Aldrich. CB-Separopore was purchased from bioWORLD. Other reagents were of analytical grade, obtained from local suppliers, and used without further treatment. Static adsorption experiments of each protein onto CB-Separopore were performed at 25 °C. The dye-ligands affinity adsorbents were equilibrated with the corresponding buffer before use in the static adsorption experiments, as described elsewhere.31 After the equilibration of the adsorption, the concentrations of protein (C) in supernatant were determined with spectrophotometer. The adsorption density (Q) of proteins will be calculated by mass balances. The adsorption isotherms of both proteins at two pH values (pH 6 and pH 7) with three different salt concentrations of sodium chloride (0, 0.05, and 0.1 M) were obtained.

3. RESULTS AND DISCUSSION 3.1. Simulation Results. A summary of the simulation results is shown in Table 1. The potential energy and hydrogen bond number in this table are average values of simulations under the same conditions. Table 1. Simulation Results of HSA, BSA, and bHb at Different pH Values and Salt Concentrationsa number of bound CB

interaction potential energy (kJ/mol)

number of hydrogen bonding

protein

pH

Cs (M)

HSA

7 7 12

0.01 0.1 0.02

11/36 19/36 11/36

−274.26 −311.95 −373.71

0.83 0.77 1.16

BSA

7 7 12

0.01 0.1 0.02

16/36 16/36 10/36

−375.71 −381.42 −353.03

1.06 1.20 0.66

bHb

7 7 12

0.01 0.1 0.02

17/36 18/36 8/36

−295.16 −313.02 −306.86

1.12 1.01 0.90

a

All simulations were repeated three times, each with 12 CB molecules. This gives a total number of 36 bound CB (maximum).

As can be seen in this table, CBs bind onto all three proteins under all conditions studied. This result is consistent with the former research works that affinity binding happens between CB and these three proteins. Obvious influence of both pH and salt concentration can be seen, which is complicated due to the complex affinity mechanisms. Difference is seen between BSA and HSA, although these two proteins have similar structures. This difference in their binding with CB has been reported by C

DOI: 10.1021/acs.iecr.7b01556 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Figure 2. Contribution percentages of five main parts of CB (see Figure 1a) in protein−CB interactions under different conditions.

Figure 3. Contact percentages of each amino acid residues on HSA, BSA, and bHb in their interactions with CB under different conditions.

albumin are very important to their interactions with the CB molecule. However, specific binding sites are also seen for both serum albumins. At pH 7 with 0.01 M ions, two specific binding sites for HSA are seen between residue 110 and residue 180. At pH 12, specific binding sites of BSA are located between residue 150 and residue 290. At pH 7, where bHb is not much charged, its binding sites scatter all over its surface. This is one reason why more CB molecules are bound onto bHb at pH 7, as discussed above. No specific binding sites are seen. It indicates the importance of electrostatic interaction on bHb− CB binding when bHb does not charge. At pH 12, the binding sites concentrated in chain D. These binding sites on proteins provide very important information to design the further adsorption simulations. Protein initial orientations will be set according to these binding sites, so that the key parts on the protein surface are placed near the ligands. It will result in effective calculation, saving much computing resources. 3.2. Experimental Results. Adsorption isotherms of BSA and bHb onto CB-Separopore at pH 6 and pH 7 with 0, 0.05, and 0.1 M NaCl are shown in Figure 4. The presented isotherms show a good comparability with reported isotherms

residues, the last 5 ns of each simulation when the binding reaches equilibration were calculated. According to the default value of GROMACS of contacting distance, the residues with an average distance in these 5 ns to CB molecules smaller than 0.6 nm were taken as contacted. The frequency of each residue of proteins, acting as binding residues for each interaction with a CB molecule, is presented in Figure 3. The difference of binding sites on each protein under different conditions indicates again the influence of pH and salt concentration on the protein−CB interactions. As discussed above, for all three of these proteins, fewer CB molecules bound at pH 12 (see Table 1). Figure 3 also shows that the binding sites of all proteins at pH 12 (blue) are concentrated, so fewer CBs are able to reach close enough to these binding sites. Comparing the distribution of binding sites on HSA and BSA, they are shown in similar positions under the same conditions. This is due to the similar structure of these two proteins. For both HSA and BSA, there are two parts on their surface acting as binding sites under most of the conditions. One is between residues 350 and 380, the other is between residues 550 and 570. It means these two parts of serum D

DOI: 10.1021/acs.iecr.7b01556 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

Figure 4. Isotherms of protein adsorption under different conditions. Lines are fitted isotherms by a Langmuir model.

decrease in its adsorption, while further increases in salt concentration does not influence the isotherm much anymore. It indicates electrostatic force is the main driving force with lower salt concentration. As the salt concentration increases, the hydrophobic interaction between CB and bHb gets stronger.

for comparable adsorption systems of CB-modified adsorbents and serum albumin, as well as bHb.14,38,39 The obtained isotherms are described with a Langmuir model. As can be seen in Figure 4, both BSA and bHb reach stronger adsorption at pH 6. At pH 6, an obvious influence of salt concentration on adsorption density is seen for both proteins. One reason is that protein is less negatively charged at pH 6, so that the repulsion with sulfonic groups on CB reduces, which benefits adsorption. An unusual trend is seen for BSA with a lower adsorption capacity without NaCl present. At pH 7, adsorption density of BSA is not affected much by the change of salt concentration. The same phenomena is seen for bHb at pH 7 when Cs > 0 M NaCl. This is consistent with the simulation results shown in Table 1, that the number of bound CBs and the interaction potential energy, as well as the number of hydrogen bonds of both proteins, do not change with salt concentration increasing at pH 7. Comparing the adsorption of these two proteins, bHb adsorbs stronger at pH 6 and pH 7. This can be explained by the simulation results in Figure 3 that bHb has more possible binding sites than BSA over its surface at pH 7. Another possible reason is the diffusion of protein in chromatographic pore. At pH ∼7, BSA charges more than bHb with stronger electrostatic repulsion between proteins. For BSA, salt concentration has different influence on its adsorption at pH 6 and pH 7. This again indicates the complexity of affinity interaction between protein and CBs. At pH 6, best adsorption happens with 0.05 M NaCl. As the salt concentration is increased, the adsorption of BSA increases first and then decreases. This is a result of a combination of hydrophobic and electrostatic interactions. For bHb, the increase of salt concentration at pH 6 leads to the decrease of its adsorption density. It indicates the importance of electrostatic interaction for bHb at pH 6 by a reduced protein− surface electrostatic attraction at high salt concentrations.31 At pH 7, the adsorption of bHb seems to be very sensible to small amounts of salt concentration. 0.05 M NaCl leads to a steep

4. CONCLUSIONS The all-atom models and force field have been used to investigate the affinity binding and adsorption of HSA, BSA, and bHb onto CB using MD simulations. It was shown that the simulations were capable to study the mechanism of protein interaction, as well as the influence of pH and salt concentrations in detail. The adsorption experiments have been performed so that the results can be compared with the simulations results. Different proteinseven both serum albuminsshow different binding mechanism with CB with different influence from pH and salt concentration. The contributions of different CB parts to the interactions with different proteins have been analyzed and compared. The key part of CB that forms hydrogen bonding with bHb is found. Important residues on the protein surface for the interactions are discussed under different conditions. Interesting phenomena have been observed and discussed. To summarize, from the former work of this project, the interaction and adsorption mechanisms of different proteins have been studied from both microscopic and macroscopic points of view. The obtained results lead to a better understanding of the complicated affinity interaction between proteins and CB. Moreover, it provides a basis for the further investigations of interaction between protein and small molecules.



AUTHOR INFORMATION

Corresponding Author

*Tel.: ++49 (0)40 42878-3141. Fax: ++49 (0)40 42878-2992. E-mail: [email protected]. E

DOI: 10.1021/acs.iecr.7b01556 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research ORCID

interferon alpha-2b with Cibacron Blue F3G-A. J. Anal. Chem. 2003, 58 (11), 1038−1041. (12) Compagnini, A.; Fisichella, S.; Foti, S.; Maccarrone, G.; Saletti, R. Isolation by gel-permeation chromatography of a non-covalent complex of Cibacron Blue F3G-A with human serum albumin. J. Chromatogr. A 1996, 736 (1−2), 115−123. (13) Andac, M.; Galaev, I.; Denizli, A. Dye attached poly(hydroxyethyl methacrylate) cryogel for albumin depletion from human serum. J. Sep. Sci. 2012, 35, 1173−1182. (14) Uzun, L.; Yavuz, H.; Say, R.; Ersöz, A.; Denizli, A. Poly(ethylene dimethacrylate-glycidyl methacrylate) Monolith as a Stationary Phase in Dye-Affinity Chromatography. Ind. Eng. Chem. Res. 2004, 43 (20), 6507−6513. (15) Andac, C. A.; Andac, M.; Denizli, A. Predicting the binding properties of cibacron blue F3GA in affinity separation systems. Int. J. Biol. Macromol. 2007, 41, 430−438. (16) Li, W.; Zhang, S. P.; Sun, Y. Modeling of the linear-gradient dyeligand affinity chromatography with a binary adsorption isotherm. Biochem. Eng. J. 2004, 22 (1), 63−70. (17) Denizli, A.; Piskin, E. Dye-ligand affinity systems. J. Biochem. Biophys. Methods 2001, 49 (1−3), 391−416. (18) Gutenwik, J.; Nilsson, B.; Axelsson, A. Coupled diffusion and adsorption effects for multiple proteins in agarose gel. AIChE J. 2004, 50 (12), 3006−3018. (19) Ruckenstein, E.; Zeng, X. F. Albumin separation with Cibacron Blue carrying macroporous chitosan and chitin affinity membranes. J. Membr. Sci. 1998, 142 (1), 13−26. (20) Gianazza, E.; Arnaud, P. Chromatogarphy of plasma proteins on immobilized Cibacron Blue F3-GA. Mechanism of the molecular interaction. Biochem. J. 1982, 203 (3), 637−641. (21) Zhang, S. P.; Sun, Y. Ionic strength dependence of protein adsorption to dye-ligand adsorbents. AIChE J. 2002, 48 (1), 178−186. (22) He, L. Z.; Gan, Y. R.; Sun, Y. Adsorption-desorption of BSA to highly substituted dye-ligand adsorbent: quantitative study of the effect of ionic strength. Bioprocess Eng. 1997, 17 (5), 301−305. (23) Boyer, P. M.; Hsu, J. T. Effects of ligand concentration on protein adsorption in dye-ligand adsorbents. Chem. Eng. Sci. 1992, 47 (1), 241−251. (24) Sereikaite, J.; Bumeliene, Z.; Bumelis, V. A. Bovine serum albumin-dye binding. Acta Chromatogr. 2005, 15, 298−307. (25) Sereikaite, J.; Bumelis, V.-A. Examination of dye-protein interaction by gel-permeation chromatography. Biomed. Chromatogr. 2006, 20 (2), 195−199. (26) Basar, N.; Uzun, L.; Gü ner, A.; Denizli, A. Spectral characterization of lysozyme adsorption on dye-affinity beads. J. Appl. Polym. Sci. 2008, 108 (6), 3454−3461. (27) Bornmann, L.; Hess, B. Interaction of Cibacron dyes with dehydrogenases and kinases. Z. Naturforsch. C 1977, 32 (9−10), 756− 759. (28) Small, D. A. P.; Lowe, C. R.; Atkinson, T.; Bruton, C. J. Affinity labeling of enzymes with triazine dyes. Isolation of a peptide in the catalytic domain of horse-liver alcohol dehydrogenase using Procion Blue MX-R as a structural probe. Eur. J. Biochem. 1982, 128 (1), 119− 123. (29) Biellmann, J. F.; Samama, J. P.; Bränden, C. I.; Eklund, H. X-ray studies of the binding of Cibacron blue F3GA to liver alcohol dehydrogenase. Eur. J. Biochem. 1979, 102 (1), 107−110. (30) Zhang, S. P.; Sun, Y. Ionic strength dependence of protein adsorption to dye-ligand adsorbents. AIChE J. 2002, 48 (1), 178−186. (31) Liang, J.; Fieg, G.; Shi, Q. H.; Sun, Y. Single and binary adsorption of proteins on ion-exchange adsorbent: The effectiveness of isothermal models. J. Sep. Sci. 2012, 35 (17), 2162−2173. (32) Zhang, L.; Sun, Y. Molecular simulation of adsorption and its implications to protein chromatography: A review. Biochem. Eng. J. 2010, 48 (3), 408−415. (33) Basconi, J. E.; Carta, G.; Shirts, M. R. Multiscale modeling of protein adsorption and transport in macroporous and polymer-grafted ion exchangers. AIChE J. 2014, 60 (11), 3888−3901.

Thomas Waluga: 0000-0001-5308-3338 Present Address †

Agilent Technologies, Inc., D-76337 Waldbronn, Germany.

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the German Research Foundation (Project No. FI 1452/7-1).



ABBREVIATIONS MD = molecular dynamics HSA = human serum albumin BSA = bovine serum albumin bHb = bovine hemoglobin CB = Cibacron Blue



NOMENCLATURE C = liquid-phase concentration of protein in the bulk phase, mmol/L Cs = salt concentration in mobile phase, M Q = solid-phase protein concentration based on the volume of wet gel, mmol/L



REFERENCES

(1) Arica, M. Y.; Testereci, H. N.; Denizli, A. Dye-ligand and metal chelate poly (2-hydroxyethylmethacrylate) membranes for affinity separation of proteins. J. Chromatogr. A 1998, 799 (1−2), 83−91. (2) Gianazza, E.; Arnaud, P. A general method for fractionation of plasma-proteins. Dye-ligand affinity chromatography on immobilized Cibacron blue F3-GA. Biochem. J. 1982, 201 (1), 129−136. (3) McCreath, G. E.; Chase, H. A.; Owen, R. O.; Lowe, C. R. Expanded bed affinity chromatography of dehydrogenases from bakers’ yeast using dye-ligand perfluoropolymer supports. Biotechnol. Bioeng. 1995, 48 (4), 341−354. (4) Subramanian, S.; Ross, P. D. Dye-ligand affinity chromatography: the interaction of Cibacron Blue F3GA with proteins and enzymes. Crc Cr. Rev. Bioch. Mol. 1984, 16 (2), 169−205. (5) Reuter, R.; Naumann, M.; Güttel, K.; Hofmann, E. Interactions of immobilized and free triazine dyes with glucose-6-phosphate dehydrogenase from yeast. Biomed. Biochim. Acta 1986, 45 (3), 273−280. (6) Tejedor, M. C.; Delgado, C.; Grupeli, M.; Luque, J. Affinity partitioning of erythrocytic phosphofructokinase in aqueous two-phase systems containing poly(ethylene glycol)-bound cibacron blue. Influence of pH, ionic strength and substrates/effectors. J. Chromatogr. 1992, 589 (1−2), 127−134. (7) Barden, R. E.; Darke, P. L.; Deems, R. A.; Dennis, E. A. Interaction of phospholipase A2 from cobra venom with Cibacron Blue F3GA. Biochemistry 1980, 19 (8), 1621−1625. (8) Hubble, J.; Mayes, A. G.; Eisenthal, R. Spectral analysis of interactions between proteins and dye ligands. Anal. Chim. Acta 1993, 279 (1), 167−177. (9) Zhang, S. P.; Sun, Y. Further studies on the contribution of electrostatic and hydrophobic interactions to protein adsorption on dye-ligand adsorbents. Biotechnol. Bioeng. 2001, 75 (6), 710−717. (10) Kirchberger, J.; Seidel, H.; Kopperschläger, G. Interaction of procion red HE-3B and other reactive dyes with alkaline phosphatase: A study by means of kinetic, difference spectroscopic and chromatographic methods. Biomed. Biochim. Acta 1987, 46 (10), 653−663. (11) Bumeliene, Z.; Sereikaite, I.; Bumelis, V. A.; Smirnovas, V.; Gedminiene, G.; Braziunaite, L.; Bajorunaite, E. Determination of the dissociation constant and stoichiometry of a complex of the protein F

DOI: 10.1021/acs.iecr.7b01556 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research (34) Liang, J.; Fieg, G.; Keil, F. J.; Jakobtorweihen, S. Adsorption of Proteins onto Ion-Exchange Chromatographic Media: A Molecular Dynamics Study. Ind. Eng. Chem. Res. 2012, 51 (49), 16049−16058. (35) Liang, J.; Fieg, G.; Jakobtorweihen, S. Molecular dynamics simulations of a binary protein mixture adsorption onto ion-exchange adsorbent. Ind. Eng. Chem. Res. 2015, 54, 2794−2802. (36) Pronk, S.; Pall, S.; Schulz, R.; Larsson, P.; Bjelkmar, P.; Apostolov, R.; Shirts, M. R.; Smith, J. C.; Kasson, P. M.; van der Spoel, D.; Hess, B.; Lindahl, E. GROMACS 4.5: A high-throughput and highly parallel open source molecular simulation toolkit. Bioinformatics 2013, 29 (7), 845−854. (37) Leatherbarrow, R. J.; Dean, P. D. G. Studies on mechanism of binding of serum albumins to immobilized Cibacron blue F3GA. Biochem. J. 1980, 189, 27−34. (38) Zhang, S.; Sun, Y. A predictive Model for Salt Effects on the Dye-Ligand Affinity Adsorption Equlibrium of Protein. Ind. Eng. Chem. Res. 2003, 42 (6), 1235−1242. (39) Kassab, A.; Yavuz, H.; Odabasi, M.; Denizli, A. Human serum albumin chromatography by Cibacron blue F3GA-derived microporous polyamide hollow fibre affinity membranes. J. Chromatogr., Biomed. Appl. 2000, 746, 123−132.

G

DOI: 10.1021/acs.iecr.7b01556 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX