The effects of arginine on protein binding and elution in hydrophobic interaction and ion-exchange chromatography

Protein Expression and Purification 54 (2007) 110–116 www.elsevier.com/locate/yprep

The effects of arginine on protein binding and elution in hydrophobic interaction and ion-exchange chromatography Tsutomu Arakawa a, Kouhei Tsumoto b, Kazuo Nagase c, Daisuke Ejima b

c,*

a Alliance Protein Laboratories, Thousand Oaks, CA 91360, USA Department of Biomolecular Engineering, Graduate School of Engineering, Tohoku University, Aoba-ku, Sendai 980-8579, Japan c Ajinomoto Co. Inc., 1-1 Suzuki-cho, Kawasaki-ku, Kawasaki 210-8681, Japan

Received 24 December 2006, and in revised form 15 February 2007 Available online 27 February 2007

Abstract Arginine is effective in suppressing aggregation of proteins and may be beneficial to be included during purification processes. We have shown that arginine reduces non-specific protein binding in gel permeation chromatography and facilitates elution of antibodies from Protein-A columns. Here we have examined the effects of arginine on binding and elution of the proteins during hydrophobic interaction (HIC) and ion-exchange chromatographies (IEC) using recombinant monoclonal antibodies (mAbs) and human interleukin-6. In the case of HIC, the proteins were bound to a phenyl-Sepharose column in the presence of ammonium sulfate (AS) with or without arginine and eluted with a descending concentration of AS. While use of 1 M AS in the loading buffer resulted in complete binding of the mAb, inclusion of 1 M arginine in loading and equilibration buffer, only when using low-substituted phenyl-Sepharose, resulted in weaker binding of the proteins. While decreasing AS concentration to 0.75 M resulted in partial elution of the mAB, elution was facilitated with inclusion of 0.5–1 M arginine. In the case of IEC, arginine was included in the loading samples. Inclusion of arginine during binding to the IEC columns resulted in a greater recovery and less aggregation even when elution was done in the absence of arginine. These results indicate that arginine enhances elution of proteins bound to the resin, suggesting its effectiveness as a solvent for elution in HIC and IEC.  2007 Elsevier Inc. All rights reserved. Keywords: Arginine; Hydrophobic interaction chromatography; Ion-exchange chromatography; Elution; Antibody; Ammonium sulfate

We and other groups have shown that arginine is effective in suppressing aggregation of proteins [1,2] and hence may be beneficial to be included at moderate concentrations during column chromatography. It is critical therefore to understand the impact arginine may have on performance of various chromatographies. We have shown that arginine has a positive impact on Protein-A [3,4], gel permeation [5] and dye-affinity chromatographies [6]; i.e., arginine facilitated elution of antibodies from Protein-A under mild pH, reduced non-specific protein binding in gel permeation chromatography and increased the recovery in dye-affinity chromatography. Here we have examined

the effects of arginine on protein binding and elution in hydrophobic interaction (HIC)1 and ion-exchange chromatography (IEC), which constitute the important analytical and preparative separation systems for proteins [7,8]. HIC exploits the weak hydrophobic surface of both native proteins and column matrix and hence requires strong salting-out solvents for binding [9–14]. In general, proteins are bound to HIC columns at high ammonium sulfate (AS) concentration followed by elution with a descending concentration of AS. Although AS is primarily used for HIC, other salts, including glutamic acid (Na salt), sodium

1

*

Corresponding author. Fax: +81 44 244 5809. E-mail address: [email protected] (D. Ejima).

1046-5928/$ - see front matter  2007 Elsevier Inc. All rights reserved. doi:10.1016/j.pep.2007.02.010

Abbreviations used: HIC, hydrophobic interaction chromatography; IEC, ion exchange chromatography; mAb, monoclonal antibody; AS, ammonium sulfate; IL-6, interleukin-6; AcONa, sodium acetate.

T. Arakawa et al. / Protein Expression and Purification 54 (2007) 110–116

sulfate, guanidinium sulfate, aspartic acid (Na salt) and NaCl, have also been used [15,16]. The most commonly used salt for elution in IEC is NaCl, although other salts can be used to modulate the ionic strength. Arginine is also effective in solubilizing proteins from insoluble pellets [17,18], increases the solubility of the proteins [19] and assists in protein refolding [20–22]. In general, arginine is removed prior to the subsequent purification steps. If arginine does not interfere with the protein binding in the subsequent steps, then it can simply be removed upon elution, which makes an extra step of arginine removal unnecessary. Purification techniques, which work in the presence of arginine, are desired. The present results show that the proteins to be purified do bind to the HIC and IEC columns in the presence of arginine and the bound proteins can be eluted either with or without arginine. In addition, it is demonstrated that inclusion of arginine increases the recovery of the proteins from the columns and results in less aggregation. Materials and methods Materials Recombinant human interleukin-6 (IL-6) was prepared according to the methods reported previously [23]. Two humanized monoclonal antibodies, IgG4-A and IgG4-B, and a mouse monoclonal antibody, IgG1, were prepared in Ajinomoto Co. All reagents are of biochemical research grade. Arginine stock solution was prepared from L-arginine hydrochloride. Methods Hydrophobic interaction chromatography: A 1 ml HiTrap low-substituted and high-substituted phenylSepharose columns (Amersham Biosciences) were equilibrated at a flow rate of 0.5 or 1 ml/min with an appropriate buffer containing AS in 10 mM phosphate. In some cases, arginine was included in the equilibration buffer and loading sample to mimic the condition at which proteins to be purified are exposed to arginine in the prior step of protein purification process. An appropriate volume of protein samples, 2 M AS and 2 M arginine were mixed to generate loading samples as indicated in each experiment. In all experiments, 4 mg of the proteins were loaded on the column. Column chromatographies were carried out at 4 C. The amount of the proteins eluted in each elution condition was spectrophotometrically determined and expressed as a percentage to the amount of total proteins loaded. Ion-exchange chromatography: Following the previous observation [23] that inclusion of sodium acetate in the loading sample increases the recovery of monomeric IL-6, the effects of replacing sodium acetate with arginine or NaCl were examined. The HiTrap CM-Sepharose IEC column (1.6 · 10 cm) was equilibrated with 10 mM sodium acetate, pH 5.8. After loading the IL-6 sample in an appro-

111

priate solvent (as indicated in the figures), the column was eluted with a linear gradient of sodium acetate. For mAb, it is customary to have IEC chromatography following Protein-A step and hence the IEC was done after elution of the mAb from the Protein-A chromatography process. The mAb was bound to the 1 ml HiTrap Protein-A column and eluted with 0.5 M arginine at pH 4. Then the conditions for binding of the Protein-A pool to the IEC column and subsequent elution from the column were examined. Results Hydrophobic interaction chromatography In the first experiment, 4 mg of IgG4-A were loaded to the phenyl-Sepharose (low-substituted) in 1 M AS, 10 mM phosphate, pH 6.0, which resulted in complete binding. Elution was initiated by decreasing AS concentration to 0.75 M AS in 10 mM phosphate, pH 6.0, resulting in elution of only a fraction of the protein applied (16%, first column) and the remaining bound proteins eluting with lower AS concentrations, as summarized in Fig. 1. It is evident that 0.75 M AS is insufficient for effective elution of the bound IgG4-A. Next, 1 M arginine was included in the 0.75 M AS elution buffer, resulting in much higher recovery (45%, second column of Fig. 1). In these experiments, arginine was not included in the equilibration buffer and loading sample. In order to mimic the condition of arginine-containing samples [11,12], the following experiments were carried out to load the same mAb in the presence of arginine. First, the HIC column was equilibrated with 1 M AS/0.5 M arginine and then the mAb in 1 M AS/0.5 M arginine loaded. Under this condition, a small amount of the protein flowed-through the column (data not shown), indicating that 0.5 M arginine weakens

70 60 50 40 30 20 10 0 None

1 M Arg

Loading without arginine

0.5 M Arg

1 M Arg

1 M Arg

Loading with arginine

Fig. 1. Elution of IgG4-A from low-substituted phenyl-Sepharose using 0.75 M AS with and without arginine. In all cases, 4 mg IgG4-A were loaded in the presence of 1 M AS. Although the elution conditions were identical for the fourth and fifth columns, binding conditions were different, i.e., 1 M AS/1 M Arg for the fourth column and 1 M AS/0.5 M Arg for the fifth column.

112

T. Arakawa et al. / Protein Expression and Purification 54 (2007) 110–116

)

60

Total recovery yield

hydrophobic interaction of the protein with the HIC column. After washing the column with 1 M AS/0.5 M arginine, the bound protein was eluted with 0.75 M AS/ 0.5 M arginine (third column of Fig. 1). The elution recovery with the 0.75 M AS/0.5 M arginine was intermediate (30%) between 0.75 M AS alone (first column) and 0.75 M AS/1 M arginine (second column), suggesting that the ability of arginine to dissociate proteins from ProteinA is concentration-dependent. Arginine concentration for loading was increased to 1 M arginine; i.e., the column was equilibrated with 1 M AS/1 M arginine and the mAb in the same buffer was loaded. The flow-through peak increased compared to that observed with 1 M AS/0.5 M arginine (data not shown), indicating that an increase in arginine concentration to 1 M further weakened binding of the mAb to the HIC column. Elution was more efficient with 0.75 M AS/1 M arginine than with 0.75 M AS/0.5 M arginine (52%, fourth column). Based on the above experiments, 0.5 M arginine (and probably less arginine) is more optimal for binding, while elution is efficient with 1 M arginine (and probably higher arginine concentrations). Therefore, an experiment shown in the fifth column was carried out, in which the mAb was bound in 1 M AS/0.5 M arginine to the HIC column equilibrated with the same buffer. The flow-through peak decreased compared to that with 1 M AS/1 M arginine, consistent with the experiment of the third column. A sharp elution peak was observed with 0.75 M AS/1 M arginine with a recovery of 66%. Based on these observations, one should adjust the arginine concentration of the loading sample to an optimal level for complete binding and achieve elution by an appropriate concentrations of AS and arginine. Such a concern appears unnecessary when using highsubstituted phenyl-Sepharose. Binding was complete even in the presence of 1 M arginine (data not shown), consis-

50

tent with higher ligand density and hence more hydrophobic surface of this resin. When the IgG4-A was bound in 1 M AS without arginine and bound IgG4-A eluted with 0.5 M AS alone, the recovery was only 4% from this column (first column, Fig. 2), consistent with the more hydrophobic surface. The elution pattern was also extremely broad, indicating that 0.5 M AS is insufficient for elution. However, elution increased by the addition of 0.5 and 1 M arginine to the 0.5 M AS elution buffer. Namely, the same mAb was bound to the column in 1 M AS and then bound protein eluted with 0.5 M AS containing 0.5 or 1 M arginine, which resulted in 25 and 55% elution (second and third column). A combination of 0.5 M AS with 1 M arginine for elution resulted in a sharp peak (see Fig. 2, last profile). Elution recovery further increased when the loading sample also contained 1 M arginine or the bound IgG4A was eluted with 10 mM phosphate alone (i.e., without AS). The recovery was 50% with 10 mM phosphate alone and increased nearly to 100% when the 10 mM phosphate elution buffer contained 0.5 or 1 M arginine (data not shown). Essentially identical results were obtained with mouse IgG1. As shown in Fig. 3, mouse IgG1 was loaded to the high-substituted phenyl-Sepharose in 1 M AS, 10 mM phosphate, pH 7.0 containing 1 M arginine, resulting in complete binding and hence again demonstrating that inclusion of arginine even at such high concentration does not interfere with binding. About 60% of the loaded material was eluted using 10 mM phosphate. The elution recovery increased to >80% when the elution buffer contained 0.5 and 1 M arginine. In this case, 0.5 M arginine appears to be sufficient for higher recovery. These results demonstrate that arginine has little impact on protein binding to HIC columns, in particular using the high-substituted phenyl-Sepharose.

40 30 20 10 0

1 M AS

Elution

0.5 M AS

1 M AS

1 M AS

0.5 M AS + 0.5 M A rg 0.5 M AS + 1.0 M A rg

Abs280

Loading

Elution Fig. 2. Elution of IgG4-A from high-substituted phenyl-Sepharose using 0.5 M AS with and without arginine. Due to the limited space, arginine is abbreviated as Arg in this and the in Fig. 3.

T. Arakawa et al. / Protein Expression and Purification 54 (2007) 110–116 100 80 60 40 20 0

Loading

1 M AS + 1 M Arg

Elution

10 mM phosphate

1 M AS + 1 M

1 M AS + 1 M Arg

10 mM phosphate + 0.5 M Arg 10 mM phosphate + 1.0 M Arg

Fig. 3. Elution of mouse IgG1 from high-substituted phenyl-Sepharose using 10 mM phosphate with and without arginine.

Ion-exchange chromatography Previously we have observed that inclusion of sodium acetate in the loading sample increased the recovery of IL-6 and decreased its aggregation during IEC [23]. We have thus examined the effects of arginine using the same protocol. IL-6 was bound to the CM-Sepharose in the absence and presence of 0.2 M salts in 20 mM sodium acetate (AcONa) and the bound protein eluted by raising the AcONa concentration to 0.5 M. Binding of IL-6 was complete even in the presence of 0.2 M arginine, indicating no interference with the binding. The elution results are summarized in Fig. 4. As has been observed before, the addition of 0.2 M sodium acetate increased the recovery by 8% (from 80 to 88%) and decreased aggregation, as determined by analytical size exclusion chromatography, by 13% (from 21 to 8%), as seen in Fig. 4 (compare first and last column). An increase in recovery (to 82%) and a decrease in aggregation (to 10%) was observed with 0.2 M NaCl (Fig. 4, third column). Recovery further increased to 90% when the sample loading was done in the presence of 0.2 M arginine (second column). More importantly the amount of aggregates decreased to less

100

50 45

80

40 35

60

30 25

40

20 15

20

10 5

0

0 No additive

0.2 M Arg

0.2 M NaCl

Additive (M)

0.2 M AcONa Total recovery yield Aggregate content

Fig. 4. Elution of IL-6 from CM-Sepharose. IL-6 was loaded with and without 0.2 M salts.

113

than 1% using 0.2 M arginine. Here not only arginine was compatible with protein binding to the IEC column, but its inclusion in the loading sample reduced aggregation. This reduction may be due to suppression of protein-protein interaction during loading of IL-6. Binding of IL-6 was not affected by the presence of 0.2 M arginine. Does arginine alter the elution when present in the elution solvent? This was tested by a salt gradient elution of IL-6 from the CM-Sepharose at pH 5.0. When eluted with NaCl alone, the protein elution consistently occurred at 0.375 M salt. When both the starting and ending elution solvents contained 0.2 M arginine, the protein elution occurred at 0.2 M NaCl. Thus, it is evident that arginine does contribute to elution as an ion, mostly as a monovalent ion at pH 5.0. The slightly higher total salt concentration, i.e., 0.4 M, required for elution suggests that the effectiveness of arginine as an ion is weaker than NaCl, which may be due to its larger size. A platform approach has been developed for purification of mAbs [24], in which IEC follows the Protein-A step in the down-stream process. We have shown that an aqueous arginine solution facilitates the elution of bound antibodies from Protein-A at higher pH, reducing deleterious effects of low pH on the antibody structure and subsequent aggregation. It is thus desirable that the eluted antibody be bound to the next column, i.e., IEC, without removing arginine. After searching for the condition for IEC, we found that 0.25 M arginine is acceptable for binding mAbs to SP-Sepharose at pH 4.2. Table 1 summarizes the recovery and aggregate content of IEC process following the elution from Protein-A. The human IgG-4A was loaded to Protein-A column at 20 mg protein per ml resin and eluted with 0.1 M citrate, pH 2.9 (first row of Table 1) or 0.5– 0.7 M arginine at pH 4 (second to fifth row); note that duplicate experimental data are shown with similar results. When using 0.1 M citrate, such a low pH was necessary for higher elution of this particular mAb from Protein-A; in fact, 83–96% recovery was observed (see fourth column). When the eluted material was subjected to a size exclusion chromatography (SEC) analysis at neutral pH, it showed 38–43% aggregates of the total eluted protein. As previously shown, the observed aggregation is most likely due to low pH-induced conformational changes, which leads to aggregation upon pH titration prior to the SEC analysis. While the solvent pH was 2.9, the eluted pool was at pH 3.1, due to strong buffer action of eluted antibodies. Conversely, the recovery of eluted protein is close to 100% when using arginine at much higher pH (e.g., 0.5– 0.7 M arginine at pH 4). The pH of the eluted pools was slightly above the pH of the elution solvents, similar to the previous observation. In addition to high recovery using arginine, there is little aggregation as analyzed by SEC (below 1% as seen experiment 2–5 in Table 1). The advantage of 0.1 M citrate is that the eluent can be applied directly to the next IEC column. However, the recovery from the IEC column was also low and the aggregate content was high. The overall yield was 57–74% and the

114

Table 1 Elution of IgG-4A from Protein-A and SP-Sepharose Experiment No.

Eluent

Protein-A pH of Total collected protein fraction (%)

Aggregate content (%)

Buffer condition of loading sample

Eluent

Total protein (%)

Aggregate Total content protein (%) (%)

Monomer (%)

Relative yield of monomer

1–1 1–2

0.1 M citrate (pH 2.9)

3.13 3.10

83.1 96.4

43.3 38.4

0.1 M citrate (pH 3.5)

0.1 M citrate + 0.25 M NaCl (pH 4.0)

68.7 76.4

11 33

57.1 73.6

50.9 49.3

1

2–1 2–2

0.5 M Arg (pH 3.9)

4.13 4.20

92.6 100

0.5 0.9

0.25 M Arg + 20 mM AcONa (pH 4.0)

0.5 M AcONa (pH 5.5)

84.9 92.5

1.1 1.2

78.6 92.5

77.7 91.4

1.55 1.82

3–1 3–2

0.7 M Arg (pH 3.9)

4.11 4.17

95.5 100

0.9 0.8

0.2 M Arg + 20 mM AcONa (pH 4.2)

0.5 M AcONa (pH 5.5)

87.9 93.0

1.1 1.4

83.9 93.0

83.0 91.7

1.66 1.83

4–1 4–2

0.7 M Arg + 50 mM AcONa (pH 4.0)

4.10 4.14

100 96.0

0.7 0.8

0.2 M Arg + 20 mM AcONa (pH 4.2)

0.5 M AcONa (pH 5.5)

92.4 91.4

1.3 1.3

92.4 87.7

91.2 86.6

1.82 1.73

5–1 5–2

0.5 M Arg (pH 3.8)

4.06 4.05

98.9 95.7

1.1 0.8

0.25 M Arg + 20 mM AcONa (pH 4.0)

0.5 M AcONa (pH 5.5)

85.7 96.7

1.6 1.3

84.8 92.5

83.4 91.3

1.67 1.82

Cation exchange

Total yield

Experiment Protein A No. Eluent

Cation exchange

Total yield

pH of collected fraction

Total protein (%)

Aggregate content (%)

Buffer condition of loading sample

Eluent

Total protein (%)

Aggregate content (%)

Total protein (%)

Monomer Relative (%) yield of monomer

0.1 M citrate + 0.25 M NaCl (pH 4.0) 0.5 M AcONa (pH 5.5)

40.2

21.1

36.6

28.9

1

84.8

1.7

76.2

74.9

2.59

0.5 M AcONa (pH 5.5)

85.6

1.7

82.0

80.6

2.79

1–1

0.1 M citrate (pH 2.9)

3.13

91.0

24.0

0.1 M citrate (pH 3.5)

2–1

0.5 M Arg (pH 3.9)

4.15

89.8

1.0

3–1

0.7 M Arg + 50 mM AcONa (pH 4.0)

4.09

95.8

1.0

0.25 M Arg + 20 mM AcONa (pH 4.0) 0.2 M Arg + 20 mM AcONa (pH 4.2)

T. Arakawa et al. / Protein Expression and Purification 54 (2007) 110–116

Table 2 Elution of IgG-4B from Protein-A and SP-Sepharose

T. Arakawa et al. / Protein Expression and Purification 54 (2007) 110–116

monomer recovery was only 50%. The aggregate content after the IEC was 11–33%. The observed lower aggregation content of one case (11%) may be due to the lower protein recovery (69%), after SP-Sepharose, due to preferential loss of aggregates in the column. Use of arginine to elute from Protein-A may be disadvantage, since at least 0.5 M is required for efficient elution above pH 4. Dilution to 0.25 M arginine was, however, found sufficient for binding to SP-Sepharose at pH 4.0– 4.2. Elution of the bound antibody from the IEC column was always 90%, leading to the overall recovery of 80–95% with the monomer recovery of 80–90%. As shown in the Table 1, the aggregate content was much less with arginine elution (slightly above 1% after IEC column) than citrate elution. A shown in the last column, the overall recovery of the monomer exceeds 150% when using arginine in the Protein-A step, relative to the recovery without arginine. The IEC experiment was also carried out with IgG4-B. About 90% of the antibody was recovered from ProteinA using 0.1 M citrate, pH 2.9, which generated 24% aggregates during SEC analysis (first row of Table 2). This was loaded to SP-Sepharose and eluted using 0.1 M citrate, 0.25 M NaCl, pH 4.0 with the recovery of only 40%. The overall yield was only 36% and the monomer recovery was 29% with aggregate content of 21%. When eluted with 0.5–0.7 M arginine, pH 4, from Protein-A, the elution recovery was similar, but with much less aggregation (1%)(second and third row). This was diluted 2-fold and then loaded to the same column at pH 4. Elution from the column resulted in the recovery of 85% and aggregate content of 1.7% after SP-Sepharose. In this case, the monomer content in the final product was greater than 250% (last column of Table 2). The observed lower amounts of aggregates in the final product using arginine is due to less damage in the Protein-A eluent; in other words, whatever the damage suffered in the Protein-A step, that is responsible for aggregation, is carried over to the final product. This means that inclusion of arginine in the SP-Sepharose step does not, positively or negatively, affect the final product of the mAbs tested here. Discussion The goal of this study was to examine the effects of arginine on protein binding to the HIC and IEC columns. It is evident that the proteins tested do bind to these columns in the presence of arginine, when its concentration during protein binding is optimized and an appropriate column is selected. We and others have shown that arginine facilitates refolding [19–22], suppresses aggregation [1,2], increases reversibility of thermal unfolding [1,2], solubilizes insoluble pellets [17,18], dissociates antibodies or Fc-fusion proteins from Protein-A [3,4] and reduces non-specific binding of proteins, in particular aggregates, during gel permeation chromatography [5]. Here we demonstrated that arginine weakens hydrophobic interactions and facili-

115

tates elution of bound proteins from phenyl-Sepharose during descending AS concentration. In addition, inclusion of arginine in the loading sample increased the recovery of the total protein and decreased the aggregation during IEC. AS has been the salt most commonly used for modulating binding and elution of proteins in HIC. Why is AS the salt of choice? Extending an earlier observation by Hofmeister and Traube on salting-out or salting-in effects of salts, Melander and Horvath [25] showed that the surface tension increments of salts in water correlate with the ability of the salts to enhance protein binding to HIC columns, with the exception of certain salts, such as MgCl2. Preferential interaction measurements confirmed the importance of surface tension effects and showed why certain salts deviate [26–28]. Based on the surface tension effects and preferential interactions with the proteins, AS is considered to be the salt of choice due to its high solubility. It is also evident from the surface tension and preferential interactions that salts, other than AS, can be used for HIC. In fact, other additives, although limited, have been used to modulate binding and elution in HIC. Here we have shown that not only HIC can be performed in the presence of arginine, but also arginine increases the recovery of the proteins. We have also shown here that both mAb and IL-6 can be bound to the IEC columns in the presence of arginine. In addition, inclusion of arginine in the loading samples enhanced elution of IL-6 and reduced its aggregation. How does arginine work in IEC and HIC? Arginine has been demonstrated to suppress aggregation [1,2] and nonspecific surface adsorption of the proteins [5]. It is thus possible that arginine also suppresses non-specific binding of IL-6 and mAb to the HIC and IEC columns and reduces aggregation of IL-6 during binding to the IEC column. What is the mechanism of arginine in suppressing protein aggregation and non-specific binding? Arginine has been implicated to bind, although to limited extent, to the proteins [29,30]. GdnHCl, which contains a guanidinium group present in arginine, has high affinity for aromatic groups [31] and interacts with tryptophan side chains [32]. It thus appears that a limited extent of arginine binding is responsible for its effect on suppressing aggregation and non-specific adsorption of the proteins. These observations open a way one can use for purifying proteins after solubilization or elution by arginine-containing solvents without its removal. Since arginine solubilizes certain proteins from insoluble pellets in the active, folded structure [17,18], HIC in the presence of arginine should be a useful technique for purification of the proteins without potential denaturation. Arginine facilitates elution of antibodies from Protein-A under mildly acidic pH. Here we have shown that HIC and IEC can be used without removing arginine as a next step to Protein-A chromatography. Of course, it would be advantageous if more chromatography options, applicable to the arginine-containing solvent, are available.

116

T. Arakawa et al. / Protein Expression and Purification 54 (2007) 110–116

References [1] K. Shiraki, M. Kudou, S. Fujiwara, T. Imanaka, M. Takagi, Biophysical effect of amino acids on the prevention of protein aggregation, J. Biochem. 132 (2002) 591–595. [2] T. Arakawa, K. Tsumoto, The effects of arginine on refolding of aggregated proteins: not facilitate refolding, but suppress aggregation, Biochem. Biophys. Res. Commun. 304 (2003) 148–152. [3] T. Arakawa, J.S. Philo, K. Tsumoto, R. Yumioka, D. Ejima, Elution of antibodies from a Protein-A column by aqueous arginine solutions, Protein Expr. Purif. 36 (2004) 244–248. [4] D. Ejima, R. Yumioka, K. Tsumoto, T. Arakawa, Effective elution of antibodies by arginine and arginine derivatives in affinity column chromatography, Anal. Biochem. 345 (2005) 250–257. [5] D. Ejima, R. Yumioka, T. Arakawa, K. Tsumoto, Arginine as an effective additive in gel permeation chromatography, J. Chromatogr. A. 1094 (2005) 49–55. [6] T. Arakawa, D. Ejima, K. Tsumoto, M. Ishibashi, M. Tokunaga, Improved performance of column chromatography by arginine: Dyeaffinity chromatography, Protein Expr. Purif. 52 (2006) 410–414. [7] J.L. Fausnaugh, E. Pfannkoch, S. Gupta, F.E. Regnier, Highperformance hydrophobic interaction chromatography of proteins, Anal. Biochem. 137 (1984) 464–472. [8] A.J. Alpert, Hydrophilic-interaction chromatography for the separation of peptides, nucleic acids and other polar compounds, J. Chromatogr. 499 (1990) 177–196. [9] S. Hjerten, Some general aspects of hydrophobic interaction chromatography, J. Chromatogr. 87 (1973) 325–331. [10] R.A. Rimerman, G.W. Hatfield, Phosphate-induced protein chromatography, Science 182 (1973) 1270–1286. [11] D.E. Comings, A.G. Miguel, B.H. Lesser, Nuclear proteins. VI. Fractionation of chromosomal non-histone proteins using hydrophobic chromatography, Biochim. Biophys. Acta 563 (1979) 253–260. [12] G.J. Doellgast, W.H. Fishman, Purification of human placental alkaline phosphatase. Salt effects in affinity chromatography, Biochem. J. 141 (1974) 103–112. [13] I. Rahimi-Laridjani, H. Grimminger, F. Lingens, Affinity chromatography of biosynthetic threonine deaminase of Escherichia coli, FEBS Lett. 30 (1973) 185–187. [14] V.A. Memoli, G.J. Doellgast, Hemoglobin & serum albumin: saltmediated hydrophobic chromatography, Biochem. Biophys. Res. Commun. 66 (1975) 1011–1016. [15] R. Srinivasan, E. Ruckenstein, Role of physical forces in hydrophobic. interaction chromatography, Sep. Purif. Methods 9 (1980) 267– 370. [16] L.O. Narhi, Y. Kita, T. Arakawa, Hydrophobic interaction chromatography in alkaline pH, Anal. Biochem. 182 (1989) 266–270. [17] T. Tsumoto, M. Umetsu, I. Kumagai, D. Ejima, T. Arakawa, Solubilization of active green fluorescent protein from insoluble particles by guanidine and arginine, Biochem. Biophys. Res. Commun. 312 (2003) 1383–1386.

[18] M. Umetsu, K. Tsumoto, S. Nitta, T. Adschiri, n D. Ejima, T. Arakawa, I. Kumagai, Nondenaturing solubilization of beta2 microglobulin from inclusion bodies by L-arginine, Biochem. Biophys. Res. Commun. 328 (2005) 189–197. [19] J.G. Ho, A.P. Middelberg, P. Ramage, H.P. Kocher, The likelihood of aggregation during protein renaturation can be assessed using the second virial coefficient, Protein Sci. 12 (2003) 708–716. [20] J. Buchner, R. Rudolph, Renaturation, purification and characterization of recombinant Fab-fragments produced in Escherichia coli, Biotechnology 9 (1991) 157–162. [21] D. Arora, N.J. Khanna, Method for increasing the yield of properly folded recombinant human gamma interferon from inclusion bodies, J. Biotechnol. 52 (1996) 127–133. [22] E.D. Clark, E. Schwarz, R. Rudolph, Inhibition of aggregation side reactions during in vitro protein folding, Methods Enzymol. 309 (1999) 217–237. [23] D. Ejima, M. Watanabe, Y. Sato, M. Date, N. Yamada, Y. Takahara, High yield refolding and purification process for recombinant human interleukin-6 expressed in Escherichia coli, Biotechnol. Bioeng. 62 (1999) 301–310. [24] A.A. Shukla, B. Hubbard, T. Tressel, S. Guhan, D. Low, Downstream processing of monoclonal antibodies-Application of platform approaches, J. Chromatogr. B Analyt Technol Biomed Life Sci. 848 (2007) 28–39. [25] W. Melander, C. Horvath, Salt effect on hydrophobic interactions in precipitation and chromatography of proteins: an interpretation of the lyotropic series, Arch. Biochem. Biophys. 183 (1977) 200– 277. [26] T. Arakawa, S.N. Timasheff, Preferential interactions of proteins with salts in concentrated solutions, Biochemistry 21 (1982) 6545– 6552. [27] T. Arakawa, S.N. Timasheff, Mechanism of protein salting in and salting out by divalent cation salts: balance between hydration and salt binding, Biochemistry 23 (1984) 5912–5923. [28] T. Arakawa, R. Bhat, S.N. Timasheff, Why preferential hydration does not always stabilize the native structure of globular proteins, Biochemistry 29 (1990) 1924–1931. [29] Y. Kita, T. Arakawa, T.-Y. Lin, S.N. Timasheff, Contribution of the surface free energy perturbation to protein-solvent interactions, Biochemistry 33 (1994) 15178–15189. [30] T.-Y. Lin, S.N. Timasheff, On the role of surface tension in the stabilization of globular proteins, Protein Sci. 5 (1996) 372–381. [31] Y. Nozaki, C. Tanford, The solubility of amino acids, diglycine, and triglycine in aqueous guanidine hydrochloride solutions, J. Biol. Chem. 245 (1970) 1648–1652. [32] Y. Tanaka, K. Tsumoto, M. Umetsu, T. Nakanishi, Y. Yasutake, N. Sakai, M. Yao, I. Tanaka, T. Arakawa, I. Kumagai, Structural evidence for guanidine-protein side chain interactions: crystal structure of CutA from Pyrococcus horikoshii in 3 M guanidine hydrochloride, Biochem. Biophys. Res. Commun. 323 (2004) 185–191.