Journal of Chromatography A, 1327 (2014) 66–72
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Solid-phase polyethylene glycol conjugation using hydrophobic interaction chromatography Jianlou Niu a,b,1 , Yanlin Zhu a,1 , Yaoyao Xie a , Lintao Song a , Lu Shi a , Junjie Lan a , Bailin Liu b , Xiaokun Li a , Zhifeng Huang a,∗ a b
School of Pharmacy, Wenzhou Medical University, Wenzhou 325035, Zhejiang, China Chengdu Institute of Organic Chemistry, Chinese Academy of Sciences, Chengdu 610041, Sichuan, China
a r t i c l e
i n f o
Article history: Received 30 October 2013 Received in revised form 11 December 2013 Accepted 14 December 2013 Available online 24 December 2013 Keywords: Solid-phase PEGylation Hydrophobic interaction chromatography Lysozyme FGF-1 mPEG-butyraldehyde
a b s t r a c t PEGylation is a widely applied approach to improve the pharmacokinetic and pharmacodynamic properties of protein therapeutics. The current solution-phase PEGylation protocols often suffer from poor yield of homogeneously PEGylated bioactive products and hence fall short of being commercially attractive. To improve upon these techniques, here we developed a novel, solid-phase PEGylation methodology using a hydrophobic interaction chromatography (HIC) resin. Two variations of the HIC-based PEGylation are described that are tailored towards conjugation of proteins with hydrophobicity index above (lysozyme) and below (fibroblast growth factor 1, FGF-1) that of the mPEG-butyraldehyde (mPEG) chain used. In the case of lysozyme, the protein was first immobilized on the HIC, and the HIC-bound protein was then conjugated by passing over the column. In the case of FGF-1, the mPEG solution was first immobilized on the HIC, and the FGF-1 solution was then passed through the column. Circular dichroism (CD) spectroscopy demonstrated HIC-based PEGylation almost retained the secondary structures of proteins. Bioactivity assay showed that the recovery of activity of HIC-based PEGylated rhFGF-1 (i.e. 92%) was higher than that of liquid-phase PEGylated rhFGF-1 (i.e. 61%), while HIC-based PEGylated lysozyme showed the same activity recovery (i.e. 7%) as the liquid-phase PEGylated form. For specific proteins, the HIC-based solidphase PEGylation maybe offer a more promising alternative than the current PEGylation methods and is expected to have a major impact in the area of protein-based therapeutics. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Recombinant protein therapeutics have emerged as effective treatments for a variety of conditions, ranging from cancer to metabolic disorders and wound diseases, but they are commonly limited by their poor pharmacokinetics and immunogenicity [1–3]. Arguably, the most widely used method for improving the pharmacokinetics of protein therapies is through polyethylene glycol conjugation (PEGylation) [4,5]. The covalent attachment of PEG chains on protein surfaces effectively increases the half-life of the protein in vivo by decreasing renal clearance and masking potentially immunogenic epitopes and protease cleavage sites [6,7]. There are several food and drug administration (FDA)-approved PEGylated drugs in the market (such as PEGylated asparaginase (Onspar), PEGylated adenosine deamidase (Adagen) and PEGylated interferon (PEG-Intron and PEG-Asys)) [8] and many currently in clinical trials [9]. PEGylation is mostly performed in solution,
∗ Corresponding author. Tel.: +86 13645770691; fax: +86 57786689983. E-mail address: [email protected] (Z. Huang). 1 These authors contributed equally to this work. 0021-9673/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.chroma.2013.12.049
which causes various complications. Solution-phase chemistry is commonly associated with difficulties in controlling the reaction selectivity and the extent of covalent conjugation, isomerization, low efficiency in the separation of various end-products and unreacted starting materials and often low yields of the desired reaction product [10–14]. In lieu of the shortcomings of the solution-phase PEGylation, solid-phase PEGylation method has been proposed to minimize the variations between batches and increase the specificity and yield of the desired reaction product [15–19]. Solid-phase PEGylation is often carried out by immobilizing the protein on a solid matrix packed in a column, followed by passing the activated PEG reagent through a column [15–19]. The PEGylated protein is then separated from the unreacted protein and undesirable PEGylated isomers by gradient elution. Utilizing the ion exchange interactions between proteins and an ion exchange column (IEX), Monkarsh et al. [20] prepared and separated various constitutional isomers of PEGylated interferon-␣-2a (PEG-INF). As one of the earliest solidphase purification techniques, ion exchange chromatography has also been applied in the PEGylation of bovine hemoglobin, human serum albumin (HSA) and staphylokinase (SAK) [16,21]. However, studies have shown that IEX-based PEGylated proteins modified on
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
67
Fig. 1. Schematic illustration of the two strategies of solid-phase PEGylation using an HIC. (A) and (B) facilitate the conjugation of proteins with a higher or lower hydrophobicity than mPEG, respectively.
an ion exchange matrix often lose some bioactivity [16,21], and the efficiency for mono-PEGylation varies between proteins and hence this method is not broadly applicable. Moreover, we found that the electrostatic interactions between proteins and the matrices of the IEX block PEGylation sites. Additionally, the basic pH (i.e., pH >7.0) at which proteins are immobilized on ion exchange matrices is not suitable for N-terminal PEGylation [15]. Our previous research developed another novel solid-phase PEGylation strategy using heparin sepharose affinity chromatography (AC) for recombinant human keratinocyte growth factor 1 (rhKGF-1) and recombinant human fibroblast growth factor 2 (rhFGF-2) [18,19]. However, this solid-phase strategy is more specific because it is only suitable for a limited number of proteins that possess an intrinsic affinity for a heparin sepharose matrix. To improve upon the currently established solid-phase PEGylation strategies, here we devised a novel approach using a hydrophobic interaction (HIC) resin as a solid phase. This strategy takes advantage of the non-specific pH-insensitive nature of hydrophobic interactions thus increasing the likelihood of an exposed N-terminus available for PEGylation [17]. In this study, we present two variations of HIC-based PEGylation tailored towards proteins with higher (lysozyme) and lower (rhFGF-1) hydrophobicity indices than the mPEG used for conjugation (Fig. 1A and B). In the case of lysozyme, the protein is first immobilized on a commercially available hydrophobic matrix, HiTrap Phenyl FF, followed by the passage of an mPEG solution in the presence of a reducing agent. For rhFGF-1, activated mPEG is first immobilized on the HiTrap Phenyl FF, followed by the passage of a protein solution through the column in the presence of a reducing agent. With the help of hydrophobic interactions between proteins and mPEG on the HIC platform, we successfully modified lysozyme and rhFGF-1. Circular dichroism (CD) spectroscopy demonstrated HIC-based PEGylation almost retained the secondary structures of proteins. Bioactivity assay showed that the recovery of activity of HIC-based PEGylated rhFGF-1 was higher than that of liquid-phase PEGylated rhFGF1, while HIC-based PEGylated lysozyme showed the same activity recovery as the liquid-phase PEGylated form. The HIC-based PEGylation offers a robust strategy for PEGylating proteins with different degrees of hydrophobicity.
2. Experimental procedures 2.1. Materials PEG20 kDa-butyraldehyde (mPEG20K, its polydispersity is about 1.1), sodium cyanoborohydride (NaBH3 CN) and lysozyme were purchased from Sigma-Aldrich (St. Louis, MO, USA). Purified rhFGF-1 was produced by the Key Laboratory of Biotechnology and Pharmaceutical Engineering of Zhejiang Province, Wenzhou Medical University, China. HiTrap Phenyl fast flow (FF) columns and carboxymethyl (CM) Sepharose FF columns were purchased from GE Healthcare (Piscataway, NJ, USA). The protein assay reagent used for quantitative protein analysis was purchased from Bio-Rad (Hercules, CA, USA). Other reagents, unless otherwise indicated, were of the highest quality commercially available. All chromatographic separations and solid-phase PEGylation reactions were performed on an ÄKTA explorer purchased from GE Healthcare (Piscataway, NJ, USA) and equipped with an ÄKTA explorer workstation (Amersham Biosciences). 2.2. Solid-phase PEGylation of lysozyme using an HIC The solid-phase PEGylation of lysozyme, which has higher hydrophobicity index than mPEG20K, entailed the following three steps: (1) Two milliliters (mL) of 1.5 mg/mL lysozyme, prepared in a binding buffer (3 M NaCl in 20 mM Phosphate Buffer (PB), pH 6.0), was passed at 0.5 mL/min through an HIC (HiTrap Phenyl FF) column (2 mL bed volume), which was pre-equilibrated with the binding buffer at 0.5 mL/min. (2) Binding buffer containing 3.75 mg/mL mPEG20K and 20 mM NaBH3 CN was passed through the column at 0.5 mL/min for 20 min, followed by elution of the excess mPEG20K with binding buffer. (3) The reaction complex was completely eluted with elution buffer (20 mM PB, pH 6.0) at a rate of 1 mL/min. The elution fractions were collected and analyzed by 12% SDS-PAGE. In compared with solid-phase PEGylation of lysozyme, liquidphase PEGylation of lysozyme was performed. The reaction mixture including 1.5 mg/mL lysozyme, 3.75 mg/mL mPEG20K and 20 mM NaBH3CN in 20 mM Phosphate Buffer (pH 6.0) was incubated at
68
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
room temperature of 4 h. The purification procedure is the same to that of solid-phase PEGylation. 2.3. Solid-phase PEGylation of rhFGF-1 using an HIC RhFGF-1 has a lower hydrophobicity index than mPEG20K, and therefore solid-phase PEGylation protocol was modified as follows. (1) mPEG20K in binding buffer (3 M NaCl in 20 mM PB, pH 6.0) was first immobilized on the HIC (HiTrap Phenyl FF) column. (2) Binding buffer containing 1.5 mg/mL rhFGF-1 and 20 mM NaBH3 CN was passed at 0.5 mL/min through the HIC for 20 min, followed by elution of the excess protein with binding buffer. (3) The reaction complex was completely eluted with elution buffer (20 mM PB, pH 6.0) at a rate of 1 mL/min. All elution fractions were collected, and the PEGylation yield was estimated by 12% SDS-PAGE and densitometry quantification. 2.4. Purification of the solid-phase PEGylated lysozyme and rhFGF-1 The lysozyme reaction mixture was applied on a CM Sepharose fast flow column (1 mL bed volume) pre-equilibrated with 15 column volumes (CVs) of binding buffer (20 mM PB, pH 7.0) at a flow rate of 0.5 mL/min. The samples were washed with 10 CVs of binding buffer and then eluted with elution buffer A (0.3 M NaCl, 20 mM PB, pH 7.0). All elution fractions were collected and analyzed by 12% SDS-PAGE. Interestingly, purification of the PEGylated rhFGF-1 can be achieved in a single step by simply eluting the reaction products from the HIC. 2.5. Mass spectrometry of the solid-phase PEGylated lysozyme and rhFGF-1 Mass spectra were acquired using an Applied Biosystems Voyager System DE PRO MALDI-TOF mass spectrometer (Carlsbad, CA, USA) with a nitrogen laser. The matrix was a saturated solution of R-cyano-4-hydroxycinnamic acid in a 50:50 mixture of acetonitrile and water containing 0.1% trifluoroacetic acid. The purified solid-phase PEGylated proteins and matrix were mixed at a ratio of 1:1, and 1 L of the mixture was spotted onto a 100-well sample plate. All spectra were acquired in positive mode over the range 600–2500 Da under reflectron conditions (20 kV accelerating voltage, 350 ns extraction delay time) and 2–100 kDa under linear conditions (25 kV accelerating voltage, 750 ns extraction delay time). 2.6. Circular dichroism spectroscopy analysis The secondary structures of the native and solid-phase PEGylated proteins were determined with a circular dichroism (CD) spectropolarimeter (Model J-810, Jasco, Japan). Far-UV CD spectra were recorded at wavelengths between 190 and 250 nm using a 0.1 cm path length cell at 25 ◦ C with a protein concentration of 20 M in 10 mM PBS at pH 7.0. Each spectrum was obtained from an average of six scans, and the CD spectra were corrected for the buffer contributions. The CD data were presented in terms of the mean residue ellipticity (MRE) as a function of wavelength. 2.7. Activity assay of the solid-phase PEGylated lysozyme and rhFGF-1 2.7.1. Lysozyme activity assay The activity of lysozyme and its solid-phase PEGylated form were measured by their bioactivity on Micrococcus lysodeikticus (M. lysodeikticus) as described previously [22].
2.7.2. Mitogenic Activity Analysis of the PEGylated rhFGF-1 We used the NIH 3T3 cell line (American Type Culture Collection, Rockville, MD) to determine the activity of proliferation of the solid-phase PEGylated rhFGF-1. First, the cells were seeded in 96well plates with 5 × 103 cells per well and incubated for 24 h. They were then washed with 1× PBS and starved in low-glucose Dulbecco’s Modified Eagle’s Medium (DMEM) containing 0.4% Fetal Bovine Serum (FBS) for 24 h. Finally, the cells were treated with 1 M rhFGF-1 or 1 M solid-phase PEGylated rhFGF-1 for 24 h, and the number of viable cells was detected by the methylthiazoletetrazolium (MTT) method as previously reported [23]. 2.8. Protein quantification Protein concentrations were measured via the Bradford method with BSA as the protein standard [24]. 3. Results and discussion Previous work on solid-phase protein PEGylation is mostly based on the immobilization of the protein on an ion exchange column (IEX) [15,16,21] or affinity column (AC), followed by conjugation of the protein with the appropriate activated PEG reagent. The specificity of mono-PEGylation in such reactions was found to be higher than that in equivalent solution-phase reactions [15,16,18,19,21]. This was primarily attributed to the occlusion of many potential PEGylation sites due to their interaction with the column matrix. Two main shortcomings of solid-phase PEGylation on an IEX include its critical dependence on the PEGylation site remaining exposed on the media-bound protein and the pH limitations of the mobile phase required for adsorbing proteins. Indeed, in our hands, the IEX-based solid-phase PEGylation proved to be ineffective for both lysozyme and rhFGF-1, as PEGylation sites were mostly blocked by the electrostatic interactions between the target proteins and the IEX media (Fig. S1). Moreover, the basic pH conditions required for adsorbing proteins on an IEX, especially acidic proteins, are not suitable for conducting site-specific PEGylation on N-terminus [15]. Solid-phase PEGylation using AC requires specific interactions between proteins and the AC matrix, and hence it is not generally applicable. By contrast, hydrophobic interactions between proteins (or PEG reagent) and the HIC matrix are generally pH independent allowing for preservation of the reactive aldehyde group of the PEG reagent and the terminal amino group of the target proteins. This prompted us to evaluate whether an HIC could be used as a solid-phase medium to perform PEGylation. According to the crystal structures of lysozyme and rhFGF-1 [25,26], the lysozyme has many more solvent-exposed hydrophobic amino acids than FGF1 suggesting that lysozyme is significantly more hydrophobic than rhFGF-1 (Fig. 2). 3.1. Solid-phase PEGylation of rhFGF-1 using an HIC We first used rhFGF-1 as a model protein to evaluate the feasibility of solid-phase PEGylation on an HIC. It is well known that the binding level of ligands to HIC media is strongly affected by the surface-exposed hydrophobic regions [27]. Three different hydrophobic resins, including HiTrap Butyl FF (weakly hydrophobic), HiTrap Octyl FF (moderately hydrophobic) and HiTrap Phenyl FF (strongly hydrophobic) were screened for binding with rhFGF1. Due to the weak hydrophobicity of rhFGF-1, rhFGF-1 could not bind with the above three types of HIC media under the binding buffer (3 M NaCl in 20 mM PB, pH 6.0). We then tried to immobilize mPEG20K on the HIC using the binding buffer (3 M NaCl in 20 mM PB, pH 6.0). It was observed that mPEG20K could only be efficiently adsorbed to the HiTrap Phenyl FF matrix, which has
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
69
Fig. 2. The exposed hydrophobic surface of lysozyme (A) (PDB ID: 1LZY) and rhFGF-1 (B) (PDB ID: 1RG8).
Fig. 3. Elution chromatogram and SDS-PAGE analysis of the solid-phase PEGylation of rhFGF-1. (A) Elution chromatogram of the PEGylated mixture from the HiTrap Phenyl FF column. (B) SDS-PAGE analysis of the fractions collected from the HiTrap Phenyl FF column. Lane con, native rhFGF-1; lane m, molecular weight standards; lane a/b/c are eluted fractions of peaks a/b/c from the HiTrap Phenyl FF column.
the strongest hydrophobic properties of the three matrices tested. Binding buffer containing 1.5 mg/mL rhFGF-1 and 20 mM NaBH3 CN was then passed at 0.5 mL/min through the HIC for 20 min, and the reaction mixture was eluted with elution buffer (20 mM PB, pH 6.0). The elution profile and SDS-PAGE analysis (Fig. 3A and B) demonstrated that during the solid-phase PEGylation, unreacted rhFGF-1 passed through the matrix, while PEGylated rhFGF-1 was adsorbed on the HIC media after covalent bioconjugation with mPEG20K. This led to a PEGylated rhFGF-1 yield up to 47%. Importantly, using this solid-phase PEGylation protocol, the PEGylated rhFGF-1 can be directly eluted from the HIC with a high purity (i.e., >95%) using the elution buffer (Fig. 3). This novel solid-phase PEGylation technique could potentially be widely used to PEGylate other proteins, especially weakly hydrophobic proteins, as it relies on the immobilization of PEG and not the protein being PEGylated. Additionally, because the PEGylated protein can be obtained by a one-step purification, this technique will significantly improve the reaction product yield and the downstream processes of protein engineering.
3.2. Validation and bioactivity assay of mono-PEGylated rhFGF-1 To confirm that rhFGF-1 was mono-PEGylated, MALDI-TOF mass spectrometry (MALDI-TOF-MS) was performed. Our MS data showed that the PEGylated rhFGF-1 has a molecular weight of 36613.8 Da (Fig. 4A), suggesting that a single 20 kDa PEG chain was conjugated to the native rhFGF-1 (15.3 kDa). The broad peak of the PEGylated rhFGF-1 was likely due to the PEG polydispersity, as previously reported. To investigate whether this solid-phase PEGylation affects the rhFGF-1 structure, the secondary structure of the PEGylated rhFGF-1 was compared with that of non-PEGylated rhFGF-1 using circular dichroism (CD) spectroscopy. The far-UV CD
spectrum recorded for the PEGylated rhFGF-1 was similar to that of non-PEGylated rhFGF-1 (Fig. 4B), hence we can propose that this solid-phase PEGylation of rhFGF-1 retained the secondary structure of rhFGF-1. To evaluate the bioactivity of the PEGylated rhFGF-1, the proliferative effect of the PEGylated rhFGF-1 on NIH 3T3 cells was compared to that of non-PEGylated rhFGF-1 using a standard MTT assay. The non-PEGylated rhFGF-1 and PEGylated rhFGF-1 induced comparable mitogenic responses in the NIH 3T3 cells, demonstrating that this solid-phase PEGylation had no adverse effect on the proliferative activity of rhFGF-1 (Table 1). Previous studies on the proliferation activity of solution-phase PEGylated rhFGF-1 reported a 39% reduction compared with native rhFGF-1 [23]. Hence solidphase PEGylation on the HIC is more favorable in preserving the bioactivity of rhFGF-1. 3.3. Solid-phase PEGylation of lysozyme using an HIC As mentioned above, lysozyme possesses many more surfaceexposed hydrophobic residues than rhFGF-1. Accordingly, we employed two strategies to PEGylate lysozyme on an HIC. Similar to rhFGF-1, mPEG20K was firstly immobilized on a HiTrap Phenyl FF column under the binding buffer (3 M NaCl in 20 mM PB, pH 6.0). Binding buffer containing 1.5 mg/mL lysozyme and Table 1 Biological activity of the native and PEGylated rhFGF-1.
rhFGF-1 Solid-phase PEGylated rhFGF-1
Mean biological activity (IU/mg)
SD (n = 3)
Relative activity (%)
5.1 × 106 4.7 × 106
6.6 × 105 6.5 × 105
100 92
70
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
Fig. 4. Identification of the solid-phase PEGylated rhFGF-1. (A) MALDI-TOF mass spectrum of the PEGylated rhFGF-1 showing the molecular mass of the PEGylated rhFGF-2 as 36613.8 Da. (B) Far-UV CD spectra of the native (black line) and solid-phase PEGylated rhFGF-1 (blue line). The ellipticities are reported as mean residue ellipticity (MRE) (deg cm2 /dmol). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
20 mM NaBH3 CN was then passed at 0.5 mL/min through the HIC for 20 min, and the reaction mixture was eluted with elution buffer (20 mM PB, pH 6.0). The SDS-PAGE analysis (Fig. 5A, lane 2) demonstrated that the PEGylated yield of lysozyme is very low compared with that of rhFGF-1. This can be explained by the strong hydrophobic interaction of lysozyme with the HIC matrix. The lysozyme can also be immobilized on the column, which most likely hinders the PEGylation reaction. In an alternative approach, we immobilized the lysozyme on the HiTrap Phenyl FF column under the binding buffer (2 M NaCl in 20 mM PB, pH 6.0). Binding buffer containing additional mPEG20K and 20 mM NaBH3 CN was then passed at 0.5 mL/min through the HIC for 20 min, and the reaction mixture was eluted with elution buffer (20 mM PB, pH 6.0). The result showed that: (1) lysozyme can be readily PEGylated, and the conjugation yield is up to 41% (Fig. 5A, lane 1); and (2) both non-PEGylated and PEGylated lysozyme can be simultaneously eluted from the HIC. The PEGylated lysozyme was then efficiently separated from the non-PEGylated lysozyme by CM ion exchange chromatography (Fig. 5B), which suggests that PEGylation alters the charge properties of the lysozyme. In addition to the properties of the specific HIC matrix, the adsorption of proteins and PEG reagents to the HIC media is affected by the salt concentration of the binding buffer. In the present study, we analyzed the effect of the salt concentration in the binding buffer, ranging from 1 M to 3 M, on the solid-phase PEGylation of
proteins on the HIC. It was found that when the salt concentration is lower than 3 M, i.e., 1 M or 2 M, mPEG20K cannot adhere to the HIC, and therefore rhFGF-1 cannot be PEGylated (data not shown). In contrast, lysozyme can be adsorbed to the HIC column in the presence of 3 or 2 M NaCl, but only under 2 M NaCl lysozyme can be efficiently PEGylated (Fig. 5A), consistent with the results that mPEG20K cannot adhere to the HIC in the presence of 2 M NaCl, which facilitates the solid-phase PEGylation of lysozyme. Additionally, in the presence of 1 M NaCl, neither rhFGF-1 nor lysozyme can be PEGylated using their respective strategies, meaning that both lysozyme and mPEG20K are unable to bind to the HiTrap Phenyl FF column when the binding buffer contains only 1 M NaCl. Based on these results, we propose that lysozyme is more hydrophobic than mPEG20K, which provides a possible explanation why the PEGylation conjugation yield was lower when lysozyme was passed through the HIC immobilized with mPEG20K. 3.4. Structural validation and bioactivity assay of the mono-PEGylated lysozyme We performed MALDI-TOF-MS of the PEGylated lysozyme to verify whether the lysozyme is modified by a single molecule of mPEG20K. The MS data show that the PEGylated lysozyme has an average molecular weight of 35,684 Da (Fig. 6A), suggesting the conjugation of a single 20 kDa PEG chain to the native
Fig. 5. Solid-phase PEGylation of lysozyme using an HIC and the separation of the solid-phase PEGylated lysozyme. (A) SDS-PAGE analysis of the solid-phase PEGylation products using two different strategies. Lane m, molecular weight standards; lane 1, the product obtained from the process where binding buffer (2 M NaCl in 20 mM PB, pH 6.0) containing additional mPEG20K and 20 mM NaBH3 CN was passed through HIC-bound lysozyme; lane 2, the product obtained from the process where binding buffer (3 M NaCl in 20 mM PB, pH 6.0) containing lysozyme and NaBH3 CN was passed through HIC-bound mPEG. (B) Separation behavior of the solid-phase PEGylated lysozyme using CM ion exchange chromatography (IEC). Inset Panel. SDS-PAGE analysis of the fractions collected from IEC: lane m, molecular weight standards; lane con, native lysozyme; lane t, solid-phase PEGylated reaction mixture; lane a, eluted fraction of Peak a from IEC; lane b, eluted fraction of Peak b from IEC.
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
71
Fig. 6. Validation and structural evaluation of the solid-phase PEGylated lysozyme. (A) MALDI-TOF mass spectrum of the PEGylated lysozyme showing the molecular mass as 35684.0 Da. (B) Far-UV CD spectra of the native (black line) and solid-phase PEGylated lysozyme (blue line). The ellipticities are reported as mean residue ellipticity (MRE) (deg cm2 /dmol). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Table 2 Enzymatic activity of the native and PEGylated lysozyme.
Native lysozyme Liquid-phase PEGylated lysozyme Solid-phase PEGylated lysozyme
Acknowledgments
Mean biological activity (IU/mg)
Relative activity (%)
2.5 × 104 1.73 × 103 1.75 × 103
100 7 7
lysozyme (14.4 kDa). The CD spectra of the PEGylated and native lysozymes were approximately overlapped in wavelength from 190 to 250 nm (Fig. 6B), which indicates that the solid-phase PEGylation of lysozyme does not change the secondary structure of the native protein. Table 2 shows the enzymatic activities of unmodified lysozyme, PEGylated mono-PEGylated lysozyme obtained by solid-phase and liquid-phase reactions. The solid-phase PEGylated lysozyme showed significantly reduced enzymatic activity compared with the unmodified enzyme. As described previously, reduction in lysozyme activity could be due to a variety of factors such as steric hindrance, which in the case of a small and compact protein like lysozyme could be quite significant [17]. In addition, based on the previous study, we know that catalytic residues Asp-52 and Glu-35 of hen egg white lysozyme play the important role for its biological activity [28]. N-terminal PEGylation probably blocked these two catalytic residues of lysozyme, which partially cause the loss of the enzymatic activity. Nevertheless, PEGylated lysozyme obtained by both liquid- and solid-phase PEGylation showed similar specific enzymatic activities (Table 2), indicating that the solid-phase reaction did not have any detrimental effect on protein. 4. Conclusions Here we report a novel solid-phase PEGylation technique using hydrophobic interaction chromatography and introduce two modifications of this technique based on the hydrophobicity of the target proteins. For proteins with a relatively strong hydrophobicity, the target protein is first adsorbed on the HIC under specific conditions, and the PEG reagent is passed through the column to achieve PEGylation. For proteins with a relatively weak hydrophobicity, the PEG reagent is first adsorbed on the HIC, and then the target protein was passed through the column to initiate the PEGylation. As proof of concept, we successfully applied our HIC-based PEGylation technique to lysozyme and rhFGF-1 which are known to be difficult to be PEGylated using IEX or AC as solid phase. The HIC-based solid-phase PEGylation is robust, facile and reliable and hence is expected to facilitate the ongoing efforts in the manufacturing of protein-based therapeutics.
The authors are thankful to Dr. Moosa Mohammadi and Belov Artur for a critical review and editing of the manuscript and thoughtful suggestions. This work was supported by grants from the Natural Science Foundation of China, 81102486 (to Z.H.) and Zhejiang Key Group Project in Scientific Innovation 2010R10042-01 (to Z.H.). Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chroma. 2013.12.049. References [1] T.H. Nguyen, S.H. Kim, C.G. Decker, D.Y. Wong, J.A. Loo, H.D. Maynard, Nat. Chem. 5 (2013) 221. [2] X. Jiang, S. Zou, B. Ye, S. Zhu, Y. Liu, J. Hu, Bone 46 (2010) 1156. [3] A. Kharitonenkov, T.L. Shiyanova, A. Koester, A.M. Ford, R. Micanovic, E.J. Galbreath, G.E. Sandusky, L.J. Hammond, J.S. Moyers, R.A. Owens, J. Gromada, J.T. Brozinick, E.D. Hawkins, V.J. Wroblewski, D.S. Li, F. Mehrbod, S.R. Jaskunas, A.B. Shanafelt, J. Clin. Invest. 115 (2005) 1627. [4] S. Jevsevar, M. Kunstelj, V.G. Porekar, Biotechnol. J. 5 (2010) 113. [5] K. Knop, R. Hoogenboom, D. Fischer, U.S. Schubert, Angew. Chem. Int. Ed. Engl. 49 (2010) 6288. [6] J. Kopitz, Z. Fik, S. Andre, K. Smetana Jr., H.J. Gabius, Mol. Pharmacol. 10 (2013) 2054. [7] F.M. Veronese, A. Mero, BioDrugs 22 (2008) 315. [8] R. Duncan, Nat. Rev. Drug Discovery 2 (2003) 347. [9] F.M. Veronese, J.M. Harris, Adv. Drug Delivery Rev. 54 (2002) 453. [10] M. Abe, P. Akbarzaderaleh, M. Hamachi, N. Yoshimoto, S. Yamamoto, Biotechnol. J. 5 (2010) 477. [11] D. Yu, X. Shang, R. Ghosh, J. Chromatogr. A. 1217 (2010) 5595. [12] Y. Yamamoto, Y. Tsutsumi, Y. Yoshioka, T. Nishibata, K. Kobayashi, T. Okamoto, Y. Mukai, T. Shimizu, S. Nakagawa, S. Nagata, T. Mayumi, Nat. Biotechnol. 21 (2003) 546. [13] P. Hazra, L. Adhikary, N. Dave, A. Khedkar, H.S. Manjunath, R. Anantharaman, H. Iyer, Biotechnol. Prog. 26 (2010) 1695. [14] H. Qiu, E. Boudanova, A. Park, J.J. Bird, D.M. Honey, C. Zarazinski, B. Greene, J.S. Kingsbury, S. Boucher, J. Pollock, J.M. McPherson, C.Q. Pan, Bioconjug. Chem. 24 (2013) 408. [15] B.K. Lee, J.S. Kwon, H.J. Kim, S. Yamamoto, E.K. Lee, Bioconjug. Chem. 18 (2007) 1728. [16] X. Suo, X. Lu, T. Hu, G. Ma, Z. Su, Biotechnol. Lett. 31 (2009) 1191. [17] X. Shang, D. Yu, R. Ghosh, Biomacromolecules 12 (2011) 2772. [18] Z. Huang, C. Ye, Z. Liu, X. Wang, H. Chen, Y. Liu, L. Tang, H. Zhao, J. Wang, W. Feng, X. Li, Bioconjug. Chem. 23 (2012) 740. [19] Z. Huang, G. Zhu, C. Sun, J. Zhang, Y. Zhang, C. Ye, X. Wang, D. Ilghari, X. Li, PLoS One 7 (2012) e36423. [20] S.P. Monkarsh, Y. Ma, A. Aglione, P. Bailon, D. Ciolek, B. DeBarbieri, M.C. Graves, K. Hollfelder, H. Michel, A. Palleroni, J.E. Porter, E. Russoman, S. Roy, Y.C. Pan, Anal. Biochem. 247 (1997) 434. [21] X. Suo, C. Zheng, P. Yu, X. Lu, G. Ma, Z. Su, Artif. Cells, Blood Substitues, Immobilization Biotechnol. 37 (2009) 147. [22] Z. Huang, S.S. Leong, J. Biotechnol. 142 (2009) 157.
72
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
[23] Z. Huang, M. Lu, G. Zhu, H. Gao, L. Xie, X. Zhang, C. Ye, Y. Wang, C. Sun, X. Li, Wound Repair Regen. 19 (2011) 633. [24] M.M. Bradford, Anal. Biochem. 72 (1976) 248. [25] K. Harata, Acta Crystallogr., Sect. D: Biol. Crystallogr. 49 (1993) 497. [26] M.J. Bernett, T. Somasundaram, M. Blaber, Proteins 57 (2004) 626.
[27] K. Mayolo-Deloisa, M.E. Lienqueo, B. Andrews, M. Rito-Palomares, J.A. Asenjo, J. Chromatogr. A 1242 (2012) 11. [28] B.A. Malcolm, S. Rosenberg, M.J. Corey, J.S. Allen, A. de Baetselier, J.F. Kirsch, Proc. Nat. Acad. Sci. U.S.A. 86 (1989) 133.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Solid-phase polyethylene glycol conjugation using hydrophobic interaction chromatography Jianlou Niu a,b,1 , Yanlin Zhu a,1 , Yaoyao Xie a , Lintao Song a , Lu Shi a , Junjie Lan a , Bailin Liu b , Xiaokun Li a , Zhifeng Huang a,∗ a b
School of Pharmacy, Wenzhou Medical University, Wenzhou 325035, Zhejiang, China Chengdu Institute of Organic Chemistry, Chinese Academy of Sciences, Chengdu 610041, Sichuan, China
a r t i c l e
i n f o
Article history: Received 30 October 2013 Received in revised form 11 December 2013 Accepted 14 December 2013 Available online 24 December 2013 Keywords: Solid-phase PEGylation Hydrophobic interaction chromatography Lysozyme FGF-1 mPEG-butyraldehyde
a b s t r a c t PEGylation is a widely applied approach to improve the pharmacokinetic and pharmacodynamic properties of protein therapeutics. The current solution-phase PEGylation protocols often suffer from poor yield of homogeneously PEGylated bioactive products and hence fall short of being commercially attractive. To improve upon these techniques, here we developed a novel, solid-phase PEGylation methodology using a hydrophobic interaction chromatography (HIC) resin. Two variations of the HIC-based PEGylation are described that are tailored towards conjugation of proteins with hydrophobicity index above (lysozyme) and below (fibroblast growth factor 1, FGF-1) that of the mPEG-butyraldehyde (mPEG) chain used. In the case of lysozyme, the protein was first immobilized on the HIC, and the HIC-bound protein was then conjugated by passing over the column. In the case of FGF-1, the mPEG solution was first immobilized on the HIC, and the FGF-1 solution was then passed through the column. Circular dichroism (CD) spectroscopy demonstrated HIC-based PEGylation almost retained the secondary structures of proteins. Bioactivity assay showed that the recovery of activity of HIC-based PEGylated rhFGF-1 (i.e. 92%) was higher than that of liquid-phase PEGylated rhFGF-1 (i.e. 61%), while HIC-based PEGylated lysozyme showed the same activity recovery (i.e. 7%) as the liquid-phase PEGylated form. For specific proteins, the HIC-based solidphase PEGylation maybe offer a more promising alternative than the current PEGylation methods and is expected to have a major impact in the area of protein-based therapeutics. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Recombinant protein therapeutics have emerged as effective treatments for a variety of conditions, ranging from cancer to metabolic disorders and wound diseases, but they are commonly limited by their poor pharmacokinetics and immunogenicity [1–3]. Arguably, the most widely used method for improving the pharmacokinetics of protein therapies is through polyethylene glycol conjugation (PEGylation) [4,5]. The covalent attachment of PEG chains on protein surfaces effectively increases the half-life of the protein in vivo by decreasing renal clearance and masking potentially immunogenic epitopes and protease cleavage sites [6,7]. There are several food and drug administration (FDA)-approved PEGylated drugs in the market (such as PEGylated asparaginase (Onspar), PEGylated adenosine deamidase (Adagen) and PEGylated interferon (PEG-Intron and PEG-Asys)) [8] and many currently in clinical trials [9]. PEGylation is mostly performed in solution,
∗ Corresponding author. Tel.: +86 13645770691; fax: +86 57786689983. E-mail address: [email protected] (Z. Huang). 1 These authors contributed equally to this work. 0021-9673/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.chroma.2013.12.049
which causes various complications. Solution-phase chemistry is commonly associated with difficulties in controlling the reaction selectivity and the extent of covalent conjugation, isomerization, low efficiency in the separation of various end-products and unreacted starting materials and often low yields of the desired reaction product [10–14]. In lieu of the shortcomings of the solution-phase PEGylation, solid-phase PEGylation method has been proposed to minimize the variations between batches and increase the specificity and yield of the desired reaction product [15–19]. Solid-phase PEGylation is often carried out by immobilizing the protein on a solid matrix packed in a column, followed by passing the activated PEG reagent through a column [15–19]. The PEGylated protein is then separated from the unreacted protein and undesirable PEGylated isomers by gradient elution. Utilizing the ion exchange interactions between proteins and an ion exchange column (IEX), Monkarsh et al. [20] prepared and separated various constitutional isomers of PEGylated interferon-␣-2a (PEG-INF). As one of the earliest solidphase purification techniques, ion exchange chromatography has also been applied in the PEGylation of bovine hemoglobin, human serum albumin (HSA) and staphylokinase (SAK) [16,21]. However, studies have shown that IEX-based PEGylated proteins modified on
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
67
Fig. 1. Schematic illustration of the two strategies of solid-phase PEGylation using an HIC. (A) and (B) facilitate the conjugation of proteins with a higher or lower hydrophobicity than mPEG, respectively.
an ion exchange matrix often lose some bioactivity [16,21], and the efficiency for mono-PEGylation varies between proteins and hence this method is not broadly applicable. Moreover, we found that the electrostatic interactions between proteins and the matrices of the IEX block PEGylation sites. Additionally, the basic pH (i.e., pH >7.0) at which proteins are immobilized on ion exchange matrices is not suitable for N-terminal PEGylation [15]. Our previous research developed another novel solid-phase PEGylation strategy using heparin sepharose affinity chromatography (AC) for recombinant human keratinocyte growth factor 1 (rhKGF-1) and recombinant human fibroblast growth factor 2 (rhFGF-2) [18,19]. However, this solid-phase strategy is more specific because it is only suitable for a limited number of proteins that possess an intrinsic affinity for a heparin sepharose matrix. To improve upon the currently established solid-phase PEGylation strategies, here we devised a novel approach using a hydrophobic interaction (HIC) resin as a solid phase. This strategy takes advantage of the non-specific pH-insensitive nature of hydrophobic interactions thus increasing the likelihood of an exposed N-terminus available for PEGylation [17]. In this study, we present two variations of HIC-based PEGylation tailored towards proteins with higher (lysozyme) and lower (rhFGF-1) hydrophobicity indices than the mPEG used for conjugation (Fig. 1A and B). In the case of lysozyme, the protein is first immobilized on a commercially available hydrophobic matrix, HiTrap Phenyl FF, followed by the passage of an mPEG solution in the presence of a reducing agent. For rhFGF-1, activated mPEG is first immobilized on the HiTrap Phenyl FF, followed by the passage of a protein solution through the column in the presence of a reducing agent. With the help of hydrophobic interactions between proteins and mPEG on the HIC platform, we successfully modified lysozyme and rhFGF-1. Circular dichroism (CD) spectroscopy demonstrated HIC-based PEGylation almost retained the secondary structures of proteins. Bioactivity assay showed that the recovery of activity of HIC-based PEGylated rhFGF-1 was higher than that of liquid-phase PEGylated rhFGF1, while HIC-based PEGylated lysozyme showed the same activity recovery as the liquid-phase PEGylated form. The HIC-based PEGylation offers a robust strategy for PEGylating proteins with different degrees of hydrophobicity.
2. Experimental procedures 2.1. Materials PEG20 kDa-butyraldehyde (mPEG20K, its polydispersity is about 1.1), sodium cyanoborohydride (NaBH3 CN) and lysozyme were purchased from Sigma-Aldrich (St. Louis, MO, USA). Purified rhFGF-1 was produced by the Key Laboratory of Biotechnology and Pharmaceutical Engineering of Zhejiang Province, Wenzhou Medical University, China. HiTrap Phenyl fast flow (FF) columns and carboxymethyl (CM) Sepharose FF columns were purchased from GE Healthcare (Piscataway, NJ, USA). The protein assay reagent used for quantitative protein analysis was purchased from Bio-Rad (Hercules, CA, USA). Other reagents, unless otherwise indicated, were of the highest quality commercially available. All chromatographic separations and solid-phase PEGylation reactions were performed on an ÄKTA explorer purchased from GE Healthcare (Piscataway, NJ, USA) and equipped with an ÄKTA explorer workstation (Amersham Biosciences). 2.2. Solid-phase PEGylation of lysozyme using an HIC The solid-phase PEGylation of lysozyme, which has higher hydrophobicity index than mPEG20K, entailed the following three steps: (1) Two milliliters (mL) of 1.5 mg/mL lysozyme, prepared in a binding buffer (3 M NaCl in 20 mM Phosphate Buffer (PB), pH 6.0), was passed at 0.5 mL/min through an HIC (HiTrap Phenyl FF) column (2 mL bed volume), which was pre-equilibrated with the binding buffer at 0.5 mL/min. (2) Binding buffer containing 3.75 mg/mL mPEG20K and 20 mM NaBH3 CN was passed through the column at 0.5 mL/min for 20 min, followed by elution of the excess mPEG20K with binding buffer. (3) The reaction complex was completely eluted with elution buffer (20 mM PB, pH 6.0) at a rate of 1 mL/min. The elution fractions were collected and analyzed by 12% SDS-PAGE. In compared with solid-phase PEGylation of lysozyme, liquidphase PEGylation of lysozyme was performed. The reaction mixture including 1.5 mg/mL lysozyme, 3.75 mg/mL mPEG20K and 20 mM NaBH3CN in 20 mM Phosphate Buffer (pH 6.0) was incubated at
68
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
room temperature of 4 h. The purification procedure is the same to that of solid-phase PEGylation. 2.3. Solid-phase PEGylation of rhFGF-1 using an HIC RhFGF-1 has a lower hydrophobicity index than mPEG20K, and therefore solid-phase PEGylation protocol was modified as follows. (1) mPEG20K in binding buffer (3 M NaCl in 20 mM PB, pH 6.0) was first immobilized on the HIC (HiTrap Phenyl FF) column. (2) Binding buffer containing 1.5 mg/mL rhFGF-1 and 20 mM NaBH3 CN was passed at 0.5 mL/min through the HIC for 20 min, followed by elution of the excess protein with binding buffer. (3) The reaction complex was completely eluted with elution buffer (20 mM PB, pH 6.0) at a rate of 1 mL/min. All elution fractions were collected, and the PEGylation yield was estimated by 12% SDS-PAGE and densitometry quantification. 2.4. Purification of the solid-phase PEGylated lysozyme and rhFGF-1 The lysozyme reaction mixture was applied on a CM Sepharose fast flow column (1 mL bed volume) pre-equilibrated with 15 column volumes (CVs) of binding buffer (20 mM PB, pH 7.0) at a flow rate of 0.5 mL/min. The samples were washed with 10 CVs of binding buffer and then eluted with elution buffer A (0.3 M NaCl, 20 mM PB, pH 7.0). All elution fractions were collected and analyzed by 12% SDS-PAGE. Interestingly, purification of the PEGylated rhFGF-1 can be achieved in a single step by simply eluting the reaction products from the HIC. 2.5. Mass spectrometry of the solid-phase PEGylated lysozyme and rhFGF-1 Mass spectra were acquired using an Applied Biosystems Voyager System DE PRO MALDI-TOF mass spectrometer (Carlsbad, CA, USA) with a nitrogen laser. The matrix was a saturated solution of R-cyano-4-hydroxycinnamic acid in a 50:50 mixture of acetonitrile and water containing 0.1% trifluoroacetic acid. The purified solid-phase PEGylated proteins and matrix were mixed at a ratio of 1:1, and 1 L of the mixture was spotted onto a 100-well sample plate. All spectra were acquired in positive mode over the range 600–2500 Da under reflectron conditions (20 kV accelerating voltage, 350 ns extraction delay time) and 2–100 kDa under linear conditions (25 kV accelerating voltage, 750 ns extraction delay time). 2.6. Circular dichroism spectroscopy analysis The secondary structures of the native and solid-phase PEGylated proteins were determined with a circular dichroism (CD) spectropolarimeter (Model J-810, Jasco, Japan). Far-UV CD spectra were recorded at wavelengths between 190 and 250 nm using a 0.1 cm path length cell at 25 ◦ C with a protein concentration of 20 M in 10 mM PBS at pH 7.0. Each spectrum was obtained from an average of six scans, and the CD spectra were corrected for the buffer contributions. The CD data were presented in terms of the mean residue ellipticity (MRE) as a function of wavelength. 2.7. Activity assay of the solid-phase PEGylated lysozyme and rhFGF-1 2.7.1. Lysozyme activity assay The activity of lysozyme and its solid-phase PEGylated form were measured by their bioactivity on Micrococcus lysodeikticus (M. lysodeikticus) as described previously [22].
2.7.2. Mitogenic Activity Analysis of the PEGylated rhFGF-1 We used the NIH 3T3 cell line (American Type Culture Collection, Rockville, MD) to determine the activity of proliferation of the solid-phase PEGylated rhFGF-1. First, the cells were seeded in 96well plates with 5 × 103 cells per well and incubated for 24 h. They were then washed with 1× PBS and starved in low-glucose Dulbecco’s Modified Eagle’s Medium (DMEM) containing 0.4% Fetal Bovine Serum (FBS) for 24 h. Finally, the cells were treated with 1 M rhFGF-1 or 1 M solid-phase PEGylated rhFGF-1 for 24 h, and the number of viable cells was detected by the methylthiazoletetrazolium (MTT) method as previously reported [23]. 2.8. Protein quantification Protein concentrations were measured via the Bradford method with BSA as the protein standard [24]. 3. Results and discussion Previous work on solid-phase protein PEGylation is mostly based on the immobilization of the protein on an ion exchange column (IEX) [15,16,21] or affinity column (AC), followed by conjugation of the protein with the appropriate activated PEG reagent. The specificity of mono-PEGylation in such reactions was found to be higher than that in equivalent solution-phase reactions [15,16,18,19,21]. This was primarily attributed to the occlusion of many potential PEGylation sites due to their interaction with the column matrix. Two main shortcomings of solid-phase PEGylation on an IEX include its critical dependence on the PEGylation site remaining exposed on the media-bound protein and the pH limitations of the mobile phase required for adsorbing proteins. Indeed, in our hands, the IEX-based solid-phase PEGylation proved to be ineffective for both lysozyme and rhFGF-1, as PEGylation sites were mostly blocked by the electrostatic interactions between the target proteins and the IEX media (Fig. S1). Moreover, the basic pH conditions required for adsorbing proteins on an IEX, especially acidic proteins, are not suitable for conducting site-specific PEGylation on N-terminus [15]. Solid-phase PEGylation using AC requires specific interactions between proteins and the AC matrix, and hence it is not generally applicable. By contrast, hydrophobic interactions between proteins (or PEG reagent) and the HIC matrix are generally pH independent allowing for preservation of the reactive aldehyde group of the PEG reagent and the terminal amino group of the target proteins. This prompted us to evaluate whether an HIC could be used as a solid-phase medium to perform PEGylation. According to the crystal structures of lysozyme and rhFGF-1 [25,26], the lysozyme has many more solvent-exposed hydrophobic amino acids than FGF1 suggesting that lysozyme is significantly more hydrophobic than rhFGF-1 (Fig. 2). 3.1. Solid-phase PEGylation of rhFGF-1 using an HIC We first used rhFGF-1 as a model protein to evaluate the feasibility of solid-phase PEGylation on an HIC. It is well known that the binding level of ligands to HIC media is strongly affected by the surface-exposed hydrophobic regions [27]. Three different hydrophobic resins, including HiTrap Butyl FF (weakly hydrophobic), HiTrap Octyl FF (moderately hydrophobic) and HiTrap Phenyl FF (strongly hydrophobic) were screened for binding with rhFGF1. Due to the weak hydrophobicity of rhFGF-1, rhFGF-1 could not bind with the above three types of HIC media under the binding buffer (3 M NaCl in 20 mM PB, pH 6.0). We then tried to immobilize mPEG20K on the HIC using the binding buffer (3 M NaCl in 20 mM PB, pH 6.0). It was observed that mPEG20K could only be efficiently adsorbed to the HiTrap Phenyl FF matrix, which has
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
69
Fig. 2. The exposed hydrophobic surface of lysozyme (A) (PDB ID: 1LZY) and rhFGF-1 (B) (PDB ID: 1RG8).
Fig. 3. Elution chromatogram and SDS-PAGE analysis of the solid-phase PEGylation of rhFGF-1. (A) Elution chromatogram of the PEGylated mixture from the HiTrap Phenyl FF column. (B) SDS-PAGE analysis of the fractions collected from the HiTrap Phenyl FF column. Lane con, native rhFGF-1; lane m, molecular weight standards; lane a/b/c are eluted fractions of peaks a/b/c from the HiTrap Phenyl FF column.
the strongest hydrophobic properties of the three matrices tested. Binding buffer containing 1.5 mg/mL rhFGF-1 and 20 mM NaBH3 CN was then passed at 0.5 mL/min through the HIC for 20 min, and the reaction mixture was eluted with elution buffer (20 mM PB, pH 6.0). The elution profile and SDS-PAGE analysis (Fig. 3A and B) demonstrated that during the solid-phase PEGylation, unreacted rhFGF-1 passed through the matrix, while PEGylated rhFGF-1 was adsorbed on the HIC media after covalent bioconjugation with mPEG20K. This led to a PEGylated rhFGF-1 yield up to 47%. Importantly, using this solid-phase PEGylation protocol, the PEGylated rhFGF-1 can be directly eluted from the HIC with a high purity (i.e., >95%) using the elution buffer (Fig. 3). This novel solid-phase PEGylation technique could potentially be widely used to PEGylate other proteins, especially weakly hydrophobic proteins, as it relies on the immobilization of PEG and not the protein being PEGylated. Additionally, because the PEGylated protein can be obtained by a one-step purification, this technique will significantly improve the reaction product yield and the downstream processes of protein engineering.
3.2. Validation and bioactivity assay of mono-PEGylated rhFGF-1 To confirm that rhFGF-1 was mono-PEGylated, MALDI-TOF mass spectrometry (MALDI-TOF-MS) was performed. Our MS data showed that the PEGylated rhFGF-1 has a molecular weight of 36613.8 Da (Fig. 4A), suggesting that a single 20 kDa PEG chain was conjugated to the native rhFGF-1 (15.3 kDa). The broad peak of the PEGylated rhFGF-1 was likely due to the PEG polydispersity, as previously reported. To investigate whether this solid-phase PEGylation affects the rhFGF-1 structure, the secondary structure of the PEGylated rhFGF-1 was compared with that of non-PEGylated rhFGF-1 using circular dichroism (CD) spectroscopy. The far-UV CD
spectrum recorded for the PEGylated rhFGF-1 was similar to that of non-PEGylated rhFGF-1 (Fig. 4B), hence we can propose that this solid-phase PEGylation of rhFGF-1 retained the secondary structure of rhFGF-1. To evaluate the bioactivity of the PEGylated rhFGF-1, the proliferative effect of the PEGylated rhFGF-1 on NIH 3T3 cells was compared to that of non-PEGylated rhFGF-1 using a standard MTT assay. The non-PEGylated rhFGF-1 and PEGylated rhFGF-1 induced comparable mitogenic responses in the NIH 3T3 cells, demonstrating that this solid-phase PEGylation had no adverse effect on the proliferative activity of rhFGF-1 (Table 1). Previous studies on the proliferation activity of solution-phase PEGylated rhFGF-1 reported a 39% reduction compared with native rhFGF-1 [23]. Hence solidphase PEGylation on the HIC is more favorable in preserving the bioactivity of rhFGF-1. 3.3. Solid-phase PEGylation of lysozyme using an HIC As mentioned above, lysozyme possesses many more surfaceexposed hydrophobic residues than rhFGF-1. Accordingly, we employed two strategies to PEGylate lysozyme on an HIC. Similar to rhFGF-1, mPEG20K was firstly immobilized on a HiTrap Phenyl FF column under the binding buffer (3 M NaCl in 20 mM PB, pH 6.0). Binding buffer containing 1.5 mg/mL lysozyme and Table 1 Biological activity of the native and PEGylated rhFGF-1.
rhFGF-1 Solid-phase PEGylated rhFGF-1
Mean biological activity (IU/mg)
SD (n = 3)
Relative activity (%)
5.1 × 106 4.7 × 106
6.6 × 105 6.5 × 105
100 92
70
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
Fig. 4. Identification of the solid-phase PEGylated rhFGF-1. (A) MALDI-TOF mass spectrum of the PEGylated rhFGF-1 showing the molecular mass of the PEGylated rhFGF-2 as 36613.8 Da. (B) Far-UV CD spectra of the native (black line) and solid-phase PEGylated rhFGF-1 (blue line). The ellipticities are reported as mean residue ellipticity (MRE) (deg cm2 /dmol). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
20 mM NaBH3 CN was then passed at 0.5 mL/min through the HIC for 20 min, and the reaction mixture was eluted with elution buffer (20 mM PB, pH 6.0). The SDS-PAGE analysis (Fig. 5A, lane 2) demonstrated that the PEGylated yield of lysozyme is very low compared with that of rhFGF-1. This can be explained by the strong hydrophobic interaction of lysozyme with the HIC matrix. The lysozyme can also be immobilized on the column, which most likely hinders the PEGylation reaction. In an alternative approach, we immobilized the lysozyme on the HiTrap Phenyl FF column under the binding buffer (2 M NaCl in 20 mM PB, pH 6.0). Binding buffer containing additional mPEG20K and 20 mM NaBH3 CN was then passed at 0.5 mL/min through the HIC for 20 min, and the reaction mixture was eluted with elution buffer (20 mM PB, pH 6.0). The result showed that: (1) lysozyme can be readily PEGylated, and the conjugation yield is up to 41% (Fig. 5A, lane 1); and (2) both non-PEGylated and PEGylated lysozyme can be simultaneously eluted from the HIC. The PEGylated lysozyme was then efficiently separated from the non-PEGylated lysozyme by CM ion exchange chromatography (Fig. 5B), which suggests that PEGylation alters the charge properties of the lysozyme. In addition to the properties of the specific HIC matrix, the adsorption of proteins and PEG reagents to the HIC media is affected by the salt concentration of the binding buffer. In the present study, we analyzed the effect of the salt concentration in the binding buffer, ranging from 1 M to 3 M, on the solid-phase PEGylation of
proteins on the HIC. It was found that when the salt concentration is lower than 3 M, i.e., 1 M or 2 M, mPEG20K cannot adhere to the HIC, and therefore rhFGF-1 cannot be PEGylated (data not shown). In contrast, lysozyme can be adsorbed to the HIC column in the presence of 3 or 2 M NaCl, but only under 2 M NaCl lysozyme can be efficiently PEGylated (Fig. 5A), consistent with the results that mPEG20K cannot adhere to the HIC in the presence of 2 M NaCl, which facilitates the solid-phase PEGylation of lysozyme. Additionally, in the presence of 1 M NaCl, neither rhFGF-1 nor lysozyme can be PEGylated using their respective strategies, meaning that both lysozyme and mPEG20K are unable to bind to the HiTrap Phenyl FF column when the binding buffer contains only 1 M NaCl. Based on these results, we propose that lysozyme is more hydrophobic than mPEG20K, which provides a possible explanation why the PEGylation conjugation yield was lower when lysozyme was passed through the HIC immobilized with mPEG20K. 3.4. Structural validation and bioactivity assay of the mono-PEGylated lysozyme We performed MALDI-TOF-MS of the PEGylated lysozyme to verify whether the lysozyme is modified by a single molecule of mPEG20K. The MS data show that the PEGylated lysozyme has an average molecular weight of 35,684 Da (Fig. 6A), suggesting the conjugation of a single 20 kDa PEG chain to the native
Fig. 5. Solid-phase PEGylation of lysozyme using an HIC and the separation of the solid-phase PEGylated lysozyme. (A) SDS-PAGE analysis of the solid-phase PEGylation products using two different strategies. Lane m, molecular weight standards; lane 1, the product obtained from the process where binding buffer (2 M NaCl in 20 mM PB, pH 6.0) containing additional mPEG20K and 20 mM NaBH3 CN was passed through HIC-bound lysozyme; lane 2, the product obtained from the process where binding buffer (3 M NaCl in 20 mM PB, pH 6.0) containing lysozyme and NaBH3 CN was passed through HIC-bound mPEG. (B) Separation behavior of the solid-phase PEGylated lysozyme using CM ion exchange chromatography (IEC). Inset Panel. SDS-PAGE analysis of the fractions collected from IEC: lane m, molecular weight standards; lane con, native lysozyme; lane t, solid-phase PEGylated reaction mixture; lane a, eluted fraction of Peak a from IEC; lane b, eluted fraction of Peak b from IEC.
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
71
Fig. 6. Validation and structural evaluation of the solid-phase PEGylated lysozyme. (A) MALDI-TOF mass spectrum of the PEGylated lysozyme showing the molecular mass as 35684.0 Da. (B) Far-UV CD spectra of the native (black line) and solid-phase PEGylated lysozyme (blue line). The ellipticities are reported as mean residue ellipticity (MRE) (deg cm2 /dmol). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Table 2 Enzymatic activity of the native and PEGylated lysozyme.
Native lysozyme Liquid-phase PEGylated lysozyme Solid-phase PEGylated lysozyme
Acknowledgments
Mean biological activity (IU/mg)
Relative activity (%)
2.5 × 104 1.73 × 103 1.75 × 103
100 7 7
lysozyme (14.4 kDa). The CD spectra of the PEGylated and native lysozymes were approximately overlapped in wavelength from 190 to 250 nm (Fig. 6B), which indicates that the solid-phase PEGylation of lysozyme does not change the secondary structure of the native protein. Table 2 shows the enzymatic activities of unmodified lysozyme, PEGylated mono-PEGylated lysozyme obtained by solid-phase and liquid-phase reactions. The solid-phase PEGylated lysozyme showed significantly reduced enzymatic activity compared with the unmodified enzyme. As described previously, reduction in lysozyme activity could be due to a variety of factors such as steric hindrance, which in the case of a small and compact protein like lysozyme could be quite significant [17]. In addition, based on the previous study, we know that catalytic residues Asp-52 and Glu-35 of hen egg white lysozyme play the important role for its biological activity [28]. N-terminal PEGylation probably blocked these two catalytic residues of lysozyme, which partially cause the loss of the enzymatic activity. Nevertheless, PEGylated lysozyme obtained by both liquid- and solid-phase PEGylation showed similar specific enzymatic activities (Table 2), indicating that the solid-phase reaction did not have any detrimental effect on protein. 4. Conclusions Here we report a novel solid-phase PEGylation technique using hydrophobic interaction chromatography and introduce two modifications of this technique based on the hydrophobicity of the target proteins. For proteins with a relatively strong hydrophobicity, the target protein is first adsorbed on the HIC under specific conditions, and the PEG reagent is passed through the column to achieve PEGylation. For proteins with a relatively weak hydrophobicity, the PEG reagent is first adsorbed on the HIC, and then the target protein was passed through the column to initiate the PEGylation. As proof of concept, we successfully applied our HIC-based PEGylation technique to lysozyme and rhFGF-1 which are known to be difficult to be PEGylated using IEX or AC as solid phase. The HIC-based solid-phase PEGylation is robust, facile and reliable and hence is expected to facilitate the ongoing efforts in the manufacturing of protein-based therapeutics.
The authors are thankful to Dr. Moosa Mohammadi and Belov Artur for a critical review and editing of the manuscript and thoughtful suggestions. This work was supported by grants from the Natural Science Foundation of China, 81102486 (to Z.H.) and Zhejiang Key Group Project in Scientific Innovation 2010R10042-01 (to Z.H.). Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chroma. 2013.12.049. References [1] T.H. Nguyen, S.H. Kim, C.G. Decker, D.Y. Wong, J.A. Loo, H.D. Maynard, Nat. Chem. 5 (2013) 221. [2] X. Jiang, S. Zou, B. Ye, S. Zhu, Y. Liu, J. Hu, Bone 46 (2010) 1156. [3] A. Kharitonenkov, T.L. Shiyanova, A. Koester, A.M. Ford, R. Micanovic, E.J. Galbreath, G.E. Sandusky, L.J. Hammond, J.S. Moyers, R.A. Owens, J. Gromada, J.T. Brozinick, E.D. Hawkins, V.J. Wroblewski, D.S. Li, F. Mehrbod, S.R. Jaskunas, A.B. Shanafelt, J. Clin. Invest. 115 (2005) 1627. [4] S. Jevsevar, M. Kunstelj, V.G. Porekar, Biotechnol. J. 5 (2010) 113. [5] K. Knop, R. Hoogenboom, D. Fischer, U.S. Schubert, Angew. Chem. Int. Ed. Engl. 49 (2010) 6288. [6] J. Kopitz, Z. Fik, S. Andre, K. Smetana Jr., H.J. Gabius, Mol. Pharmacol. 10 (2013) 2054. [7] F.M. Veronese, A. Mero, BioDrugs 22 (2008) 315. [8] R. Duncan, Nat. Rev. Drug Discovery 2 (2003) 347. [9] F.M. Veronese, J.M. Harris, Adv. Drug Delivery Rev. 54 (2002) 453. [10] M. Abe, P. Akbarzaderaleh, M. Hamachi, N. Yoshimoto, S. Yamamoto, Biotechnol. J. 5 (2010) 477. [11] D. Yu, X. Shang, R. Ghosh, J. Chromatogr. A. 1217 (2010) 5595. [12] Y. Yamamoto, Y. Tsutsumi, Y. Yoshioka, T. Nishibata, K. Kobayashi, T. Okamoto, Y. Mukai, T. Shimizu, S. Nakagawa, S. Nagata, T. Mayumi, Nat. Biotechnol. 21 (2003) 546. [13] P. Hazra, L. Adhikary, N. Dave, A. Khedkar, H.S. Manjunath, R. Anantharaman, H. Iyer, Biotechnol. Prog. 26 (2010) 1695. [14] H. Qiu, E. Boudanova, A. Park, J.J. Bird, D.M. Honey, C. Zarazinski, B. Greene, J.S. Kingsbury, S. Boucher, J. Pollock, J.M. McPherson, C.Q. Pan, Bioconjug. Chem. 24 (2013) 408. [15] B.K. Lee, J.S. Kwon, H.J. Kim, S. Yamamoto, E.K. Lee, Bioconjug. Chem. 18 (2007) 1728. [16] X. Suo, X. Lu, T. Hu, G. Ma, Z. Su, Biotechnol. Lett. 31 (2009) 1191. [17] X. Shang, D. Yu, R. Ghosh, Biomacromolecules 12 (2011) 2772. [18] Z. Huang, C. Ye, Z. Liu, X. Wang, H. Chen, Y. Liu, L. Tang, H. Zhao, J. Wang, W. Feng, X. Li, Bioconjug. Chem. 23 (2012) 740. [19] Z. Huang, G. Zhu, C. Sun, J. Zhang, Y. Zhang, C. Ye, X. Wang, D. Ilghari, X. Li, PLoS One 7 (2012) e36423. [20] S.P. Monkarsh, Y. Ma, A. Aglione, P. Bailon, D. Ciolek, B. DeBarbieri, M.C. Graves, K. Hollfelder, H. Michel, A. Palleroni, J.E. Porter, E. Russoman, S. Roy, Y.C. Pan, Anal. Biochem. 247 (1997) 434. [21] X. Suo, C. Zheng, P. Yu, X. Lu, G. Ma, Z. Su, Artif. Cells, Blood Substitues, Immobilization Biotechnol. 37 (2009) 147. [22] Z. Huang, S.S. Leong, J. Biotechnol. 142 (2009) 157.
72
J. Niu et al. / J. Chromatogr. A 1327 (2014) 66–72
[23] Z. Huang, M. Lu, G. Zhu, H. Gao, L. Xie, X. Zhang, C. Ye, Y. Wang, C. Sun, X. Li, Wound Repair Regen. 19 (2011) 633. [24] M.M. Bradford, Anal. Biochem. 72 (1976) 248. [25] K. Harata, Acta Crystallogr., Sect. D: Biol. Crystallogr. 49 (1993) 497. [26] M.J. Bernett, T. Somasundaram, M. Blaber, Proteins 57 (2004) 626.
[27] K. Mayolo-Deloisa, M.E. Lienqueo, B. Andrews, M. Rito-Palomares, J.A. Asenjo, J. Chromatogr. A 1242 (2012) 11. [28] B.A. Malcolm, S. Rosenberg, M.J. Corey, J.S. Allen, A. de Baetselier, J.F. Kirsch, Proc. Nat. Acad. Sci. U.S.A. 86 (1989) 133.
