Protein Expression and Purification 95 (2014) 211–218
Contents lists available at ScienceDirect
Protein Expression and Purification journal homepage: www.elsevier.com/locate/yprep
Soluble expression and partial purification of recombinant human erythropoietin from E. coli Taeck-Hyun Jeong a,1, Young-Jin Son b,1, Han-Bong Ryu a,1, Bon-Kyung Koo a, Seung-Mi Jeong a, Phuong Hoang a, Bich Hang Do a, Jung-A Song a, Seon-Ha Chong a, Robert Charles Robinson c,d,e, Han Choe a,⇑ a
Department of Physiology and Bio-Medical Institute of Technology, University of Ulsan College of Medicine, Seoul 138-736, South Korea Department of Pharmacy, Sunchon National University, Suncheon, Jeonnam 540-742, South Korea Institute of Molecular and Cell Biology, Proteos, 61 Biopolis Drive, Singapore 138673, Singapore d Department of Biochemistry, National University of Singapore, 8 Medical Drive, Singapore 117597, Singapore e School of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, Singapore 637551, Singapore b c
a r t i c l e
i n f o
Article history: Received 23 March 2012 and in revised form 1 January 2014 Available online 9 January 2014 Keywords: rhEpo Escherichia coli expression system Maltose binding protein (MBP) Therapeutic protein
a b s t r a c t Human erythropoietin (hEpo) is an essential regulator of erythrocyte production that induces the division and differentiation of erythroid progenitor cells in the bone marrow into mature erythrocytes. It is widely used for the treatment of anemia resulting from chronic kidney disease, chemotherapy, and cancerrelated therapies. Active hEpo, and hEpo analogs, have been purified primarily from mammalian cells, which has several disadvantages, including low yields and high production costs. Although an Escherichia coli (E. coli) expression system may provide economic production of therapeutic proteins, it has not been used for the production of recombinant hEpo (rhEpo) because it aggregates in inclusion bodies in the E. coli cytoplasm and is not modified post-translationally. We investigated the soluble overexpression of active rhEpo with various protein tags in E. coli, and found out that several tags increased the solubility of rhEpo. Among them maltose binding protein (MBP)-tagged rhEpo was purified using affinity and gel filtration columns. Non-denaturing electrophoresis and MALDI-TOF MS analysis demonstrated that the purified rhEpo had two intra-disulfide bonds identical to those of the native hEpo. An in vitro proliferation assay showed that rhEpo purified from E. coli had similar biological activity as rhEpo derived from CHO cells. Therefore, we report for the first time that active rhEpo was overexpressed as a soluble form in the cytoplasm of E. coli and purified it in simple purification steps. We hope that our results offer opportunities for progress in rhEpo therapeutics. Ó 2014 Elsevier Inc. All rights reserved.
Introduction Human erythropoietin (hEpo),2 a 34-kDa glycoprotein produced primarily by the kidney, is the principal factor regulating erythrocyte production [1]. hEpo induces erythroid progenitor cell proliferation, differentiation, and maturation and inhibits apoptosis, resulting in increased erythrocyte production [2]. hEpo achieves its effects by causing homodimerization of the hEpo receptor with subsequent autophosphorylation of the tyrosine kinase JAK2 [3] and phosphorylation of the receptor itself, as well as various substrate proteins, leading to up-regulation of a number of signaling pathways and activation of gene transcription [4]. Natural hEpo is glycoprotein which ⇑ Corresponding author. Tel.: +82 2 3010 4292; fax: +82 2 3010 8148. E-mail address: [email protected] (H. Choe). These authors contributed equally to the work. 2 Abbreviations used: hEpo; human erythropoietin; rhEpo, recombinant human Epo; MBP, maltose binding protein; His, histidine; GST, glutathione S-transferase; TRX, thioredoxin; IPTG, isopropyl-b-D-thiogalactoside; GPC, gel permeation chromatography; IL-3, interleukin 3; PEG, polyethylene glycol. 1
1046-5928/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.pep.2014.01.001
has three sites of N-glycosylation (N24, N38, and N83) and one of O-glycosylation (S126). E. coli is the most frequently used host cell for the production of recombinant protein. Its expression system may allow economic and fast production of large amounts of heterologous protein. However, there were some disadvantages of E. coli production for recombinant human protein. The recombinant protein is usually expressed as an inclusion body and non-glycosylated form [5]. Recombinant human Epo (rhEpo) has been used clinically to treat anemia resulting from chronic kidney disease, chemotherapy, and cancer-related therapies [6]. In addition to these well-established applications, it also has therapeutic potential in the treatment of acute brain, heart, and kidney injuries [7]. According to some previous reports, glycosylation of hEpo is important in maintaining its stability and hematopoietic activity in vivo [8–11]. In order to produce active rhEpo, it has been typically produced in mammalian cell cultures [12–17]. In this study, we demonstrate for the first time that the expression of soluble active maltose binding protein (MBP)-tagged rhEpo
212
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
in E. coli, partial purification, and cell-based biological activity. This study offers opportunities for progress in rhEpo therapeutics. Materials and methods Histidine (His), glutathione S-transferase (GST), thioredoxin (TRX), NusA, or MBP tagged rhEpo constructs The pET22b-based destination vector, which allows inclusion of various protein tags at the N-terminus of the encoded protein, was used as the expression vector. To amplify the hEpo gene, two primers (forward 50 -GCGCTGGGCGCGCAGAAAGAAGCTATCAGTC-30 and reverse 50 -GACTGATAGCTTCTTTCTGCGCGCCCAGCGC-30 ) were used. For construction of the entry vector, DNA sequences encoding the 166 amino acids of mature human Epo (NP_000790.2) were synthesized and codon-optimized for E. coli expression by a gene synthesis service (Epoch Life Science Inc., TX, USA). The codonoptimized rhEpo gene was inserted into the pDONR 207 vector using the BP recombination cloning system (Invitrogen, CA, USA). The TEV protease recognition site (ENLYFQ;G) was inserted in front of the amino acid sequence of rhEpo. To obtain the expression vector, the rhEpo gene in pENTR-rhEpo was cloned into pHGWA (His6), pHGGWA (GST), pHXGWA (TRX), pHNGWA (NusA), and pHMGWA (MBP) vectors using the LR recombination cloning system (Invitrogen). After confirming the DNA sequence of rhEpo, the resulting plasmids were introduced into E. coli BL21 (DE3). Expression of rhEpo with various protein tags To analyze the expression level of rhEpo with variant protein tags, expression clones were pre-grown overnight in 3 mL LB broth supplemented with 50 lg/mL ampicillin at 37 °C. Pre-cultured samples were inoculated in 1 L new LB broth containing an appropriate antibiotic and grown to mid-log phase (OD600 = 0.4–0.5). Then, isopropyl-b-d-thiogalactoside (IPTG) was added to a final concentration of 1 mM. To improve the soluble expression level, the temperature of cultivation was changed to 30 °C. After 3 h, the cells were harvested by centrifugation (3500 rpm, 30 min). Cell pellets were suspended in 10 mL/g cell buffer A (50 mM Tris–HCl, pH 8.0, 0.5 mM EDTA, 5% glycerol) and disrupted by sonication. After centrifugation (12,000 rpm, 30 min), soluble and insoluble fractions were collected. The expression levels of rhEpo with various protein tags were assessed using 10% Tris–glycine SDS–PAGE, followed by image analysis using the ImageJ program (http:// imagej.nih.gov/ij). Purification of rhEpo Eighty-one milliliters of clarified supernatant was loaded onto a MBP Trap HP containing 20 mL volume resin (GE Life Sciences, NJ, USA) equilibrated with 5 column volume of buffer A using ÄKTA Prime Plus (GE Life Sciences). The column was then washed with buffer A until the UV (280 nm) value reached baseline. The bound protein was then eluted with buffer B (50 mM Tris–HCl, pH 8.0, 0.5 mM EDTA, 5% glycerol, 20 mM maltose monohydrate; Junsei Chemical Co., Tokyo, Japan). The eluted MBP-rhEpo fusion protein was pooled and its size was confirmed by 10% Tris–glycine SDS– PAGE analysis. The eluted MBP-rhEpo was filtered by 0.45 lM filter. The protein concentration of MBP-rhEpo was analyzed by Bradford assay, and then diluted with TEV standard buffer (50 mM Tris–HCl, pH 8.0, 0.5 mM EDTA, 5 mM dithiothreitol (DTT), 5% (v/v) glycerol) at 0.5 mg/mL. The MBP-tagged rhEpo eluted from the MBP affinity column was treated with partially purified rTEV protease in TEV standard buffer. The reaction condition for cutting fusion protein was
30 mg of MBP-rhEpo solution per 1 mg of rTEV protease crude solution. The reaction solution was incubated at room temperature overnight [18]. To further purify rhEpo, the sample treated with TEV protease was loaded on a HiLoad 26/100 Superdex200 HR gel filtration chromatography column (GE Life Sciences, NJ, USA) equilibrated with buffer A, supplemented with 5 mM DTT, 0.01% Triton X-100, and 0.5 M NaCl. The flow rate was 1 mL/min. After exchanging buffer A with PBS supplemented by 5% glycerol using an 3000 NMWL Centricon (Millipore, MA, USA), the protein was analyzed using 10% Tris–glycine SDS–PAGE and quantified using the ImageJ program. The final purified rhEpo was stored at 20 °C. Cation exchange and the second gel filtration chromatography of rhEpo In order to check the impurity profile of rhEpo after ion exchange chromatography, cation exchange chromatography was carried out on AKTA™ avant chromatography system (GE Life Sciences, NJ, USA) using High S column (5 mL packing resin, GE Life Sciences, NJ, USA). Equilibrium buffer was 6 M urea and 0.25 M acetic acid, pH 2.5. Elution buffer was 6 M urea and 0.25 M acetic acid, 1 M NaCl, pH 2.5. The flow rate was 2.5 mL/min and the elution started with 100% of equilibrium buffer, then followed by linear gradient of 0–100% elution buffer for 40 min. Elution was monitored with UV detector at 280 nm. A second gel filtration chromatography step was performed in an attempt to further purify the rhEpo. It was carried out on AKTA™ avant chromatography system (GE Life Sciences, NJ, USA) using Superose™ 12 10/300 GL column (GE Life Sciences, NJ, USA). Mobile phase was 50 mM sodium phosphate and 0.15 M NaCl, pH 7.0. The flow rate was 0.5 mL/min. The elution was monitored with UV detector at 280 nm. Endotoxin assay of purified rhEpo To determine the amount of the remaining endotoxin, ToxinSensor™ chromogenic LAL Endotoxin Assay Kit (GenScript, Piscataway, NJ, USA) was used. Endotoxin standard (100 lL) and samples (100 lL) were added into endotoxin-free vials. The pHs of standard and samples were adjusted at pH 7.0. Limulus amebocyte lysate (LAL) reagent was added into each vial and mixed thoroughly. The vials were incubated at 37 °C for 45 min. After incubation, 100 lL of chromogenic substrate solution was added to each vial and incubated at 37 °C for 6 min. Five hundred microliters of stop solution and 500 lL of color-stabilizer (#2 and #3) were added into each vial and mixed well. The absorbance of each reaction at 545 nm was analyzed by spectrophotometer. After analyzing the absorbance, we calculated endotoxin unit with the standard curve obtained by standard solution. Expression and purification of rTEV protease Expression clones for rTEV were cultured in 3 mL LB broth supplemented with 50 lg/mL ampicillin at 37 °C. When the optical density at 600 nm was 0.5, isopropyl-b-D-thiogalactoside (IPTG) was added to a final concentration of 1 mM in order to express rTEV. After 3 h, the cells were harvested by centrifugation (3500 rpm, 30 min). Cell pellets were suspended in 10 mL/g cell buffer A (50 mM Tris–HCl, pH 8.0, 0.5 mM EDTA, 5% glycerol) and disrupted by sonication. After centrifugation (12,000 rpm, 30 min), soluble fraction was collected. Then the soluble fraction was filtered with 0.45 lM filter. Ten milliliters of clarified supernatant was loaded onto a nickel column containing 5 mL volume resin (GE Life Sciences, NJ, USA) equilibrated with 2 column volumes of binding buffer (20 mM Tris–HCl, pH 8.0, 0.5 M NaCl, 5% glycerol).
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
213
Fig. 1. rhEpo hybrid protein design and expression level studies. (A) A schematic protocol for the vector construction. (B) The proteins produced by the five different rhEpo expression vectors: His6-rhEpo, TRX-rhEpo, GST-rhEpo, MBP-rhEpo, and NusA-rhEpo fusion protein. (C) Lane 1: molecular weight standards; C: negative control; T: total cell lysate; S: supernatant. All strains were grown under the same conditions. C(negative control) means the cultured broth before IPTG induction. (D) Cell growth curve of MBPrhEpo. Cells were incubated for 2.5 h and then induced by addition of IPTG to 1 mM. Left panel: SDS–PAGE analysis of MBP-tagged rhEpo expression. Lane 1: molecular weight standards; C: control; I: total cell lysate at induction (0 h); 1: supernatant after induction for 1 h; 2: 2 h; 3: 3 h; 4: 4 h. Right panel: verification of effective conditions for the expression of MBP-tagged rhEpo. A sample was collected at 3 h after induction and then analyzed using 10% SDS–PAGE. Lane 1: molecular weight standards; T: total cell lysate of induced MBP-tagged rhEpo; P: pellet; S: supernatant.
The column was then washed with the five column volumes of washing buffer (20 mM Tris–HCl, pH 8.0, 0.5 M NaCl, 5% glycerol, 100 mM imidazole). The bound protein was then eluted with elution buffer (20 mM Tris–HCl, pH 8.0, 0.5 M NaCl, 5% glycerol, 1 M imidazole). The eluted rTEV protease was pooled and desalted with 3000 NMWL Centricon (Millipore, MA, USA). The protein concentration of final purified rTEV protease was analyzed by Bradford assay. The TEV protease was stored at 4 °C before use.
Western blot analysis To verify the identify rhEpo, western blot analysis was carried out using an anti-hEpo antibody (Abcam, Cambridge, UK). Samples were loaded onto 10% Tris–glycine SDS–PAGE gels. Resolved proteins were transferred to a nitrocellulose membrane and the membrane was blocked with 5% skimmed milk in TBST (10 mM Tris–HCl, pH 7.4, 150 mM NaCl, 0.05% Tween 20) for 30 min at
214
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
Fig. 1 (continued)
room temperature. The membrane was then incubated with a mouse mAb to hEpo (Abcam, Cambridge, UK) in TBST (1:2000) at Table 1 Expression level and solubility of rhEpo with various protein tags.
a
Tagging protein
Size (kDa)
Expression levela (%)
Solubility (%)
His6 TRX GST MBP NusA
22.4 34.2 48.0 62.7 77.3
47 48 7 36 60
17 34 – 92 85
Expression level refers to the percentage of expressed rhEpo in reference to the total cellular protein levels.
4 °C overnight and washed five times with TBST for 10 min. The washed membrane was incubated with peroxidase-labeled antimouse IgG antibody (Vector Laboratories, MA, USA) in TBST (1:5000) at room temperature for 1 h and washed 5 times in TBST. After washing, the membrane was developed using picoEPD (Elpis Biotech, Daejeon, Korea).
MALDI-TOF MS analysis of disulfide linkages in rhEpo To identify disulfide linkages in purified rhEpo, MALDI-TOF MS analysis was performed under non-reducing and reducing conditions. After treatment of 1 lg protein with or without 10 mM
215
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218 Table 2 Partial purification of rhEpo.
a b c
Purification step
Volume (mL)
Pellet weighta(g)
Total protein (mg)
Purity in SDS– PAGE (%)
Estimated fusion protein (mg)
Estimated rhEpob (mg)
Step yield (%)
Overall yield (%)
Specific activity (U/mg)
Bacterial culture SPNT (after lysis)c MBP eluate TEV treatment GPC eluate
1000 81 122 50 20
4.05 – – – –
– 324 212 130 7.6
– 46 62 29 95
148 130 – –
44.6 39.2 37.3 7.2
– 88 95 19
100 88 84 16
– – – – 5.4 104
Pellet weight means the weight of cell pellet after centrifugation. Estimated rhEpo could be calculated with this equation ‘Estimated fusion protein X 0.3 = Estimated rhEpo’. SPNT was the supernatant of cell lysis broth after the collected cells were resuspended.
human GM-CSF (Prospec, NJ, USA), and 10% FBS at 37 °C in a humidified atmosphere containing 5% CO2. For the bioassays, cells were washed and resuspended at a concentration of 1 104 cells/ mL in RPMI 1640 medium containing 2.2 g/L sodium bicarbonate (Bioshop), 100 lg/mL penicillin/streptomycin antibiotics, and 10% FBS. 100 lL of cells were distributed to each well of a 96-well plate followed by the addition of 100 lL of a dilution series of rhEpo. After incubation at 37 °C in 5% CO2 for 48 h, cellular growth was determined using the WST-1 assay [21]. These experiments were repeated in triplicate. Results Vector construction and expression of rhEpo
DTT for 5 min at room temperature, the reduced and the non-reduced rhEpo samples were precipitated at a final concentration of 10% TCA, and 0.1 mg of precipitated rhEpo was resuspended with 12 lL of IAA buffer (25 mM iodoacetamide, 0.5 M Tris–HCl, pH 8.0, 100 mM NaCl, 1 mM EDTA, 2% SDS, 5% glycerol) and incubated at room temperature in the dark for 1 h. Samples without DTT were separated on a 10% non-reducing Tris–tricine SDS–PAGE gel. MALDI-TOF MS analysis was carried out as described previously [19].
rhEpo has been reported to express as inclusion bodies in E. coli and its correct refolding is problematic [22,23]. Thus, for the soluble overexpression of rhEpo in E. coli, we selected five commonly used protein tags: His, GST, TRX, NusA, and MBP. For construction of vectors encoding rhEpo as hybrid proteins with MBP, the LR recombination cloning system was used (Fig. 1A) and the same LR method was used to construct vectors encoding rhEpo with their various protein tags (Fig. 1B). The expression level and solubility of each tagged protein was analyzed by 10% SDS–PAGE gel and quantified using the ImageJ analysis program. Fig. 1C and Table 1 show that all of the tagged proteins with rhEpo were expressed, with the exception of the GST tagged version. rhEpo when fused with MBP and NusA showed greatest solubility. From these results, we concluded that MBP and NusA were useful in making highly soluble rhEpo in E. coli. For the effective purification of rhEpo, MBP-tagged rhEpo was selected because of its high solubility and a readily available affinity purification system. To determine the time course of expression, samples were collected at multiple time points after induction. Analysis by SDS– PAGE revealed that expression of MBP-tagged rhEpo increased significantly at 3 h after induction (Fig. 1D). The size of expressed MBP-tagged rhEpo was around 63 kDa. A yield of 4.05 g wet cells/ L was obtained under these conditions. The total protein amount of supernatant was 324 mg. The estimated amount of MBP-tagged rhEpo (fusion protein) was 148 mg and the estimated rhEPO was 44.6 mg (Table 2).
In vitro cell proliferation assay
Partial purification of rhEpo
The human TF-1 cell line [20] (ATCC, VA, USA) was maintained in RPMI 1640 (Gibco, NY, USA), supplemented with 2.2 g/L sodium bicarbonate (Bioshop, Ontario, Canada), 100 lg/mL penicillin/ streptomycin antibiotics (Gibco, NY, USA), 2 ng/mL recombinant
Eighty-one milliliters of clarified supernatant, from the lysed MBP-tagged rhEpo expressing E. coli, was loaded onto an MBP affinity column equilibrated with buffer A. The MBP-tagged rhEpo was eluted with buffer B. As shown in Fig. 2A, lanes 2–4, most of
Fig. 2. SDS–PAGE and western blot analyses of rhEpo during purification. (A) SDS– PAGE analysis. Lane M: molecular weight standards; 1: total cell lysate; 2: supernatant; 3: flow-through of the MBP affinity column; 4: the eluted MBP-tagged rhEpo; 5: TEV protease-treated sample; 6: final product eluted from GPC. (B) Western blot analysis. 1: total cell lysate; 2: supernatant; 3: flow-through of the MBP affinity column; 4: the eluted MBP-tagged rhEpo; 5: TEV protease-treated sample; 6: final product eluted from GPC in reducing conditions.
216
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
Fig. 3. MALDI-TOF MS analysis for disulfide linkages of rhEpo. (A) Tryptic peptide map of rhEpo. (B) Reducing and non-reducing PAGE analysis of rhEpo. After TCA was added to a final concentration of 10% to the purified rhEpo, samples in reducing and non-reducing conditions were loaded on a 10% Tris–tricine SDS–PAGE gel. After the gel was stained, the band corresponding to rhEpo was cut out and analyzed by MALDI-TOF MS. The arrows point at the correct MW for the rhEpo. Note the reduced sample runs at a slightly higher MW due to modification. (C and D) The mass spectra are shown for the reducing (C) and non-reducing conditions (D).
the impurities flowed through the column. One hundred twentytwo milliliter of MBP-tagged rhEpo was eluted and the protein containing fractions were pooled. The protein concentration of the eluate of MBP affinity column was 1.74 mg/mL, and the amount of total protein was 212 mg. The estimated yield of fusion protein was 130 mg, which corresponds to 39.2 mg of rhEpo (Table 2). To separate MBP and rhEpo, TEV protease was used to treat the sample eluted from the MBP affinity column. The weight ratio of TEV protease/MBP-tagged rhEpo was 1:30. The rhEpo with MBP was analyzed by SDS–PAGE to show the purity of rhEpo. It was confirmed that the digestion of rhEpo by TEV protease was not complete (Fig. 2A, lane 5), the MBP-tagged rhEpo still remained after cutting by rTEV protease. The size was of rhEpo was 18 kDa, and severed MBP was 44 kDa. After TEV treatment, gel permeation chromatography (GPC) was performed to purify rhEpo using commonly used mobile phases, such as phosphate-buffered saline or buffer A. However, rhEpo did not purify in these general buffers. After assessing many reagents and pH conditions for the purification of rhEpo, it was determined that high salt and a low concentration of detergent were important factors for the purification of rhEpo (data not shown). Based on these data, the TEV-treated sample was loaded on a HiLoad 26/100 Superdex-200 HR column equilibrated with buffer
A, adding 5 mM DTT, 0.01% Triton X-100, and 0.5 M NaCl. SDS– PAGE analysis showed that pure rhEpo was obtained in one of the three major peaks (Fig. 2A, lane 6). Results of the purification procedure are summarized in Table 2. The final yield of the purified rhEpo was approximately 7.2 mg per 1 L bacterial culture broth and its purity was comparatively high (Table 2). But the purified rhEpo was not completely pure. There were two faint bands when it was loaded on SDS–PAGE gel with high concentration. It was inferred that these two bands were unsevered MBP-tagged rhEpo and severed MBP fragment according to SDS–PAGE result (Data not shown). The identity of rhEpo was confirmed using western blot analysis (Fig. 2B). In order to purify rhEpo further and to remove endotoxins, we performed cation exchange chromatography (High S resin) and the gel filtration chromatography (Superose 12 resin) after the TEV treatment step. These further steps improved little for the purity of the rhEpo (data not shown), however the cation exchange chromatography step minimized the endotoxin level to below 20 EU/mg. Characterization of rhEpo by MALDI-TOF MS analysis For identification of the purified rhEpo, MALDI-TOF MS analysis was carried out with reduced as control and finally purified rhEpo
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
Fig. 4. In vitro cell proliferation assay. TF-1 cells were incubated with various concentrations of the rhEpo purified from CHO cells (s) as a positive control (PCEpo) and rhEpo purified from E. coli (d) (rhEpo). After incubation, the biological activity was measured using the WST-1 assay for cell proliferation. The results are the mean ± SE of triplicates. The PCEpo and purified rhEpo show similar biological activity in this in vitro assay. The ED50 of purified rhEpo was 0.37 ng/mL and the specific activity was 5.4 104 U/mg.
as mentioned in Materials and methods. For characterization of the disulfide linkages of the purified rhEpo, MALDI-TOF MS analysis was carried out under reducing and non-reducing conditions as mentioned in Materials and methods. We identified 11 matching peptides of a total of 15 tryptic peptides from a tryptic peptide map of rhEpo (Fig. 3A). According to the peptide map, reduced rhEpo showed a more than 73% match with human Epo and four cysteine residues were modified with IAA, T2+IAA (763.8 Da), T15+IAA (970.2 Da), and T5+2 IAA (2803.7 Da) (Fig. 3C). It was assumed that the purified rhEpo consisted with similar form, such as monomeric form, when it was analyzed under non-reducing conditions (Fig. 3B). Comparison of C and D in Fig. 3, reveals that the purified rhEpo contains two intra-disulfide bonds, Cys13–Cys167 (T2+T15, 1617.9 Da) and Cys35–Cys39 (T5, 2689.2 Da). These data show that the rhEpo purified from E. coli using MBP as a tagging protein had the two correct disulfide bonds. In vitro cell proliferation assay Human TF-1 cells are erythroleukemia cells, requiring hEpo, interleukin 3 (IL-3), or granulocyte–macrophage colony-stimulating factor (GM-CSF) for growth and survival. To analyze the proliferative effects of purified rhEpo on TF-1 cells, 100 lL of 1 104 cells/mL was added to wells of a 96-well plate, and serial 10-fold dilutions of the samples were added in triplicate to wells. After incubation, the biological activity of rhEpo was measured in this proliferation assay. The purified rhEpo showed higher cell proliferation-stimulating activity at a protein concentration of 0.2 ng/ mL than that of the negative control. Compared with the positive control (PCEpo), the purified rhEpo showed similar biological activity in vitro. The ED50 of the purified rhEpo was 0.37 ng/mL and the specific activity was 5.4 104 U/mg (Fig. 4 and Table 2). Discussion In this study, we demonstrated that rhEpo could be solubly overexpressed in the cytoplasm of E. coli and the protein could be purified in simple chromatographic steps and the purified protein showed similar activity as the rhEpo expressed in mammalian cells. For the expression of soluble protein in E. coli, rhEpo was
217
fused with various protein tags. As shown in Fig. 1C, the solubility of MBP- and NusA-tagged rhEpo was much higher than that of His-, TRX-, or GST-tagged rhEpo. The MBP-tagged rhEpo was selected, over NusA-tagged rhEpo due to the greater percentage of soluble protein (Table 1). It has been suggested that MBP acts as a general molecular chaperone [24], for the properly folded form of rhEpo has several clusters of hydrophobic residues on the surface that could serve as binding sites [25]. An MBP affinity column and GPC were used to purify rhEpo (Fig. 2). Initial GPC purification trials were unsuccessful because the rhEpo coeluted with uncleaved proteins and MBP. After screening various buffer conditions, it was found that a combination of high salt concentration and low detergent concentration were important factors for the purification of rhEpo (data not shown). This result suggests that ionic interactions are an important factor in the isolation of rhEpo. The final yield of purified rhEpo was approximately 7.2 mg per L bacterial cultivation and the purity was greater than 95% in SDS–PAGE (Table 2). However, the protein purity after these two chromatography steps was not as good as previous reports [22,26,27]. So, we carried out cation exchange chromatography and a second gel filtration chromatography with a Superose 12 column to improve the purity and also to reduce the endotoxin level. Although this cation exchange chromatography step improved little in terms of purity of rhEpo, it reduced the endotoxin level to 20 EU/mg. The second gel filtration chromatography was not effective in improving the purity of rhEpo (data not shown). Therefore using other ion exchange chromatography and polishing the purification steps are necessary in further study. To determine the identity of the purified rhEpo and the formation of disulfide linkages, MADLI-TOF MS analysis was conducted under reducing and non-reducing conditions (Fig. 3). As shown in Fig. 3C, the purified rhEpo have a pattern highly similar to native hEpo. Additionally, Fig. 3B showed that the purified rhEpo had two intra-disulfide bonds, at sites Cys13–Cys167 and Cys35– Cys39 (Fig. 3C and D), which is consistent with the disulfide bond pattern of hEpo reported previously [28,29]. Natural hEpo has three N-glycosylation and one O-glycosylation sites. In this study, the purified rhEpo showed biological activity similar to that of rhEpo purified from mammalian cells (Fig. 4). Based on the early studies, glycosylation was important for the stability of hEpo [9] by decreasing the clearance rate, but not for its binding affinity to the hEpo receptor [30]. Because of the complexity of the glycosylation, rhEpo and its analogs typically have been produced in mammalian cell cultures [12–17]. However, rhEpo expressed by mammalian cell has several disadvantages, such as low yields and high costs of production. Although E. coli expression system also has disadvantages such as the formation of inclusion bodies and no glycosylation, it can be better expression system compared to mammalian system. The disadvantages of E. coli expression system can be overcome by fusing solubilizing protein tags [25] and/or by conjugating covalent polyethylene glycol (PEG) [31]. Several attempts have been made to express rhEpo in E. coli for structural or functional studies [9,32,33], but none of them showed a soluble expression of the protein in E. coli and few details about the purification procedures were revealed. Because rhEpo derived from E. coli is not glycosylated, a modification such as PEGylation can increase the half-life of the recombinant protein to the level of glycosylated rhEpo in vivo [22,23]. Additionally, in a specific kidney disease, a high amount but a short half-life hEpo is required [34]. In such cases, a non-glycosylated rhEpo from E. coli may be helpful. In conclusion, we reported for the first time that rhEpo can be expressed as a soluble form in E. coli and partially purified in two simple steps. We demonstrated that the purified rhEpo had two correct intra-disulfide bonds. This rhEpo showed similar biological
218
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
activity to rhEpo derived from mammalian cells. Thus, this study may offer opportunities for progress in rhEpo therapeutics.
Acknowledgments This work was supported by the MRC grant (2008-0062286) and the Priority Research Center Program (2009-0094054) funded by the Ministry of Education, Science and Technology, South Korea. RCR is funded by the Agency for Science, Technology and Research (A⁄STAR), Singapore.
References [1] S.B. Krantz, Erythropoietin, Blood 77 (1991) 419–434. [2] M. Silva, D. Grillot, A. Benito, C. Richard, G. Nunez, J.L. Fernandez-Luna, Erythropoietin can promote erythroid progenitor survival by repressing apoptosis through Bcl-XL and Bcl-2, Blood 88 (1996) 1576–1582. [3] B.A. Witthuhn, F.W. Quelle, O. Silvennoinen, T. Yi, B. Tang, O. Miura, J.N. Ihle, JAK2 associates with the erythropoietin receptor and is tyrosine phosphorylated and activated following stimulation with erythropoietin, Cell 74 (1993) 227–236. [4] J.E. Damen, G. Krystal, Early events in erythropoietin-induced signaling, Exp. Hematol. 24 (1996) 1455–1459. [5] C. Bachran, S. Abdelazim, R.J. Fattah, S. Liu, S.H. Leppla, Recombinant expression and purification of a tumor-targeted toxin in Bacillus anthracis, Biochem. Biophys. Res. Commun. 430 (2013) 150–155. [6] A. Markham, H.M. Bryson, Epoetin alfa. A review of its pharmacodynamic and pharmacokinetic properties and therapeutic use in nonrenal applications, Drugs 49 (1995) 232–254. [7] E.J. Sharples, C. Thiemermann, M.M. Yaqoob, Novel applications of recombinant erythropoietin, Curr. Opin. Pharmacol. 6 (2006) 184–189. [8] M.S. Dordal, F.F. Wang, E. Goldwasser, The role of carbohydrate in erythropoietin action, Endocrinology 116 (1985) 2293–2299. [9] L.O. Narhi, T. Arakawa, K.H. Aoki, R. Elmore, M.F. Rohde, T. Boone, T.W. Strickland, The effect of carbohydrate on the structure and stability of erythropoietin, J. Biol. Chem. 266 (1991) 23022–23026. [10] K. Yamaguchi, K. Akai, G. Kawanishi, M. Ueda, S. Masuda, R. Sasaki, Effects of site-directed removal of N-glycosylation sites in human erythropoietin on its production and biological properties, J. Biol. Chem. 266 (1991) 20434–20439. [11] M. Higuchi, M. Oh-eda, H. Kuboniwa, K. Tomonoh, Y. Shimonaka, N. Ochi, Role of sugar chains in the expression of the biological activity of human erythropoietin, J. Biol. Chem. 267 (1992) 7703–7709. [12] J.H. Park, C. Kim, W.B. Kim, Y.K. Kim, S.Y. Lee, J.M. Yang, Efficiency of promoter and cell line in high-level expression of erythropoietin, Biotechnol. Appl. Biochem. 32 (Pt 3) (2000) 167–172. [13] N. Irani, A.J. Beccaria, R. Wagner, Expression of recombinant cytoplasmic yeast pyruvate carboxylase for the improvement of the production of human erythropoietin by recombinant BHK-21 cells, J. Biotechnol. 93 (2002) 269–282. [14] M.D. Wang, M. Yang, N. Huzel, M. Butler, Erythropoietin production from CHO cells grown by continuous culture in a fluidized-bed bioreactor, Biotechnol. Bioeng. 77 (2002) 194–203. [15] J.C. Egrie, E. Dwyer, J.K. Browne, A. Hitz, M.A. Lykos, Darbepoetin alfa has a longer circulating half-life and greater in vivo potency than recombinant human erythropoietin, Exp. Hematol. 31 (2003) 290–299.
[16] D. Zanette, A. Soffientini, C. Sottani, E. Sarubbi, Evaluation of phenylboronate agarose for industrial-scale purification of erythropoietin from mammalian cell cultures, J. Biotechnol. 101 (2003) 275–287. [17] K. Schriebl, E. Trummer, C. Lattenmayer, R. Weik, R. Kunert, D. Muller, H. Katinger, K. Vorauer-Uhl, Biochemical characterization of rhEpo-Fc fusion protein expressed in CHO cells, Protein Expr. Purif. 49 (2006) 265–275. [18] L.J. Lucast, R.T. Batey, J.A. Doudna, Large-scale purification of a stable form of recombinant tobacco etch virus protease, Biotechniques 30 (2001) 544–546. [19] J.W. Lee, J.D. Helmann, The PerR transcription factor senses H2O2 by metalcatalysed histidine oxidation, Nature 440 (2006) 363–367. [20] T. Kitamura, T. Tange, T. Terasawa, S. Chiba, T. Kuwaki, K. Miyagawa, Y.F. Piao, K. Miyazono, A. Urabe, F. Takaku, Establishment and characterization of a unique human cell line that proliferates dependently on GM-CSF, IL-3, or erythropoietin, J. Cell. Physiol. 140 (1989) 323–334. [21] A.R. Mire-Sluis, R. Thorpe, Laboratory protocols for the quantitation of cytokines by bioassay using cytokine responsive cell lines, J. Immunol. Methods 211 (1998) 199–210. [22] Y.J. Wang, Y.D. Liu, J. Chen, S.J. Hao, T. Hu, G.H. Ma, Z.G. Su, Efficient preparation and PEGylation of recombinant human non-glycosylated erythropoietin expressed as inclusion body in E. coli, Int. J. Pharm. 386 (2010) 156–164. [23] Y.J. Wang, S.J. Hao, Y.D. Liu, T. Hu, G.F. Zhang, X. Zhang, Q.S. Qi, G.H. Ma, Z.G. Su, PEGylation markedly enhances the in vivo potency of recombinant human non-glycosylated erythropoietin: a comparison with glycosylated erythropoietin, J. Control. Release 145 (2010) 306–313. [24] H. Bach, Y. Mazor, S. Shaky, A. Shoham-Lev, Y. Berdichevsky, D.L. Gutnick, I. Benhar, Escherichia coli maltose-binding protein as a molecular chaperone for recombinant intracellular cytoplasmic single-chain antibodies, J. Mol. Biol. 312 (2001) 79–93. [25] R.B. Kapust, D.S. Waugh, Escherichia coli maltose-binding protein is uncommonly effective at promoting the solubility of polypeptides to which it is fused, Protein Sci. 8 (1999) 1668–1674. [26] S.M. Singh, A.K. Panda, Solubilization and refolding of bacterial inclusion body proteins, J. Biosci. Bioeng. 99 (2005) 303–310. [27] B. Batas, C. Schiraldi, J.B. Chaudhuri, Inclusion body purification and protein refolding using microfiltration and size exclusion chromatography, J. Biotechnol. 68 (1999) 149–158. [28] J.C. Cheetham, D.M. Smith, K.H. Aoki, J.L. Stevenson, T.J. Hoeffel, R.S. Syed, J. Egrie, T.S. Harvey, NMR structure of human erythropoietin and a comparison with its receptor bound conformation, Nat. Struct. Biol. 5 (1998) 861–866. [29] D. Wen, J.P. Boissel, T.E. Tracy, R.H. Gruninger, L.S. Mulcahy, J. Czelusniak, M. Goodman, H.F. Bunn, Erythropoietin structure-function relationships: high degree of sequence homology among mammals, Blood 82 (1993) 1507–1516. [30] R.J. Darling, U. Kuchibhotla, W. Glaesner, R. Micanovic, D.R. Witcher, J.M. Beals, PGlycosylation of erythropoietin affects receptor binding kinetics: role of electrostatic interactions, Biochemistry (Mosc) 41 (2002) 14524–14531. [31] F.M. Veronese, G. Pasut, PEGylation, successful approach to drug delivery, Drug Discov. Today 10 (2005) 1451–1458. [32] L.O. Narhi, T. Arakawa, K. Aoki, J. Wen, S. Elliott, T. Boone, J. Cheetham, Asn to Lys mutations at three sites which are N-glycosylated in the mammalian protein decrease the aggregation of Escherichia coli-derived erythropoietin, Protein Eng. 14 (2001) 135–140. [33] J.P. Boissel, W.R. Lee, S.R. Presnell, F.E. Cohen, H.F. Bunn, Erythropoietin structure-function relationships. Mutant proteins that test a model of tertiary structure, J. Biol. Chem. 268 (1993) 15983–15993. [34] Y. Yokomaku, T. Sugimoto, S. Kume, S. Araki, K. Isshiki, M. Chin-Kanasaki, M. Sakaguchi, N. Nitta, M. Haneda, D. Koya, T. Uzu, A. Kashiwagi, Asialoerythropoietin prevents contrast-induced nephropathy, J. Am. Soc. Nephrol. 19 (2008) 321–328.
Contents lists available at ScienceDirect
Protein Expression and Purification journal homepage: www.elsevier.com/locate/yprep
Soluble expression and partial purification of recombinant human erythropoietin from E. coli Taeck-Hyun Jeong a,1, Young-Jin Son b,1, Han-Bong Ryu a,1, Bon-Kyung Koo a, Seung-Mi Jeong a, Phuong Hoang a, Bich Hang Do a, Jung-A Song a, Seon-Ha Chong a, Robert Charles Robinson c,d,e, Han Choe a,⇑ a
Department of Physiology and Bio-Medical Institute of Technology, University of Ulsan College of Medicine, Seoul 138-736, South Korea Department of Pharmacy, Sunchon National University, Suncheon, Jeonnam 540-742, South Korea Institute of Molecular and Cell Biology, Proteos, 61 Biopolis Drive, Singapore 138673, Singapore d Department of Biochemistry, National University of Singapore, 8 Medical Drive, Singapore 117597, Singapore e School of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, Singapore 637551, Singapore b c
a r t i c l e
i n f o
Article history: Received 23 March 2012 and in revised form 1 January 2014 Available online 9 January 2014 Keywords: rhEpo Escherichia coli expression system Maltose binding protein (MBP) Therapeutic protein
a b s t r a c t Human erythropoietin (hEpo) is an essential regulator of erythrocyte production that induces the division and differentiation of erythroid progenitor cells in the bone marrow into mature erythrocytes. It is widely used for the treatment of anemia resulting from chronic kidney disease, chemotherapy, and cancerrelated therapies. Active hEpo, and hEpo analogs, have been purified primarily from mammalian cells, which has several disadvantages, including low yields and high production costs. Although an Escherichia coli (E. coli) expression system may provide economic production of therapeutic proteins, it has not been used for the production of recombinant hEpo (rhEpo) because it aggregates in inclusion bodies in the E. coli cytoplasm and is not modified post-translationally. We investigated the soluble overexpression of active rhEpo with various protein tags in E. coli, and found out that several tags increased the solubility of rhEpo. Among them maltose binding protein (MBP)-tagged rhEpo was purified using affinity and gel filtration columns. Non-denaturing electrophoresis and MALDI-TOF MS analysis demonstrated that the purified rhEpo had two intra-disulfide bonds identical to those of the native hEpo. An in vitro proliferation assay showed that rhEpo purified from E. coli had similar biological activity as rhEpo derived from CHO cells. Therefore, we report for the first time that active rhEpo was overexpressed as a soluble form in the cytoplasm of E. coli and purified it in simple purification steps. We hope that our results offer opportunities for progress in rhEpo therapeutics. Ó 2014 Elsevier Inc. All rights reserved.
Introduction Human erythropoietin (hEpo),2 a 34-kDa glycoprotein produced primarily by the kidney, is the principal factor regulating erythrocyte production [1]. hEpo induces erythroid progenitor cell proliferation, differentiation, and maturation and inhibits apoptosis, resulting in increased erythrocyte production [2]. hEpo achieves its effects by causing homodimerization of the hEpo receptor with subsequent autophosphorylation of the tyrosine kinase JAK2 [3] and phosphorylation of the receptor itself, as well as various substrate proteins, leading to up-regulation of a number of signaling pathways and activation of gene transcription [4]. Natural hEpo is glycoprotein which ⇑ Corresponding author. Tel.: +82 2 3010 4292; fax: +82 2 3010 8148. E-mail address: [email protected] (H. Choe). These authors contributed equally to the work. 2 Abbreviations used: hEpo; human erythropoietin; rhEpo, recombinant human Epo; MBP, maltose binding protein; His, histidine; GST, glutathione S-transferase; TRX, thioredoxin; IPTG, isopropyl-b-D-thiogalactoside; GPC, gel permeation chromatography; IL-3, interleukin 3; PEG, polyethylene glycol. 1
1046-5928/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.pep.2014.01.001
has three sites of N-glycosylation (N24, N38, and N83) and one of O-glycosylation (S126). E. coli is the most frequently used host cell for the production of recombinant protein. Its expression system may allow economic and fast production of large amounts of heterologous protein. However, there were some disadvantages of E. coli production for recombinant human protein. The recombinant protein is usually expressed as an inclusion body and non-glycosylated form [5]. Recombinant human Epo (rhEpo) has been used clinically to treat anemia resulting from chronic kidney disease, chemotherapy, and cancer-related therapies [6]. In addition to these well-established applications, it also has therapeutic potential in the treatment of acute brain, heart, and kidney injuries [7]. According to some previous reports, glycosylation of hEpo is important in maintaining its stability and hematopoietic activity in vivo [8–11]. In order to produce active rhEpo, it has been typically produced in mammalian cell cultures [12–17]. In this study, we demonstrate for the first time that the expression of soluble active maltose binding protein (MBP)-tagged rhEpo
212
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
in E. coli, partial purification, and cell-based biological activity. This study offers opportunities for progress in rhEpo therapeutics. Materials and methods Histidine (His), glutathione S-transferase (GST), thioredoxin (TRX), NusA, or MBP tagged rhEpo constructs The pET22b-based destination vector, which allows inclusion of various protein tags at the N-terminus of the encoded protein, was used as the expression vector. To amplify the hEpo gene, two primers (forward 50 -GCGCTGGGCGCGCAGAAAGAAGCTATCAGTC-30 and reverse 50 -GACTGATAGCTTCTTTCTGCGCGCCCAGCGC-30 ) were used. For construction of the entry vector, DNA sequences encoding the 166 amino acids of mature human Epo (NP_000790.2) were synthesized and codon-optimized for E. coli expression by a gene synthesis service (Epoch Life Science Inc., TX, USA). The codonoptimized rhEpo gene was inserted into the pDONR 207 vector using the BP recombination cloning system (Invitrogen, CA, USA). The TEV protease recognition site (ENLYFQ;G) was inserted in front of the amino acid sequence of rhEpo. To obtain the expression vector, the rhEpo gene in pENTR-rhEpo was cloned into pHGWA (His6), pHGGWA (GST), pHXGWA (TRX), pHNGWA (NusA), and pHMGWA (MBP) vectors using the LR recombination cloning system (Invitrogen). After confirming the DNA sequence of rhEpo, the resulting plasmids were introduced into E. coli BL21 (DE3). Expression of rhEpo with various protein tags To analyze the expression level of rhEpo with variant protein tags, expression clones were pre-grown overnight in 3 mL LB broth supplemented with 50 lg/mL ampicillin at 37 °C. Pre-cultured samples were inoculated in 1 L new LB broth containing an appropriate antibiotic and grown to mid-log phase (OD600 = 0.4–0.5). Then, isopropyl-b-d-thiogalactoside (IPTG) was added to a final concentration of 1 mM. To improve the soluble expression level, the temperature of cultivation was changed to 30 °C. After 3 h, the cells were harvested by centrifugation (3500 rpm, 30 min). Cell pellets were suspended in 10 mL/g cell buffer A (50 mM Tris–HCl, pH 8.0, 0.5 mM EDTA, 5% glycerol) and disrupted by sonication. After centrifugation (12,000 rpm, 30 min), soluble and insoluble fractions were collected. The expression levels of rhEpo with various protein tags were assessed using 10% Tris–glycine SDS–PAGE, followed by image analysis using the ImageJ program (http:// imagej.nih.gov/ij). Purification of rhEpo Eighty-one milliliters of clarified supernatant was loaded onto a MBP Trap HP containing 20 mL volume resin (GE Life Sciences, NJ, USA) equilibrated with 5 column volume of buffer A using ÄKTA Prime Plus (GE Life Sciences). The column was then washed with buffer A until the UV (280 nm) value reached baseline. The bound protein was then eluted with buffer B (50 mM Tris–HCl, pH 8.0, 0.5 mM EDTA, 5% glycerol, 20 mM maltose monohydrate; Junsei Chemical Co., Tokyo, Japan). The eluted MBP-rhEpo fusion protein was pooled and its size was confirmed by 10% Tris–glycine SDS– PAGE analysis. The eluted MBP-rhEpo was filtered by 0.45 lM filter. The protein concentration of MBP-rhEpo was analyzed by Bradford assay, and then diluted with TEV standard buffer (50 mM Tris–HCl, pH 8.0, 0.5 mM EDTA, 5 mM dithiothreitol (DTT), 5% (v/v) glycerol) at 0.5 mg/mL. The MBP-tagged rhEpo eluted from the MBP affinity column was treated with partially purified rTEV protease in TEV standard buffer. The reaction condition for cutting fusion protein was
30 mg of MBP-rhEpo solution per 1 mg of rTEV protease crude solution. The reaction solution was incubated at room temperature overnight [18]. To further purify rhEpo, the sample treated with TEV protease was loaded on a HiLoad 26/100 Superdex200 HR gel filtration chromatography column (GE Life Sciences, NJ, USA) equilibrated with buffer A, supplemented with 5 mM DTT, 0.01% Triton X-100, and 0.5 M NaCl. The flow rate was 1 mL/min. After exchanging buffer A with PBS supplemented by 5% glycerol using an 3000 NMWL Centricon (Millipore, MA, USA), the protein was analyzed using 10% Tris–glycine SDS–PAGE and quantified using the ImageJ program. The final purified rhEpo was stored at 20 °C. Cation exchange and the second gel filtration chromatography of rhEpo In order to check the impurity profile of rhEpo after ion exchange chromatography, cation exchange chromatography was carried out on AKTA™ avant chromatography system (GE Life Sciences, NJ, USA) using High S column (5 mL packing resin, GE Life Sciences, NJ, USA). Equilibrium buffer was 6 M urea and 0.25 M acetic acid, pH 2.5. Elution buffer was 6 M urea and 0.25 M acetic acid, 1 M NaCl, pH 2.5. The flow rate was 2.5 mL/min and the elution started with 100% of equilibrium buffer, then followed by linear gradient of 0–100% elution buffer for 40 min. Elution was monitored with UV detector at 280 nm. A second gel filtration chromatography step was performed in an attempt to further purify the rhEpo. It was carried out on AKTA™ avant chromatography system (GE Life Sciences, NJ, USA) using Superose™ 12 10/300 GL column (GE Life Sciences, NJ, USA). Mobile phase was 50 mM sodium phosphate and 0.15 M NaCl, pH 7.0. The flow rate was 0.5 mL/min. The elution was monitored with UV detector at 280 nm. Endotoxin assay of purified rhEpo To determine the amount of the remaining endotoxin, ToxinSensor™ chromogenic LAL Endotoxin Assay Kit (GenScript, Piscataway, NJ, USA) was used. Endotoxin standard (100 lL) and samples (100 lL) were added into endotoxin-free vials. The pHs of standard and samples were adjusted at pH 7.0. Limulus amebocyte lysate (LAL) reagent was added into each vial and mixed thoroughly. The vials were incubated at 37 °C for 45 min. After incubation, 100 lL of chromogenic substrate solution was added to each vial and incubated at 37 °C for 6 min. Five hundred microliters of stop solution and 500 lL of color-stabilizer (#2 and #3) were added into each vial and mixed well. The absorbance of each reaction at 545 nm was analyzed by spectrophotometer. After analyzing the absorbance, we calculated endotoxin unit with the standard curve obtained by standard solution. Expression and purification of rTEV protease Expression clones for rTEV were cultured in 3 mL LB broth supplemented with 50 lg/mL ampicillin at 37 °C. When the optical density at 600 nm was 0.5, isopropyl-b-D-thiogalactoside (IPTG) was added to a final concentration of 1 mM in order to express rTEV. After 3 h, the cells were harvested by centrifugation (3500 rpm, 30 min). Cell pellets were suspended in 10 mL/g cell buffer A (50 mM Tris–HCl, pH 8.0, 0.5 mM EDTA, 5% glycerol) and disrupted by sonication. After centrifugation (12,000 rpm, 30 min), soluble fraction was collected. Then the soluble fraction was filtered with 0.45 lM filter. Ten milliliters of clarified supernatant was loaded onto a nickel column containing 5 mL volume resin (GE Life Sciences, NJ, USA) equilibrated with 2 column volumes of binding buffer (20 mM Tris–HCl, pH 8.0, 0.5 M NaCl, 5% glycerol).
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
213
Fig. 1. rhEpo hybrid protein design and expression level studies. (A) A schematic protocol for the vector construction. (B) The proteins produced by the five different rhEpo expression vectors: His6-rhEpo, TRX-rhEpo, GST-rhEpo, MBP-rhEpo, and NusA-rhEpo fusion protein. (C) Lane 1: molecular weight standards; C: negative control; T: total cell lysate; S: supernatant. All strains were grown under the same conditions. C(negative control) means the cultured broth before IPTG induction. (D) Cell growth curve of MBPrhEpo. Cells were incubated for 2.5 h and then induced by addition of IPTG to 1 mM. Left panel: SDS–PAGE analysis of MBP-tagged rhEpo expression. Lane 1: molecular weight standards; C: control; I: total cell lysate at induction (0 h); 1: supernatant after induction for 1 h; 2: 2 h; 3: 3 h; 4: 4 h. Right panel: verification of effective conditions for the expression of MBP-tagged rhEpo. A sample was collected at 3 h after induction and then analyzed using 10% SDS–PAGE. Lane 1: molecular weight standards; T: total cell lysate of induced MBP-tagged rhEpo; P: pellet; S: supernatant.
The column was then washed with the five column volumes of washing buffer (20 mM Tris–HCl, pH 8.0, 0.5 M NaCl, 5% glycerol, 100 mM imidazole). The bound protein was then eluted with elution buffer (20 mM Tris–HCl, pH 8.0, 0.5 M NaCl, 5% glycerol, 1 M imidazole). The eluted rTEV protease was pooled and desalted with 3000 NMWL Centricon (Millipore, MA, USA). The protein concentration of final purified rTEV protease was analyzed by Bradford assay. The TEV protease was stored at 4 °C before use.
Western blot analysis To verify the identify rhEpo, western blot analysis was carried out using an anti-hEpo antibody (Abcam, Cambridge, UK). Samples were loaded onto 10% Tris–glycine SDS–PAGE gels. Resolved proteins were transferred to a nitrocellulose membrane and the membrane was blocked with 5% skimmed milk in TBST (10 mM Tris–HCl, pH 7.4, 150 mM NaCl, 0.05% Tween 20) for 30 min at
214
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
Fig. 1 (continued)
room temperature. The membrane was then incubated with a mouse mAb to hEpo (Abcam, Cambridge, UK) in TBST (1:2000) at Table 1 Expression level and solubility of rhEpo with various protein tags.
a
Tagging protein
Size (kDa)
Expression levela (%)
Solubility (%)
His6 TRX GST MBP NusA
22.4 34.2 48.0 62.7 77.3
47 48 7 36 60
17 34 – 92 85
Expression level refers to the percentage of expressed rhEpo in reference to the total cellular protein levels.
4 °C overnight and washed five times with TBST for 10 min. The washed membrane was incubated with peroxidase-labeled antimouse IgG antibody (Vector Laboratories, MA, USA) in TBST (1:5000) at room temperature for 1 h and washed 5 times in TBST. After washing, the membrane was developed using picoEPD (Elpis Biotech, Daejeon, Korea).
MALDI-TOF MS analysis of disulfide linkages in rhEpo To identify disulfide linkages in purified rhEpo, MALDI-TOF MS analysis was performed under non-reducing and reducing conditions. After treatment of 1 lg protein with or without 10 mM
215
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218 Table 2 Partial purification of rhEpo.
a b c
Purification step
Volume (mL)
Pellet weighta(g)
Total protein (mg)
Purity in SDS– PAGE (%)
Estimated fusion protein (mg)
Estimated rhEpob (mg)
Step yield (%)
Overall yield (%)
Specific activity (U/mg)
Bacterial culture SPNT (after lysis)c MBP eluate TEV treatment GPC eluate
1000 81 122 50 20
4.05 – – – –
– 324 212 130 7.6
– 46 62 29 95
148 130 – –
44.6 39.2 37.3 7.2
– 88 95 19
100 88 84 16
– – – – 5.4 104
Pellet weight means the weight of cell pellet after centrifugation. Estimated rhEpo could be calculated with this equation ‘Estimated fusion protein X 0.3 = Estimated rhEpo’. SPNT was the supernatant of cell lysis broth after the collected cells were resuspended.
human GM-CSF (Prospec, NJ, USA), and 10% FBS at 37 °C in a humidified atmosphere containing 5% CO2. For the bioassays, cells were washed and resuspended at a concentration of 1 104 cells/ mL in RPMI 1640 medium containing 2.2 g/L sodium bicarbonate (Bioshop), 100 lg/mL penicillin/streptomycin antibiotics, and 10% FBS. 100 lL of cells were distributed to each well of a 96-well plate followed by the addition of 100 lL of a dilution series of rhEpo. After incubation at 37 °C in 5% CO2 for 48 h, cellular growth was determined using the WST-1 assay [21]. These experiments were repeated in triplicate. Results Vector construction and expression of rhEpo
DTT for 5 min at room temperature, the reduced and the non-reduced rhEpo samples were precipitated at a final concentration of 10% TCA, and 0.1 mg of precipitated rhEpo was resuspended with 12 lL of IAA buffer (25 mM iodoacetamide, 0.5 M Tris–HCl, pH 8.0, 100 mM NaCl, 1 mM EDTA, 2% SDS, 5% glycerol) and incubated at room temperature in the dark for 1 h. Samples without DTT were separated on a 10% non-reducing Tris–tricine SDS–PAGE gel. MALDI-TOF MS analysis was carried out as described previously [19].
rhEpo has been reported to express as inclusion bodies in E. coli and its correct refolding is problematic [22,23]. Thus, for the soluble overexpression of rhEpo in E. coli, we selected five commonly used protein tags: His, GST, TRX, NusA, and MBP. For construction of vectors encoding rhEpo as hybrid proteins with MBP, the LR recombination cloning system was used (Fig. 1A) and the same LR method was used to construct vectors encoding rhEpo with their various protein tags (Fig. 1B). The expression level and solubility of each tagged protein was analyzed by 10% SDS–PAGE gel and quantified using the ImageJ analysis program. Fig. 1C and Table 1 show that all of the tagged proteins with rhEpo were expressed, with the exception of the GST tagged version. rhEpo when fused with MBP and NusA showed greatest solubility. From these results, we concluded that MBP and NusA were useful in making highly soluble rhEpo in E. coli. For the effective purification of rhEpo, MBP-tagged rhEpo was selected because of its high solubility and a readily available affinity purification system. To determine the time course of expression, samples were collected at multiple time points after induction. Analysis by SDS– PAGE revealed that expression of MBP-tagged rhEpo increased significantly at 3 h after induction (Fig. 1D). The size of expressed MBP-tagged rhEpo was around 63 kDa. A yield of 4.05 g wet cells/ L was obtained under these conditions. The total protein amount of supernatant was 324 mg. The estimated amount of MBP-tagged rhEpo (fusion protein) was 148 mg and the estimated rhEPO was 44.6 mg (Table 2).
In vitro cell proliferation assay
Partial purification of rhEpo
The human TF-1 cell line [20] (ATCC, VA, USA) was maintained in RPMI 1640 (Gibco, NY, USA), supplemented with 2.2 g/L sodium bicarbonate (Bioshop, Ontario, Canada), 100 lg/mL penicillin/ streptomycin antibiotics (Gibco, NY, USA), 2 ng/mL recombinant
Eighty-one milliliters of clarified supernatant, from the lysed MBP-tagged rhEpo expressing E. coli, was loaded onto an MBP affinity column equilibrated with buffer A. The MBP-tagged rhEpo was eluted with buffer B. As shown in Fig. 2A, lanes 2–4, most of
Fig. 2. SDS–PAGE and western blot analyses of rhEpo during purification. (A) SDS– PAGE analysis. Lane M: molecular weight standards; 1: total cell lysate; 2: supernatant; 3: flow-through of the MBP affinity column; 4: the eluted MBP-tagged rhEpo; 5: TEV protease-treated sample; 6: final product eluted from GPC. (B) Western blot analysis. 1: total cell lysate; 2: supernatant; 3: flow-through of the MBP affinity column; 4: the eluted MBP-tagged rhEpo; 5: TEV protease-treated sample; 6: final product eluted from GPC in reducing conditions.
216
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
Fig. 3. MALDI-TOF MS analysis for disulfide linkages of rhEpo. (A) Tryptic peptide map of rhEpo. (B) Reducing and non-reducing PAGE analysis of rhEpo. After TCA was added to a final concentration of 10% to the purified rhEpo, samples in reducing and non-reducing conditions were loaded on a 10% Tris–tricine SDS–PAGE gel. After the gel was stained, the band corresponding to rhEpo was cut out and analyzed by MALDI-TOF MS. The arrows point at the correct MW for the rhEpo. Note the reduced sample runs at a slightly higher MW due to modification. (C and D) The mass spectra are shown for the reducing (C) and non-reducing conditions (D).
the impurities flowed through the column. One hundred twentytwo milliliter of MBP-tagged rhEpo was eluted and the protein containing fractions were pooled. The protein concentration of the eluate of MBP affinity column was 1.74 mg/mL, and the amount of total protein was 212 mg. The estimated yield of fusion protein was 130 mg, which corresponds to 39.2 mg of rhEpo (Table 2). To separate MBP and rhEpo, TEV protease was used to treat the sample eluted from the MBP affinity column. The weight ratio of TEV protease/MBP-tagged rhEpo was 1:30. The rhEpo with MBP was analyzed by SDS–PAGE to show the purity of rhEpo. It was confirmed that the digestion of rhEpo by TEV protease was not complete (Fig. 2A, lane 5), the MBP-tagged rhEpo still remained after cutting by rTEV protease. The size was of rhEpo was 18 kDa, and severed MBP was 44 kDa. After TEV treatment, gel permeation chromatography (GPC) was performed to purify rhEpo using commonly used mobile phases, such as phosphate-buffered saline or buffer A. However, rhEpo did not purify in these general buffers. After assessing many reagents and pH conditions for the purification of rhEpo, it was determined that high salt and a low concentration of detergent were important factors for the purification of rhEpo (data not shown). Based on these data, the TEV-treated sample was loaded on a HiLoad 26/100 Superdex-200 HR column equilibrated with buffer
A, adding 5 mM DTT, 0.01% Triton X-100, and 0.5 M NaCl. SDS– PAGE analysis showed that pure rhEpo was obtained in one of the three major peaks (Fig. 2A, lane 6). Results of the purification procedure are summarized in Table 2. The final yield of the purified rhEpo was approximately 7.2 mg per 1 L bacterial culture broth and its purity was comparatively high (Table 2). But the purified rhEpo was not completely pure. There were two faint bands when it was loaded on SDS–PAGE gel with high concentration. It was inferred that these two bands were unsevered MBP-tagged rhEpo and severed MBP fragment according to SDS–PAGE result (Data not shown). The identity of rhEpo was confirmed using western blot analysis (Fig. 2B). In order to purify rhEpo further and to remove endotoxins, we performed cation exchange chromatography (High S resin) and the gel filtration chromatography (Superose 12 resin) after the TEV treatment step. These further steps improved little for the purity of the rhEpo (data not shown), however the cation exchange chromatography step minimized the endotoxin level to below 20 EU/mg. Characterization of rhEpo by MALDI-TOF MS analysis For identification of the purified rhEpo, MALDI-TOF MS analysis was carried out with reduced as control and finally purified rhEpo
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
Fig. 4. In vitro cell proliferation assay. TF-1 cells were incubated with various concentrations of the rhEpo purified from CHO cells (s) as a positive control (PCEpo) and rhEpo purified from E. coli (d) (rhEpo). After incubation, the biological activity was measured using the WST-1 assay for cell proliferation. The results are the mean ± SE of triplicates. The PCEpo and purified rhEpo show similar biological activity in this in vitro assay. The ED50 of purified rhEpo was 0.37 ng/mL and the specific activity was 5.4 104 U/mg.
as mentioned in Materials and methods. For characterization of the disulfide linkages of the purified rhEpo, MALDI-TOF MS analysis was carried out under reducing and non-reducing conditions as mentioned in Materials and methods. We identified 11 matching peptides of a total of 15 tryptic peptides from a tryptic peptide map of rhEpo (Fig. 3A). According to the peptide map, reduced rhEpo showed a more than 73% match with human Epo and four cysteine residues were modified with IAA, T2+IAA (763.8 Da), T15+IAA (970.2 Da), and T5+2 IAA (2803.7 Da) (Fig. 3C). It was assumed that the purified rhEpo consisted with similar form, such as monomeric form, when it was analyzed under non-reducing conditions (Fig. 3B). Comparison of C and D in Fig. 3, reveals that the purified rhEpo contains two intra-disulfide bonds, Cys13–Cys167 (T2+T15, 1617.9 Da) and Cys35–Cys39 (T5, 2689.2 Da). These data show that the rhEpo purified from E. coli using MBP as a tagging protein had the two correct disulfide bonds. In vitro cell proliferation assay Human TF-1 cells are erythroleukemia cells, requiring hEpo, interleukin 3 (IL-3), or granulocyte–macrophage colony-stimulating factor (GM-CSF) for growth and survival. To analyze the proliferative effects of purified rhEpo on TF-1 cells, 100 lL of 1 104 cells/mL was added to wells of a 96-well plate, and serial 10-fold dilutions of the samples were added in triplicate to wells. After incubation, the biological activity of rhEpo was measured in this proliferation assay. The purified rhEpo showed higher cell proliferation-stimulating activity at a protein concentration of 0.2 ng/ mL than that of the negative control. Compared with the positive control (PCEpo), the purified rhEpo showed similar biological activity in vitro. The ED50 of the purified rhEpo was 0.37 ng/mL and the specific activity was 5.4 104 U/mg (Fig. 4 and Table 2). Discussion In this study, we demonstrated that rhEpo could be solubly overexpressed in the cytoplasm of E. coli and the protein could be purified in simple chromatographic steps and the purified protein showed similar activity as the rhEpo expressed in mammalian cells. For the expression of soluble protein in E. coli, rhEpo was
217
fused with various protein tags. As shown in Fig. 1C, the solubility of MBP- and NusA-tagged rhEpo was much higher than that of His-, TRX-, or GST-tagged rhEpo. The MBP-tagged rhEpo was selected, over NusA-tagged rhEpo due to the greater percentage of soluble protein (Table 1). It has been suggested that MBP acts as a general molecular chaperone [24], for the properly folded form of rhEpo has several clusters of hydrophobic residues on the surface that could serve as binding sites [25]. An MBP affinity column and GPC were used to purify rhEpo (Fig. 2). Initial GPC purification trials were unsuccessful because the rhEpo coeluted with uncleaved proteins and MBP. After screening various buffer conditions, it was found that a combination of high salt concentration and low detergent concentration were important factors for the purification of rhEpo (data not shown). This result suggests that ionic interactions are an important factor in the isolation of rhEpo. The final yield of purified rhEpo was approximately 7.2 mg per L bacterial cultivation and the purity was greater than 95% in SDS–PAGE (Table 2). However, the protein purity after these two chromatography steps was not as good as previous reports [22,26,27]. So, we carried out cation exchange chromatography and a second gel filtration chromatography with a Superose 12 column to improve the purity and also to reduce the endotoxin level. Although this cation exchange chromatography step improved little in terms of purity of rhEpo, it reduced the endotoxin level to 20 EU/mg. The second gel filtration chromatography was not effective in improving the purity of rhEpo (data not shown). Therefore using other ion exchange chromatography and polishing the purification steps are necessary in further study. To determine the identity of the purified rhEpo and the formation of disulfide linkages, MADLI-TOF MS analysis was conducted under reducing and non-reducing conditions (Fig. 3). As shown in Fig. 3C, the purified rhEpo have a pattern highly similar to native hEpo. Additionally, Fig. 3B showed that the purified rhEpo had two intra-disulfide bonds, at sites Cys13–Cys167 and Cys35– Cys39 (Fig. 3C and D), which is consistent with the disulfide bond pattern of hEpo reported previously [28,29]. Natural hEpo has three N-glycosylation and one O-glycosylation sites. In this study, the purified rhEpo showed biological activity similar to that of rhEpo purified from mammalian cells (Fig. 4). Based on the early studies, glycosylation was important for the stability of hEpo [9] by decreasing the clearance rate, but not for its binding affinity to the hEpo receptor [30]. Because of the complexity of the glycosylation, rhEpo and its analogs typically have been produced in mammalian cell cultures [12–17]. However, rhEpo expressed by mammalian cell has several disadvantages, such as low yields and high costs of production. Although E. coli expression system also has disadvantages such as the formation of inclusion bodies and no glycosylation, it can be better expression system compared to mammalian system. The disadvantages of E. coli expression system can be overcome by fusing solubilizing protein tags [25] and/or by conjugating covalent polyethylene glycol (PEG) [31]. Several attempts have been made to express rhEpo in E. coli for structural or functional studies [9,32,33], but none of them showed a soluble expression of the protein in E. coli and few details about the purification procedures were revealed. Because rhEpo derived from E. coli is not glycosylated, a modification such as PEGylation can increase the half-life of the recombinant protein to the level of glycosylated rhEpo in vivo [22,23]. Additionally, in a specific kidney disease, a high amount but a short half-life hEpo is required [34]. In such cases, a non-glycosylated rhEpo from E. coli may be helpful. In conclusion, we reported for the first time that rhEpo can be expressed as a soluble form in E. coli and partially purified in two simple steps. We demonstrated that the purified rhEpo had two correct intra-disulfide bonds. This rhEpo showed similar biological
218
T.-H. Jeong et al. / Protein Expression and Purification 95 (2014) 211–218
activity to rhEpo derived from mammalian cells. Thus, this study may offer opportunities for progress in rhEpo therapeutics.
Acknowledgments This work was supported by the MRC grant (2008-0062286) and the Priority Research Center Program (2009-0094054) funded by the Ministry of Education, Science and Technology, South Korea. RCR is funded by the Agency for Science, Technology and Research (A⁄STAR), Singapore.
References [1] S.B. Krantz, Erythropoietin, Blood 77 (1991) 419–434. [2] M. Silva, D. Grillot, A. Benito, C. Richard, G. Nunez, J.L. Fernandez-Luna, Erythropoietin can promote erythroid progenitor survival by repressing apoptosis through Bcl-XL and Bcl-2, Blood 88 (1996) 1576–1582. [3] B.A. Witthuhn, F.W. Quelle, O. Silvennoinen, T. Yi, B. Tang, O. Miura, J.N. Ihle, JAK2 associates with the erythropoietin receptor and is tyrosine phosphorylated and activated following stimulation with erythropoietin, Cell 74 (1993) 227–236. [4] J.E. Damen, G. Krystal, Early events in erythropoietin-induced signaling, Exp. Hematol. 24 (1996) 1455–1459. [5] C. Bachran, S. Abdelazim, R.J. Fattah, S. Liu, S.H. Leppla, Recombinant expression and purification of a tumor-targeted toxin in Bacillus anthracis, Biochem. Biophys. Res. Commun. 430 (2013) 150–155. [6] A. Markham, H.M. Bryson, Epoetin alfa. A review of its pharmacodynamic and pharmacokinetic properties and therapeutic use in nonrenal applications, Drugs 49 (1995) 232–254. [7] E.J. Sharples, C. Thiemermann, M.M. Yaqoob, Novel applications of recombinant erythropoietin, Curr. Opin. Pharmacol. 6 (2006) 184–189. [8] M.S. Dordal, F.F. Wang, E. Goldwasser, The role of carbohydrate in erythropoietin action, Endocrinology 116 (1985) 2293–2299. [9] L.O. Narhi, T. Arakawa, K.H. Aoki, R. Elmore, M.F. Rohde, T. Boone, T.W. Strickland, The effect of carbohydrate on the structure and stability of erythropoietin, J. Biol. Chem. 266 (1991) 23022–23026. [10] K. Yamaguchi, K. Akai, G. Kawanishi, M. Ueda, S. Masuda, R. Sasaki, Effects of site-directed removal of N-glycosylation sites in human erythropoietin on its production and biological properties, J. Biol. Chem. 266 (1991) 20434–20439. [11] M. Higuchi, M. Oh-eda, H. Kuboniwa, K. Tomonoh, Y. Shimonaka, N. Ochi, Role of sugar chains in the expression of the biological activity of human erythropoietin, J. Biol. Chem. 267 (1992) 7703–7709. [12] J.H. Park, C. Kim, W.B. Kim, Y.K. Kim, S.Y. Lee, J.M. Yang, Efficiency of promoter and cell line in high-level expression of erythropoietin, Biotechnol. Appl. Biochem. 32 (Pt 3) (2000) 167–172. [13] N. Irani, A.J. Beccaria, R. Wagner, Expression of recombinant cytoplasmic yeast pyruvate carboxylase for the improvement of the production of human erythropoietin by recombinant BHK-21 cells, J. Biotechnol. 93 (2002) 269–282. [14] M.D. Wang, M. Yang, N. Huzel, M. Butler, Erythropoietin production from CHO cells grown by continuous culture in a fluidized-bed bioreactor, Biotechnol. Bioeng. 77 (2002) 194–203. [15] J.C. Egrie, E. Dwyer, J.K. Browne, A. Hitz, M.A. Lykos, Darbepoetin alfa has a longer circulating half-life and greater in vivo potency than recombinant human erythropoietin, Exp. Hematol. 31 (2003) 290–299.
[16] D. Zanette, A. Soffientini, C. Sottani, E. Sarubbi, Evaluation of phenylboronate agarose for industrial-scale purification of erythropoietin from mammalian cell cultures, J. Biotechnol. 101 (2003) 275–287. [17] K. Schriebl, E. Trummer, C. Lattenmayer, R. Weik, R. Kunert, D. Muller, H. Katinger, K. Vorauer-Uhl, Biochemical characterization of rhEpo-Fc fusion protein expressed in CHO cells, Protein Expr. Purif. 49 (2006) 265–275. [18] L.J. Lucast, R.T. Batey, J.A. Doudna, Large-scale purification of a stable form of recombinant tobacco etch virus protease, Biotechniques 30 (2001) 544–546. [19] J.W. Lee, J.D. Helmann, The PerR transcription factor senses H2O2 by metalcatalysed histidine oxidation, Nature 440 (2006) 363–367. [20] T. Kitamura, T. Tange, T. Terasawa, S. Chiba, T. Kuwaki, K. Miyagawa, Y.F. Piao, K. Miyazono, A. Urabe, F. Takaku, Establishment and characterization of a unique human cell line that proliferates dependently on GM-CSF, IL-3, or erythropoietin, J. Cell. Physiol. 140 (1989) 323–334. [21] A.R. Mire-Sluis, R. Thorpe, Laboratory protocols for the quantitation of cytokines by bioassay using cytokine responsive cell lines, J. Immunol. Methods 211 (1998) 199–210. [22] Y.J. Wang, Y.D. Liu, J. Chen, S.J. Hao, T. Hu, G.H. Ma, Z.G. Su, Efficient preparation and PEGylation of recombinant human non-glycosylated erythropoietin expressed as inclusion body in E. coli, Int. J. Pharm. 386 (2010) 156–164. [23] Y.J. Wang, S.J. Hao, Y.D. Liu, T. Hu, G.F. Zhang, X. Zhang, Q.S. Qi, G.H. Ma, Z.G. Su, PEGylation markedly enhances the in vivo potency of recombinant human non-glycosylated erythropoietin: a comparison with glycosylated erythropoietin, J. Control. Release 145 (2010) 306–313. [24] H. Bach, Y. Mazor, S. Shaky, A. Shoham-Lev, Y. Berdichevsky, D.L. Gutnick, I. Benhar, Escherichia coli maltose-binding protein as a molecular chaperone for recombinant intracellular cytoplasmic single-chain antibodies, J. Mol. Biol. 312 (2001) 79–93. [25] R.B. Kapust, D.S. Waugh, Escherichia coli maltose-binding protein is uncommonly effective at promoting the solubility of polypeptides to which it is fused, Protein Sci. 8 (1999) 1668–1674. [26] S.M. Singh, A.K. Panda, Solubilization and refolding of bacterial inclusion body proteins, J. Biosci. Bioeng. 99 (2005) 303–310. [27] B. Batas, C. Schiraldi, J.B. Chaudhuri, Inclusion body purification and protein refolding using microfiltration and size exclusion chromatography, J. Biotechnol. 68 (1999) 149–158. [28] J.C. Cheetham, D.M. Smith, K.H. Aoki, J.L. Stevenson, T.J. Hoeffel, R.S. Syed, J. Egrie, T.S. Harvey, NMR structure of human erythropoietin and a comparison with its receptor bound conformation, Nat. Struct. Biol. 5 (1998) 861–866. [29] D. Wen, J.P. Boissel, T.E. Tracy, R.H. Gruninger, L.S. Mulcahy, J. Czelusniak, M. Goodman, H.F. Bunn, Erythropoietin structure-function relationships: high degree of sequence homology among mammals, Blood 82 (1993) 1507–1516. [30] R.J. Darling, U. Kuchibhotla, W. Glaesner, R. Micanovic, D.R. Witcher, J.M. Beals, PGlycosylation of erythropoietin affects receptor binding kinetics: role of electrostatic interactions, Biochemistry (Mosc) 41 (2002) 14524–14531. [31] F.M. Veronese, G. Pasut, PEGylation, successful approach to drug delivery, Drug Discov. Today 10 (2005) 1451–1458. [32] L.O. Narhi, T. Arakawa, K. Aoki, J. Wen, S. Elliott, T. Boone, J. Cheetham, Asn to Lys mutations at three sites which are N-glycosylated in the mammalian protein decrease the aggregation of Escherichia coli-derived erythropoietin, Protein Eng. 14 (2001) 135–140. [33] J.P. Boissel, W.R. Lee, S.R. Presnell, F.E. Cohen, H.F. Bunn, Erythropoietin structure-function relationships. Mutant proteins that test a model of tertiary structure, J. Biol. Chem. 268 (1993) 15983–15993. [34] Y. Yokomaku, T. Sugimoto, S. Kume, S. Araki, K. Isshiki, M. Chin-Kanasaki, M. Sakaguchi, N. Nitta, M. Haneda, D. Koya, T. Uzu, A. Kashiwagi, Asialoerythropoietin prevents contrast-induced nephropathy, J. Am. Soc. Nephrol. 19 (2008) 321–328.
