Interferons (IFNs) are divided into three types: І, Π, and III, all belonging to the cytokines family. These proteins are all structurally and functionally related, demonstrating antiviral and antiproliferative effects on various cell types, and possessing a variety of potent immunomodulatory effects (1-5). Beta interferon (IFN-b), a member of type І IFN family, has been greatly utilized as an effective treatment for numbers of diseases such as multiple sclerosis and several other diseases such as cancers, and cold sores. Therefore, its large-scale production via recombinant DNA technology and obtaining a pure product is of great importance especially in pharmaceutical industries.
IFN-β is encoded by a single gene in fibroblasts. Since large-cale production of this interferon from fibroblasts is difficult, molecular cloning technology has now paved the way for massive production using varieties of host cells (6-11).
IFN-b-1b is a type of recombinant human IFN-β in which the cysteine at position 17 is replaced with serine via a T to A transition in the first base of the codon. This mutant IFN is a non-glycosylated protein with an approximate molecular weight of 18.500 kDa and 165 amino acids (12).
Heterologous expression of foreign genes in prokaryotic expression systems, especially Escherichia coli often leads to the production of the expressed proteins as insoluble inclusion bodies (IBs) (13). Inclusion bodies enclose not only recombinant protein but also host proteins such as RNA polymerase, outer membrane proteins, enzymes, and other molecules, like 16S and 23S rRNA, and plasmids. These impurities can affect the protein refolding yield during the purification process (14). Therefore, purification of IBs from all such adhering impurities is of great interest and importance. The first step in the separation of IBs, is washing them with appropriate solutions (15). To recover an active protein after the washing step, two important steps must be conducted; solubilization and refolding.
Solubilization inevitably leads to monomolecular dispersion and minimum intra- or inter-chain interactions. The choice of solublizing agents, such as urea and guanidine hydrochloride plays a key role in the efficiency of this step, and in subsequent refolding process (16, 17). Protein refolding is a difficult task, needing a trial and error process to find the right conditions. Since protein refolding competes with reactions, such as misfolding and aggregation, the composition of the refolding buffer is crucial in preventing the formation of inactive proteins. In this regard, many additives to the refolding buffer has been tested. Polyethylene glycol (18), cyclodextrin (19), L-arginine (20, 21), and proline (22) have been reported to suppress aggregation (23, 24). Among these additives, arginine has been widely used to assist refolding of various recombinant proteins (25-28). In fact, L-arginine is known as a refolding enhancer and can improve the renaturation yield (25, 29-33).
This paper describes the optimization of the initial steps to recover rh-IFN-b-1b from Escherichia coli. The washing and solubilization steps were enhanced using effective solutions. Renaturation was investigated using refolding buffer in the presence and absence of L-arginine.
The aim of this research was process optimization, focusing on the refolding of a recombinant interferon-beta.
3. Materials and Methods
3.1. Culture and Growth Conditions
Escherichia coli BL21 (K12) was provided by the Fermentation Group at the National Institute of Genetic Engineering and Biotechnology (NIGEB), Iran. Expression vector PYG87 of IFN-b has been described thoroughly by Ghane et al. (34).
To cultivate higher cell densities, a 3-1 bioreactor (New Brunswick, Biofilo 3000, USA) was used, and the fermentation process was carried out according to the method by Ghane et al. (34).
3.2. Harvesting and Isolation of Inclusion Bodies
The culture medium was harvested by centrifugation at 1789g, for 30 min at 4ºC (Sigma, 3K30, USA). The collected cells were suspended in 1500 ml of lysis buffer (Table 1). The cells were disrupted by a homogenizer (Niro Sovani, Italy) at 600 bar for two cycles in an ice bath. The lysate was centrifuged at 18894g for 30 min and 4ºC, and the resulting inclusion bodies (IBs) were obtained in pellet form.
3.3. Washing of Recombinant Human Interferon βIBs
The IB pellet was resuspended in different detergent buffers such as 2% and 0.5% (v/v) Triton-X100, and sodium deoxycholate in 50 mM Tris–base pH 8.0 containing 100 mM NaCl and 1 mM EDTA. The resulting suspensions were washed by stirring with a magnetic agitator for 1 h. The washing process was conducted at room temperature using 1 g IBs/15 ml of detergent buffer. After centrifugation at 18407g for 1 hour and 4ºC, an impurity profile was established for the inclusion bodies by applying SDS-PAGE electrophoresis and subsequent densitometric analysis using Alpha Ease FC software (Genetic Technologies, Inc., USA). The recovery yield after each washing step was defined as below:
Recovery yield (%) =
Two denaturing agents, Guanidine HCl (4-8 M) and urea (6 and 8 M) were investigated. The most effective reducing agents, i.e., b-mercaptoethanol and dithiothreitol were also examined. Washed IBs (1 g) were dissolved in 100 ml of various denaturing buffers (Table 1). Solubilization was conducted at room temperature with continuous stirring for 1 h, until complete denaturation and reduction of IBs were achieved. The insoluble particles were removed by centrifugation at 18407g for 30 min and 4ºC. The resultant supernatant, containing denatured rhIFN-b was diluted in. The total protein concentration for all supernatants was measured by the Bradford assay (35). Solubilization yield was defined as mentioned below:
Solubilization yield (%) =
3.5. Refolding by Dilution
Two sets of experiments were carried out. One experiment involved diluting the denatured protein 20-fold in refolding buffers, A1-A4 (Table 1). However, in the second set, 40-fold dilution of the denatured protein was carried out with the same buffers. Accordingly, refolding was initiated by adding the denatured protein slowly to a glass container filled with the refolding buffers, A1-A4, in a drop-wise fashion. For the first few minutes, a fast and efficient mixing was implemented, after which the solution was incubated at 4°C overnight without further agitation. Finally the solution was centrifuged at 18407 g for 20 min at 4°C, to remove precipitated materials and the resulting supernatant was used in the subsequent analyses. Turbidity was measured at 340 nm using a UV visible spectrophotometer (BECKMAN DU 530, USA), to determine aggregation during refolding.
Refolding by dilution in the presence of L-arginine was performed in the same way as mentioned above, except that the buffer in this case contained 0.5 M L-arginine.
3.6. Purification of Recombinant Human Interferon b by Affinity Chromatography
Purification was performed using a 1×10 cm column packed with 7 ml of the Blue-Sepharose Fast Flow gel (Pharmacia, USA). Three buffers were used to carry out the purification process. Prior to sample application, the column was equilibrated with buffer A (Table 1). The refolded protein was loaded on the column, five bed volumes of Buffer B (Table 1) and 10 bed volumes of buffer C (Table 1) were used consecutively. Buffer D (Table 1) was applied for elution of target protein from the column. The column temperature and flow rate were maintained at 20°C and 1 ml.min-1, respectively, throughout the process. A chromatogram was generated as a result of measuring the absorbance of the collected fractions at 280 nm.
3.7.Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE)
In order to study the protein pattern, electrophoresis was carried out under denaturing conditions (35). Protein bands were developed by the Coomassie blue staining method. Two standards were applied during the reducing SDS-PAGE procedure: a primary standard of rhIFN-b (pharmaceutical interferon b from Shering, Germany) and a molecular weight marker composed of several denatured proteins (MW between 14.4 and 94 kDa, Roche, Germany). Loading buffer (0.15 M Tris-base, 15% v/v glycerol, 0.05% v/v bromophenol blue, 10% v/v b-mercaptoethanol) was added to samples and standards, prior to boiling them for 3 min. They were loaded onto the gel and ran at a constant voltage of 150 v for 4 h (36).
After staining and destaining, the gels were photographed in a scanner. This purity was defined as the ratio of the density of the rhIFN-b band to the total density of all bands in the gel.
3.9. Biological Activity of Refolded Protein
To evaluate the biological activity of the interferon, the most common assay based on the antiviral property of this cytokine was used (37). Hela cells and the vesicular stomatitis virus (VSV) as the lytic virus were used in assay. Hela cells were grown in a 96-well microplate, at 37°C in a CO2 incubator. After Hela cells were reached confluency, refolded interferon and the purchased standard were added to the cells, separately. Two rows of wells (sixteen) were considered for the cell and virus control samples. Cells (with the exception of the control cells) were inoculated with VSV at 102 CCID50.ml-1 (cell culture infective dose) following 24 h. The cells were monitored for the cytopathic effect induced by viruses, using an inverted microscope (Axiovert 25, zeiss, Germany).
3.10. Western Blotting
Proteins in the samples were electrophoretically separated by SDS-PAGE and transferred to a PVDF membrane. Western blot was performed as described previously with rabbit anti human interferon-β (prepared by NIGEB, Iran) and horseradish peroxidase-conjugated sheep anti-rabbit immunoglobulin (Serotec, UK) (38). The developed blots were photographed by the Eagle Eye system (cell phone C6_01, Nokia).
4.1.Washing of rhIFN-b inclusion bodies
The yield resulting from the washing process was detected by SDS-PAGE and the dry weight of the pellet remaining after each wash was measured.
The three washing buffers did not have any significant effects on the removal of impurities. However, the best protein recovery was achieved with detergent buffer A (76.3%), which was higher than buffer B (62%) and C (21%) (Table 2). The presence of sodium deoxycholate and an increased concentration of Triton X-100 in buffers B and C, made the solution more viscose and improper for solubilization of IBs. After washing with buffer A, the IBs were dissolved in denaturing buffer, followed by centrifugation, yielding a supernatant containing the denatured rhIFN-b. To find the optimal combination of denaturant/reducing agent in the denaturing buffer, a set of experiments was carried out. Yield of recovery and purity, measured via Bradford assay and SDS-PAGE, differed significantly when urea was replaced with Guanidine HCl. Solubilization yields in guanidine hydrochloride were higher than urea (Figure 1).The use of b-mercaptoethanolas a reducing agent instead of DTT, further improved the recovery of protein during the solubilization step. The highest recovery value (9.1%) was obtained when the highest concentration of Guanidine HCl (8 M) was used (Figure 1). A lower concentration of Guanidine HCl significantly decreased the recovery of the protein (data not shown). Purity of solubilized rhIFN-b (up to 48%) was achieved at 8 M Guanidine HCl+ b-mercaptoethanol (b-ME) (Figure 1C).
Multiple chromatography steps were followed to remove remaining contaminations within both denatured and solubilized rhIFN-b proteins (0.820 mg.ml-1).
4.2. Investigation of Factors Affecting on Refolding
Figure 2 shows the effect of dilution in the final concentrations of protein and recovery. In the case of the 20-fold dilution, protein concentration reached 40 mg.ml-1, and precipitation happened in the first few minutes after the addition of the denatured protein to the refolding buffers, A1-A4. In fact, maximum turbidity (0.15) occurred in the presence of buffer A1 (without any urea; Figure 2A). The presence of urea in the other refolding buffers, A2-A4, significantly decreased the negative effects of high protein concentration; a minimum turbidity of 0.05 was obtained in the presence of the buffer A4, which contained 2 M urea (Figure 2A). SDS-PAGE results of the supernatant sample showed all the soluble proteins had aggregated during refolding and there was no band of interferon on the gel (Figure 3A, lanes 1-4). Moreover, no remarkable suppressive impact on aggregate formation was observed after adding different concentrations of urea to the refolding buffer. Meanwhile, the amount of precipitation decreased to some extent in the presence of 2 M urea (buffer A4; Figure 3A, lane 10).
Recent reports have demonstrated that arginine can suppress formation of insoluble aggregates during the refolding of rhIFN-g. In this study, the effects of arginine on the aggregation process during the refolding of rhIFN-β were also investigated. Accordingly, in order to increase the refolding yield and minimize aggregation during the refolding process, L-arginine was applied to the refolding buffers, B1-B4 (Table 1). The absorbance of protein solution was significantly reduced and reached zero at 340 nm, when buffer B4 was used (Figure 2B). Refolding yield was remarkably enhanced following arginine application. Additionally, sharp bands associated with soluble interferon were observed in the supernatant samples (Figure 3C, lanes 7-10), along with a noticeable decrease in precipitate (Figure 3C, lanes 1-4).
4.3.Purification of interferon
Purification of rhIFN-b was carried out by a simple step of Blue Sepharose CL-6B chromatography, achieving a protein purity of 93.5%. Purified sample was separated on SDS-PAGE (data not shown) and transferred onto a membrane and analyzed via specific antibody (Figure 4). The blot was indicative of the presence of purified rhIFN-b.
4.4.Biological Activity of rhIFN-b
The biological assay of the interferon on HeLa cells showed that rhIFN could inactivate the replication cycle of the virus. However the inactivation effect was smaller once compared with the standard rhIFN-b. Control cells showed no change in size and shape (Figure 5A), while HeLa cells treated with virus (VSV) demonstrated cytopathic effect (Figure 5B) on the contrary, no cytopathic effect was observed for the inoculated cells with VSV after treatment with standard IFN (Figure 5C). HeLa cells inoculated with VSV after exposure to rhIFN (40 ng.ml-1) had no detectable cytopathic effect (Figure 5D).
Eukaryotic proteins expressed in bacteria in large quantities, are insoluble products. One of the most important steps in purification is separation of insoluble bodies. The most difficult contaminants to remove from the crude inclusion body protein were membrane-associated proteins that were released upon cell disruption. In order to remove such contaminants, washing step was performed. The use of the different buffers demonstrated that the presence of sodium deoxycholate and Triton X-100 is essential to achieve maximum purity (28). Already has used sodium deoxycholate and Triton X-100 for removing membrane proteases (39). Dasari also showed that the increase in the concentration of Triton X-100 causes IBs to precipitate via increasing buffer viscosity (40). In this research, the best efficiency of IBs recovery was found at a concentration 0.5 M of Triton X-100 (without the use of Sodium DeoxyCholate in the second phase of the washing stage).
Optimum buffering conditions were defined for protein resolvation similar to Valent et al. (2005) (38). Buffer solution contained 8 M Guanidine HCl and 100 mM b-2ME. The dissolving phase is usually performed in the neutral pH (pH 7-8) and in a buffering system that is compatible with the protein of interest.
In this research pH 10.5 showed most effective, once pH ranges of 6.5-10.5 were applied to find out the best buffering system to bring rhIFN-b into solution. For the refolding of dissolved IBs, it is necessary to remove reducing agents such as Urea and Guanidine HCl. Several variables have an important role in this process, including components of refolding buffer, final concentration of desired protein, and refolding procedure (41). Normally low concentrations (0.1-2 M) of chaotropic agents such as urea next to some other additives like L-Arg are being used in refolding buffer (41). It has proven that L-Arg is capable to inhibit of precipitation during the refolding, although the echanism has remained elusive. Dialysis or dilution method is typically used to remove denaturants.
Here, refolding buffer contained 2 M Urea +0.6 M L-Arg.
It is obvious that a high final protein concentration would inevitably produce aggregates. Increasing the protein concentration, leaves an insufficient volume of buffer, thus leading to increased local protein concentration, whereby some protein molecules come into contact with one another in the process of refolding. The denatured rhIFN-b has many hydrophobic amino acid residues, which are liable to aggregation by hydrophobic interactions.
Under refolding conditions, unfolded proteins or intermediate components are extremely prone to aggregation.
In this study, with regard to the 40-fold dilution of protein solution, when the final protein concentration decreased, aggregation also declined. So the minimum value of OD340 (0.032) was obtained when buffer A4 was used (Figure 3A). SDS-PAGE data was illustrative of lower protein concentration favors protein refolding (Figure 3B, lanes 1-4). The effect of the urea concentration on rhIFN-β refolding is shown in Figure 3B. The weak interferon band associated with the pellet sample (Figure 3B, lane 10) indicates that 2 M urea causes protein refolding and lowers aggregation. Urea concentrations lower than 2 M would not restrain the aggregation of proteins, efficiently (Figure 3B, lanes 7-9).
This study determined the capability of arginine to suppress the formation of insoluble precipitants throughout the rhIFN-b refolding process. The ability of arginine has also been proved for other proteins (21). Here, the use of refolding buffer B2 containing a low urea concentration can be considered as an optimized protocol for refolding of IFN-b.
The authors would like to acknowledge Mr. Habib Ghomi for helping with the fermentation.
Results presented in this paper were extracted from the data of project 298 supported by National Institute of Genetic Engineering and Biotechnology (NIGEB) of Islamic Republic of Iran.