The hematopoietic system is a complex structure with multipotent stem cells which produce mature hematopoietic cells that circulate in peripheral blood. Hematopoiesis is a highly regulated process and all of its stages operate under the control of specific factors called cytokines. Majority of the cytokines are glycoproteins that regulate survival, proliferation and differentiation of hematopoietic progenitor cells and also affect the function of mature blood cells. Colony stimulating factors (CSFs) are a family of cytokines which consist of five types of glycoproteins consisting of M (macrophage)-CSF, G (granulocyte)-CSF, E (eosinophil)-CSF, GM (granulocyte macrophage)-CSF and multi-CSF or Il-3 (Creighton, 1999).
Granulocyte colony stimulating factor (G-CSF) is a major member of this family that is produced by bone marrow stromal cells, endothelial cells, macrophages and fibroblasts. This factor induces proliferation and differentiation of neutrophil progenitor cells as well as activation of mature granulocytes for more efficient immune responses. hG-CSF is a glycoprotein consisting of 174 amino acid residues (hG-CSFb, 18.8 KD) or 177 amino acid residues (hG-CSFa, 19.6 KD) with one O-linked glycosyl group on Threonine (Thr) at position 133. These two different G-CSFs are encoded by two different mRNAs produced by alternative splicing from a single precursor RNA. Even though these different forms have similar biological activities, hG-CSFb is 20 times more active than hG-CSFa (Nagata 1989; Nagata et al., 1986 a, b).
The molecular structure of this cytokine is an up-up-down-down, antiparallel, left-handed four a-helical bundle without any b-sheets (Fig. 1). There are two disulfide bonds residing between Cysteine (Cys) 36/42 and Cys 64/74 of this molecule (Werner et al., 1994; Hill et al., 1993).
The two forms of hG-CSF cDNA, hG-CSFa and hG-CSFb, have already been cloned by Nagata et al. (1986 a) and Souza et al. (1986), respectively. Currently, hG-CSFb is synthesized by biotechnological methods and used for treatment of neutropenia arising from chemotherapy and radiotherapy and in patients who have undergone bone marrow transplantation (Fernández-Varón and Villamayor, 2007; Klingemann, 1989). So far, several researchers have attempted to improve G-CSF biological activity, stability and shelf life by using different mutagenesis techniques (Luo et al., 2002; Bishop et al., 2001; Lu et al., 1999; Ishikawa et al., 1993 and 1992; Devlin et al., 1988).
The aim of this study was to design and produce an engineered recombinant G-CSF, using site-directed mutagenesis, in order to improve its biological characteristics in comparison to those of native hG-CSF.
PCR based site-directed mutagenesis by the overlap extension method (Sambrook and Russel, 2001) was performed in three stages. Sequence, polarity and position of the primer pairs as well as restriction enzyme cutting sites are presented in Table 1. The primers were designed according to the codon usage frequency of E. coli by using the GeneRunner software version 3.05 (Hasting software, Inc.) Nde I and Xho I specific sites and additional nucleotides on either sides (on 3´ end of F and 5´ end of R primer) were created in the F and R designated primers, respectively. The His hexamer (His6-tag) sequence as well as the ATTA stop sequence was designed at the R primer 5´ overhang, all conforming to the cloning and expression strategies of this study.
PCR reactions were carried out in a total reaction mixture volume of 25 ml containing Pfu buffer 1X (20 mM Tris-HCl (pH 8.8), 10 mM (NH4)2SO4, 10 mM KCl, 0.1% (v/v) Triton X-100, 0.1 mg/ml of BSA, 2 mM MgSO4) 0.2 mM of each dNTP (Cinnagen, Iran), 10 pmol of each forward and reverse primers, 50-150 ng of template DNA and 1 U of Pfu DNA polymerase (Fermentas, Lithuania).
In the first and second PCR stages, a clone containing native hG-CSF cDNA (Saeedinia et al., 2003) was used as template. The first round was performed using F and R1 primers as shown schematically in Figure 2 under the following conditions: initial denaturation at 94ºC for 3 min followed by 35 cycles of denaturation at 94ºC for 45 sec, primer annealing at 60ºC for 45 sec and extension at 72ºC for 1 min followed by a final extension step at 72ºC for 8 min. During the second PCR round, F1 and R primers were used in the abovementioned PCR mixture and thermocycle profile. In the third PCR stage, the amplicons obtained through the first and second rounds were diluted by 1/100, and used together as template. F and R primers were used to produce and amplify the full length mutant G-CSF (muG-CSF) cDNA using the following thermal cycle profile: initial denaturation at 94ºC for 3 min followed by 35 cycles of denaturation at 94ºC for 45 sec, primer annealing at 60ºC for 45 sec and extension at 72ºC for 2.5 min followed by a final extension step at 72ºC for 10 min.
The PCR product, mutant G-CSF (muG-CSF) amplicon, was gel purified and cloned into the EcoR V linearized pBluescript SK(-) plasmid. Restriction analysis of this intermediate plasmid was carried out to confirm cloning fidelity. In order to confirm the desired mutations, the recombinant vector was sequenced in a bidirectional manner, using T3 and T7 universal primers.
The muG-CSF fragment was exposed to Nde I/Xho I double digestion and subcloned into the pET-21a(+) vector (Novagen, USA) that was linearized with the same restriction endonucleases. All DNA manipulations including restriction digestion, T4 ligation and agarose gel electrophoresis techniques were carried out as described by Sambrook and Russel (2001).
The recombinant construct was used to transform E. coli BL21 (DE3) competent cells. After selection and verification of recombinant colonies, cells harboring pET21-muG-CSF were cultivated in LB broth medium (Scharlau, Spain). The expression of muG-CSF was induced by addition of Isopropyl b-D-1-thiogalactopyranoside (IPTG) at a final concentration of 1 mM when the cells had reached an optical density of 0.7 at 600 nm (OD600). The cells were then incubated for further 5 h. Bacterial cells were harvested at 0, 0.5, 1, 2, 3, 4 and 5 h after induction. To get the total cell protein extract, cells from 1 ml culture samples were separated by centrifugation at 6000 ×g for 10 min at 4°C, and then lysed directly by adding 100 ml of sample buffer (80 mM Tris-HCl (pH 6.8), 2% (w/v) SDS, 5% (v/v) b-mercaptoethanol, 10% (v/v) glycerol, 0.001% (w/v) bromophenol blue and boiled for 10 min. 25 ml of the prepared protein samples were electrophoresed on a 12.5% (w/v) SDS-polyacrylamide gel. Commercial recombinant hG-CSF (Filgrastim-neupogen, Roche, Germany) was used as a positive control. Electrophoresis was carried out at 20 mA, for 2.5 h.
Protein bands on the SDS-polyacrylamide gel were transferred to a polyvinylidene difluorid (PVDF) filter. The filter was blocked with 3% (w/v) bovine serum albumin (BSA) (Merck, Germany) and incubated with 0.5 µg/ml of murine monoclonal anti-human G-CSF (Sigma, USA) for 1 h. After washing, it was incubated with 1/2000 dilution of anti-mouse IgG conjugated with Horse Radish Peroxidase (HRP) (Sigma, USA) as secondary antibody for 1h. 0.5 mg/ml 3,3-Diaminobenzidine reagent and 0.1% (v/v) H2O2 in tris buffer saline (TBS) were used as the HRP substrate for color development.
During the first round of PCR, a 95 bp DNA fragment encoding the first 30 N-terminus amino acids of the mutant G-CSF was amplified. In the second round, a 515 bp DNA fragment was amplified which covers nucleotides 44 to 560 of the full length mutant G-CSF cDNA, encoding amino acids 14 to 180 of muG-CSF (including His6-tag). In the third reaction, these amplicons joined each other to produce a 560 bp full length muG-CSF cDNA (Fig. 3a). This final fragment was flanked by the Nde I and Xho I restriction sites at its 5´ and 3´ termini, respectively. The fragment was then cloned into the pET-21a(+) via the same restriction sites in the correct orientation and under the control of the T7 promoter. Sequencing results confirmed the desired mutations in muG-CSF cDNA and cloning fidelity. Restriction analysis was also carried out to demonstrate the latter (Fig. 3b).
As demonstrated in Figure 4, the expressed muG-CSF with a size of approximately 19 kD was detected in total cell protein samples obtained from induced cells harboring pET21-muG-CSF. The intensity of the target band gradually grown increased over times, during which most of the recombinant muG-CSF was observed 5 h after induction.
Western-blot analysis using monoclonal antibody against hG-CSF proved the specificity of the detected band in SDS-PAGE. Figure 5 illustrates the result of Western-blot analysis carried out with monoclonal antibody.
Mutational analysis studies of hG-CSF have revealed that internal and C-terminal regions of this protein play an essential role in interaction with its cell surface receptor and manipulation of these evolutionally-conserved sequences mostly results in significant loss of biological activity (Young et al., 1997). Conversely, mutagenesis studies on N-terminal regions have shown a delicate relationship between the non-essential N-terminal structure and the G-CSF biological activity via a mechanism which dose not directly change the backbone of the molecular structure. In this regard, Kuga et al. (1989) and Okabe at al. (1990) have shown that substitution of Thr 1, leucine (Leu) 3, glysine (Gly) 4, proline (Pro) 5 and Cys 17 with alanine (Ala), Thr, tyrosine (Tyr), arginine (Arg) and serine (Ser) will result in 2-4 times more biological activity than that of the wild type protein. Cys 17 is not involved in structural disulfide bonds, and its substitution with Ser or Ala does not have any impact on its biological activity. Therefore many researchers have considered these substitutions in amino acid sequence to avoid the formation of unwanted disulfide bonds through protein folding process. Moreover this alteration results in a more thermodynamically stable G-CSF structure. (Yamasaki et al., 1998; Reidhaar-Olson et al., 1996; Ishikawa et al., 1993; Lu et al., 1992; Wingfield et al., 1988).
Other mutagenesis studies have proved that substitution of each Gly residue at positions 26, 28, 149 and 150 singly or in combination, with Ala will result in proteins with dramatically enhanced stability while retaining wild type levels of biological activity (Bishop et al., 2001). Cys 17 and Gly 28 substitution with Ala, has also led to a 5-fold improvement in G-CSF storage stability and shelf life (Luo et al., 2002).
Imaginary structural analysis using the Swiss-Pdb Viewer (v3.7) software of the crystal structure of hG-CSF (protein data bank (PDB) record 1GNC) (Zink et al., 1994), demonstrated that all the abovementioned substitutions on the same oligopeptide do not affect the 3-dimensional molecular structure, which may still be left functional.
Here we combined these favorable mutations in the same coding sequence of G-CSF by performing a two step site-directed mutagenesis (Table 1). The main objective was a combination of the results of abovementioned studies to create a new mutant G-CSF, developed both physically and biologically, with enhanced efficacy as a recombinant drug. In summary, the pET system, a well known prokaryotic expression vector, was used to express the muG-CSF recombinant protein. The cloning strategies employed in this study were: i) Using the ATG triplets within the Nde I recognition site to encode the N-terminal Met base-paired to the AUG start codon in the transcript. This restriction site was engineered into flanking reverse primer (Sambrook and Russel, 2001). ii) To avoid adding two unwanted amino acid residues at the C-terminus of muG-CSF, the pET carrying His6-tag sequence was not used, instead, the six His triplet codons followed by a stop sequence was engineered into the flanking reverse primer (Sambrook and Russel, 2001). Such a fusion facilitates detection of the expressed muG-CSF and will be very useful in the purification procedures of later experiments. It has also been shown that conjugating the unrelated small oligopeptides to the C-terminus of hG-CSF does not have any effect upon its biological activity (Oshima et al., 2000).
With respect to SDS-PAGE and western-blot analysis, the bands corresponding to muG-CSF were observed at a level slightly higher than that of the commercial G-CSF (Figs. 4 and 5) because the addition of the His hexamer to the polypeptide increases molecular weight by approximately 0.7 KD.
In this research, a new formulation of mutant G-CSF has been designed and expressed in prokaryotic system that would be an improved version of native hG-CSF and an appropriate candidate for pharmaceutical purposes after verifying its function, biological activity and safety.
The authors would like to thank Mohamad R. Masoumian, Hesam Barjesteh, Hossein A. Sami, Kaveh Baghai and mehdi zeinoddini for their excellent technical assistance.