Appl Biochem Biotechnol (2014) 173:421–432 DOI 10.1007/s12010-014-0849-7
Expression and Purification of Bioactive High-Purity Recombinant Mouse SPP1 in Escherichia coli Yunsheng Yuan & Xiyuan Zhang & Shunyan Weng & Wen Guan & Di Xiang & Jin Gao & Jingjing Li & Wei Han & Yan Yu
Received: 19 October 2013 / Accepted: 5 March 2014 / Published online: 25 March 2014 # Springer Science+Business Media New York 2014
Abstract Secreted phosphoprotein 1 (SPP1) is a phosphorylated acidic glycoprotein. It is broadly expressed in a variety of tissues, and it is involved in a number of physiological and pathological events, including cancer metastasis, tissues remodeling, pro-inflammation regulation, and cell survival. SPP1 has shown its function of protecting tissues and organs against injury and wound, giving itself potentials to become a therapy target or giving its antibodies of other counter-acting reagents potentials to become drug candidates. Non-tagged (native) recombinant SPP1 would be valuable in therapeutic and pharmaceutical researches. In our study, mouse Spp1 DNA fragment without signal peptide was built in pET28a(+) vector and transformed into Escherichia coli BL21 (DE3). The recombinant mouse SPP1 (rmSPP1) was then expressed in bacteria upon induction by isopropyl β-D-thiogalactopyranoside (IPTG). The abundance of rmSPP1 was increased using isoelectric precipitation and ammonium sulfate fractionation methods, and anion and cation exchange chromatography was employed to
Electronic supplementary material The online version of this article (doi:10.1007/s12010-014-0849-7) contains supplementary material, which is available to authorized users.
Y. Yuan : X. Zhang : S. Weng : W. Guan : Y. Yu Shanghai Key Laboratory of Veterinary Biotechnology, College of Agriculture and Biology, Shanghai Jiao Tong University, 800 Dongchuan Rd, 200240 Shanghai, People’s Republic of China Y. Yuan e-mail: [email protected] X. Zhang e-mail: [email protected] S. Weng e-mail: [email protected] W. Guan e-mail: [email protected] Y. Yuan : Y. Yu (*) Key Laboratory of Urban Agriculture (South) Ministry of Agriculture, College of Agriculture and Biology, Shanghai Jiao Tong University, 800 Dongchuan Rd, 200240 Shanghai, People’s Republic of China e-mail: [email protected]
422
Appl Biochem Biotechnol (2014) 173:421–432
further purify rmSPP1. Finally, we got rmSPP1 product with 12.8 % productivity, 97 % purity, satisfactory bioactivity, and low endotoxin content. Keywords Secreted phosphoprotein 1 . Chromatography . Protein purification . Chemotaxis
Introduction Secreted phosphoprotein 1 (SPP1), also named osteopontin (OPN), is a phosphorylated acidic glycoprotein existing in mammals. It is involved in many physiological and pathological processes, including bone metabolism, inflammation progress, tumor metastasis, injury repair, and so on [1–3]. An amount of studies have demonstrated that SPP1 is not only a biomarker in many diseases, but it is also involved in relieving tissue injury caused by stress or toxin. For example, SPP1 could attenuate cardiac ischemia–reperfusion injury [4], hyperoxia-induced lung injury [5], and liver injury [6, 7]. SPP1 could also protect islet cell and improve survival of patients with pancreatic adenocarcinoma [8, 9]. In terms of structure, SPP1 has several special characteristics, including a RGD sequence, a Ca2+ binding site, and two heparin binding domains [10]. It could be cleaved by several proteases, such as thrombin, enterokinase, or matrix metalloproteinases (MMPs) [11, 12]. A number of investigators have addressed that the activity of SPP1 in the regulation of cell adhesion, protein–protein interaction, or bone resorption is independent from its degree of posttranslational modification [11]. To investigate the structure and biological functions of SPP1 in vitro, full-length SPP1 or its fragments have been isolated from milk [13, 14] or urine [15]. SPP1 could also be secreted by cultured rat smooth muscle cells [16]. Acutely, GST or six times His-tagged SPP1, expressed in Escherichia coli and purified with affinity chromatography, has been widely used to investigate the structure and biological functions of SPP1 [17, 18]. However, these tagged proteins could not be used to investigate pharmaceutical functions of SPP1 in animal disease models. So, it is important to develop a new method of purifying native recombinant SPP1 protein, for studies on the structure, stability, or therapeutic use of SPP1. Recombinant SPP1 would be useful to further study on the functions of SPP1 in animal diseases model or preclinical researches.
X. Zhang Department of Tumor Biology, Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, E407 New Research Building, 3970 Reservoir Road NW, Washington, DC 20057, USA D. Xiang Children’s Hospital Oakland Research Institute, 5700 Martin Luther King Jr Way, Oakland, CA 94609, USA e-mail: [email protected] J. Gao : J. Li : W. Han Laboratory of Regeneromics, School of Pharmacy, Shanghai Jiaotong University, 800 Dongchuan Rd, 200240 Shanghai, People’s Republic of China J. Gao e-mail: [email protected] J. Li e-mail: [email protected] W. Han e-mail: [email protected]
Appl Biochem Biotechnol (2014) 173:421–432
423
In this study, we created a novel method to produce and purify native recombinant mouse SPP1 (rmSPP1) in E. coli. Spp1 DNA sequence was amplified from mouse liver cDNA and inserted into pET28a(+) vector (Merck, German). rmSPP1 was expressed in BL21 E. coli. Biological active rmSPP1 was obtained at 97 % purity using a three-step procedure.
Materials and Methods Construction of the rmSPP1 Expression Vector Total RNA was isolated from C57BL/C mouse liver using TRIzol reagent (Invitrogen, USA), and cDNA was reversely transcribed with M-MuLV Reverse Transcriptase (Fermentas). Two PCR reactions were employed to clone mouse Spp1 gene. DNA fragment containing the open reading frame (ORF) of mouse Spp1 (167–1,184 bp, NCBI mRNA sequence no. NM_001204201) was amplified in the first PCR reaction. One pair of primers was used in this reaction as follows: the forward primer 5′-ACCTGACAAGACATCAACTGTGCC-3′ and the reverse primer 5′-TTCCTGCTTAACCCTCACTAACAC-3′ (Invitrogen, Shanghai). PCR was performed in a 50 μl reaction system containing 1× PCR buffer, MgCl2 (1.5 mM), dNTPs (0.2 mM), 5 U of Taq plus DNA polymerase (BBI, Canada), 2 μl of cDNA as template, and 120 nM each of the forward and reverse primers. After denaturing the cDNA for 3 min at 95 °C, amplification parameters were set up as follows: denaturation, 30 s at 95 °C; annealing, 40 s at 68 °C; and extension, 2 min at 72 °C. The DNA fragment was inserted into a T vector (Takara, Japan). Having its DNA sequence confirmed by sequencing, the construct was used as the template of the next PCR reaction to create rmSPP1 expression construct. The second PCR reaction was used to amplify mature mouse SPP1 DNA (226 to 1,059 bp), and a pair of primers was used in the reaction as follows: the forward primer 5′-AGTCCCATGGCTCTCC CGGTGAAAGTGACTG-3′ and the reverse primer 5′ -AGTCCTCGAGTTAGTTGACCTC AGA AGATG -3′ (underlined letters indicate NcoI and XhoI cleavage sites, respectively). The protein corresponding to this ORF would be a native mouse SPP1 which contains an Ala codon (GTC) at the 5′ end of the DNA fragment. The PCR parameters were the same as the first PCR reaction, except for the fact that annealing temperature was changed to 55 °C. The amplified DNA was digested with NcoI and XhoI and ligated with pET28a(+) vector which has been digested with NcoI and XhoI as well. The recombinant plasmid containing the mouse SPP1 gene was named pET28a-Spp1 (Fig. 1), which was verified by sequencing assay. Expression of rmSPP1 E. coli BL21 (DE3) cells were transformed with the plasmid pET28a-Spp1 and plated onto Luria–Bertani (LB) agar plate with kanamycin (50 μg/ml). A single colony was selected and inoculated in a 50-ml LB medium supplemented with kanamycin (50 μg/ml), with overnight shaking at 250 rpm, 37 °C. Thirty milliliters of overnight culture was added to the 3-l fresh LB medium and continued to be incubated until OD600 reached 1.0. Then, isopropyl β-Dthiogalactopyranoside (IPTG, 1 mM) was added into the medium to induce rmSPP1 expression in bacteria. Cells were harvested by centrifugation for 10 min (10,000g, 4 °C); after, bacteria were incubated for 4 h after IPTG was added (200 rpm, 37 °C). Cells were washed with 1× phosphate-buffered saline (PBS) (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4; pH 7.4) and resuspended with 30 ml of lysis buffer (1× PBS, pH 7.4; 1 mM EDTA; 0.1 mM phenylmethylsulfonyl fluoride (PMSF)). After 30 cycles of ultrasonication
424
Appl Biochem Biotechnol (2014) 173:421–432
Fig. 1 Map of Spp1 expression plasmid in E. coli. The underlined sequences are primers for PCR; the italicized letters represent restricted sits; the sequence between Start and Stop is translated to the recombinant mouse Spp1
(Sonics, USA) on ice (3 s of sonication and 7 s of resting for each cycle), the supernatant was collected by centrifugation (18,000g, 4 °C, 30 min). Purification of rmSPP1 We used isoelectric precipitation and ammonium sulfate ((NH4)2SO4) fractionation methods to exclude as much of bacterial protein as possible before going through ion exchange chromatography. The supernatant then went through a series of pH adjustment to pH values of 6.0, 5.0, and 4.3 with 1 M HCl solution. After each step of pH adjustment, the sample was kept under 4 °C for 30 min, for the thorough sedimentation of bacterial proteins whose isoelectric points (PIs) are close to these pH points. The resulting insoluble bacterial proteins were removed by centrifugation for 15 min (18,000g, 4 °C). Crude rmSPP1 was precipitated with 20 % saturated (NH4)2SO4 at pH 4.3 and solubilized in 3 ml of 1× PBS. The solution containing crude rmSPP1 then went through a tenfold dilution by 20 mM Tris–HCl (1 mM EDTA, 0.1 mM PMSF, pH 8.0) and was loaded onto a Q-Sepharose column (GE Healthcare, USA) attached to an AKTA Explorer 10/100 system (Amersham Pharmacia Biotech, Sweden) at the rate of 1 ml/min. After loading was finished, the column was washed with 4 column
Appl Biochem Biotechnol (2014) 173:421–432
425
volumes of buffer 1 (20 mM Tris–HCl, 1 mM EDTA, 25 mM NaCl, pH 8.0). rmSPP1 was eluted using a mixed solution of buffer 1 coming from sample pump A and buffer 2 (20 mM Tris–HCl, 1 mM EDTA, 25 mM NaCl, pH 8.0) from sample pump B. The mixed solution goes through the Q-Sepharose column with the speed of 1 ml/min, while the percentage of solution coming from sample pump B is elevated from 10 to 70 % with a fixed rate of 1 % elevation per min, thus creating a consecutive NaCl concentration gradient. Proteins would be eluted at different points in such gradient, according to the difference of their binding ability to the QSepharose matrix under pH 8.0. Fractions were collected according to peaks of UV absorption and conductivity curve. Fractions containing rmSPP1 were pooled and dialyzed overnight in 100-fold volume of buffer 3 (20 mM Na citrate, 20 mM NaCl, 1 mM EDTA, pH 3.2). Dialyzed protein was then loaded onto the S-Sepharose column (GE Healthcare, USA) attached to an AKTA Explorer 10/100 system at a rate of 1 ml/min, and the column was washed with 4 column volumes of buffer 3 at the same speed. The column-bound protein was eluted with a similar mixed solution containing a consecutive NaCl concentration gradient, while the percentage of solution coming from sample pump B is elevated from 12 to 72 % this time. Fractions were collected according to UV absorption and conductivity curves. The protein concentration in fractions was detected with Bradford assay. Sodium Dodecyl Sulfate–Polyacrylamide Gel Electrophoresis, Western Blotting, High-Performance Liquid Chromatography–Size Exclusion Chromatography Analysis, and Endotoxin Assay To estimate the purity of rmSpp1, 5 μg of protein was loaded onto 15 % polyacrylamide gel ran with the PowerPac Basic (Bio-Rad, USA) and stained with Coomassie bright blue R-250 (Beyotime, China). To estimate the purity and productivity of rmSPP1, the bands on gels were scanned with Tannon Gel Image System (Tannon 2500, China), and intensity of each band was recorded and analyzed with GIS software 1.0 (Tannon, China). Western blotting was performed as previously described [19]. Monoclonal anti-mouse Spp1 antibody (Santa Cruz, USA), horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG antibody (Boster, China), and DAB kit (Boster, China) were used. The purity of protein was further analyzed with high-performance liquid chromatography– size exclusion chromatography (HPLC–SEC) with TSK-GEL G2000SWXL column (Tosoh, Japan) on LC-2010HAT (Shimadzu, Japan). Twenty microliters of rmSPP1 at 0.5 mg/ml in citrate buffer (pH 6.5) was loaded onto the column. The protein was eluted by citrate buffer (0.1 M citrate, 0.1 M NaCl, pH 6.5) at speed of 0.5 ml/min. Absorbance was read at 280 nm, and profile was analyzed. Endotoxin in the rmSPP1 product was analyzed by kinetic turbidimetric LAL kit (Xiamen, China) according to the manufacture’s instruction. In Vitro Chemotaxis Assay The chemotaxis assay was performed with a 24-well transwell chamber (Costar 3422, 8 μm pore; Corning, USA). Peritoneal macrophages isolation and the chemotaxis assay were performed according to previous descriptions [20]. Briefly, peritoneal macrophages were collected and washed using RPMI 1640 (Gibco, USA) medium. Cells (1×105) in 200 μl RPMI 1640 were seeded to the top chamber of a 24-well transwell chamber plate. Different concentrations of rmSPP1 (2, 5, 10, and 20 ng/ml) were employed to test the chemoattractant effects with PBS as a negative control. After 2 h, cells on the upper surface of the membrane were cleared, and the membrane was fixed with 4 % formaldehyde (Sinopharm, China)
426
Appl Biochem Biotechnol (2014) 173:421–432
solution for 10 min and stained with hematoxylin (Sinopharm, China) for 10 min. The membrane was washed three times with distilled water and carefully placed on a glass slide. Cells that have migrated through the membranes and attached to their lower surfaces were counted in five random sights from each membrane, using a light microscope (Nikon, Japan). The experiment has been repeated for three times. Results of chemotaxis were calculated as the number of migrated cells at different conditions versus that of the control group. Data were analyzed with Student’s t test and presented as mean±SD.
Results The Mature Form of Mouse SPP1 Was Expressed in E. coli The mouse SPP1 precursor has 294 amino acid (aa) residues (GenBank accession no. J04806.1). The mature mouse SPP1 only has 278 aas (L17-N294), while the signal peptide expanding the first 16 aas of the precursor at the N-terminus is removed during its maturation [14, 17]. In the first PCR reaction, a long DNA sequence (1,018 bp) (Fig. 2a) has been cloned from cDNA library of liver tissue, and rmSpp1 cDNA (837 bp) (Fig. 2b) was amplified from the long DNA sequence by PCR. After verifying the sequence, the plasmid was transformed into E. coli, and expression of rmSPP1 was induced by IPTG. For the optimization of expression conditions, a series of IPTG concentrations (0, 0.2, 0.6, and 1 mM) were used to induce rmSpp1 expression. rmSpp1 productivity was positively correlated with IPTG concentration (Fig. S1A). We also found that the productivity of rmSpp1 was not significantly different between 4 and 6 h post 1 mM IPTG induction (Fig. S1B) (6.6±1.38 vs 7.7± 0.98 %). So, 1 mM of IPTG and 4 h of incubation after induction were applied to the following experiments. The expected protein product should expand from MRL17 to N294 with a theoretical molecular weight of about 30 kDa. Interestingly, the migration pattern of SPP1 in sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) always looks like that of a 60-kDa protein. In our study, the strong band around 60 kDa was observed in the supernatant of cell lysate upon induction (Fig. 3a, lane 3). To identify this protein, we performed Western blotting with commercial anti-mSPP1 monoclonal antibody. As the result
Fig. 2 Electrophoresis of Spp1 cloning and subcloning DNA fragments by PCR. a DNA fragment of Spp1 cloning. b DNA fragment of Spp1 subcloning for pET28a(+) by PCR
Appl Biochem Biotechnol (2014) 173:421–432
427
Fig. 3 Prime purification of rmSPP1 and rmSPP1 was detected by Western blot. a MW molecular weight; lanes indicate the following: 1 before induced, 2 whole cell, 3 supernatant after sonication, 4 supernatant after pH 5.0, 5 supernatant after pH 4.3, and 6 crude rmSpp1. b rmSpp1 was detected by Western blot; lane 1 indicates loading 5 μl bacterial lysate before induce and lane 2 loading 5 μl bacterial lysate after being induced by IPTG
was presented, two specific blotting bands close to each other were shown on the membrane. The size of the band was in accordance with the induced protein (Fig. 3b). So, we conclude that rmSPP1 could be expressed in E. coli after being induced by 1 mM IPTG. Purification of rmSPP1 After isoelectric precipitation and ammonium sulfate ((NH4)2SO4) fractionation, the purity of crude rmSPP1 was about 45 % which was estimated by SDS–PAGE (Table 1). Such product could not be used for its low purity. The purification strategy was created, starting with isoelectric precipitation and ammonium sulfate fractionation methods. The theoretical PI of the protein is pH 4.31, but the protein still remained in the supernatant after pH being adjusted through pH 7.4 to 4.3 with HCl (Fig. 3a, lane 4). Anion exchange chromatography was employed to further purify rmSPP1, because rmSPP1 is an acidic protein, and it could easily bind to both columns when dissolved in corresponding solutions. There were two UV peaks in the elution phase, and rmSPP1 existed in the first peak, as SDS–PAGE results indicate (Fig. 4). The purity of rmSPP1 was increased to 79.00 % after using anion exchange chromatography (Table 1). Applying cation exchange chromatography could further increase the purity, and since endotoxin could not bind to cation exchange matrix, we expected to remove endotoxin as well. For cation exchange chromatography, rmSPP1 was eluted in the second peak (Fig. 5a, b). The yielding purity of protein was about 97 % as analyzed with HPLC–SEC (Fig. 5c). Finally, the protein yield of a representative batch was shown in Table 1, and the productivity of rmSPP1 was 12.80 %. Concentration of endotoxin in rmSPP1 product was less than 0.2 EU/mg.
Table 1 Result of rmSPP1 purification at different stages. These data were from a 3-l medium Supernatant of cell lysate Prime purification Q-Sepharose eluate S-Sepharose eluate Total protein (mg)
671
22.26
9.45
6.20
Purity (%)
7.00
45.00
79.00
97.00
rmSPP1 (mg)
46.97
10.02
7.47
6.01
21.32
15.90
12.80
Yield of rmSPP1 (%)
428
Appl Biochem Biotechnol (2014) 173:421–432
Fig. 4 Purification of rmSPP1 by Q-Sepharose column. a Elution profile of rmSpp1 by Q-Sepharose column. b SDS–PAGE analysis of different fractions collected by Q-Sepharose; lanes 2–8 indicate loading nos. 4–10 fractions, respectively
The Purified rmSPP1 Was Biologically Active Biological function parameters are very important for recombinant protein products. Recently, many researchers have reported that SPP1 is involved in metastasis/ invasiveness of different cancers and associated with tissue inflammation. SPP1 could bind to integrin or CD44 receptors through RGD sequence and regulate cell migration in vitro. So, analysis on both the percentage of SPP1 binding to cells and its effect on cell migration could be performed to analyze the biological activity of SPP1 [3, 14]. In vitro macrophage migration assay is a good model for the biological activity assay of SPP1. Murine peritoneal macrophages were used to measure rmSPP1 activity. Both 5 and 10 ng/ml rmSPP1 in cell culture medium would effectively attract macrophages to migrate through the membrane, as compared with the control group in which PBS was added to cell culture medium instead of rmSPP1. Theoretically, the highest cell migration ratio would be obtained at some point between 5 and 10 ng/ml (Fig. 6a–e), while 20 ng/ml rmSPP1 would slightly reduce migration ratio (Fig. 6f). This unique dosage-dependent pattern is in accordance with multiple chemoattractant assay reports [19, 20]. These data indicated that rmSPP1 expressed and purified according to our strategy had biological activity in vitro.
Appl Biochem Biotechnol (2014) 173:421–432
429
Fig. 5 Purification profile of rmSPP1 by ion exchange chromatography. a Elution profile of rmSPP1 by SSepharose column. b SDS–PAGE analysis of different fractions collected by S-Sepharose. lane 1 indicates loading no. 12 fraction. c Fraction 12 was analyzed with HPLC–SEC, and purity of protein was over 97 %
Discussion In this study, we reported a simple and novel method to purify non-tagged recombinant SPP1, utilizing isoelectric precipitation, ammonium sulfate fractionation, Q-Sepharose and S-Sepharose chromatography. rmSPP1 could be specifically recognized by monoclonal anti-mouse osteopontin antibody (Fig. 3b) [21] and could actively increase chemoattractive migration of primary mouse peritoneal macrophages in vitro (Fig. 6) [3]. Interestingly, the actual electrophoretic mobility of SPP1 from multiple species always appears to be close to a protein with approximately twice the theoretical molecular weight of SPP1. This phenomenon has been observed on all forms of SPP1 fragments and full-length SPP1, of course, either expressed in bacteria or mammalian cells [15–18]. Mouse SPP1 and rat SPP1 amino acid sequences are 84 % identical to each other, while recombinant rat SPP1 appeared as two bands close to each other at approximately 60 and 63 kDa on SDS– PAGE gel [22], and this results is according to our data (Figs. 3b and 5b). Because protein could not be phosphorylated in E. coli, many researchers hypothesized that the anomalous migration pattern of SPP1 would most likely result from its relatively high content of acidic amino acids [17].
430
Appl Biochem Biotechnol (2014) 173:421–432
Fig. 6 Chemotaxis of rmSPP1 for mouse macrophages. a–e Zero, 2, 5, 10, and 20 ng/ml of rmSPP1 were used to induce migration of cells, respectively. *p
Expression and Purification of Bioactive High-Purity Recombinant Mouse SPP1 in Escherichia coli Yunsheng Yuan & Xiyuan Zhang & Shunyan Weng & Wen Guan & Di Xiang & Jin Gao & Jingjing Li & Wei Han & Yan Yu
Received: 19 October 2013 / Accepted: 5 March 2014 / Published online: 25 March 2014 # Springer Science+Business Media New York 2014
Abstract Secreted phosphoprotein 1 (SPP1) is a phosphorylated acidic glycoprotein. It is broadly expressed in a variety of tissues, and it is involved in a number of physiological and pathological events, including cancer metastasis, tissues remodeling, pro-inflammation regulation, and cell survival. SPP1 has shown its function of protecting tissues and organs against injury and wound, giving itself potentials to become a therapy target or giving its antibodies of other counter-acting reagents potentials to become drug candidates. Non-tagged (native) recombinant SPP1 would be valuable in therapeutic and pharmaceutical researches. In our study, mouse Spp1 DNA fragment without signal peptide was built in pET28a(+) vector and transformed into Escherichia coli BL21 (DE3). The recombinant mouse SPP1 (rmSPP1) was then expressed in bacteria upon induction by isopropyl β-D-thiogalactopyranoside (IPTG). The abundance of rmSPP1 was increased using isoelectric precipitation and ammonium sulfate fractionation methods, and anion and cation exchange chromatography was employed to
Electronic supplementary material The online version of this article (doi:10.1007/s12010-014-0849-7) contains supplementary material, which is available to authorized users.
Y. Yuan : X. Zhang : S. Weng : W. Guan : Y. Yu Shanghai Key Laboratory of Veterinary Biotechnology, College of Agriculture and Biology, Shanghai Jiao Tong University, 800 Dongchuan Rd, 200240 Shanghai, People’s Republic of China Y. Yuan e-mail: [email protected] X. Zhang e-mail: [email protected] S. Weng e-mail: [email protected] W. Guan e-mail: [email protected] Y. Yuan : Y. Yu (*) Key Laboratory of Urban Agriculture (South) Ministry of Agriculture, College of Agriculture and Biology, Shanghai Jiao Tong University, 800 Dongchuan Rd, 200240 Shanghai, People’s Republic of China e-mail: [email protected]
422
Appl Biochem Biotechnol (2014) 173:421–432
further purify rmSPP1. Finally, we got rmSPP1 product with 12.8 % productivity, 97 % purity, satisfactory bioactivity, and low endotoxin content. Keywords Secreted phosphoprotein 1 . Chromatography . Protein purification . Chemotaxis
Introduction Secreted phosphoprotein 1 (SPP1), also named osteopontin (OPN), is a phosphorylated acidic glycoprotein existing in mammals. It is involved in many physiological and pathological processes, including bone metabolism, inflammation progress, tumor metastasis, injury repair, and so on [1–3]. An amount of studies have demonstrated that SPP1 is not only a biomarker in many diseases, but it is also involved in relieving tissue injury caused by stress or toxin. For example, SPP1 could attenuate cardiac ischemia–reperfusion injury [4], hyperoxia-induced lung injury [5], and liver injury [6, 7]. SPP1 could also protect islet cell and improve survival of patients with pancreatic adenocarcinoma [8, 9]. In terms of structure, SPP1 has several special characteristics, including a RGD sequence, a Ca2+ binding site, and two heparin binding domains [10]. It could be cleaved by several proteases, such as thrombin, enterokinase, or matrix metalloproteinases (MMPs) [11, 12]. A number of investigators have addressed that the activity of SPP1 in the regulation of cell adhesion, protein–protein interaction, or bone resorption is independent from its degree of posttranslational modification [11]. To investigate the structure and biological functions of SPP1 in vitro, full-length SPP1 or its fragments have been isolated from milk [13, 14] or urine [15]. SPP1 could also be secreted by cultured rat smooth muscle cells [16]. Acutely, GST or six times His-tagged SPP1, expressed in Escherichia coli and purified with affinity chromatography, has been widely used to investigate the structure and biological functions of SPP1 [17, 18]. However, these tagged proteins could not be used to investigate pharmaceutical functions of SPP1 in animal disease models. So, it is important to develop a new method of purifying native recombinant SPP1 protein, for studies on the structure, stability, or therapeutic use of SPP1. Recombinant SPP1 would be useful to further study on the functions of SPP1 in animal diseases model or preclinical researches.
X. Zhang Department of Tumor Biology, Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, E407 New Research Building, 3970 Reservoir Road NW, Washington, DC 20057, USA D. Xiang Children’s Hospital Oakland Research Institute, 5700 Martin Luther King Jr Way, Oakland, CA 94609, USA e-mail: [email protected] J. Gao : J. Li : W. Han Laboratory of Regeneromics, School of Pharmacy, Shanghai Jiaotong University, 800 Dongchuan Rd, 200240 Shanghai, People’s Republic of China J. Gao e-mail: [email protected] J. Li e-mail: [email protected] W. Han e-mail: [email protected]
Appl Biochem Biotechnol (2014) 173:421–432
423
In this study, we created a novel method to produce and purify native recombinant mouse SPP1 (rmSPP1) in E. coli. Spp1 DNA sequence was amplified from mouse liver cDNA and inserted into pET28a(+) vector (Merck, German). rmSPP1 was expressed in BL21 E. coli. Biological active rmSPP1 was obtained at 97 % purity using a three-step procedure.
Materials and Methods Construction of the rmSPP1 Expression Vector Total RNA was isolated from C57BL/C mouse liver using TRIzol reagent (Invitrogen, USA), and cDNA was reversely transcribed with M-MuLV Reverse Transcriptase (Fermentas). Two PCR reactions were employed to clone mouse Spp1 gene. DNA fragment containing the open reading frame (ORF) of mouse Spp1 (167–1,184 bp, NCBI mRNA sequence no. NM_001204201) was amplified in the first PCR reaction. One pair of primers was used in this reaction as follows: the forward primer 5′-ACCTGACAAGACATCAACTGTGCC-3′ and the reverse primer 5′-TTCCTGCTTAACCCTCACTAACAC-3′ (Invitrogen, Shanghai). PCR was performed in a 50 μl reaction system containing 1× PCR buffer, MgCl2 (1.5 mM), dNTPs (0.2 mM), 5 U of Taq plus DNA polymerase (BBI, Canada), 2 μl of cDNA as template, and 120 nM each of the forward and reverse primers. After denaturing the cDNA for 3 min at 95 °C, amplification parameters were set up as follows: denaturation, 30 s at 95 °C; annealing, 40 s at 68 °C; and extension, 2 min at 72 °C. The DNA fragment was inserted into a T vector (Takara, Japan). Having its DNA sequence confirmed by sequencing, the construct was used as the template of the next PCR reaction to create rmSPP1 expression construct. The second PCR reaction was used to amplify mature mouse SPP1 DNA (226 to 1,059 bp), and a pair of primers was used in the reaction as follows: the forward primer 5′-AGTCCCATGGCTCTCC CGGTGAAAGTGACTG-3′ and the reverse primer 5′ -AGTCCTCGAGTTAGTTGACCTC AGA AGATG -3′ (underlined letters indicate NcoI and XhoI cleavage sites, respectively). The protein corresponding to this ORF would be a native mouse SPP1 which contains an Ala codon (GTC) at the 5′ end of the DNA fragment. The PCR parameters were the same as the first PCR reaction, except for the fact that annealing temperature was changed to 55 °C. The amplified DNA was digested with NcoI and XhoI and ligated with pET28a(+) vector which has been digested with NcoI and XhoI as well. The recombinant plasmid containing the mouse SPP1 gene was named pET28a-Spp1 (Fig. 1), which was verified by sequencing assay. Expression of rmSPP1 E. coli BL21 (DE3) cells were transformed with the plasmid pET28a-Spp1 and plated onto Luria–Bertani (LB) agar plate with kanamycin (50 μg/ml). A single colony was selected and inoculated in a 50-ml LB medium supplemented with kanamycin (50 μg/ml), with overnight shaking at 250 rpm, 37 °C. Thirty milliliters of overnight culture was added to the 3-l fresh LB medium and continued to be incubated until OD600 reached 1.0. Then, isopropyl β-Dthiogalactopyranoside (IPTG, 1 mM) was added into the medium to induce rmSPP1 expression in bacteria. Cells were harvested by centrifugation for 10 min (10,000g, 4 °C); after, bacteria were incubated for 4 h after IPTG was added (200 rpm, 37 °C). Cells were washed with 1× phosphate-buffered saline (PBS) (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4; pH 7.4) and resuspended with 30 ml of lysis buffer (1× PBS, pH 7.4; 1 mM EDTA; 0.1 mM phenylmethylsulfonyl fluoride (PMSF)). After 30 cycles of ultrasonication
424
Appl Biochem Biotechnol (2014) 173:421–432
Fig. 1 Map of Spp1 expression plasmid in E. coli. The underlined sequences are primers for PCR; the italicized letters represent restricted sits; the sequence between Start and Stop is translated to the recombinant mouse Spp1
(Sonics, USA) on ice (3 s of sonication and 7 s of resting for each cycle), the supernatant was collected by centrifugation (18,000g, 4 °C, 30 min). Purification of rmSPP1 We used isoelectric precipitation and ammonium sulfate ((NH4)2SO4) fractionation methods to exclude as much of bacterial protein as possible before going through ion exchange chromatography. The supernatant then went through a series of pH adjustment to pH values of 6.0, 5.0, and 4.3 with 1 M HCl solution. After each step of pH adjustment, the sample was kept under 4 °C for 30 min, for the thorough sedimentation of bacterial proteins whose isoelectric points (PIs) are close to these pH points. The resulting insoluble bacterial proteins were removed by centrifugation for 15 min (18,000g, 4 °C). Crude rmSPP1 was precipitated with 20 % saturated (NH4)2SO4 at pH 4.3 and solubilized in 3 ml of 1× PBS. The solution containing crude rmSPP1 then went through a tenfold dilution by 20 mM Tris–HCl (1 mM EDTA, 0.1 mM PMSF, pH 8.0) and was loaded onto a Q-Sepharose column (GE Healthcare, USA) attached to an AKTA Explorer 10/100 system (Amersham Pharmacia Biotech, Sweden) at the rate of 1 ml/min. After loading was finished, the column was washed with 4 column
Appl Biochem Biotechnol (2014) 173:421–432
425
volumes of buffer 1 (20 mM Tris–HCl, 1 mM EDTA, 25 mM NaCl, pH 8.0). rmSPP1 was eluted using a mixed solution of buffer 1 coming from sample pump A and buffer 2 (20 mM Tris–HCl, 1 mM EDTA, 25 mM NaCl, pH 8.0) from sample pump B. The mixed solution goes through the Q-Sepharose column with the speed of 1 ml/min, while the percentage of solution coming from sample pump B is elevated from 10 to 70 % with a fixed rate of 1 % elevation per min, thus creating a consecutive NaCl concentration gradient. Proteins would be eluted at different points in such gradient, according to the difference of their binding ability to the QSepharose matrix under pH 8.0. Fractions were collected according to peaks of UV absorption and conductivity curve. Fractions containing rmSPP1 were pooled and dialyzed overnight in 100-fold volume of buffer 3 (20 mM Na citrate, 20 mM NaCl, 1 mM EDTA, pH 3.2). Dialyzed protein was then loaded onto the S-Sepharose column (GE Healthcare, USA) attached to an AKTA Explorer 10/100 system at a rate of 1 ml/min, and the column was washed with 4 column volumes of buffer 3 at the same speed. The column-bound protein was eluted with a similar mixed solution containing a consecutive NaCl concentration gradient, while the percentage of solution coming from sample pump B is elevated from 12 to 72 % this time. Fractions were collected according to UV absorption and conductivity curves. The protein concentration in fractions was detected with Bradford assay. Sodium Dodecyl Sulfate–Polyacrylamide Gel Electrophoresis, Western Blotting, High-Performance Liquid Chromatography–Size Exclusion Chromatography Analysis, and Endotoxin Assay To estimate the purity of rmSpp1, 5 μg of protein was loaded onto 15 % polyacrylamide gel ran with the PowerPac Basic (Bio-Rad, USA) and stained with Coomassie bright blue R-250 (Beyotime, China). To estimate the purity and productivity of rmSPP1, the bands on gels were scanned with Tannon Gel Image System (Tannon 2500, China), and intensity of each band was recorded and analyzed with GIS software 1.0 (Tannon, China). Western blotting was performed as previously described [19]. Monoclonal anti-mouse Spp1 antibody (Santa Cruz, USA), horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG antibody (Boster, China), and DAB kit (Boster, China) were used. The purity of protein was further analyzed with high-performance liquid chromatography– size exclusion chromatography (HPLC–SEC) with TSK-GEL G2000SWXL column (Tosoh, Japan) on LC-2010HAT (Shimadzu, Japan). Twenty microliters of rmSPP1 at 0.5 mg/ml in citrate buffer (pH 6.5) was loaded onto the column. The protein was eluted by citrate buffer (0.1 M citrate, 0.1 M NaCl, pH 6.5) at speed of 0.5 ml/min. Absorbance was read at 280 nm, and profile was analyzed. Endotoxin in the rmSPP1 product was analyzed by kinetic turbidimetric LAL kit (Xiamen, China) according to the manufacture’s instruction. In Vitro Chemotaxis Assay The chemotaxis assay was performed with a 24-well transwell chamber (Costar 3422, 8 μm pore; Corning, USA). Peritoneal macrophages isolation and the chemotaxis assay were performed according to previous descriptions [20]. Briefly, peritoneal macrophages were collected and washed using RPMI 1640 (Gibco, USA) medium. Cells (1×105) in 200 μl RPMI 1640 were seeded to the top chamber of a 24-well transwell chamber plate. Different concentrations of rmSPP1 (2, 5, 10, and 20 ng/ml) were employed to test the chemoattractant effects with PBS as a negative control. After 2 h, cells on the upper surface of the membrane were cleared, and the membrane was fixed with 4 % formaldehyde (Sinopharm, China)
426
Appl Biochem Biotechnol (2014) 173:421–432
solution for 10 min and stained with hematoxylin (Sinopharm, China) for 10 min. The membrane was washed three times with distilled water and carefully placed on a glass slide. Cells that have migrated through the membranes and attached to their lower surfaces were counted in five random sights from each membrane, using a light microscope (Nikon, Japan). The experiment has been repeated for three times. Results of chemotaxis were calculated as the number of migrated cells at different conditions versus that of the control group. Data were analyzed with Student’s t test and presented as mean±SD.
Results The Mature Form of Mouse SPP1 Was Expressed in E. coli The mouse SPP1 precursor has 294 amino acid (aa) residues (GenBank accession no. J04806.1). The mature mouse SPP1 only has 278 aas (L17-N294), while the signal peptide expanding the first 16 aas of the precursor at the N-terminus is removed during its maturation [14, 17]. In the first PCR reaction, a long DNA sequence (1,018 bp) (Fig. 2a) has been cloned from cDNA library of liver tissue, and rmSpp1 cDNA (837 bp) (Fig. 2b) was amplified from the long DNA sequence by PCR. After verifying the sequence, the plasmid was transformed into E. coli, and expression of rmSPP1 was induced by IPTG. For the optimization of expression conditions, a series of IPTG concentrations (0, 0.2, 0.6, and 1 mM) were used to induce rmSpp1 expression. rmSpp1 productivity was positively correlated with IPTG concentration (Fig. S1A). We also found that the productivity of rmSpp1 was not significantly different between 4 and 6 h post 1 mM IPTG induction (Fig. S1B) (6.6±1.38 vs 7.7± 0.98 %). So, 1 mM of IPTG and 4 h of incubation after induction were applied to the following experiments. The expected protein product should expand from MRL17 to N294 with a theoretical molecular weight of about 30 kDa. Interestingly, the migration pattern of SPP1 in sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) always looks like that of a 60-kDa protein. In our study, the strong band around 60 kDa was observed in the supernatant of cell lysate upon induction (Fig. 3a, lane 3). To identify this protein, we performed Western blotting with commercial anti-mSPP1 monoclonal antibody. As the result
Fig. 2 Electrophoresis of Spp1 cloning and subcloning DNA fragments by PCR. a DNA fragment of Spp1 cloning. b DNA fragment of Spp1 subcloning for pET28a(+) by PCR
Appl Biochem Biotechnol (2014) 173:421–432
427
Fig. 3 Prime purification of rmSPP1 and rmSPP1 was detected by Western blot. a MW molecular weight; lanes indicate the following: 1 before induced, 2 whole cell, 3 supernatant after sonication, 4 supernatant after pH 5.0, 5 supernatant after pH 4.3, and 6 crude rmSpp1. b rmSpp1 was detected by Western blot; lane 1 indicates loading 5 μl bacterial lysate before induce and lane 2 loading 5 μl bacterial lysate after being induced by IPTG
was presented, two specific blotting bands close to each other were shown on the membrane. The size of the band was in accordance with the induced protein (Fig. 3b). So, we conclude that rmSPP1 could be expressed in E. coli after being induced by 1 mM IPTG. Purification of rmSPP1 After isoelectric precipitation and ammonium sulfate ((NH4)2SO4) fractionation, the purity of crude rmSPP1 was about 45 % which was estimated by SDS–PAGE (Table 1). Such product could not be used for its low purity. The purification strategy was created, starting with isoelectric precipitation and ammonium sulfate fractionation methods. The theoretical PI of the protein is pH 4.31, but the protein still remained in the supernatant after pH being adjusted through pH 7.4 to 4.3 with HCl (Fig. 3a, lane 4). Anion exchange chromatography was employed to further purify rmSPP1, because rmSPP1 is an acidic protein, and it could easily bind to both columns when dissolved in corresponding solutions. There were two UV peaks in the elution phase, and rmSPP1 existed in the first peak, as SDS–PAGE results indicate (Fig. 4). The purity of rmSPP1 was increased to 79.00 % after using anion exchange chromatography (Table 1). Applying cation exchange chromatography could further increase the purity, and since endotoxin could not bind to cation exchange matrix, we expected to remove endotoxin as well. For cation exchange chromatography, rmSPP1 was eluted in the second peak (Fig. 5a, b). The yielding purity of protein was about 97 % as analyzed with HPLC–SEC (Fig. 5c). Finally, the protein yield of a representative batch was shown in Table 1, and the productivity of rmSPP1 was 12.80 %. Concentration of endotoxin in rmSPP1 product was less than 0.2 EU/mg.
Table 1 Result of rmSPP1 purification at different stages. These data were from a 3-l medium Supernatant of cell lysate Prime purification Q-Sepharose eluate S-Sepharose eluate Total protein (mg)
671
22.26
9.45
6.20
Purity (%)
7.00
45.00
79.00
97.00
rmSPP1 (mg)
46.97
10.02
7.47
6.01
21.32
15.90
12.80
Yield of rmSPP1 (%)
428
Appl Biochem Biotechnol (2014) 173:421–432
Fig. 4 Purification of rmSPP1 by Q-Sepharose column. a Elution profile of rmSpp1 by Q-Sepharose column. b SDS–PAGE analysis of different fractions collected by Q-Sepharose; lanes 2–8 indicate loading nos. 4–10 fractions, respectively
The Purified rmSPP1 Was Biologically Active Biological function parameters are very important for recombinant protein products. Recently, many researchers have reported that SPP1 is involved in metastasis/ invasiveness of different cancers and associated with tissue inflammation. SPP1 could bind to integrin or CD44 receptors through RGD sequence and regulate cell migration in vitro. So, analysis on both the percentage of SPP1 binding to cells and its effect on cell migration could be performed to analyze the biological activity of SPP1 [3, 14]. In vitro macrophage migration assay is a good model for the biological activity assay of SPP1. Murine peritoneal macrophages were used to measure rmSPP1 activity. Both 5 and 10 ng/ml rmSPP1 in cell culture medium would effectively attract macrophages to migrate through the membrane, as compared with the control group in which PBS was added to cell culture medium instead of rmSPP1. Theoretically, the highest cell migration ratio would be obtained at some point between 5 and 10 ng/ml (Fig. 6a–e), while 20 ng/ml rmSPP1 would slightly reduce migration ratio (Fig. 6f). This unique dosage-dependent pattern is in accordance with multiple chemoattractant assay reports [19, 20]. These data indicated that rmSPP1 expressed and purified according to our strategy had biological activity in vitro.
Appl Biochem Biotechnol (2014) 173:421–432
429
Fig. 5 Purification profile of rmSPP1 by ion exchange chromatography. a Elution profile of rmSPP1 by SSepharose column. b SDS–PAGE analysis of different fractions collected by S-Sepharose. lane 1 indicates loading no. 12 fraction. c Fraction 12 was analyzed with HPLC–SEC, and purity of protein was over 97 %
Discussion In this study, we reported a simple and novel method to purify non-tagged recombinant SPP1, utilizing isoelectric precipitation, ammonium sulfate fractionation, Q-Sepharose and S-Sepharose chromatography. rmSPP1 could be specifically recognized by monoclonal anti-mouse osteopontin antibody (Fig. 3b) [21] and could actively increase chemoattractive migration of primary mouse peritoneal macrophages in vitro (Fig. 6) [3]. Interestingly, the actual electrophoretic mobility of SPP1 from multiple species always appears to be close to a protein with approximately twice the theoretical molecular weight of SPP1. This phenomenon has been observed on all forms of SPP1 fragments and full-length SPP1, of course, either expressed in bacteria or mammalian cells [15–18]. Mouse SPP1 and rat SPP1 amino acid sequences are 84 % identical to each other, while recombinant rat SPP1 appeared as two bands close to each other at approximately 60 and 63 kDa on SDS– PAGE gel [22], and this results is according to our data (Figs. 3b and 5b). Because protein could not be phosphorylated in E. coli, many researchers hypothesized that the anomalous migration pattern of SPP1 would most likely result from its relatively high content of acidic amino acids [17].
430
Appl Biochem Biotechnol (2014) 173:421–432
Fig. 6 Chemotaxis of rmSPP1 for mouse macrophages. a–e Zero, 2, 5, 10, and 20 ng/ml of rmSPP1 were used to induce migration of cells, respectively. *p
