A double epitope tag for quantification of recombinant protein using fluorescence resonance energy transfer

Analytical Biochemistry 380 (2008) 249–256

Contents lists available at ScienceDirect

Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio

A double epitope tag for quantification of recombinant protein using fluorescence resonance energy transfer Koji Enomoto a,*, Ken-Ichiro Uwabe b, Shoichi Naito b, Jyunji Onoda b, Akira Yamauchi b, Yoshito Numata a, Hiroshi Takemoto a a b

Discovery Research Laboratories, Shionogi & Co., Ltd., 5-12-4, Sagisu, Fukushima-ku, Osaka 553-0002, Japan Developmental Research Laboratories, Shionogi & Co., Ltd., 3-1-1, Futaba-cho, Toyonaka, Osaka 561-0825, Japan

a r t i c l e

i n f o

Article history: Received 24 March 2008 Available online 3 June 2008 Keywords: Rare earth cryptate Homogeneous time-resolved fluorescence Affinity tag Protein expression Protein purification Recombinant protein Immunoassay

a b s t r a c t The expression of recombinant proteins is a well-accepted technology, but their detection and purification often require time-consuming and complicated processes. This paper describes the development of a novel double epitope tag (GEPGDDGPSGAEGPPGPQG) for rapid and accurate quantification of recombinant protein by a homogeneous immunoassay based on fluorescence resonance energy transfer. In our double epitope tagging system, recombinant proteins can be simply measured on a microtiter plate by addition of a pair of fluorophore-labeled monoclonal antibodies (their epitopes; GEPGDDGPS and GPPGPQG). The sensitivity of the immunoassay with an incubation time of only 5 min is almost equal to that of labor-intensive Western blotting. In addition, culture media and extracts of host cells generally used for protein expression have little effect on this immunoassay. To investigate the utility of our proposed tag for protein production, several different proteins containing this tag were practically expressed and purified. The data presented demonstrate that the double epitope tag is a reliable tool that can alleviate the laborious and troublesome processes of protein production. Ó 2008 Elsevier Inc. All rights reserved.

In the post genomic era, intense scientific research has been directed toward elucidating the functions of newly found genes and identifying therapeutically relevant genes. In order to characterize specific gene products of interest, it is important to express and purify recombinant proteins in a variety of expression hosts as quickly and efficiently as possible. In the pharmaceutical industries, in particular, with the development of many high-throughput technologies for lead discovery, the bottleneck has quickly shifted to fast and efficient preparation of a sufficient amount of active protein [1]. Such recombinant proteins are required for the production of antibodies, development of functional assays, identification of interacting proteins, and characterization of their native structures. Over the last two decades, epitope tagging and epitope-specific antibodies have been used for the analysis, identification, and purification of recombinant proteins, especially of novel proteins for which there are no specific antibodies [2,3]. The most widely used tags are small peptides such as polyhistidine [4], FLAG [5], and MYC [6], and soluble proteins such as glutathione S-transferase (GST)1 [7] and malt-

ose-binding protein (MBP) [8], which are placed at either the amino- or the carboxy-terminus of recombinant proteins. Expression levels of recombinant proteins vary with the target proteins, DNA constructs, host cells, and gene transfer methods. Therefore, it is important to determine the optimal conditions of expression and purification of recombinant proteins for their efficient output. Electrophoretic and immunoanalytical technologies, such as SDS-PAGE, enzyme-linked immunosorbent assay (ELISA), and Western blotting, can be used for the optimization, but these assays are time consuming and involve complicated steps. In this report, we describe the development of a novel double epitope tag composed of 19 amino acids (GEPGDDGPSG AEGPPGPQG) for rapid and accurate quantification of recombinant protein with a homogeneous sandwich immunoassay. We first generated two monoclonal antibodies, TAG-6G4 (epitope; GEPGDDGPS) and TAG-2E6 (epitope; GPPGPQG), that could simultaneously bind to a double epitope tag. We next developed a homogeneous immunoassay to measure the double epitope tag

* Corresponding author. Fax: +81 6 6458 0987. E-mail address: [email protected] (K. Enomoto). 1 Abbreviations used: biotin-HPDP, N-(6-(biotinamido)hexyl)-30 -(20 -pyridyldithio)-propionamide; BSA, bovine serum albumin; ELISA, enzyme-linked immunosorbent assay; FRET, fluorescence resonance energy transfer; GST, glutathione S-transferase; HTRF, homogeneous time-resolved fluorescence; KLH, keyhole limpet hemocyanine; MBP, maltosebinding protein; maleimide PEO2-biotin, (+)-biotinyl-3-maleimidopropionamidyl-3,6-dioxaoctanediamine; MTA, 50 -deoxy-50 -methylthioadenosine; PBS, phosphate-buffered saline; scFv, single chain antibody; SPDS, human spermidine synthase; sulfo-EMCS, N-(6-maleimidocaproyloxy) sulfosuccinimide; sulfo-NHS-LC-biotin, sulfosuccinimidyl-6(biotinamido) hexanoate; sulfo-SMCC, sulfo-succinimidyl 4-(N-maleimidomethyl)-cyclohexane-1-carboxylate. 0003-2697/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2008.05.043

250

Double epitope tag for protein quantitation / K. Enomoto et al. / Anal. Biochem. 380 (2008) 249–256

using HTRF technology [9]. HTRF is based on fluorescence resonance energy transfer (FRET) from europium cryptate (fluorescent donor) to cross-linked allophycocyanin (XL665, fluorescent acceptor). When both fluorescence molecules are in the proximity, XL665 emits long-lived fluorescence at 665 nm. Time-resolved measurement of this long-lived fluorescence makes it possible to omit the nonspecific short-lived fluorescence. Ratio measurement of the fluorescence at 665 and 620 nm emitted from cryptate as an internal reference helps correct the quenching of fluorescence by media absorbency. These features of HTRF enable us to detect molecules without bound/free separation. By using the double epitope tag, recombinant protein can be simply measured on a microtiter plate by addition of two monoclonal antibodies, TAG-6G4 and TAG-2E6, which are labeled with europium cryptate and XL665, respectively. We evaluated the double epitope tag for its ability to aid the detection and purification of recombinant proteins. Our results demonstrate that the novel double epitope tag makes possible a simple, rapid, and sensitive method of detecting recombinant protein to alleviate the laborious and troublesome processes of protein production.

Materials and methods Reagents Europium cryptate and XL665 were purchased from CIS Bio International (Marcoule, France). N-(6-Maleimidocaproyloxy) sulfosuccinimide (sulfo-EMCS), sulfo-succinimidyl 4-(N-maleimidomethyl)-cyclohexane-1-carboxylate (sulfo-SMCC), N-(6-(biotinamido)- hexyl)- 30 -(20 -pyridyldithio)-propionamide (biotin-HPDP), sulfosuccinimidyl-6-(biotinamido) hexanoate (sulfo-NHS-LC-biotin), (+)-biotinyl-3-maleimidopropionamidyl-3,6-dioxaoctanediamine (maleimide PEO2-biotin), keyhole limpet hemocyanine (KLH), and streptavidin were from Pierce (Rockford, IL). Lipofectamine 2000 was from Invitrogen (Carlsbad, CA), bovine serum albumin (BSA) from Shibayagi (Gunma, Japan), and Block-Ace from Dainippon Pharmaceutical (Osaka, Japan). Haptenic peptide I (GEPGDDAPSC) and collagen partial peptide (GEPGDDAPSGAEGP* PGPQG, P* was hydroxylated) were synthesized by BioSynthesis (Lewisville, TX). Haptenic peptide II (CGPP*GPQG, P* was hydroxylated) and haptenic peptide III (L-Cys-L-Ala-D-isoGlu-L-Lys-D-Ala-D-Ala) were synthesized by Peptide Institute (Osaka, Japan) and Greiner Bio-One (Frickenhausen, Germany), respectively. Assay buffer, wash solution, enhancement solution, and europium-labeled streptavidin used for DELFIA were purchased from Perkin-Elmer (Waltham, MA). Goat anti-mouse IgG, protease inhibitor cocktail, and T4 DNA ligase were from Roche (Penzberg, Germany). Anti-mouse IgG-horseradish peroxidase conjugate and TMB + substrate chromogen were acquired from GE Healthcare (Buckinghamshire, UK) and Daco Cytomation (Carpentaria, CA), respectively. Biotinylation of peptides Haptenic peptide I (0.21 lmol) or II (0.28 lmol) was dissolved in 100 ll of coupling buffer (0.1 M phosphate buffer, pH 6.0, 5 mM EDTA) and then mixed with 2-fold molar excess of biotin-HPDP dissolved in DMF (100 ll). Collagen partial peptide (0.1 lmol) was dissolved in 170 ll of 0.1 M phosphate buffer, pH 7.4, and then mixed with 10-fold molar excess of sulfoNHS-LC-biotin solution (90 ll). After these mixtures had been incubated for over 2 h at room temperature, all of the biotinylated peptides were purified by reverse-phase HPLC (column: YMC-Pack, ODS-AM, 4.6 id  150 mm; eluting conditions: metha-

nol/0.1% TFA, 0–60%, 30 min, linear gradient; flow rate 1 ml/min; detection wavelength: 220 nm). Haptenic peptide III (420 nmol) was diluted in the coupling buffer (100 ll) after being dissolved in a small amount of methanol. Next, the peptide solution was mixed with 3-fold molar excess of maleimide PEO2-biotin in methanol and incubated overnight at 4 °C. Biotinylated haptenic peptide III was also purified by reverse-phase HPLC under the same conditions. Generation of monoclonal antibodies Haptenic peptides I and II were conjugated with KLH (Immunogen I) and BSA (Immunogen II), respectively. Immunogen I was prepared as follows. KLH (20 mg) was dissolved in phosphate-buffered saline (PBS; 2 ml) and then sulfo-SMCC (3.3 mg) was added to the protein solution. After the mixture had been left for 1 h at room temperature, maleimide-activated KLH was separated in the coupling buffer by gel filtration using a PD-10 column (GE Healthcare). To the KLH solution (10 mg) was added haptenic peptide I (2.1 mg) dissolved in the conjugation buffer (0.1 ml), and the mixture was incubated for 3 h at room temperature. After incubation, the conjugate was dialyzed against distilled water and then freeze-dried. For preparation of Immunogen II, BSA (20 mg) was dissolved in 0.5 ml of 0.1 M phosphate buffer, pH 7.0, and then sulfo-EMCS (5.8 mg) was added to the protein solution. After the mixture had been left for 1 h at room temperature, maleimideactivated BSA was separated by gel filtration. To the maleimideactivated BSA solution (11 mg) was added haptenic peptide II (3.8 mg) dissolved in the coupling buffer, and the mixture was incubated for 16 h at 4 °C. The conjugate was also dialyzed and freeze-dried. A/J jms Slc mice were immunized with intraperitoneal injections of Immunogen I or II in Freund’s adjuvant at 3-week intervals over a period of 3 months. Fusion of spleen cells from the immunized mouse with mouse myeloma cells, p3X63-Ag8.UI, was performed in accordance with a conventional procedure [10]. Culture supernatants of hybridomas were screened by DELFIA as follows. Each well of 96-well microtiter plates (Sumitomo Bakelite, Tokyo, Japan) was filled with 150 ll of goat anti-mouse IgG (10 lg/ ml) dissolved in 50 mM Tris–HCl, pH 7.4, and incubated overnight at 4 °C. After removal of the antibody solution, the wells were filled with 200 ll of 2% Block-Ace in the same buffer and incubated for 2 h at room temperature. After removal of the solution, 50-ll portions of assay buffer and hybridoma supernatants were consecutively added to the wells. Labeled antigen solutions, which were composed of mixtures of biotinylated haptenic peptide I or II (20 ng/ml), and europium-labeled streptavidin (200 ng/ml) in assay buffer were prepared and 50 ll of the appropriate mixture was added to the wells. The plates were incubated at 4 °C for 16 h and then washed twice with wash solution. After the washing, 150 ll of enhancement solution was added to the wells, and the specific antibodies were detected by measuring time-resolved fluorescence (Ex 340 nm, Em 615 nm) with a Victor Multilabel Counter (Perkin-Elmer). Positive cells were cloned by a limiting dilution technique. These hybridomas were expanded intraperitoneally in Balb/c mice, and monoclonal antibodies TAG-6G4 and TAG-2E6 binding to haptenic peptides I and II, respectively, were purified from ascitic fluids using Affi-Gel Protein A MAPS II Kit (Bio-Rad Laboratories, Hercules, CA). The isotypes of these antibodies were determined with the Mouse Monoclonal Antibody Isotyping ELISA Kit (BD Pharmingen, San Diego, CA). To compare the affinities of the monoclonal antibodies for the collagen partial peptide, competitive immunoassays were also performed with biotinylated collagen partial peptide and serially diluted free peptide solution in the screening assay.

Double epitope tag for protein quantitation / K. Enomoto et al. / Anal. Biochem. 380 (2008) 249–256

251

Homogeneous immunoassay for double epitope-tagged recombinant protein

Expression of double epitope-tagged single chain antibody (scFv) fragment

Monoclonal antibodies TGA-6G4 and TAG-2E6 were labeled with europium cryptate and XL665, respectively, as previously reported [11,12]. XL665-hapten I conjugate was prepared using XL665 instead of KLH in the same manner as Immunogen I. A 10ll portion of test sample was added to each well of a 384-well assay plate (black, low volume, Corning, Acton, MA). To detect the tagged protein with sandwich immunoassay, 10 ll of a mixture of 200 ng/ml cryptate-labeled TAG-6G4 and 5000 ng/ml XL665-labeled TAG-2E6 diluted in HTRF buffer (50 mM Tris–HCl, pH 7.4, 150 mM NaCl, 0.1% BSA, 800 mM potassium fluoride) was added to the wells. For detection using competitive immunoassay, 5 ll of 400 ng/ml cryptate-labeled TAG-6G4 and 5 ll of 10,000 ng/ml XL665-hapten I conjugate diluted in HTRF buffer were successively added to the wells. After the plate had been incubated at room temperature for an appropriate time, the HTRF value ([ratio of time-resolved fluorescence at 665 nm to that at 620 nm]  10,000) in each well was measured using Rubystar (BMG Labtech, Offenburg, Germany). In order to investigate the effects of biological media on the immunoassay, the following sample diluents were used: TBS (50 mM Tris–HCl, pH 7.4, 150 mM NaCl) containing 0.1% BSA, culture supernatant of Spodoptera frugiperda Sf9 cell (Sf900 II serum free medium, Invitrogen), Sf9 cell lysate prepared with RIPA buffer [13], YPD medium (yeast extract/peptone/dextrose) from Saccharomyces cerevisiae BJ1191 strain culture and Escherichia coli BL21 strain extract prepared by sonication in PBS, 293T cell culture supernatant, and their lysate prepared as described below.

Haptenic peptide III was conjugated to KLH, and the hybridoma producing monoclonal antibody against the peptide was produced in the same manner as TAG-6G4 described above. Total RNA was extracted from the hybridoma (5  106 cells) with the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The first strands of cDNA corresponding to the variable regions of the heavy-chain (VH) and light-chain (VL) of the immunoglobulin were synthesized from purified RNA by reversetranscriptase reaction using the SMART RACE cDNA Amplification Kit (BD Biosciences Clontech, Mountain View, CA) coupled with the 30 specific primers based on the nucleotide sequences in the constant regions of the heavy chain (50 -TAG AGT CAC CGA GGA GCC AGT TGT-30 ) and the light chain (50 -GAC TGA GGC ACC TCC AGA TGT TAA-30 ). The VH and VL genes were PCR-amplified separately using 30 specific primers, 50 -AGG GGC CAG TGG ATA GAC CGA TGG GGC TGT-30 and 50 -GGA TGG TGG GAA GAT GGA TAC AGT TGG TGC AGC-30 , respectively. The PCR products were subcloned into pCR2.1 vector (Invitrogen) and sequences of VH and VL genes were determined. To assemble the 30 end of the VL and the 50 end of the VH with oligonucleotide sequence encoding three repeats of Gly4Ser linker and to add an oligonucleotide sequence encoding the double epitope tag to the 30 end of the VH gene, PCR were performed as follows. The VL gene was amplified using the Expand High Fidelity PCR System (Roche) and the oligonucleotide primers 50 -TTC TAT GCG GCC CAG CCG GCC ATG GCC GAT GTT TTG ATG ACC CAA-30 (sense, SfiI site underlined) and 50 -CCA CCG CCG GAT CCA CCT CCG CCT GAA CCG CCT CCG CCA TCA GCC CGT TTG ATT TC-30 (antisense, BamHI site underlined and in bold, the sequence encoding the linker), and then digested with SfiI and BamHI. The VH gene was amplified in the same manner as the VL gene with the oligonucleotide primers 50 -GCG GAG GTG GAT CCG GCG GTG GCG GAT CGC AAG TTA CTC TAA AAG AG-30 (sense, BamHI site underlined and in bold, the sequence encoding the linker) and 50 -TTA TGA TAG CGG CCG CTC AAC CCT GGG GAC CTG GTG GAC CTT CGG CAC CAG AGG GAC CGT CAT CTC CAG GCT CTC CTG AGG AGA CGG TGA C-30 (antisense, NotI site underlined and in bold, the sequence encoding the double epitope tag), and then digested with BamHI and NotI digested PCR products were inserted between the SfiI and NotI sites of the predigested pCANTAB-5E expression vector (GE Healthcare) with T4 DNA ligase. Following transformation of competent Escherichia coli TG1 (Invitrogen) and selective plating, the appropriate clone was selected by sequencing inserts. For expression of the scFv fragment, TG1 was transformed with the pCANTAB-5E construct and grown at 30 °C in 2X YT medium (yeast extract/tryptone) containing 2% glucose and 100 lg/ml ampicillin. The overnight culture was diluted 10 times in 100 ml of 2X YT medium containing 2% glucose and shaken at 30 °C until the absorbance at 600 nm was around 0.6. The culture was then centrifuged at 1500g for 20 min, and the pellet was resuspended in 100 ml of 2X YT medium containing 1 mM IPTG and 100 lg/ ml ampicillin. The cells were allowed to incubate at 30 °C for 8 h with shaking and then the culture supernatant was harvested by centrifugation for 30 min at 1500g. The cell pellet was resuspended in 2 ml of ice-cold 1X TES buffer (0.2 M Tris–HCl, pH 8.0, 0.5 mM EDTA, 0.5 M sucrose) and then diluted by addition of 3.3 ml of ice-cold 0.2X TES buffer. The extract was incubated for 30 min on ice and then centrifuged for 30 min at 10,000g to remove cell debris. Finally, the crude periplasmic fraction was passed through a 0.22-lm filter. The activity of the double epitope-tagged scFv fragment was confirmed by ELISA as follows. Flat-bottomed 384-well Maxisorp plates (Nunc, Roskilde, Denmark) were incubated overnight with streptavidin (10 lg/ml PBS) at 4 °C and any remaining protein-

Expression of double epitope-tagged human spermidine synthase Expression vector of full-length human spermidine synthase (SPDS) tagged with the double epitope at the carboxy-terminus was constructed by modification of the original vector used for expression of SPDS tagged with 3X FLAG in the amino-terminus as described in our previous study [12]. To assemble the oligonucleotide sequence encoding the double epitope tag at the 30 end of the SPDS gene, PCR was performed using the original vector as a template and primers: 50 -TCA TTC GGG CCC CGC CGA AAG TCT CTT CAA GGA GTC-30 (sense, ApaI site underlined) and 50 -TTA TGA TAG CGG CCG CTC AAC CCT GGG GAC CTG GTG GAC CTT CGG CAC CAG AGG GAC CGT CAT CTC CAG GCT CTC CGC TCA CAT CAT TCA G-30 (antisense, NotI site underlined and in bold, the sequence encoding the double epitope tag). The PCR product was then subcloned into ApaI and NotI sites of the original vector. Expression of the double epitope-tagged SPDS was also performed according to our previous report [12]. Briefly, a modified vector was transfected into 293T cells (American Type Culture Collection, Manassas, VA) with Lipofectamine 2000 and the cells were maintained in Dulbecco’s modified eagle’s medium (Sigma) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT) for 3 days. Next, the cells were pelleted by centrifugation and resuspended in cold TBS containing protease inhibitor cocktail. The cell suspension was sonicated and centrifuged and then the supernatants were collected. The activity of the double epitope-tagged SPDS was measured by the competitive immunoassay described previously [12]. In brief, purified SPDS was incubated with 0.65 mM putrescine and 0.05 mM decarboxylated S-adenosylmethionine in enzyme buffer (0.1 M potassium phosphate, pH 7.4, 0.1% BSA, 0.2 mM DTT, 0.2 mM adenine). The reaction was stopped by the addition of 4-methylcyclohexylamine and then cryptate-labeled anti-50 -deoxy-50 -methylthioadenosine (MTA) antibody and XL665-MTA conjugate were added to the reaction mixture. Finally, the amount of MTA was quantified by measuring the HTRF value.

252

Double epitope tag for protein quantitation / K. Enomoto et al. / Anal. Biochem. 380 (2008) 249–256

binding sites were blocked by incubation with 3% BSA in PBS. Each reagent of the ELISA was diluted in PBS containing 1% BSA and 0.01% Tween 20. Between the addition of different reagents, wells were washed three times with PBS. Biotinylated haptenic peptide III (300 ng/ml, 40 ll) was added to the wells and incubated for 1 h. Purified double epitope-tagged scFv fragment (40 ll) was added and the plate was incubated for 2 h. A mixture (40 ll) of TAG-6G4 monoclonal antibody (8 lg/ml) and anti-mouse IgG-horseradish peroxidase conjugate (diluted 1/1000) was added to the wells, and the plate was incubated for 2 h. TMB + substrate chromogen (40 ll) was added to the wells and after 10 min, the reaction was stopped by adding 4 M sulphuric acid (40 ll). The absorbance of each well was measured at 450 nm using an Envision Multilabel Reader (Perkin-Elmer). Immunoaffinity chromatography All of the chromatography processes were performed at 4 °C. For purification utilizing the double epitope tag, monoclonal antibody TAG-2E6 was covalently conjugated to agarose gel of Aminolink Immobilization Kit (Pierce) according to the manufacturer’s instructions. The biological medium including double epitopetagged recombinant protein was applied to the column. The column was extensively washed with PBS and then tagged protein was eluted with 0.1 M glycine buffer, pH 3.0. Eluted fractions were immediately neutralized with 1 M Tris buffer, pH 9.0. For purification exploiting 3X FLAG, anti-FLAG M2 immunoaffinity column and 3X FLAG peptide (Sigma) were used for binding and elution, respectively. SDS-PAGE and Western blotting Samples were electrophoresed with a 5–20% acrylamide precast gel (Atto Corporation, Tokyo, Japan). Proteins on the gel were visualized with SimplyBlue SafeStain or SilverXpress Silver Staining Kit (Invitrogen). For Western blotting, electrophoresed proteins were transferred from the gel to PVDF membrane (Atto Corporation) and incubated sequentially with either monoclonal antibody TAG-6G4 or TAG-2E6, and anti-mouse IgG-horseradish peroxidase conjugate. Bands were visualized by ECL Advance (GE Healthcare) and captured on film.

double epitope-tagged protein as a calibrator for quantitation, human spermidine synthase-tagged doubly with 3X FLAG in the amino-terminus and the double epitope in the carboxy-terminus was expressed in 293T cells and immunoaffinity-purified with antiFLAG column. Addition of both cryptate-labeled TAG-6G4 and XL665-labeled TAG-2E6 to the double epitope-tagged SPDS solution caused stable fluorescence based on nonradiative energy transfer. Cryptate-labeled TAG-6G4 was used at the concentration of 100 ng/ml to compensate for the variability of optical quality of sample by ratio methodology (fluorescence intensity at 620 nm = 40,000) [9]. The optimal concentration of XL665-labeled TAG-2E6 was 2500 ng/ml to obtain an adequate range of quantitation. Representative calibration curves based on the double epitope-tagged SPDS calibrators of 3.9–2000 ng/ml are shown in Fig. 1. The fluorescence signals showed a dose–response relationship to double epitope-tagged SPDS concentration. Biological media such as culture supernatants and extracts of mammalian cells, insect cells, bacteria, and yeast had no or little effect on the calibration curve (Fig. 1). These results suggest that the homogeneous sandwich immunoassay can be used for detection of double epitope-tagged recombinant proteins in a wide variety of expression hosts whether the tagged proteins are secretory or nonsecretory. On the other hand, the sensitivity of the immunoassay using a pair of cryptate-labeled TAG-2E6 and XL665-labeled TAG-6G4 was much lower than that of the first developed immunoassay (data not shown). Therefore, we decided to use the sandwich immunoassay using cryptate-labeled TAG-6G4 and XL665-labeled TAG-2E6 for subsequent studies. Additionally, much lower readouts were observed if free collagen partial peptide (not fused to protein) was used as a calibrator. Presumably, the formation of an immune complex in the homogeneous solution was prevented by the threedimensional structure characteristic of the proline-rich 19-amino acid peptide fragment. Comparison of homogeneous immunoassay with Western blotting for detection of double epitope-tagged recombinant protein In the process of producing recombinant proteins, Western blotting is utilized to trace proteins and roughly estimate their amounts during the expression and purification steps. However, Western blotting requires many steps such as electrophoresis, protein transfer, and antibody reaction. Moreover, this technique is semiquantitative. We tried to compare the performance of our

Results and discussion 12000

Development of homogeneous immunoassay for double epitopetagged recombinant protein HTRF (Ratio)

Double epitope tag (GEPGDDGPSGAEGPPGPQG) was designed on the basis of a partial sequence of human type II collagen containing a neoepitope produced by matrix metalloprotease [14]. In order to develop a sandwich immunoassay measuring the double epitope tag, hybridomas secreting monoclonal antibodies TAG6G4 (IgG1j) and TAG-2E6 (IgG2bj) binding to amino-terminal region (GEPGDDGPS) and carboxyl-terminal region (GPPGPQG) of the double epitope tag, respectively, were established. The apparent affinities of these monoclonal antibodies for the double epitope tag were roughly decided by competitive ELISAs. The concentrations of free-collagen partial peptide (GEPGDDAPSGAEGP* PGPQG, P* was hydroxylated) for 50% inhibition of binding of labeled collagen partial peptide to TAG-6G4 and TAG-2E6 immobilized on polystyrene microtiter wells were 31.9 and 371.3 nM, respectively (data not shown). This result shows that TAG-6G4 has about 10 times greater affinity for the collagen partial peptide than TAG2E6. Monoclonal antibodies TAG-6G4 and TAG-2E6 were labeled with europium cryptate and XL665, respectively. To prepare a

10000 8000 6000 4000 2000 0 1

10

100

1000

10000

Double epitope tagged SPDS (ng/ml) Fig. 1. (A) Calibration curves of double epitope-tagged protein with homogeneous sandwich immunoassay based on fluorescence resonance energy transfer. Double epitope-tagged SPDS protein was serially diluted in Tris-BSA (s), Escherichia coli BL21 strain extract (D), culture supernatant of Saccharomyces cerevisiae BJ1191 strain (N), mammalian 293T cell culture supernatant (h), 293T cell extract (j), Spodoptera frugiperda Sf9 cell culture supernatant (e), and Sf9 cell extract (), and then measured with cryptate-labeled TAG-6G4 and XL665-labeled TAG-2E6. Each point represents a mean ± SD of three determinations.

Double epitope tag for protein quantitation / K. Enomoto et al. / Anal. Biochem. 380 (2008) 249–256

homogeneous immunoassay with that of Western blotting for detecting recombinant proteins. The double epitope-tagged SPDS was serially diluted twofold and then detected by both the homogeneous sandwich immunoassay and the Western blotting using the same antibodies, TAG-6G4 and TAG-2E6. As a detection reagent for Western blotting, GE Healthcare ECL Advance, which has a relatively high sensitivity, was used for comparison. Western blotting showed a single band with a molecular mass of 35 kDa corresponding to the double epitope-tagged SPDS protein in both blots (Fig. 2A). The detection limits of Western blotting using TAG-6G4 and TAG-2E6 were 313 and 625 pg/lane (31.3 and 62.5 ng/ml as sample protein concentrations), respectively. On the other hand, the fluorescence signal of the homogeneous immunoassay was elevated with increasing tagged protein concentration, but varied depending on the incubation time after addition of antibodies (Fig. 2B). To investigate the quantitation limit of the immunoassay, the imprecision profile of the immunoassay was calculated by 15 replicate measurements of the tagged SPDS solution in a single run at several incubation periods (Fig. 2C). The lowest tagged SPDS concentrations that were measured with an imprecision of <15% (quantitation limit) decreased with prolonging of the incubation time after addition of antibodies (15.6 ng/ml for 5 min, 7.8 ng/ml for 30 min, 3.9 ng/ml for 60 min, and 2.0 ng/ml for 120 min, respectively). If incubation time was further extended up to 24 h, the quantitation limit of the immunoassay did not differ from that for a 120-min incubation. Surprisingly, even if incubation time was only 5 min, the quantitation limit of the immunoassay was almost equal to the detection limit of Western blotting. These results suggest that the combination of double epitope tagging with homogeneous immunoassay detection is useful for rapid determination of the amount of recombinant protein in each step of protein production. Comparison of sandwich format and competitive format for detection of double epitope-tagged protein by homogeneous immunoassay The competitive immunoassay was also developed with XL665hapten I conjugate and cryptate-labeled TAG-6G4 for measuring

253

double epitope-tagged proteins. Mixing XL665-hapten I conjugate with cryptate-labeled TAG-6G4 caused signals based on FRET, which decreased on addition of the double epitope-tagged SPDS in dose-dependent manner (Fig. 3A). Another competitive immunoassay using XL665-hapten II and cryptate-labeled TAG-2E6 was also developed but was less sensitive (data not shown). Presumably, the difference of sensitivities between the two immunoassays was due to different affinities of their antibodies for the corresponding haptenic antigens. In order to compare the quantitation limit of the competitive format using XL665-hapten I conjugate and cryptate-labeled TAG-6G4 with that of the sandwich format, the imprecision profiles of the competitive immunoassay were also determined by the same method as the sandwich format. The imprecision profiles of the competitive format showed that 5min incubation could not determine even 1000 ng/ml tagged protein with an imprecision <15% (Fig. 3B). Even if the incubation time was prolonged to 24 h, the quantitation limit determined with the profile only reached the level equal to the detection limit of Western blotting (62.5 ng/ml, Fig. 3B). These results show that the incubation time required for the competitive format is longer than that for the sandwich format and the sensitivity of the competitive format is much lower than that of the sandwich format for detecting the double epitope-tagged proteins. Short polypeptide tags such as polyhistidine and FLAG probably cannot be measured by sandwich immunoassay because they are not thought to have two or more epitopes to which different antibodies can bind simultaneously. Such short tags would probably be difficult to measure sensitively using homogeneous immunoassay based on FRET. For large protein tags such as GST and MBP, sandwich immunoassays can be developed. However, as the efficiency of FRET is dependent on the distance between the fluorescent donor and the acceptor [15], sensitive detection based on FRET should not be easy unless two antibodies binding to the protein are in close proximity of each other. Although our double epitope tag is composed of only 19 amino acids, two different antibodies can simultaneously bind to the tag through their epitopes. Probably, the proximity of two monoclonal antibodies binding to the tag induces highly efficient FRET. Conse-

Double epitope tagged SPDS (ng/ml)

A

0

3.9

7.8

15.6

31.3

62.5

125

250

500

1000

TAG-6G4 TAG-2E6

B

C

14000 4000

30

10000

2000

8000 00

6000

1

10

100

C.V. (%)

HTRF (Ratio)

12000

15

4000 2000 0 0.1 1 10 100 1000 10000 Double epitope tagged SPDS (ng/ml)

0 0.1 1 10 100 1000 Double epitope tagged SPDS (ng/ml)

Fig. 2. Comparison of detection limit of homogeneous sandwich immunoassay with that of Western blotting. (A) Serially diluted double epitope-tagged SPDS protein was detected by Western blotting using TAG-6G4 or TAG-2E6. (B) The same samples were measured by the homogeneous sandwich immunoassay. Fluorescence signals were plotted at 5 min (s), 30 min (d), 1 h (D), 2 h (N), and 24 h (h) after addition of antibodies. Each point represents a mean ± SD of three determinations. (C) Imprecision profiles of the homogeneous sandwich immunoassay. The precision profile of the immunoassay at each incubation period was calculated from 15 determinations of each point in one assay. Symbols represent the same periods as B.

Double epitope tag for protein quantitation / K. Enomoto et al. / Anal. Biochem. 380 (2008) 249–256

A

7000

B

30

C.V. (%)

254

15

HTRF (Ratio)

6000 5000 4000 3000 2000 1000 0 1

10

100

1000

0 10

10000

Double epitope tagged SPDS (ng/ml)

100

1000

10000

Double epitope tagged SPDS (ng/ml)

Fig. 3. Homogeneous competitive immunoassay for double epitope-tagged protein. (A) Calibration curves. Serially diluted double epitope-tagged SPDS proteins were measured with cryptate-labeled TAG-6G4 and XL665-hapten I conjugate. Each point of the curves represents a mean ± SD of three determinations. (B) Imprecision profiles. The precision profiles of the competitive immunoassay were determined by the same method as Fig. 2C. Symbols represent the same periods as Fig. 2B.

quently, our homogeneous sandwich immunoassay makes possible more rapid and sensitive detection of double epitope-tagged proteins than Western blotting. Expression, detection, and purification of recombinant SPDS using the double epitope tag In order to investigate the utility of the double epitope tag in producing recombinant protein, SPDS was expressed and purified with this tag. The expression vector of SPDS tagged with the double epitope at the C-terminus was transfected into 293T cells (1  107 cells), and then 1 ml of the cell extract was harvested after 48 h. To measure the expression level of intracytoplasmic tagged SPDS, both cryptate-labeled TAG-6G4 and XL665-labeled TAG2E6 were added to a 100-fold diluted portion of the cell extract and fluorescence was measured after 30-min incubation. HTRF values of the extracts of mock and SPDS-transfected cells were 482 and 8946, respectively. The background signal determined using only the lysis buffer as a sample was 488. From the calibration curve, the concentration of the tagged SPDS protein in the original extract was estimated to be 70 lg/ml. These results demonstrate that double epitope tagging allows easy and prompt estimation of the expression level of recombinant protein. Next, purification

A

of the tagged SPDS was attempted from the cell extract by immunoaffinity chromatography using TAG-2E6. All of the cell extract of transfected cells was loaded onto the TAG-2E6-affinity column at near-neutral pH. After washing the column thoroughly, the bound proteins were eluted by changing the pH from near-neutral to acidic. All of the fractions collected during immunoaffinity chromatography (0.5 ml each) were monitored with both UV absorbance and the homogeneous sandwich immunoassay (30-min incubation). Increase of absorbance was observed only in the fractions passing through the column at near-neutral pH, whereas the fluorescence signal of the immunoassay increased only in a couple of fractions containing proteins eluted from the column just after changing the pH to acidic (Fig. 4A). Coomassie staining of SDSPAGE showed a single band corresponding to the molecular mass of the tagged SPDS in the fractions where a relatively high fluorescence was observed in the homogeneous immunoassay (Fig. 4B). Consequently, 63 lg of the tagged SPDS protein in 2.5 ml eluate (19.0 ml to 21.5 ml of elution volume of the chromatogram shown in Fig. 4A) had been purified in 90% yield. These results show that the immunoaffinity chromatography works well and the homogeneous immunoassay allows rapid monitoring of the elution of recombinant protein using the double epitope tagging. TAG-6G4 conjugated to agarose gel has not been considered suitable for

B

10000

0.20

1

2

3

6000 0.10 4000 0.05 2000

HTRF (Ratio)

Absorbance (280nm)

8000 0.15

kDa 97 66 45

0.00

0 0

30

5 10 15 18 19 20 21 22 Elution volume (ml)

Fig. 4. Immunoaffinity chromatography of double epitope-tagged SPDS protein. (A) Chromatograms of absorbance (s) and the homogeneous sandwich immunoassay (30min incubation, d) of column elution. An arrow indicates the point where pH was changed from 7.4 to 3.0. (B) SDS-PAGE analysis by Coomassie staining. Lane 1, molecular mass markers; Lane 2, elution fraction passing through the affinity column at pH 7.4; Lane 3, acidic elution fraction where the highest fluorescence signal was observed in the immunoassay.

Double epitope tag for protein quantitation / K. Enomoto et al. / Anal. Biochem. 380 (2008) 249–256

nal intensity of the immunoassay predicted a very low expression level of the scFv fragment, increase of the fluorescence signal of the immunoassay was observed in the acidic eluate of the immunoaffinity chromatography. In the elution fraction where the highest fluorescence signal was detected by the immunoassay, SDS-PAGE with the silver staining method showed a main band of 28.5 kDa, which was identical to the molecular mass of the scFv fragment (Fig. 5B). Sequentially, about 270 ng of the tagged scFv fragment in 3 ml eluate (22–24 ml of elution volume of the chromatogram shown in Fig. 5A) had been purified for 35% yield. Finally, the immunoreactivity of the purified scFv fragment was evaluated by enzyme immunoassay using the antigenic peptide immobilized on polystyrene microtiter wells. Background absorbance was 0.07 while the absorbance of the elution fraction reached over 2.0. From the results, 28.5-kDa protein purified with immunoaffinity chromatography was confirmed to be the active double epitope-tagged scFv fragment.

immunoaffinity support from the finding that elution of the tagged protein binding to the gel was insufficient, probably due to its high affinity for the double epitope tag (data not shown). Finally, the enzyme activity of the purified SPDS protein was estimated using our previously reported method by measuring 50 -deoxy-50 -methylthioadenosine, which was a by-product of the SPDS reaction [12]. The specific activity of the purified SPDS protein was 17.6 ± 1.4 nmol/ min/mg (n = 3), which was comparable to the activity of 3X FLAG-tagged SPDS which was purified mildly by competition of the binding to an anti-FLAG column between the tagged SPDS and 3X FLAG peptide (13.1 ± 0.3 nmol/min/mg, n = 3). These findings demonstrate that our double epitope tagging is useful for the production of recombinant SPDS protein without interfering with the functionality of the protein. Expression, detection, and purification of scFv fragment using the double epitope tag As another example of protein production, the scFv fragment recognizing a pentapeptide (L-Ala-D-isoGlu-L-Lys-D-Ala-D-Ala) related to bacterial cell wall biosynthesis was produced with the double epitope tag. Genes of variable regions of heavy chain and light chain were amplified from a hybridoma producing monoclonal antibody (IgG2aj) against the pentapeptide, and then the expression vector of the scFv fragment tagged with the double epitope at the C-terminus of the variable region of the heavy chain was constructed. Using the construct, the scFv fragment was expressed in Escherichia coli (100-ml culture), and then the amounts of the tagged scFv fragment contained in culture supernatant and periplasmic fraction (5.3 ml) were measured using the homogeneous sandwich immunoassay (60-min incubation). The fluorescence signal of the culture supernatant was equal to the background level as well as the fresh unused medium. On the other hand, the fluorescence signal of the periplasmic fraction was 4043, which suggested that the concentration of the scFv fragment in the fraction was 130 ng/ml (determined using the tagged SPDS protein as a calibrator). Immunoaffinity purification of the scFv fragment from the periplasmic fraction was performed in the same manner as the SPDS protein (fractionated to every 1 ml). Chromatograms generated by monitoring the absorbance and the immunoassay (60-min incubation) showed the same patterns as the purification of the double epitope-tagged SPDS protein (Fig. 5A). Although sig-

A

Concluding remarks A simple, rapid, and sensitive detection method of recombinant proteins has been successively developed using double epitope tagging in combination with HTRF technology. A noble affinity tag (GEPGDDGPSGAEGPPGPQG) was designed on the basis of a partial sequence of human type II collagen by consideration of the following features: (a) including two different epitopes which enables detection with sandwich immunoassay; (b) choosing as short a sequence as possible to obtain high FRET efficiency in homogeneous detection; (c) causing a minimal effect on tertiary structure and biological activity of the target protein. In our double epitope tagging system, the expressed protein can be sensitively detected by only 5-min incubation just after addition of a mixture of two kinds of fluorophore-labeled antibodies (their epitopes; GEPGDDGPS and GPPGPQG) to the sample. The quantitation limit of the 5-min immunoassay is comparable to the detection limit of Western blotting which requires labor-intensive processes. Another advantage is that components of the culture media and extracts of host cells generally used for protein expression have little effect on the immunoassay. Therefore, double epitope tagging should be useful for a variety of expression hosts. In this study, active SPDS protein and scFv fragment expressed in mammalian cells and E. coli,

5.0

B

8000

3.0 4000 2.0 2000 1.0

0.0

0 5

9

13

17

21

1

2 3 4

kDa

6000 HTRF (Ratio)

Absorbance (280nm)

4.0

1

255

98 64 50 36 22

25

Elution volume (ml) Fig. 5. Immunoaffinity chromatography of double epitope-tagged scFv fragment. (A) Chromatograms of absorbance (e) and the homogeneous sandwich immunoassay (60min incubation, ). An arrow indicates the point where the pH changed to acidic. (B) SDS-PAGE analysis by silver staining. Lanes 1 and 4, molecular mass markers; Lane 2, elution fraction not binding to the affinity column at pH 7.4; Lane 3, acidic elution fraction having the highest fluorescence signal in the immunoassay.

256

Double epitope tag for protein quantitation / K. Enomoto et al. / Anal. Biochem. 380 (2008) 249–256

respectively, were produced by a combination of the immunoassay and immunoaffinity chromatography with the double epitope tagging. If recombinant protein coexists with free double epitope peptide or any other protein containing the same sequence as the tag in an expression system, our immunoassay will probably overestimate the amount of the tagged protein. Comparing fluorescence signal of the cells transfected with a plasmid-encoding protein of interest with that of mock transfectant is effective in avoiding estimation error. The shortcoming of our tag detection system is that information cannot be obtained about the molecular mass of the fused protein. Even if the amount of the double epitope-tagged protein can be traced with the immunoassay, Western blotting and SDSPAGE are necessary for confirmation of the molecular size and purity of the protein at least once during the overall process. However, the double epitope tagging makes it possible to alleviate laborious and troublesome processes of conventional protein production. References [1] M. Forstner, L. Leder, L.M. Mayr, Optimization of protein expression systems for modern drug discovery, Exp. Rev. Proteomics 4 (2007) 67–78. [2] K. Terpe, Overview of tag protein fusions: from molecular and biochemical fundamentals to commercial systems, Appl. Microbiol. Biotechnol. 60 (2003) 523–533. [3] J. Arnau, C. Lautitzen, G.E. Petersen, J. Pedersen, Current strategies for the use of affinity tags and tag removal for the purification of recombinant proteins, Protein Expression Purif. 48 (2006) 1–13. [4] E. Hochuli, H. Dobeli, A. Schacher, New metal chelate absorbent selective for proteins and peptides containing neighbouring histidine residues, J. Chromatogr. 422 (1987) 177–184.

[5] B.L. Brizzard, R.G. Chubet, D.L. Vizard, Immunoaffinity purification of FLAG peptide-tagged bacterial alkaline phosphatase using novel monoclonal antibody and peptide elution, Biotechniques 16 (1994) 730–735. [6] G.I. Evan, G.K. Lewis, G. Ramsay, J.M. Bishop, Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product, Mol. Cell. Biol. 5 (1985) 3610–3616. [7] D.B. Smith, K.S. Johnson, Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione-S-transferase, Gene 67 (1988) 32–40. [8] C. di Guan, P. Li, P.D. Riggs, H. Inouye, Vectors that facilitate the expression and purification of foreign peptides in Escherichia coli by fusion to maltose-binding protein, Gene 67 (1988) 21–30. [9] H. Bazin, M. Préaudat, E. Trinquet, G. Mathis, Homogeneous time resolved fluorescence resonance energy transfer using rare earth cryptates as a tool for probing molecular interactions in biology, Spectrochim. Acta A 57 (2001) 2197–2211. [10] G. Kohler, C. Milstein, Continuous cultures of fused cells secreting antibody of predefined specificity, Nature 256 (1975) 495–497. [11] K. Enomoto, A. Araki, T. Nakajima, H. Ohta, K. Dohi, M. Préaudat, P. Seguin, G. Mathis, R. Suzuki, G. Kominami, H. Takemoto, High-throughput miniaturized immunoassay for human interleukin-13 secreted from NK3.3 cells using homogenous time-resolved fluorescence, J. Pharm. Biomed. Anal. 28 (2002) 73–79. [12] K. Enomoto, T. Nagasaki, A. Yamauchi, J. Onoda, K. Sakai, T. Yoshida, K. Maekawa, Y. Kinoshita, I. Nishino, S. Kikuoka, T. Fukunaga, K. Kawamoto, Y. Numata, H. Takemoto, K. Nagata, Development of high-throughput spermidine synthase activity assay using homogeneous time-resolved fluorescence, Anal. Biochem. 351 (2006) 229–240. [13] F. Cao, M.P. Badtke, L.M. Metzger, E. Yao, B. Adeyemo, Y. Gong, J.E. Tavis, Identification of an essential molecular contact point on the duck hepatitis B virus reverse transcriptase, J. Virol. 79 (2005) 10164–10170. [14] R.C. Billinghurst, L. Dahlberg, M. Ionescu, A. Reiner, R. Bourne, C. Rorabeck, P. Mitchell, J. Hambor, O. Diekmann, H. Tschesche, J. Chen, H.V. Wart, A.R. Poole, Enhanced cleavage of Type II collagen by collagenases in osteoarthritic articular cartilage, J. Clin. Invest. 99 (1997) 1534–1545. [15] T. Förster, Zwischen molekulare Energiewanderung und Fluoreszenz, Ann. Phys. 2 (1948) 55–75.