- Email: [email protected]
A fusion protein expression analysis using surface plasmon resonance imaging
ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 330 (2004) 251–256 www.elsevier.com/locate/yabio
A fusion protein expression analysis using surface plasmon resonance imaging Jin-Mi Jung,a,b Yong-Beom Shin,a Min-Gon Kim,a Hyeon-Su Ro,a Hee-Tae Jung,b and Bong Hyun Chunga,¤ a
Korea Research Institute of Bioscience and Biotechnology, BioNanotechnology Research Center, P.O. Box 115, Yuseong, Daejeon 305-600,Republic of Korea b Laboratory of Organic Opto Electronic Materials, Korea Advanced Institute of Science and Technology, Daejeon 305-701, Republic of Korea Received 24 December 2003 Available online 25 May 2004
Abstract A surface plasmon resonance (SPR) imaging system was constructed and used to detect the aYnity-tagged recombinant proteins expressed in Escherichia coli. With regards to model proteins, the hexahistidine-ubiquitin-tagged human growth hormone (His6-UbhGH), glutathione S-transferase-tagged human interleukin-6 (GST-hIL6), and maltose-binding protein-tagged human interleukin-6 (MBP-hIL6) expressed in E. coli were analyzed. The cell lysates were spotted on gold thin Wlms coated with 11-mercaptoundecanol (MUOH)/dextran derivatized with Ni(II)-iminodiacetic acid (IDA-Ni(II)), glutathione, or cyclodextrin. After a brief washing of the gold chip, SPR imaging measurements were carried out in order to detect the bound aYnity-tagged fusion proteins. Using this new approach, rapid high-throughput expression analysis of the aYnity-tagged proteins were obtained. The SPR imaging protein chip system used to measure the expression of aYnity-tagged proteins in a high-throughput manner is expected to be an attractive alternative to traditional laborious and time-consuming methods, such as SDS–polyacrylamide gel electrophoresis (SDS–PAGE) and Western blots. 2004 Elsevier Inc. All rights reserved. Keywords: AYnity-tagged protein; Expression analysis; Gold chip; Surface plasmon resonance (SPR) imaging
While the completion of the human genome sequencing project has provided a huge amount of genetic data, nowadays more attention is being paid to protein than to the gene itself. High-throughput strategies, such as rapid protein expression and puriWcation in parallel, are essential when conducting functional and structural studies of proteins derived from a number of genes [1]. Moreover, for high-throughput expression and puriWcation, aYnity fusion tags are commonly used to facilitate the puriWcation of a number of proteins expressed simultaneously. Such fusion proteins can often be puriWed to near homogeneity from crude mixtures using a singlestep aYnity chromatography [2]. In addition to the rapid puriWcation of proteins, the fast detection of expressed proteins is also of enormous importance in high¤
Corresponding author. Fax: +82-42-879-8594. E-mail address: [email protected] (B.H. Chung).
0003-2697/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2004.02.009
throughput protein expression. However, many still rely on traditional methods such as SDS–PAGE and Western blots, which are both time consuming and labor intensive, and sometimes use hazardous materials. Surface plasmon resonance (SPR)1 is an optical technique used to detect the speciWc binding of unlabeled biomolecules onto molecules attached to chemically modiWed gold thin Wlms by measuring changes in the index of refraction upon adsorption [3,4]. Furthermore, it has also been used as a valuable tool to investigate molecular interactions in real time without the use of 1 Abbreviations used: IDA, iminodiacetic acid; GST-hIL6, glutathione S-transferase-tagged human interleukin-6; His6-Ub-hGH, hexahistidine-ubiquitin-tagged human growth hormone; IPTG, isopropyl -D-thiogalactopyranoside; MBP-hIL6, maltose-binding proteintagged human interleukin-6; MUOH, 11-mercaptoundecanol; SDS– PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; SPR, surface plasmon resonance; BSA, bovine serum albumin.
252
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
labels [5]. SPR sensors have also been used for the rapid monitoring of recombinant proteins [6]. The SPR technique can be expanded to SPR imaging and used for the high-throughput analysis of bioaYnity interactions by fabricating protein arrays on gold surfaces. SPR imaging utilizes collimated illumination of the entire surface, with the reXected beam imaged onto a two-dimensional array detector [7]. Since the SPR imaging protein array requires no labels for the detection of capture proteins, it is considered as an ideal surface-sensitive optical technique that directly detects the multiple interactions of molecules on a two-dimensional gold surface. SPR imaging has successfully been applied to the measurements of bioaYnity interactions using DNA [8–10], peptide [11], protein [12], and carbohydrate arrays [13]. In this study, we introduce an eYcient and robust technique to rapidly detect the hexahistidine (His6)-, glutathione-S-transferase (GST)-, and maltose-binding protein (MBP)-fused proteins expressed in Escherichia coli using SPR imaging measurement.
Materials and methods Construction of plasmids A gene encoding human growth hormone was ampliWed using a polymerase chain reaction (PCR) from a human cDNA library. The 3⬘ terminus was designed to contain a SalI restriction enzyme cleavage site. The PCR product was puriWed using a DNA puriWcation kit (Qiagen) and digested with the restriction enzyme SalI. To introduce hexahistidine and ubiquitin tags at the Nterminus of the human growth hormone (His6-UbhGH), a gene encoding ubiquitin was ampliWed using the 5⬘ primer (CAT ATG CAC CAC CAC CAC CAC CAC CAA ATT TTC GTC AAA ACT CTA ACA) and the 3⬘ primer (ACC ACC CCT CAA CCT CAA GAC GAG GTG AAG AGT) from the genomic DNA of Saccaromyces cerevisiae (Invitrogen). The PCR product was puriWed and digested with the restriction enzyme NdeI. The resulting DNA fragments were ligated with a pET22b vector (Novagen), that was predigested with the same restriction enzymes, using a ligation kit (Takara, Japan) (pEHUb-hGH) (Fig. 1). Meanwhile, to construct a MBP-fused human interleukin-6 (MBP-hIL6) expression vector, a gene encoding MBP was PCR-ampliWed from a pMal-c2x plasmid (New England Biolab.) using the primers containing the NdeI site at the N-terminus and the EcoRI site at the C-terminus. The PCR product was puriWed and digested with the restriction enzymes NdeI and EcoRI. The resulting DNA fragment was coligated into the NdeI and SalI sites of the pET22b vector with a DNA fragment encoding human interleukin6, which was PCR-ampliWed from pBKS-hIL6 (Cytokine Bank, Chunbuk Univ., Korea) using the primers
Fig. 1. Physical map of the plasmids.
containing the EcoRI site at the N-terminus and the SalI site at the C-terminus (pEMBP-hIL6) (Fig. 1). The GST gene was obtained from the pET41b vector (Novagen) by digesting the vector with the restriction enzymes NdeI and EcoRI. The GST-fused human interleukin-6 (GSThIL6) expression vector was constructed by replacing the NdeI and EcoRI fragment of pEMBP-hIL6 with the GST-coding DNA fragment obtained through the digestion of a pET41b vector (Novagen, USA) with the NdeI and EcoRI (pEGST-hIL6) (Fig. 1). Expression of recombinant fusion proteins The transformed cells, Escherichia coli BL21(DE3)/ pEHUb-hGH, E. coli BL21(DE3)/pEMBP-hIL6, and E.coli BL21(DE3)/pEGST-hIL6, were cultured in Luria– Bertani (LB) media and allowed to grow at 30 °C to OD 0.6 before 1 mM isopropyl -D-thiogalactopyranoside (IPTG) was added. After IPTG induction, the cells were cultured for an additional 4 h followed by centrifugation at 5000 rpm for 5 min. The resulting pellets were suspended in 50 mM Tris–HCl buVer (pH 8.0) and broken up by sonication for 5 min. After sonication, the cell lysate was again centrifuged at 5000 rpm for 5 min. SDS–PAGE analyses of fusion proteins The sample solution (10 l) was mixed with 10 l of the sample buVer (2£ containing 125 mM Tris–HCl (pH 6.8), 4% SDS, 20% glycerol, 10% -mercaptoethanol, and 1 g/ml of bromophenol blue (BPB). The mixture was boiled for 3 min and loaded to the well of a polyacrylamide gel. The proteins were separated using a 12% SDS–polyacrylamide gel electrophoresis and visualized by Coomassie brilliant blue R-250 staining.
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
Surface modiWcation of gold chips for the speciWc binding of fusion proteins A gold chip (2 nm of chromium as an adhesion layer and 45 nm of gold deposited on an 18 £ 18 £ 0.3 mm glass) was obtained from K-MAC (Korea). The gold chip was cleaned with a solution of H2SO4/H2O2 (3 v/v) at 50 °C for 30 min. It was then immersed in a 10 mM ethanolic solution of 11-mercaptoundecanol (MUOH) at room temperature for 20–24 h to assure the formation of a MUOH self-assembled monolayer on the gold surface. It was then washed with ethanol and water and sonicated for 5 min in ethanol. To activate the hydroxyl group of the MUOH, the gold chip was treated with 0.6 M epichlorohydrin in a 1:1 mixture of 0.4 M NaOH and 2-methoxyethyl ether for 4 h. After being washed with ethanol and water, the gold chip was immersed in a 0.3 mg/ml dextran solution in 0.1 M NaOH for 20 h. The dextran-coated surface was once again treated with the 0.6 M solution of the epichlorohydrin in the 1:1 mixture of the 0.4 M NaOH and 2-methoxyethyl ether for 4 h to activate the dextran’s hydroxyl group. This epichlorohydrin-activated dextran chip was further functionalized in order to give it a speciWc aYnity for the recombinant fusion proteins. It was treated with a 1.7 M solution of iminodiacetic acid (IDA) in 2 M sodium carbonate at 60 °C for 20 h, a solution of 0.1 mg/ml L-glutathione (reduced form) in pH 7.0, and 44 mM phosphate buVer at 40 °C for 20 h (glutathione chip), and a 70 mg/ml solution of -cyclodextrin in 0.1 M NaOH at 40 °C for 20 h (cyclodextrin chip), respectively. The IDA-functionalized chip, after being immersed in a 50 mM nickel(II) chloride solution for 3–4 h and washed with water, was used as a hexahistidine fusion protein-binding chip (IDA-Ni(II) chip). The glutathione and cyclodextrin chips, after having been treated with 1 M ethanolamine at 37 °C for 3–4 h and washed with water, were used respectively for the detection of GST and MBP fusion proteins. Fabrication microarray
and
detection
of
recombinant
protein
Recombinant fusion proteins, as a cell lysate or in a partially puriWed form, were diluted with a spotting buVer (50 mM Tris, 20 % (v/v) glycerol or ethylene glycol, pH 8.0). A microarrayer (Proteogen, Korea) was then used to print the recombinant fusion proteins on the surface-modiWed gold chip. Printing was carried out in an atmospherically controlled chamber with a relative humidity of 70–80% at room temperature. A Stealth pin (Model SMP-10, TeleChem.) was used for arraying spots of 335 m in diameter. After spotting, the gold slides were incubated for 30 min in the microarrayer and rinsed with a washing solution and DW. The washing solution was a 20 mM phosphate buVer containing 0.5 M NaCl
253
and 20 mM imidazole at pH 8.0 for the IDA-Ni(II) chip, a PBS buVer at pH 7.2 for the glutathione chip, and a 20 mM Tris buVer containing 200 mM NaCl and 1 mM EDTA at pH 7.4 for the cyclodextrin chip. After the gold chip was dried through centrifugation, it was then coupled with BK7 prism using index-matching oil and placed on the center of the goniometer. The CCD camera took reXected images at a Wxed incident angle. Construction of an SPR imaging system A two-dimensional surface plasmon resonance imaging system was constructed to obtain microarray images of the protein spots. A 150-W quartz tungsten-halogen lamp (Schott, Germany) was used as the light source and the light was delivered to a goniometer arm (Physik Instrumente, Germany) using a liquid light guide (Oriel Instruments, USA). The light collimated by the lenses was passed through a narrow interference Wlter (647.1 nm, D 1 nm; Oriel Instruments) and a polarizer (Newport, USA) in order to convert a monochromatic and a linear polarized beam, respectively. The diameter of this beam was about 2 cm. The gold sensor chip on which proteins were spotted was optically coupled with a prism coupler (Korea Electro-optics, Korea) via an index matching oil (nD D 1.517) and placed on the center of the goniometer. The goniometer was electronically controlled by a DC Servo motor controller (Physik Instrumente). The reXected images from the gold chip were taken by a 1/2 inch charge coupled device (CCD) camera (Sony, Japan), while the contrast images were monitored by a personal computer. A combination of lenses was placed in front of the bare CCD chip in order to obtain a clear image. The images were then stored digitally in a personal computer using a B/W frame grabber (National Instrument, USA).
Results and discussion The hexahistidine-ubiquitin-tagged hGH (His6-UbhGH), which was expressed in a soluble form, was partially puriWed using a one-step metal aYnity chromatography, and analyzed using both SPR imaging and SDS–PAGE (Fig. 2). The SPR images were captured at an incident angle that was lower than the SPR angle of the background surface. The brighter spots are indications of the aYnity binding of the target proteins on the IDA-Ni(II) chip. Protein microarray images with varying protein concentrations are shown in Fig. 2B. The brightness of the SPR image increased as the protein concentration increased. As a control, bovine serum albumin (BSA) with a concentration of 100 g/ml was spotted on the IDA-Ni(II) gold chip. As shown in Fig. 2B, the control spots are almost invisible on the SPR image, thus indicating that the protein binding is
254
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
Fig. 2. SDS–PAGE (A) and SPR imaging (B) analyses of partially puriWed His6-Ub-hGH fusion protein at diVerent concentrations: lane M, protein maker; lane 1, 0.5 mg/ml; lane 3, 0.25 mg/ml; lane 4, 0.1 mg/ml; lane 5, 0.05 mg/ml; lane 6, 0.01 mg/ml; lane 7, 0.005 mg/ml; lane 8, 0.0025 mg/ml; lane 9, 0.001 mg/ml.
Fig. 3. Plot of relative spot intensity versus His6-Ub-hGH concentration.
speciWc. The intensity diVerence, which was the average value of the spot intensity minus the background intensity calculated from the image in Fig. 2B using an image analysis software (Scion, USA), was plotted against the concentration of His6-Ub-hGH (Fig. 3). The linear dynamic range was found to be 10–75 g/ml, which was relatively narrow compared to a SPR sensor (1–200 g/ ml) [6]. The lowest visible protein concentration was found to be about 10 g/ml in the SPR imaging, whereas the protein band in the SDS–PAGE was visible at 5 g/ ml. This indicates that the SPR imaging measurement is slightly less sensitive than the SDS–PAGE. The sample volumes used for the SDS–PAGE and SPR imaging were 20 l and 3.9 nl (data from catalogue of TeleChem.), respectively. The detection limit, which is based on the protein mass, of the SPR imaging measurement is 0.04 ng/protein spot, which is much less than that the
(100 ng/protein band) of the SDS–PAGE. The location of the protein in the polyacrylamide gel is determined by either Coomassie blue staining or silver staining. In general, the former is more frequently used because of its ease of handling and rapid detection. However, silver staining methods are used to detect smaller amounts of proteins because they are considerably more sensitive. The detection limit of silver staining is 2 to 5 ng/protein band [2], which is still much higher than that of the SPR imaging. The SPR imaging sensor based on intensity measurement was calculated to exhibit 1/20 detection sensitivity to an angular interrogation-based SPR sensor [15]. The SPR sensor is known to be able to detect less than 1 g/ ml of protein [6]. Thus the detection limit achieved by our SPR imaging system (10 g/ml) is comparable to the theoretically calculated sensitivity. Low sensitivity can limit the use of SPR imaging, especially in cases where a low-level expression of recombinant proteins is present. In general, however, the concentration of recombinant proteins expressed in E. coli is usually enough to be analyzed by the SPR imaging sensor. Furthermore, the SPR imaging system has a great advantage over the existing SPR sensors in that it can detect multiple molecular interactions in array formats on a two-dimensional gold thin Wlm. Using SPR imaging, a direct analysis of the recombinant proteins without puriWcation was carried out. After recombinant E. coli cells expressing His6-Ub-hGH were lysed by sonication, the cell lysates were directly spotted onto an IDA-Ni(II) chip. After 30 min incubation of the spotted gold chip to induce the aYnity binding between His6-Ub-hGH and IDA-Ni(II), the gold chip was washed with the washing solution and deionized distilled water. After the gold chip was dried, the expressed protein spots bound on the IDA-Ni(II) chip were measured
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
Fig. 4. SDS–PAGE (A) and SPR imaging (B) analyses of E. coli cell lysates expressing His6-Ub-hGH fusion protein: lane M, protein marker; lane 1, untransformed cell lysates; lane 2 and 2⬘, transformed cell lysates, uninduced; lane 3 and 3⬘, transformed cell lysates, induced with IPTG addition.
using SPR imaging. Fig. 4 shows the result of the SDS– PAGE and SPR imaging analyses. Compared to the partially puriWed His6-Ub-hGH (Fig. 2A), a number of protein bands were displayed on the SDS–PAGE gel. Lanes 2 and 3 in the SDS–PAGE correspond to lanes 2⬘ and 3⬘ in the SPR imaging, respectively. Uninduced cells were also found to produce a small amount of His6-Ub-hGH because of the leaky translation that was often observed in the T7 promoter system. The brighter SPR spot images were observed with induced cell lysates, whereas the dim images were observed with uninduced cell lysates. Along with hexahistidine residues, GST and MBP are also widely used as an aYnity tag and oVer the possibility of purifying the recombinant proteins through the use of aYnity chromatographies. When the target protein is fused to a highly soluble partner such as the GST or MBP moiety, correctly folded and soluble heterologous proteins are produced intracellularly in the bacterial cytoplasm [14]. Human interleukin 6 fused with GST (GST-hIL6) and MBP (MBP-hIL6) was expressed in E. coli in a soluble form, and then the cell lysates were directly spotted onto a glutathione or cyclodextrin-functionalized gold chip. Figs. 5 and 6 show the SDS–PAGE and SPR imaging analyses for GST-hIL6 and MBPhIL6, respectively. Lanes 2 and 3 in the SDS–PAGE correspond to lanes 2⬘ and 3⬘ in the SPR imaging, respectively. Unlike the His6-Ub-hGH expression, the GST-hIL6 and MBP-hIL6 expression in uninduced cells was hardly detected on the SDS–PAGE (Lane 2 in Figs. 5A and 6A). The brighter SPR images were clearly observed with induced cell lysates, whereas no visible images were obtained with uninduced cell lysates (Figs. 5B and 6B). This demonstrates that the SPR imaging analysis corresponds to the SDS–PAGE quantitatively. The sequencing of the genomes of various organisms has led to the concept of analyzing protein function on a genome-wide scale. Recently, rapid progress in highthroughput and parallel approaches in protein expres-
255
Fig. 5. SDS–PAGE (A) and SPR imaging (B) analyses of E. coli cell lysates expressing GST-hIL6 fusion protein: lane M, protein marker; lane 1, untransformed cell lysates; lane 2 and 2⬘, transformed cell lysates, uninduced; lane 3 and 3⬘, transformed cell lysates, induced with IPTG addition.
Fig. 6. SDS–PAGE (A) and SPR imaging (B) analyses of E. coli cell lysates expressing MBP-hIL6 fusion protein: lane M, protein marker; lane 1, untransformed cell lysates; lane 2 and 2⬘, transformed cell lysates, uninduced; lane 3 and 3⬘, transformed cell lysates, induced with IPTG addition.
sion, puriWcation, and detection has been made to study proteins on a global scale [16,17]. So far, many researches have focused on the development of robust techniques for high throughput and the parallel expression and puriWcation of multiple proteins. However, the detection of multiple proteins, which are often expressed in 96-well microtiter plates for automation, relies on traditional methods such as the SDS–PAGE and Western blots. In this study, therefore, we introduce a novel approach to the rapid and simultaneous detection of multiple recombinant proteins expressed in E. coli with aYnity tags to facilitate puriWcation. This method enables on-chip puriWcation, label-free detection, and on-chip quantitative analyses of target proteins. It is thus expected that the SPR imaging measurement of protein expression can be an attractive alternative to traditional methods such as the SDS–PAGE and Western blots.
256
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
Acknowledgments This research was supported by a grant from the KRIBB Initiative Research Program. This work was also supported in part by a grant from KOSEF through the Center for Advanced Bioseparation Technology (BSEP).
References [1] J.S. Albala, K. Franke, I.R. McConnel, K.L. Pak, P.A. Folta, B. Karlak, B. Rubinfeld, A.H. Davies, G.G. Lennon, R. Clarket, From genes to proteins: high-throughput expression and puriWcation of the human proteome, J. Cell. Biochem. 80 (2000) 187–191. [2] F.M. Ausubel, R. Brent, R.E. Kingston, et al., Short protocols in molecular biology, second ed., Wiley, New York, 1992. [3] S.R. Karlsen, K.S. Johnston, R.C. Jorgenson, S.S. Yee, Simultaneous determination of refractive index and absorbance spectra of chemical samples using surface plasmon resonance, Sens. Actuators B 25 (1995) 747–749. [4] D. Neri, S. Montigiani, P.M. Kirkham, Biophysical methods for the determination of antibody-antigen aYnities, Trends Biotechnol. 14 (1996) 465–470. [5] M. Fivash, E.M. Towler, R.J. Fisher, BIAcore for macromolecular interaction, Curr. Opin. Biotechnol. 9 (1998) 97–101. [6] K.N. Baker, M.H. Rendall, A. Patel, P. Boyd, M. Hoare, R.B. Freedman, D.C. James, Rapid monitoring of recombinant protein products: a comparison of current technologies, Trends Biotechnol. 20 (2002) 149–155.
[7] C.E. Jordan, R.M. Corn, Surface plasmon resonance imaging measurements of electrostatic biopolymer adsorption onto chemically modiWed gold surfaces, Anal. Chem. 69 (1997) 1449–1456. [8] E.A. Smith, M.G. Erickson, A.T. Ulijasz, B. Weisblum, R.M. Corn, Surface plasmon resonance imaging of transcription factor proteins: interactions of bacterial response regulators with DNA arrays on gold Wlms, Langmuir 19 (2003) 1486–1492. [9] E.A. Smith, M. Kyo, H. Kumasawa, K. Nakatani, I. Saito, R.M. Corn, Chemically induced hairpin formulation in DNA monolayers, J. Am. Chem. Soc. 124 (2002) 6810–6811. [10] H.J. Lee, T.T. Goodrich, R.M. Corn, SPR imaging measurement of 1-D and 2-D DNA microarrays created from microXuidic channels on gold thin Wlms, Anal. Chem. 73 (2001) 5525–5531. [11] G.J. Wegner, H.J. Lee, R.M. Corn, Characterization and optimization of peptide arrays for the study of epitope-antibody interactions using surface plasmon resonance imaging, Anal. Chem. 74 (2002) 5161–5168. [12] G.J. Wegner, H.J. Lee, G. Marriott, R.M. Corn, Fabrication of histidine-tagged fusion protein arrays for surface plasmon resonance imaging studies of protein-protein and protein-DNA interactions, Anal. Chem. 75 (2003) 4740–4746. [13] E.A. Smith, W.D. Thomas, L.L. Kiessling, R.M. Corn, Surface plasmon resonance imaging studies of protein-carbohydrate interactions, J. Am. Chem. Soc. 125 (2003) 6140–6148. [14] P.A. Nygren, S. Stahl, M. Uhlen, Engineering proteins to facilitate bioprocessing, Trends Biotechnol. 12 (1994) 184–188. [15] J. Homola, S.S. Yee, G. Gauglitz, Surface plasmon resonance sensors: review, Sens. Actuators B 54 (1999) 3–15. [16] S.P. Chamber, High-throughput protein expression for the postgenomic era, Drug Discovery Today 7 (2002) 759–765. [17] S. Yokoyama, Protein expression system for structural genomics and proteomics, Curr. Opin. Chem. Biol. 7 (2003) 39–43.
A fusion protein expression analysis using surface plasmon resonance imaging Jin-Mi Jung,a,b Yong-Beom Shin,a Min-Gon Kim,a Hyeon-Su Ro,a Hee-Tae Jung,b and Bong Hyun Chunga,¤ a
Korea Research Institute of Bioscience and Biotechnology, BioNanotechnology Research Center, P.O. Box 115, Yuseong, Daejeon 305-600,Republic of Korea b Laboratory of Organic Opto Electronic Materials, Korea Advanced Institute of Science and Technology, Daejeon 305-701, Republic of Korea Received 24 December 2003 Available online 25 May 2004
Abstract A surface plasmon resonance (SPR) imaging system was constructed and used to detect the aYnity-tagged recombinant proteins expressed in Escherichia coli. With regards to model proteins, the hexahistidine-ubiquitin-tagged human growth hormone (His6-UbhGH), glutathione S-transferase-tagged human interleukin-6 (GST-hIL6), and maltose-binding protein-tagged human interleukin-6 (MBP-hIL6) expressed in E. coli were analyzed. The cell lysates were spotted on gold thin Wlms coated with 11-mercaptoundecanol (MUOH)/dextran derivatized with Ni(II)-iminodiacetic acid (IDA-Ni(II)), glutathione, or cyclodextrin. After a brief washing of the gold chip, SPR imaging measurements were carried out in order to detect the bound aYnity-tagged fusion proteins. Using this new approach, rapid high-throughput expression analysis of the aYnity-tagged proteins were obtained. The SPR imaging protein chip system used to measure the expression of aYnity-tagged proteins in a high-throughput manner is expected to be an attractive alternative to traditional laborious and time-consuming methods, such as SDS–polyacrylamide gel electrophoresis (SDS–PAGE) and Western blots. 2004 Elsevier Inc. All rights reserved. Keywords: AYnity-tagged protein; Expression analysis; Gold chip; Surface plasmon resonance (SPR) imaging
While the completion of the human genome sequencing project has provided a huge amount of genetic data, nowadays more attention is being paid to protein than to the gene itself. High-throughput strategies, such as rapid protein expression and puriWcation in parallel, are essential when conducting functional and structural studies of proteins derived from a number of genes [1]. Moreover, for high-throughput expression and puriWcation, aYnity fusion tags are commonly used to facilitate the puriWcation of a number of proteins expressed simultaneously. Such fusion proteins can often be puriWed to near homogeneity from crude mixtures using a singlestep aYnity chromatography [2]. In addition to the rapid puriWcation of proteins, the fast detection of expressed proteins is also of enormous importance in high¤
Corresponding author. Fax: +82-42-879-8594. E-mail address: [email protected] (B.H. Chung).
0003-2697/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2004.02.009
throughput protein expression. However, many still rely on traditional methods such as SDS–PAGE and Western blots, which are both time consuming and labor intensive, and sometimes use hazardous materials. Surface plasmon resonance (SPR)1 is an optical technique used to detect the speciWc binding of unlabeled biomolecules onto molecules attached to chemically modiWed gold thin Wlms by measuring changes in the index of refraction upon adsorption [3,4]. Furthermore, it has also been used as a valuable tool to investigate molecular interactions in real time without the use of 1 Abbreviations used: IDA, iminodiacetic acid; GST-hIL6, glutathione S-transferase-tagged human interleukin-6; His6-Ub-hGH, hexahistidine-ubiquitin-tagged human growth hormone; IPTG, isopropyl -D-thiogalactopyranoside; MBP-hIL6, maltose-binding proteintagged human interleukin-6; MUOH, 11-mercaptoundecanol; SDS– PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; SPR, surface plasmon resonance; BSA, bovine serum albumin.
252
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
labels [5]. SPR sensors have also been used for the rapid monitoring of recombinant proteins [6]. The SPR technique can be expanded to SPR imaging and used for the high-throughput analysis of bioaYnity interactions by fabricating protein arrays on gold surfaces. SPR imaging utilizes collimated illumination of the entire surface, with the reXected beam imaged onto a two-dimensional array detector [7]. Since the SPR imaging protein array requires no labels for the detection of capture proteins, it is considered as an ideal surface-sensitive optical technique that directly detects the multiple interactions of molecules on a two-dimensional gold surface. SPR imaging has successfully been applied to the measurements of bioaYnity interactions using DNA [8–10], peptide [11], protein [12], and carbohydrate arrays [13]. In this study, we introduce an eYcient and robust technique to rapidly detect the hexahistidine (His6)-, glutathione-S-transferase (GST)-, and maltose-binding protein (MBP)-fused proteins expressed in Escherichia coli using SPR imaging measurement.
Materials and methods Construction of plasmids A gene encoding human growth hormone was ampliWed using a polymerase chain reaction (PCR) from a human cDNA library. The 3⬘ terminus was designed to contain a SalI restriction enzyme cleavage site. The PCR product was puriWed using a DNA puriWcation kit (Qiagen) and digested with the restriction enzyme SalI. To introduce hexahistidine and ubiquitin tags at the Nterminus of the human growth hormone (His6-UbhGH), a gene encoding ubiquitin was ampliWed using the 5⬘ primer (CAT ATG CAC CAC CAC CAC CAC CAC CAA ATT TTC GTC AAA ACT CTA ACA) and the 3⬘ primer (ACC ACC CCT CAA CCT CAA GAC GAG GTG AAG AGT) from the genomic DNA of Saccaromyces cerevisiae (Invitrogen). The PCR product was puriWed and digested with the restriction enzyme NdeI. The resulting DNA fragments were ligated with a pET22b vector (Novagen), that was predigested with the same restriction enzymes, using a ligation kit (Takara, Japan) (pEHUb-hGH) (Fig. 1). Meanwhile, to construct a MBP-fused human interleukin-6 (MBP-hIL6) expression vector, a gene encoding MBP was PCR-ampliWed from a pMal-c2x plasmid (New England Biolab.) using the primers containing the NdeI site at the N-terminus and the EcoRI site at the C-terminus. The PCR product was puriWed and digested with the restriction enzymes NdeI and EcoRI. The resulting DNA fragment was coligated into the NdeI and SalI sites of the pET22b vector with a DNA fragment encoding human interleukin6, which was PCR-ampliWed from pBKS-hIL6 (Cytokine Bank, Chunbuk Univ., Korea) using the primers
Fig. 1. Physical map of the plasmids.
containing the EcoRI site at the N-terminus and the SalI site at the C-terminus (pEMBP-hIL6) (Fig. 1). The GST gene was obtained from the pET41b vector (Novagen) by digesting the vector with the restriction enzymes NdeI and EcoRI. The GST-fused human interleukin-6 (GSThIL6) expression vector was constructed by replacing the NdeI and EcoRI fragment of pEMBP-hIL6 with the GST-coding DNA fragment obtained through the digestion of a pET41b vector (Novagen, USA) with the NdeI and EcoRI (pEGST-hIL6) (Fig. 1). Expression of recombinant fusion proteins The transformed cells, Escherichia coli BL21(DE3)/ pEHUb-hGH, E. coli BL21(DE3)/pEMBP-hIL6, and E.coli BL21(DE3)/pEGST-hIL6, were cultured in Luria– Bertani (LB) media and allowed to grow at 30 °C to OD 0.6 before 1 mM isopropyl -D-thiogalactopyranoside (IPTG) was added. After IPTG induction, the cells were cultured for an additional 4 h followed by centrifugation at 5000 rpm for 5 min. The resulting pellets were suspended in 50 mM Tris–HCl buVer (pH 8.0) and broken up by sonication for 5 min. After sonication, the cell lysate was again centrifuged at 5000 rpm for 5 min. SDS–PAGE analyses of fusion proteins The sample solution (10 l) was mixed with 10 l of the sample buVer (2£ containing 125 mM Tris–HCl (pH 6.8), 4% SDS, 20% glycerol, 10% -mercaptoethanol, and 1 g/ml of bromophenol blue (BPB). The mixture was boiled for 3 min and loaded to the well of a polyacrylamide gel. The proteins were separated using a 12% SDS–polyacrylamide gel electrophoresis and visualized by Coomassie brilliant blue R-250 staining.
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
Surface modiWcation of gold chips for the speciWc binding of fusion proteins A gold chip (2 nm of chromium as an adhesion layer and 45 nm of gold deposited on an 18 £ 18 £ 0.3 mm glass) was obtained from K-MAC (Korea). The gold chip was cleaned with a solution of H2SO4/H2O2 (3 v/v) at 50 °C for 30 min. It was then immersed in a 10 mM ethanolic solution of 11-mercaptoundecanol (MUOH) at room temperature for 20–24 h to assure the formation of a MUOH self-assembled monolayer on the gold surface. It was then washed with ethanol and water and sonicated for 5 min in ethanol. To activate the hydroxyl group of the MUOH, the gold chip was treated with 0.6 M epichlorohydrin in a 1:1 mixture of 0.4 M NaOH and 2-methoxyethyl ether for 4 h. After being washed with ethanol and water, the gold chip was immersed in a 0.3 mg/ml dextran solution in 0.1 M NaOH for 20 h. The dextran-coated surface was once again treated with the 0.6 M solution of the epichlorohydrin in the 1:1 mixture of the 0.4 M NaOH and 2-methoxyethyl ether for 4 h to activate the dextran’s hydroxyl group. This epichlorohydrin-activated dextran chip was further functionalized in order to give it a speciWc aYnity for the recombinant fusion proteins. It was treated with a 1.7 M solution of iminodiacetic acid (IDA) in 2 M sodium carbonate at 60 °C for 20 h, a solution of 0.1 mg/ml L-glutathione (reduced form) in pH 7.0, and 44 mM phosphate buVer at 40 °C for 20 h (glutathione chip), and a 70 mg/ml solution of -cyclodextrin in 0.1 M NaOH at 40 °C for 20 h (cyclodextrin chip), respectively. The IDA-functionalized chip, after being immersed in a 50 mM nickel(II) chloride solution for 3–4 h and washed with water, was used as a hexahistidine fusion protein-binding chip (IDA-Ni(II) chip). The glutathione and cyclodextrin chips, after having been treated with 1 M ethanolamine at 37 °C for 3–4 h and washed with water, were used respectively for the detection of GST and MBP fusion proteins. Fabrication microarray
and
detection
of
recombinant
protein
Recombinant fusion proteins, as a cell lysate or in a partially puriWed form, were diluted with a spotting buVer (50 mM Tris, 20 % (v/v) glycerol or ethylene glycol, pH 8.0). A microarrayer (Proteogen, Korea) was then used to print the recombinant fusion proteins on the surface-modiWed gold chip. Printing was carried out in an atmospherically controlled chamber with a relative humidity of 70–80% at room temperature. A Stealth pin (Model SMP-10, TeleChem.) was used for arraying spots of 335 m in diameter. After spotting, the gold slides were incubated for 30 min in the microarrayer and rinsed with a washing solution and DW. The washing solution was a 20 mM phosphate buVer containing 0.5 M NaCl
253
and 20 mM imidazole at pH 8.0 for the IDA-Ni(II) chip, a PBS buVer at pH 7.2 for the glutathione chip, and a 20 mM Tris buVer containing 200 mM NaCl and 1 mM EDTA at pH 7.4 for the cyclodextrin chip. After the gold chip was dried through centrifugation, it was then coupled with BK7 prism using index-matching oil and placed on the center of the goniometer. The CCD camera took reXected images at a Wxed incident angle. Construction of an SPR imaging system A two-dimensional surface plasmon resonance imaging system was constructed to obtain microarray images of the protein spots. A 150-W quartz tungsten-halogen lamp (Schott, Germany) was used as the light source and the light was delivered to a goniometer arm (Physik Instrumente, Germany) using a liquid light guide (Oriel Instruments, USA). The light collimated by the lenses was passed through a narrow interference Wlter (647.1 nm, D 1 nm; Oriel Instruments) and a polarizer (Newport, USA) in order to convert a monochromatic and a linear polarized beam, respectively. The diameter of this beam was about 2 cm. The gold sensor chip on which proteins were spotted was optically coupled with a prism coupler (Korea Electro-optics, Korea) via an index matching oil (nD D 1.517) and placed on the center of the goniometer. The goniometer was electronically controlled by a DC Servo motor controller (Physik Instrumente). The reXected images from the gold chip were taken by a 1/2 inch charge coupled device (CCD) camera (Sony, Japan), while the contrast images were monitored by a personal computer. A combination of lenses was placed in front of the bare CCD chip in order to obtain a clear image. The images were then stored digitally in a personal computer using a B/W frame grabber (National Instrument, USA).
Results and discussion The hexahistidine-ubiquitin-tagged hGH (His6-UbhGH), which was expressed in a soluble form, was partially puriWed using a one-step metal aYnity chromatography, and analyzed using both SPR imaging and SDS–PAGE (Fig. 2). The SPR images were captured at an incident angle that was lower than the SPR angle of the background surface. The brighter spots are indications of the aYnity binding of the target proteins on the IDA-Ni(II) chip. Protein microarray images with varying protein concentrations are shown in Fig. 2B. The brightness of the SPR image increased as the protein concentration increased. As a control, bovine serum albumin (BSA) with a concentration of 100 g/ml was spotted on the IDA-Ni(II) gold chip. As shown in Fig. 2B, the control spots are almost invisible on the SPR image, thus indicating that the protein binding is
254
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
Fig. 2. SDS–PAGE (A) and SPR imaging (B) analyses of partially puriWed His6-Ub-hGH fusion protein at diVerent concentrations: lane M, protein maker; lane 1, 0.5 mg/ml; lane 3, 0.25 mg/ml; lane 4, 0.1 mg/ml; lane 5, 0.05 mg/ml; lane 6, 0.01 mg/ml; lane 7, 0.005 mg/ml; lane 8, 0.0025 mg/ml; lane 9, 0.001 mg/ml.
Fig. 3. Plot of relative spot intensity versus His6-Ub-hGH concentration.
speciWc. The intensity diVerence, which was the average value of the spot intensity minus the background intensity calculated from the image in Fig. 2B using an image analysis software (Scion, USA), was plotted against the concentration of His6-Ub-hGH (Fig. 3). The linear dynamic range was found to be 10–75 g/ml, which was relatively narrow compared to a SPR sensor (1–200 g/ ml) [6]. The lowest visible protein concentration was found to be about 10 g/ml in the SPR imaging, whereas the protein band in the SDS–PAGE was visible at 5 g/ ml. This indicates that the SPR imaging measurement is slightly less sensitive than the SDS–PAGE. The sample volumes used for the SDS–PAGE and SPR imaging were 20 l and 3.9 nl (data from catalogue of TeleChem.), respectively. The detection limit, which is based on the protein mass, of the SPR imaging measurement is 0.04 ng/protein spot, which is much less than that the
(100 ng/protein band) of the SDS–PAGE. The location of the protein in the polyacrylamide gel is determined by either Coomassie blue staining or silver staining. In general, the former is more frequently used because of its ease of handling and rapid detection. However, silver staining methods are used to detect smaller amounts of proteins because they are considerably more sensitive. The detection limit of silver staining is 2 to 5 ng/protein band [2], which is still much higher than that of the SPR imaging. The SPR imaging sensor based on intensity measurement was calculated to exhibit 1/20 detection sensitivity to an angular interrogation-based SPR sensor [15]. The SPR sensor is known to be able to detect less than 1 g/ ml of protein [6]. Thus the detection limit achieved by our SPR imaging system (10 g/ml) is comparable to the theoretically calculated sensitivity. Low sensitivity can limit the use of SPR imaging, especially in cases where a low-level expression of recombinant proteins is present. In general, however, the concentration of recombinant proteins expressed in E. coli is usually enough to be analyzed by the SPR imaging sensor. Furthermore, the SPR imaging system has a great advantage over the existing SPR sensors in that it can detect multiple molecular interactions in array formats on a two-dimensional gold thin Wlm. Using SPR imaging, a direct analysis of the recombinant proteins without puriWcation was carried out. After recombinant E. coli cells expressing His6-Ub-hGH were lysed by sonication, the cell lysates were directly spotted onto an IDA-Ni(II) chip. After 30 min incubation of the spotted gold chip to induce the aYnity binding between His6-Ub-hGH and IDA-Ni(II), the gold chip was washed with the washing solution and deionized distilled water. After the gold chip was dried, the expressed protein spots bound on the IDA-Ni(II) chip were measured
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
Fig. 4. SDS–PAGE (A) and SPR imaging (B) analyses of E. coli cell lysates expressing His6-Ub-hGH fusion protein: lane M, protein marker; lane 1, untransformed cell lysates; lane 2 and 2⬘, transformed cell lysates, uninduced; lane 3 and 3⬘, transformed cell lysates, induced with IPTG addition.
using SPR imaging. Fig. 4 shows the result of the SDS– PAGE and SPR imaging analyses. Compared to the partially puriWed His6-Ub-hGH (Fig. 2A), a number of protein bands were displayed on the SDS–PAGE gel. Lanes 2 and 3 in the SDS–PAGE correspond to lanes 2⬘ and 3⬘ in the SPR imaging, respectively. Uninduced cells were also found to produce a small amount of His6-Ub-hGH because of the leaky translation that was often observed in the T7 promoter system. The brighter SPR spot images were observed with induced cell lysates, whereas the dim images were observed with uninduced cell lysates. Along with hexahistidine residues, GST and MBP are also widely used as an aYnity tag and oVer the possibility of purifying the recombinant proteins through the use of aYnity chromatographies. When the target protein is fused to a highly soluble partner such as the GST or MBP moiety, correctly folded and soluble heterologous proteins are produced intracellularly in the bacterial cytoplasm [14]. Human interleukin 6 fused with GST (GST-hIL6) and MBP (MBP-hIL6) was expressed in E. coli in a soluble form, and then the cell lysates were directly spotted onto a glutathione or cyclodextrin-functionalized gold chip. Figs. 5 and 6 show the SDS–PAGE and SPR imaging analyses for GST-hIL6 and MBPhIL6, respectively. Lanes 2 and 3 in the SDS–PAGE correspond to lanes 2⬘ and 3⬘ in the SPR imaging, respectively. Unlike the His6-Ub-hGH expression, the GST-hIL6 and MBP-hIL6 expression in uninduced cells was hardly detected on the SDS–PAGE (Lane 2 in Figs. 5A and 6A). The brighter SPR images were clearly observed with induced cell lysates, whereas no visible images were obtained with uninduced cell lysates (Figs. 5B and 6B). This demonstrates that the SPR imaging analysis corresponds to the SDS–PAGE quantitatively. The sequencing of the genomes of various organisms has led to the concept of analyzing protein function on a genome-wide scale. Recently, rapid progress in highthroughput and parallel approaches in protein expres-
255
Fig. 5. SDS–PAGE (A) and SPR imaging (B) analyses of E. coli cell lysates expressing GST-hIL6 fusion protein: lane M, protein marker; lane 1, untransformed cell lysates; lane 2 and 2⬘, transformed cell lysates, uninduced; lane 3 and 3⬘, transformed cell lysates, induced with IPTG addition.
Fig. 6. SDS–PAGE (A) and SPR imaging (B) analyses of E. coli cell lysates expressing MBP-hIL6 fusion protein: lane M, protein marker; lane 1, untransformed cell lysates; lane 2 and 2⬘, transformed cell lysates, uninduced; lane 3 and 3⬘, transformed cell lysates, induced with IPTG addition.
sion, puriWcation, and detection has been made to study proteins on a global scale [16,17]. So far, many researches have focused on the development of robust techniques for high throughput and the parallel expression and puriWcation of multiple proteins. However, the detection of multiple proteins, which are often expressed in 96-well microtiter plates for automation, relies on traditional methods such as the SDS–PAGE and Western blots. In this study, therefore, we introduce a novel approach to the rapid and simultaneous detection of multiple recombinant proteins expressed in E. coli with aYnity tags to facilitate puriWcation. This method enables on-chip puriWcation, label-free detection, and on-chip quantitative analyses of target proteins. It is thus expected that the SPR imaging measurement of protein expression can be an attractive alternative to traditional methods such as the SDS–PAGE and Western blots.
256
J.-M. Jung et al. / Analytical Biochemistry 330 (2004) 251–256
Acknowledgments This research was supported by a grant from the KRIBB Initiative Research Program. This work was also supported in part by a grant from KOSEF through the Center for Advanced Bioseparation Technology (BSEP).
References [1] J.S. Albala, K. Franke, I.R. McConnel, K.L. Pak, P.A. Folta, B. Karlak, B. Rubinfeld, A.H. Davies, G.G. Lennon, R. Clarket, From genes to proteins: high-throughput expression and puriWcation of the human proteome, J. Cell. Biochem. 80 (2000) 187–191. [2] F.M. Ausubel, R. Brent, R.E. Kingston, et al., Short protocols in molecular biology, second ed., Wiley, New York, 1992. [3] S.R. Karlsen, K.S. Johnston, R.C. Jorgenson, S.S. Yee, Simultaneous determination of refractive index and absorbance spectra of chemical samples using surface plasmon resonance, Sens. Actuators B 25 (1995) 747–749. [4] D. Neri, S. Montigiani, P.M. Kirkham, Biophysical methods for the determination of antibody-antigen aYnities, Trends Biotechnol. 14 (1996) 465–470. [5] M. Fivash, E.M. Towler, R.J. Fisher, BIAcore for macromolecular interaction, Curr. Opin. Biotechnol. 9 (1998) 97–101. [6] K.N. Baker, M.H. Rendall, A. Patel, P. Boyd, M. Hoare, R.B. Freedman, D.C. James, Rapid monitoring of recombinant protein products: a comparison of current technologies, Trends Biotechnol. 20 (2002) 149–155.
[7] C.E. Jordan, R.M. Corn, Surface plasmon resonance imaging measurements of electrostatic biopolymer adsorption onto chemically modiWed gold surfaces, Anal. Chem. 69 (1997) 1449–1456. [8] E.A. Smith, M.G. Erickson, A.T. Ulijasz, B. Weisblum, R.M. Corn, Surface plasmon resonance imaging of transcription factor proteins: interactions of bacterial response regulators with DNA arrays on gold Wlms, Langmuir 19 (2003) 1486–1492. [9] E.A. Smith, M. Kyo, H. Kumasawa, K. Nakatani, I. Saito, R.M. Corn, Chemically induced hairpin formulation in DNA monolayers, J. Am. Chem. Soc. 124 (2002) 6810–6811. [10] H.J. Lee, T.T. Goodrich, R.M. Corn, SPR imaging measurement of 1-D and 2-D DNA microarrays created from microXuidic channels on gold thin Wlms, Anal. Chem. 73 (2001) 5525–5531. [11] G.J. Wegner, H.J. Lee, R.M. Corn, Characterization and optimization of peptide arrays for the study of epitope-antibody interactions using surface plasmon resonance imaging, Anal. Chem. 74 (2002) 5161–5168. [12] G.J. Wegner, H.J. Lee, G. Marriott, R.M. Corn, Fabrication of histidine-tagged fusion protein arrays for surface plasmon resonance imaging studies of protein-protein and protein-DNA interactions, Anal. Chem. 75 (2003) 4740–4746. [13] E.A. Smith, W.D. Thomas, L.L. Kiessling, R.M. Corn, Surface plasmon resonance imaging studies of protein-carbohydrate interactions, J. Am. Chem. Soc. 125 (2003) 6140–6148. [14] P.A. Nygren, S. Stahl, M. Uhlen, Engineering proteins to facilitate bioprocessing, Trends Biotechnol. 12 (1994) 184–188. [15] J. Homola, S.S. Yee, G. Gauglitz, Surface plasmon resonance sensors: review, Sens. Actuators B 54 (1999) 3–15. [16] S.P. Chamber, High-throughput protein expression for the postgenomic era, Drug Discovery Today 7 (2002) 759–765. [17] S. Yokoyama, Protein expression system for structural genomics and proteomics, Curr. Opin. Chem. Biol. 7 (2003) 39–43.