You are on page 1of 6

Anal Bioanal Chem (2006) 386:1321–1326

DOI 10.1007/s00216-006-0794-6

ORIGINAL PAPER

Reversible immobilization of proteins with streptavidin
affinity tags on a surface plasmon resonance biosensor chip
Yong-Jin Li & Li-Jun Bi & Xian-En Zhang &
Ya-Feng Zhou & Ji-Bin Zhang & Yuan-Yuan Chen &
Wei Li & Zhi-Ping Zhang

Received: 7 June 2006 / Revised: 5 August 2006 / Accepted: 21 August 2006 / Published online: 28 September 2006
# Springer-Verlag 2006

Abstract Dissociation of biotin from streptavidin is very
difficult due to their high binding affinity. The re-use of
streptavidin-modified surfaces is therefore almost impossible, making devices containing them (e.g. surface plasmon
resonance (SPR) sensor chips) expensive. This paper
describes a new protocol for reversible and site-directed
immobilization of proteins with streptavidin affinity tags on
the streptavidin-coated SPR biosensor chip (SA chip). Two
streptavidin affinity tags, nano-tag and streptavidin-binding
peptide (SBP tag), were applied. They both can specifically
interact with streptavidin but have weaker binding force
compared to the biotin–streptavidin system, thus allowing
association and dissociation under controlled conditions.
The SA chip surface could be regenerated repeatedly
without loss of activity by injection of 50 mM NaOH
solution. The fusion construct of a SBP tag and a
single-chain antibody to mature bovine prion protein
(scFv-Z186-SBP) interacts with the SA chip, resulting in
a single-chain-antibody-modified surface. The chip showed
kinetic response to the prion antigen with equilibrium
dissociation constant KD≈4.01×10−7. All results indicated
that the capture activity of the SA chip has no irreversible

Y.-J. Li : L.-J. Bi : X.-E. Zhang (*) : Y.-F. Zhou : J.-B. Zhang :
Y.-Y. Chen : W. Li : Z.-P. Zhang
Joint research group on analytical biotechnology of State Key
Laboratory of Macromolecules, Institute of Biophysics,
Chinese Academy of Sciences, Beijing 100101, China
and State Key Laboratory of Virology,
Wuhan Institute of Virology, Chinese Academy of Sciences,
Wuhan 430071, China
e-mail: zhangxe@most.cn
Y.-J. Li
Graduate School, Chinese Academy of Sciences,
Beijing, People’s Republic of China

loss after repeated immobilization and regeneration cycles.
The method should be of great benefit to various biosensors,
biochips and immunoassay applications based on the
streptavidin capture surface.
Keywords Surface plasmon resonance . Streptavidin affinity
tag . Nano-tag . Site-directed immobilization . Regeneration

Introduction
BIAcore SPR biosensor is an increasingly popular
platform for the analysis of biomolecular interactions in
real time. One of a pair of interactants, usually called the
capture molecule (or ligand), must be immobilized within
the biosensor flow cell before monitoring the interaction
of the mobile-phase partner. A number of biosensor chip
surfaces have been designed for different immobilization
purposes. As a common chip surface, carboxymethylated
dextran (CM) was covalently attached to a gold surface
[1]. Molecules carrying amine, carboxyl, aldehyde or thiol
groups can be covalently coupled to the CM chip surface.
However, this method may generate a heterogeneous
surface, as one protein or a peptide ligand may possess
many reactive groups distributed on its surface, which
usually leads to a random-orientating immobilization that,
in the worse case, might block the active site (or centre)
of the protein. To prepare a homogeneous sensor surface,
site-directed coupling or orientation-controlled immobilization of the ligand is always preferable to non-specific
coupling [2, 3]. Nitrilotriacetic acid (NTA)-modified
surfaces can be used to capture the protein with polyHis tag [4]. Two problems are encountered using the NTA
chip: a) there is usually strong non-specific adsorption of
other metal-binding proteins to the solid phase; and b) the

Five minutes after the cleaning process when the sensorgram reached a stable baseline. strep-tactin. multiple injections were performed. Nano-tag15 and Nano-tag9. the C-terminal) of an enzyme protein by gene fusion. a mercapto reagent is usually applied to block the rest of the open area of the gold surface. site-directed and reversible manner. Schmid et al. which is composed of eight amino acids (Trp-Ser-His-ProGln-Phe-Glu-Lys). Immobilization of streptavidin affinity tag fusions on the SA chip surface The SA chip was first cleaned with three consecutive 1-min injections of 40 μL of a solution of 1 M NaCl in 50 mM NaOH before the immobilization procedure. the –SH chemistry is not suited for protein immobilization in most cases as cysteine is a low-abundance amino acid in proteins and rarely appears on the protein surface. However. Lamla and Erdmann presented a new streptavidin-binding peptide for both the purification and the detection of recombinant proteins [12]. To determine the maximum immobilization level of streptavidin affinity tag fusions on this surface. respectively. The same procedure was performed for immobilization of nano-tagged peptide. Mannelli et al. The method has been widely used in the preparation of DNA chips [8]. All buffers for experiments were filtered (pore size 0. SBP tag. ScFV-Z186 (35 kDa) with SBP tag (abbreviated as scFV-Z186-SBP) and its antigen bovine mature prion protein were constructed and expressed in our laboratory. which could be solved by using other antibody-affinitive molecules. Tween 20 and Tris base were purchased from Sigma. The affinity of this interaction has been improved by a streptavidin mutant. The SA chip with streptavidin covalently immobilized on a carboxymethylated dextran matrix was used. which bind to streptavidin with dissociation constants of 4 nM and 17 nM. All experiments were performed at 25 °C within the same flow cell. Uppsala.e. Concentrations of the proteins were determined with an Eppendorf Biophotometer. the procedure for which is to be presented in another study [15]. Protein A showing different affinities to different antibodies may be a problem. thus decreasing the possibility of binding the protein in its native fold to the solid phase [5]. The interaction of strep tagII and streptavidin is fairly weak (72 μM) [10]. The nano-tags have two types. can be fused to protein at either C-terminal or N-terminal.4. One side effect of this treatment is that mercapto group may affect protein activity through interaction with the cysteine residue of the ligand. so as to stop the nonspecific adsorption of unwanted proteins. To solve this problem we had previously modified a cysteine residue onto the specific site (i. (Shanghai. the mechanism could be applied to build a new SPR biosensor surface biochemistry. diluted in running buffer to 20 μg/mL. Sweden). such as protein G. thereby allowing repeated regenerations of the SA chip. After immobilization of the ligand. recently derivatized the sensor chip surface with protein A for site-directed immobilization of an antibody [6]. which binds the strep tagII with an affinity constant of about 1 μM [14]. During recent years. The main problem is that it is almost impossible to remove the protein immobilized in the same flow cell due to the strong interaction between the streptavidin and biotin (KD≈10−15 M). . China). ScFv-Z186-SBP. exploited Au–S bond formation by the reaction of thiolated DNA probes and a bare gold chip [7]. except that the nano-tagged peptide was diluted in running buffer to 200 μg/mL. Proteins Pentapeptide (Phe-Leu-Asp-Glu-Val) with nano-tag9 was synthesized by GL Biochem (Shanghai) Ltd. The strep tagII. was injected for 7 min using a flow rate of 5 μL/min.5 nM [11]. Materials and methods Instrument and reagents The instrument used was the BIAcore 3000 (BIAcore AB. is another streptavidin-binding peptide that binds to streptavidin with a dissociation constant of 2. Running buffer for all experiments was TrisHCl (pH 7.005% (v/v) Tween 20. and thus makes the experiment very expensive. The experiment results are presented herein. which causes difficulties in the optimization process of the measurements and repeat use of the sensor chip. This type of sensor chip is very useful in the capture of antibodies. So far.22 μm) and degassed before use. Anal Bioanal Chem (2006) 386:1321–1326 Since the interactions between these two new ligands and streptavidin are moderate. This principle is illustrated in Fig.10 mM) containing 150 mM NaCl and 0. The peptide possesses nanomolar affinity for streptavidin and therefore was termed Nano-tag.1322 poly-His tag is very hydrophobic and in some cases buried in the fusion partner. some streptavidin affinity tags have been discovered and widely applied [9–13]. designed for biomolecular interaction analysis (BIA) in real time. through which a controlled orientation of protein on the gold surface was achieved [4]. 1. the streptavidin-coated biosensor chip (SA chip) seems to be the best choice for site-directed immobilization of biotinylated ligand in the SPR experiment. a 38-aminoacid-length peptide. which permits immobilization of various proteins on the SA chip sensor surface in a stable.

The injection . 10 mM glycine-HCl (pH 2. Either incomplete regeneration or reduced activity of the sensors surface will have a negative effect on the performance of assay and the useful lifetime of the sensor chip. 3). 30. Data management and calculation of kinetics were performed using the BIA evaluation 4. After each injection of antigen. a binding kinetics experiment of scFv-Z186-SBP and its antigen mature bovine prion protein was performed. 2). Those NaOH solutions with concentrations below 50 mM and glycineHCl buffer could not do it well. To test whether the capture activity of the SA chip was harmed or affected by the regeneration process.300 RU. III removal of targets by running buffer.1 software (Biacore). and the blank (injection of buffer only) was subtracted to the other five response curves to take in account the slightly decreasing baseline. A single test cycle seldom gave sufficient information for establishing regenerations. IV regeneration of the SA chip with NaOH solution. repeated interlaced injections of scFv-Z186-SBP sample and regeneration solution (50 mM NaOH) were performed. insert SPR sensorgram corresponding to steps I–IV 1323 (III) (I) (IV) Ligand-scFv surface Ligand-scFv surface Time Targets (antigen) (II) (III) scFv-target binding Regeneration of SA chip surfaces To the scFv-Z186-SBP bound surface. 40 and 50 mM). where the binding sites of the SA chip surface were believed to be nearly saturated by scFv-Z186-SBP. The same process was applied to the nano-tagged peptidebound SA chip surface for evaluation of regeneration. Results indicated that only 50 mM NaOH solution could remove the SBP-tagged protein completely from the SA chip. Finally. II injection of targets (antigens) and binding of the antigens to the ligand–scFv.0) with a flow rate of 30 μL/min. Here we tried three kinds of solutions: 10 mM glycine-HCl (pH 2.430 nM. Results and discussion Results Immobilization of scFv-Z186-SBP and regeneration of the SA chip surface A response of approximately 2. so repeated injections of scFv-Z186-SBP were performed to test whether the surface was still active and could bind the same amount of proteins as before (Fig. Continuous injections of scFv-Z186SBP (20 μg/mL) increased the immobilization response level up to 3. which seemed to be the maximum level for this 35-kDa protein (Fig. The antigen concentrations of the sample solutions ranged from 0 to 1. the binding surface was washed with the running buffer (Tris-HCl) for 30 min to remove the bond antigen.Anal Bioanal Chem (2006) 386:1321–1326 (I) (IV) Streptavidin surface SPR sensor chip Ligand-scFv (II) Signal change Fig. A blank flow cell was used as reference surface to eliminate the bulk effect. 20. Binding kinetics of scFv-Z186-SBP and its antigen prion protein To confirm the practicability of the method. 10 mM glycineHCl (pH 2. Efficient regeneration is crucial to a successful assay. whose concentration was 20 μg/mL. was obtained by the capture method as described above. Approximately 150 response units of scFv-Z186-SBP.650 RU (response units) was obtained by the first injection of scFv-Z186-SBP (5 μL/min. 7 min). All measurements were carried out at 25°C with a flow rate of 20 μL/min. The scFv-Z186-SBP-modified SA chip was then ready for the next measurement.0) and NaOH (10. followed by 1-min injection of 10 mM glycine-HCl (pH 2.5) was first injected for 1 min at a flow rate of 30 μL/min. 50 mM NaOH was injected for 1 min at a rate 30 μL/min.5). 1 Illustration of bindings and regeneration of the streptavidin-modified SPR sensor chip (SA chip): I injection of the ligand–scFv fusion onto the surface of the SA chip and binding of the ligands to the SA chip.

56. d a further 30-μL injection of the ligand gave almost no increase of signal.6.7 RU. response signals were 653. which seemed to be the saturation binding level for this 2.300 RU. it implies that response unit (RU) of capture of the nano-tagged peptide on the SA chip surface would be low.2. Response (RU) 22500 (a) (b) (c) (d) The nano-tagged peptide used in this study was a pentapeptide with a molecular mass of 2. b and c followed two consecutive 30-μL injections (5 μL/ min. The immobilization signal levels for each cycle were 653. 112.110 to 3.700-RU response. 94. 3 Reversed immobilization process of scFv-Z186-SBP surface: a–e five cycles of the scFv-Z186-SBP immobilization process.6%) and 6 RU (4. before and after regeneration.3. 1 min) . Relative to the first run. 130 and 136 RU for injections a–f. The baseline changed little after regeneration Fig. nor the nano-tagged peptide binding signals changed significantly. f–j five cycles of the scFv-Z186-SBP surface regeneration with two consecutive injections of 50 mM NaOH (30 μL/min. 657 and 650 RU. 2 Maximum immobilization level of scFv-Z186-SBP on the SA chip surface: a injection of a 35-μL volume led to a 2. 50 mM NaOH may work well to regenerate the surface. 1-min injection of 50 mM NaOH could remove the captured nano-tagged peptide surface completely from the chip surface.2 kDa. Neither of the baseline. indicating the SA chip surface was almost saturated by the ligand. 6 min) resulting in an increase of the immobilization level from 3. 657 and 650 RU respectively. 57.5 nM). 4. 658.1324 Anal Bioanal Chem (2006) 386:1321–1326 Response (RU) 25000 (a) Immobilization of nano-tagged peptide and regeneration of the SA chip surface (d) (c) (b) 24000 23000 22000 21000 1000 2000 3000 Time (s) 4000 Fig. the maximum immobilization level of the last run decreased by 20 RU (or 0. the flow rate was adjusted to 40 μL/min. The baseline before and after regeneration and the scFv-Z186-SBP binding signal changed little. Five cycles of immobilization and regeneration processes are shown in Fig.6. Since the molecular weight detection limit of the BIAcore3000 machine is 200 Da. respectively.2-kDa peptide on the SA chip. (e) 22000 21500 21000 20500 20000 (g) (f) 0 1000 (h) (i) 2000 3000 Time (s) (j) 4000 Fig. The reversed immobilization of both affinity tag fusions was performed for about 100 times discontinuously in 5 days. Signals were 30. respectively. the total signal change of the immobilization was at maximum 136 RU.6. respectively. 655.5 and 57. total signal change was 136 RU. 5. Since the interaction between nano-tag and streptavidin (KD≈ 4 nM) is weaker than that of SBP tag and streptavidin (KD≈2. 658.5.4%) for SBP-tagged ligand and nano-tagged ligand. At the flow rate of 30 μL/min. Two consecutive 1-min injections of 50 mM NaOH could regenerate the chip surface well. by continuous injections six times. The immobilization response signals for each cycle were 55. g represents the regeneration with 50 mM NaOH (30 μL/min. Concentration of scFv-Z186-SBP was 20 μg/mL volume of scFv-Z186-SBP was 5 μL (1 min) and the immobilization signal level was 650 RU. As shown in Fig. 53. respectively. 50. while the baseline drift was less than 15 RU after each regeneration process. 1 min) for each regeneration process. 4 Maximum immobilization level of nano-tagged peptide on the SA chip surface by six successive injections. After a 2-min injection.6. 655.

To apply the affinity tag–streptavidin interaction systems to SPR sensors. which helps the fusion protein fold correctly so as to form reasonable Table 1 Kinetic parameters of scFv-Z186-SBP and its antigen in two independent experiments Experiment Ka (104 M−1 s−1) Kd (10−3 s−1) KD (10−7 M) x2 1 2 Mean 1. 715.4±0. 57.21 6. A reversible surface biochemistry of the streptavidin-modified chip will be great of benefit to SPR technology. This can be easily achieved by gene fusion.21 6. particularly in biosensors. respectively. The sensorgram showed that the interaction between the single-chain antibody and prion antigen is a rapid association and slow dissociation process (Fig. a linker peptide is usually inserted between the fusion partners. higher than that of biotin (femtomolar level) and lower than that of strep tagII (micromolar level). One or two 1-min injections of NaOH (50 mM) could completely remove the bound tags from streptavidin. 358. 5 Reversed immobilization process of the nano-tagged peptide bound surface: a–e five cycles of the nano-tagged peptide immobilization process.13 6.56±0. The spikes at the beginning and the end of the injection phase are created by subtraction of a reference cell caused by the change of mobile phase in the flow cell. biochips and immunoassay.64 3. f–j five cycles of the surface regeneration. multiple measurements and experiment optimization are essential. In SPR experiments. 6 Sensorgram of the kinetic response of the scFv-Z186-SBPmodified SPR sensors to bovine mature prion antigen. 90 and 0 nM. and such once-only use of the SA chip is not economically viable for many studies. disposable) formats. One injection of 50 mM NaOH (30 μL/min.38 3. separation of biotin and streptavidin is almost impossible due to their strong binding. also through gene fusion. allowing successive regeneration of the SA chip without irreversible loss of its activity.55±0. Specific recognition and high-affinity interaction between biotin and streptavidin has found many applications in analytical biotechnology. all analytical means based on the 100 200 Time (s) 300 400 Fig. So far. Sensorgrams of them on the SA chip surface showed typical rapid association kinetics.21 4. All kinetic constants were derived from the response data and calculated by BIA evaluation 4.32 . Nevertheless.1 software.3. To remove the bond prion proteins completely from the scFv-Z186-SBP surface. 53. The results are summarized in Table 1. were determined from six sensorgrams of the antibody–antigen binding experiment.430.02 4. 6).8±0. Concentrations of the antigen were 1. Both nano-tag and SBP tag have moderate binding properties to streptavidin.65±0. respectively from top to bottom biotin–streptavidin interaction system are intended as single-use (i. the affinity tags must be fused on the ligand. which is a mature technology. Their equilibrium dissociation constants were at the nanomolar level.5. site-directed immobilization or orientation control of the ligand is crucial to obtain a homogeneous surface and to keep the bioactivity of the ligand.Anal Bioanal Chem (2006) 386:1321–1326 1325 (b) (a) (c) (d) 21500 21000 20500 20000 (f) 0 1000 (g) (h) 250 (e) Response (RU) Resonance (RU) 22000 (i) 2000 3000 Time (s) (j) 200 150 100 50 0 -50 0 4000 Fig. 1 min) was performed for each regeneration process Response kinetics of scFv-Z186-SBP-modified SPR biosensor to prion antigen The SA chip surface was pretreated with the single-chain antibody fusion protein scFv-Z186-SBP.5 and 57.69±0.14 1. Discussion Although there are many types of sensor chip surfaces available for different ligands.35 1. responses were 55. The linker could be approximately 5–15 amino acids in length or longer. 179. injection of the running buffer was performed for 30 min until the baseline reached the initial level.e.46±0.51 3.7 RU. The kinetic rate constants (ka and kd).01±0.5±0. as well as equilibrium association constant (KA) and equilibrium dissociation constant (KD). Considering the potential loss of activities of the fusion partners. 56.

Zhang ZP (2001) Bioconjugate Chem 12:924–931 . Herne TM. the scFv-Z186-SBP fusion construct had activities for both ligand–receptor binding and immunoreaction. 1). Schmid AH. Deng JY. Mannelli I. Shoseyov O (2001) J Peptide Sci 7:50–57 6. Wang R. Lindquist G (1991) Anal Biochem 198:268– 277 2. Zhang JB (2006) PhD Thesis. We have successfully demonstrated this approach in the construction of numerous fusion protein systems [13. Schmidt TG. Erdmann VA (2004) Protein Expr Purif 33:39–47 13. Skerra A (1993) Protein Eng 6:109–122 10. Kwon Y. Zhang XE. The mechanism may serve as a low cost and reliable platform for functional studies of proteins and high precision immunoassay. Liu H. Thampia KR. Tombelli S. Zhang XE (2004) Biosens Bioelectron 20:807–813 17. Shi JX. Lamla T. Zhou YF. Zhang XE. Voss S. Coleman MA (2004) J Am Chem Soc 126:14730–14731 3. Levy I. Tarlov MJ (1997) J Am Chem Soc 119:8916–8920 8. Zhang XE. Lofas S. 16–18]. orientationcontrolled. Pang DW. Zhang CG (2003) Anal Chem 75:4113–4119 14. complete reversible immobilization of the ligand with streptavidin affinity tags on the streptavidin-modified SPR biosensor chip. Zhou YF. Gershon PD. Zhang ZL. Wang SH. Skerra A (1996) J Mol Biol 255:753–766 11. Minunni M. Xie WH. Chen J. The fusion was immobilized on the SA chip through its SBP ligand (Fig. Zhang ZP. Liu H. Wilson DS. pp 32–92 16. with the scFv remaining open to the bulk solution and being ready to react with the antigen. Szostak JW (2001) Protein Expr Purif 23:440–446 12. Xie B. Camarero JA. Khilko S (1995) J Immunol Methods 183:65–76 5. as well as many other biosensor or biochip purposes. Frank R. Shao WH. Deng JY. Wen JK. Keefe AD. Zhou YF. Zhang ZP. The most commonly used linker contains repeat glycine or serine residues that provides flexibility and enhance the hydrophilicity of the peptide backbone. Bi LJ. Seelig B. Thakur MS. Suri CR (2006) Sens Actuators B 113:297–303 7. Skerra A (1997) Protein Eng 10:975–982 15. Mascini M (2005) Bioelectrochemistry 66:129–138 9.1326 orientations that retain activities of the reactive sites of the fusion partners. Michela Spiriti M. Stanca SE. The design was achieved by use of solid experimental data. The configuration ensures sensitive detection of the prion antigen. Koepke J. Cass AEG. Schmidt TGM. Zhang ZP. Zhang ZP (2000) Bioconjugate Chem 11:822–826 18. Zhang CG (2004) Anal Chem 76:632–638 4. Yang RF. Johnsson B. Zhang JB. Zhou YF. Acknowledgements The study is supported by the Chinese Academy of Sciences and the Ministry of Science and Technology. Anal Bioanal Chem (2006) 386:1321–1326 References 1. Zhang XE. Conclusion A new protocol is proposed for site-directed. In this investigation.