|Home | About | Journals | Submit | Contact Us | Français|
A large-surface biosensor technique using surface plasmon resonance (SPR) was tested for protein purification by recovery of a monoclonal antibody against human proinsulin C-peptide. Notably, both reversible attachment/desorption and actual purification of the antibody from a multi-component protein mixture was shown. For initial chip attachment of the peptide ligand, C-peptide was biotinylated and attached to neutravidin on plastic chips with a large gold surface (effective area 26 mm2). Antibody binding and desorption was monitored in real-time SPR, and for elution different conditions were employed. Five percent formic acid (in contact with the chip surface for 3 min) in a 60-μl segment between air bubbles was efficient for subsequent analysis. In this manner, protein amounts up to 35 pmoles were recovered in a single capture/elution cycle. Evaluation by SDS-PAGE showed essentially no carryover between fractions in this elution process, and also not with other proteins in the mixture after purification. Compared to existing commercial instruments, this technique gives higher recovery and makes it possible to monitor monitor protein binding/desorption. Recovery of affinity partners at the multi-pmole level is demonstrated for protein purification in SPR approaches.
Biosensor chip technologies based on the principle of surface plasmon resonance (SPR)1 can be used for online, real-time monitoring of protein-protein interactions,2–4 as well as recovery of affinity-purified proteins by elution from the chip5 and subsequent characterization by mass spectrometry (MS). Matrix-assisted laser desorption/ionization (MALDI) MS and electrospray ionization tandem MS (ESI-MS/MS) of proteins and proteolytic peptides after elution from the sensor chip have been presented.5–8 However, the amount of material recovered is limited by the size of the active chip area and frequently is in the femtomole range with conventional chip areas.9,10 For multiple structural and biological characterizations, recovery at a larger scale (tens of pmoles) is desirable. A large active surface area is then beneficial.
In this study, a high-capacity plastic biosensor chip with a gold surface area of 400 mm2 (effective area 26 mm2) (VirChip, Vir Biosensor, Taastrup, Denmark) was investigated for use in affinity capture of an antibody against human proinsulin C-peptide.11,12 For chip attachment of the peptide partner, we used the neutravidin/biotin system,13 since it is versatile and reliable. This injection-molded, disposable chip has an integrated SPR optical system14 and a surface technology supporting multiple analyses. Using a two-bubble system for elution,5 the bound protein can be recovered in amounts (several tens of pmoles) exceeding those conventionally obtained in a single capture/elution cycle, and multiple characterizations are possible. Purification of the antibody from a protein mixture was also demonstrated with this methodology.
High-capacity chips with gold surface (area 400 mm2) were supplied by Vir Biosensor (Taastrup, Denmark); neutravidin was from Pierce; Tween-20, dithiothreitol (DTT), and iodoacetamide were from Sigma; and trypsin (porcine, modified) was from Promega. Biotinylated human C-peptide was obtained from PolyPeptides (Wolfenbuttel, Germany), the 150 kDa monoclonal antibody directed against human C-peptide was from Medix Biochemica (Kauniainen, Finland) at 5.1 mg/mL in citrate/NaCl (pH 6) with 0.1% NaN3, and the protein mixture used was from Amersham Biosciences, UK. In separate experiments, the antibody C-peptide binding properties were shown to have a Kd value of 50 nM.15 All chemicals used were of analytical grade, and the water was from a Millipore purification system.
For preparation of the affinity partner bound to the chip, the neutravidin was first adsorbed to the gold surface. The chip was initially rinsed in a soft stream of water for a few seconds to remove dust and particles. It was then submerged in 2% hypochlorite solution in a 40-mL beaker for 4 min using a magnetic stirrer, rinsed in a soft stream of water for a few seconds, submerged in water for 4 min, and dried by tapping the edge of the chip on a clean, dust-free paper tissue. A biosensor instrument (Cobi R100, Vir Biosensor, Taastrup, Denmark) was prepared with a cleaning chip positioned in the flow cell, and the pump was adjusted to a flow rate of 300 μL/min. The tubing and flow cell was washed with 0.1 M NaOH and water, each for 4 min. The gold chip was then inserted into the instrument. The pump was started with the inlet tube submerged in the calibration solution (40% glycerol) to normalize the gold surface. Cell calibration was continued until the SPR image curve was horizontal and stable. Following calibration, the flow cell was rinsed with 0.1 M NaOH and water, for 4 min each, to remove glycerol from the tubes and the chip. The gold surface of the chip was equilibrated with 0.05 M sodium acetate (pH 5) for 4 min, after which 10 μg neutravidin/mL in 0.05 M sodium acetate (pH 5) was injected for 20 min. Unbound neutravidin was removed by washing with 0.05 M sodium acetate (pH 5) complemented with 0.005% Tween-20 for 4 min. After this treatment, the chip was washed with 10 mM Tris-buffered saline (TBS, pH 7.4), and biotinylated C-peptide at 20 μg/mL in TBS was then pumped into the flow cell for 10 min. Finally, the chip was washed with TBS containing 0.005% Tween-20.
The chip with neutravidin and biotinylated C-peptide was equilibrated with TBS for 4 min, after which the antibody solution (50 μg/mL) alone or in combination with other proteins was injected at 300 μL/min for 30 min, which was sufficient to reach saturation (Figure 11).). Unbound proteins were washed off with TBS for 4 min, then with 1 M NaCl for 4 min, and finally with TBS. The protein bound to immobilized C-peptide was eluted with aqueous base (50 mM NaOH) or acid (5% formic acid). For subsequent analysis, the latter was optimal, using a two-bubble system (modified from reference 5) in which first an air bubble is pumped into the flow cell, followed by 60 μL 5% formic acid, and finally a second bubble. The formic acid between the bubbles was kept in contact with the chip for 3 min with the pump stopped. The pump was then re-started and the fraction between the bubbles was collected. The chip was then re-treated with 5% formic acid to check that elution was complete, after which the chip was washed with TBS (Figure 11)) and was then available for re-use, directly or after renewed treatment at the hypochlorite stage (see neutravidin adsorption step above).
Antibody recovery after elution from the chip was evaluated by amino acid analysis in a ninhydrin-based analyzer (Biochrom 20, Amersham Pharmacia Biotech) of aliquots hydrolyzed for 24 h at 110°C in sealed, evacuated tubes with 6 M HCl containing 0.5% (w/v) phenol. The fractions collected were also analyzed by SDS-PAGE, using 12% Tris-glycine gels and silver staining, and by subsequent MALDI mass spectrometry peptide fingerprinting after tryptic digestion.
Protein bands were excised manually from the silver-stained SDS-PAGE gels and digested using a MassPREP robotic protein-handling system (Waters). Gel pieces were washed twice with 100 μL 50 mM ammonium bicarbonate/50% (v/v) acetonitrile at 40°C for 10 min, and the proteins were reduced using 10 mM DTT in 100 mM ammonium bicarbonate for 30 min, after which the gels were shrunk in acetonitrile and the proteins alkylated for 20 min with 55 mM iodoacetamide in 100 mM ammonium bicarbonate. Trypsin (25 μL of a 12 ng/μL solution in 50 mM ammonium bicarbonate) was added, and incubation was carried out for 5 h at 37°C. Peptides were extracted with 30 μL 1% formic acid/2% acetonitrile, followed by extraction with 2 × 15 μL 50% acetonitrile.
The tryptic peptide extracts from in-gel digestion were mixed at 1:1 (v/v) with a saturated solution of α-cyano-4-hydroxycinnamic acid in 75% aqueous acetonitrile with 0.1% trifluoroacetic acid, and dried on a standard steel target plate followed by MALDI MS (Voyager DE Pro, Applied Biosystems).
The identification of binding proteins to bioactive peptides is important and was tested in this study by use of a biosensor chip with a large gold surface. For initial chip attachment of the peptide ligand, neutravidin was immobilized to the gold surface through direct binding, followed by coupling of biotinylated human C-peptide. As C-peptide binding principle, reversible binding of an anti-C-peptide antibody was tested. The chip gold surface area was 400 mm2 (effective area 26 mm2), which is larger than the area of conventional chips, and the complete process of binding and elution of the C-peptide binding partner was monitored in real time, allowing for precise collection of the eluted protein (Figure 11).
To check the coupling efficiency of biotinylated C-peptide to the chip, we followed the SPR binding curve upon injection of the anti-C-peptide antibody (Figure 2A2A).). Similar measurement without previous coupling of biotinylated C-peptide to the chip, showed no detectable interaction (Figure 2B2B),), demonstrating that the antibody was not bound to the neutravidin or the gold surface.
Elution of the bound antibody from the biosensor chip was tested, varying both the elution solvent and mode. Regarding solvent, we found that the antibody could be eluted with 50 mM NaOH as well as with 5% formic acid, the latter better for subsequent analyses. Regarding elution mode, we compared two different approaches. In the first, fractions were collected at the outlet of the instrument, two drops at a time; in the second, the bound antibody was collected in a single fraction using a two-bubble system. Two air bubbles with an intervening volume of 60 μL 5% formic acid were then introduced into the tubing to the flow cell. This solution was kept in contact with the surface of the chip for 3 min by stopping the pump. The first approach, with traditional elution, was inefficient, since the analyte was spread over several fractions and diluted (Figure 3B3B),), while the second approach, with elution between two bubbles, was efficient (Figure 3A3A),), with only minor amounts in the fraction immediately after the bubble segment. The recovery of antibody was found to be at the multi-pmole level—9, 18, 30, and 35 pmoles (determined as recovered antibody using amino acid analysis), in four experiments using three different chips. This amount of protein eluted from the biosensor chip is about 30-fold that commonly achieved in a single capture/elution cycle.9,10 To illustrate the multi-analytical possibilities with the recovered antibody at this level, only 5% of the material eluted from a single experiment was submitted to SDS/polyacrylamide gel electrophoresis and silver staining. The band corresponding to the 50-kDa antibody chain (Figure 3A3A)) was excised and in-gel digested with trypsin for peptide analysis by MALDI mass spectrometry (Figure 44).). The results showed mass maps with good signal-to-noise ratios (average 7:1) and identified the C-peptide reactive antibody. Even at this level, using only a fraction of the material and silver staining, a sequence coverage of 34% was obtained, demonstrating that the remaining material from the SPR elution is sufficient for more or other analyses. Once we had shown these possibilities with the antibody recovered, we wished to prove that the system could also be used for affinity purification of the antibody in a protein mixture. The antibody was then mixed with six other proteins and injected into the biosensor instrument. After washing and final elution with the two-bubble system, recovery of the antibody essentially free from the other proteins was demonstrated (Figure 55).
In conclusion, we show that a large-surface biosensor chip active area provides recovery of bound protein at the pmole level using formic acid in a two-bubble elution approach. Recovery of up to 35 pmoles of an antibody bound to a peptide on the chip demonstrates that the preparative capability of this system, exceeds the capacity of previous systems. This method should be possible to use also for other systems and binding partners, provided affinities are as high as or higher than here, and the eluted proteins stable to weak acids/bases for the time of analysis. Binding of other proteins to C-peptide is of interest in further approaches, since this peptide has been suggested to be endocrinologically active,12 but a receptor or binding partner has still not been characterized.
This work was supported by grants from the Swedish Research Council (projects 03X-3532 and K5104-20005891), the Wallenberg Consortium North (WCN), the Juvenile Diabetes Foundation (project JDFI-4-99-647), and EC (LSHC-CT-2003-503297).