CC BY-NC-ND
0New Biotechnology• Volume 00,Number 00• December 2015
RESEARCH PAPER
■cJfetHaaBMB.
ELSEVIER
Advancing the global proteome survey platform by using an oriented single chain antibody fragment immobilization approach
Anna Sall1, Helena Persson12, Mats Ohlin1, Carl A.K. Borrebaeck1 and Christer Wingren1
1 Department of Immunotechnology, Lund University, Medicon Village (House 406), SE-223 81 Lund, Sweden
2 Science for Life Laboratory, Royal Institute of Technology, Stockholm, Sweden
Increasing the understanding of a proteome and how its protein composition is affected by for example different diseases, such as cancer, has the potential to improve strategies for early diagnosis and therapeutics. The Global Proteome Survey or GPS is a method that combines mass spectrometry and affinity enrichment with the use of antibodies. The technology enables profiling of complex proteomes in a species independent manner. The sensitivity of GPS, and other methods relying on affinity enrichment, is largely affected by the activity of the exploited affinity reagent. We here present an improvement of the GPS platform by utilizing an antibody immobilization approach which ensures a controlled immobilization process of the antibody to the magnetic bead support. More specifically, we make use of an antibody format that enables site-directed biotinylation and use this in combination with streptavidin coated magnetic beads. The performance of the expanded GPS platform was evaluated by profiling yeast proteome samples. We demonstrate that the oriented antibody immobilization strategy increases the ability of the GPS platform and results in larger fraction of functional antibodies. Additionally, we show that this new antibody format enabled in-solution capture, i.e. immobilization of the antibodies after sample incubation. A workflow has been established that permit the use of an oriented immobilization strategy for the GPS platform.
Introduction
The proteome is a snapshot of the protein composition of a specific sample type present at a certain state and time point. Investigation of the human proteome in well established samples has the possibility to increase our understanding of different human diseases, such as cancer, and contribute to the development of approaches for early diagnosis and therapeutics [1,2]. The bio-marker discovery phase has for a long time mainly been carried out by standard mass spectrometry (MS) technologies [3], but very few candidate biomarkers have been transformed into clinical use [4,5]. This suggests that there is a need for better technologies
Corresponding author: Borrebaeck, Carl A.K. (Carl.Borrebaeck@immun.lth.se), Wingren, C. (Christer.Wingren@immun.lth.se)
both concerning the discovery phase and for validation of identified biomarkers. Some of the difficulties in MS have been associated with assay sensitivity and reproducibility as well as with sample complexity in regards to the large dynamic range of proteins [6,7]. MS and antibody based technologies both have distinct advantages and disadvantages. Different approaches have therefore been made into combining MS and affinity enrichment with the use of antibodies [3,8-12].
We have previously described a method for profiling complex proteomes in a species independent manner denoted Global Proteome Survey or GPS [11,13-15] in a way that combines the power of MS and antibody technology. GPS explores single chain antibodies (scFv) directed against short peptide motifs and MS. As illustrated in Fig. 1 (left panel), the general workflow of GPS involves five
http://dx.doi.org/10.1016/j.nbt.2015.12.001 www.elsevier.com/locate/nbt 1871-6784/© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1 Please cite this article in press as: Sall, A. et al., Advancing the global proteome survey platform by using an oriented single chain antibody fragment immobilization approach, New Biotechnol. (2015), http://dx.doi.Org/10.1016/j.nbt.2015.12.001_
RESEARCH PAPER
FIGURE 1
Schematic outline of the three different capture versions evaluated for the GPS platform. In capture 1,the CIMSscFvsare chemically coupled to magnetic beads. In capture 2, biotinylated CIMS scFvs are coupled in an oriented manner to streptavidin magnetic beads. In capture 3, the biotinylated CIMS scFvs are allowed to bind tryptic peptides before coupling to streptavidin magnetic beads.
different main steps; (i) coating of scFv fragments to the surface of magnetic beads, (ii) incubation with tryptic digest, (iii) washing, (iv) elution of bound peptides and (v) detection of eluted peptides by shot-gun MS. One of the powers of GPS lays in its ability to target a large number of proteins with the use of a small number of scFv antibodies. In affinity proteomics approaches the one to one relationship between the protein of interest and its targeting affinity probe is a large technology bottleneck. Not only must the target of interest be known in advance but there is a lack of availability of these specific reagents as well as a very substantial requirement in the specificity of the affinity probes. In GPS the scFv antibodies, termed context independent motif specific (CIMS) antibodies, are directed towards sets of short peptide fragments present in up to a few hundreds of proteins. More specifically, the CIMS scFvs targets 4-6 amino acid long C-terminal sequence motifs of tryptic peptides and are termed context independent since they enables analysis of any proteome regardless of species. This means that the GPS platform holds great potential in targeting any proteome in a discovery directed manner. Theoretically, the use of about 100 of these CIMS
2 www.elsevier.com/locate/nbt
scFvs would cover 50% of the nonredundant human proteome [11,16]. The GPS platform has recently been used to characterize breast cancer tissue samples and it has successfully described protein signatures correlated to histological grades of breast cancer [15].
In immunoassays, maintaining the biological activity of the antibody during immobilization is of great importance [17,18]. The sensitivity of the assay is to a large extent influenced by the orientation of the antibody as well as the surface density of the immobilized antibody onto the solid-support [17]. By orienting the antibody on the surface, the antigen binding sites are exposed and the accessibility to the antigen is increased [19]. Consequently, a key step in GPS is the immobilization of the CIMS scFvs onto paramagnetic particles that are used to capture peptides from complex samples. This has previously been achieved by conventional chemical covalent coupling of the CIMS scFvs to activated carboxylic acid groups on the magnetic beads [11]. Here we have further advanced the GPS platform by adding a second approach of antibody immobilization. A system is used that provides the CIMS scFvs with an biotin acceptor domain (BAD) [20,21]. This enables
New Biotechnology• Volume 00,Number 00• December 2015
site-directed biotinylation of the CIMS scFvs during production and further on immobilization in an oriented manner on strepta-vidin magnetic beads.
The different versions of the GPS platform exploiting both chemically coupled and tag-oriented scFv were used for profiling yeast cell lysate samples in a parallel manner to compare how the immobilization of the CIMS scFv affects the peptide capture ability of the platform. Furthermore, the possibility of performing immobilization of the CIMS scFvs after peptide capture was evaluated.
Experimental procedures
CIMS antibodies
The human recombinant scFv antibodies CIMS17-C08, CIMS17-E02 and CIMS33-3D-F06 had been selected from the phage display library, n-CoDeR [22], as previously described [11]. These three CIMS scFv were included based on good performance in previously performed GPS capture investigations. CIMS17-C08 and CIMS17-E02 were developed against target peptide motif 17 (amino acid sequence: SSAYSR) and CIMS33-3D-F06 against target peptide motif 33 (amino acid sequence: LSADHR).
Production and purification of soluble CIMS antibody fragments
The three scFv were produced in 15 ml Escherichia coli cultures. Briefly, 15 ml TB medium (Becton, Dickinson and Company, Franklin Lakes, NJ, USA) supplemented with 0.2 M sucrose and 100 mg/ml carbenicillin (Saveen & Werner AB, Sweden) was used for production. ScFv production was induced with the use of 1 mM isopropyl thiogalactioside (IPTG). Cultures were grown over night at 30°C. Bacteria were harvested by centrifugation and scFvs were purified from the periplasmic space with MagneHisTM Protein Purification System (Promega Corporation, Madison, WI, USA) with the use of a KingFisher Flex robot. Both the purity and integrity of the scFvs were verified with SDS-PAGE.
Production and purification of soluble biotinylated CIMS antibody fragments
A vector format (pHP2-19) [23] that allows for the production of the scFvs with a biotin acceptor domain (BAD) preceding the hexahistidine tag at the C-terminus was used for production and purification of a biotinylated version of the scFvs. The DNA encoding each scFv was digested with SfiI and AvrII (New England Biolabs, Ipswich, MA, USA) according to manufacture's instructions and re-ligated into the pHP2-19 vector and transformed into AVB101 (Avidity LLC, Aurora, CO, USA), an E. coli B strain containing pBirAcm, encoding biotin ligase (BirA) and chloramphenicol resistance (Avidity LLC). Biotinylated soluble scFvs were produced and purified as described above with the exception that 10 mg/ml chloramphenicol (Saveen & Werner AB) was also added and that a final concentration of 1.5 mM of IPTG together with 50 mM biotin was used.
Functionality assay of CIMS antibodies
The functionality of the scFvs in the two different vector formats was evaluated against respective target peptide (JPT Peptide Technologies GmbH, Berlin, Germany) used for development of the scFvs in ELISA. Streptavidin were immobilized at a concentration of 1 mg/ml in 384-well plates (Corning Inc., New York, NY, USA) and incubated over night at 4°C. Unbound protein was removed
by washing the plate once with 154 mM NaCl and 0.05% (v/v) Tween 20 (wash buffer). Biotinylated peptides (biotin-SGSGSSAYSR and biotin-SGSGLSADHR) were added at a concentration of 1 mg/ml in PBS. The plate was incubated for 1 h at room temperature (RT) with mixing and then washed with wash buffer four times. The plates were blocked with 0.5% (w/v) bovine serum albumin (BSA) and 0.05% (v/v) Tween 20 in PBS (PBT) containing 2 mg/ml free biotin for 40 minutes at RT with mixing. After four washing steps with wash buffer, purified scFvs were added and incubated with target antigen for 2 h at RT with shake. Once again the plate was washed four times followed by addition of horseradish peroxide (HRP)-labelled streptavidin (Thermo Scientific Pierce, Rockford, IL, USA) or HRP-labelled anti-His6 monoclonal antibody (Roche Diagnostics GmbH, Mannheim, Germany) in PBT and incubated for 1 h at RT with mixing. Bound scFv were, after washing, detected with the use of TMB-ELISA (Thermo Scientific Pierce) as chromogen. The reaction was stopped with sulfuric acid (1 M) and the absorbance was measured at 450 nM.
Preparation of trypsin digested yeast proteome
Whole-cell protein extract from Saccharomyces cerevisiae (Promega Coporation) (1 mg) was reduced, alkylated and digested with trypsin. After thawing the protein extract at RT the sample was reduced with DTT (Saveen & Werner AB), at a final concentration of 5 mM, during 30 minutes at 37°C. Subsequently, the sample was alkylated for 30 minutes in the dark at RT with iodoacetamide (Sigma-Aldrich Corp., St. Louis, MO, USA) at a final concentration of 15 mM. Finally, sequenced-grade modified trypsin (Promega Corporation) was added at 20 mg per mg of protein and incubated for 16 h at 37°C. Next, the samples were aliquoted and stored at —80°C until use.
Coupling of magnetic beads
Two different bead systems were used for coupling of the purified scFvs: Dynabeads M-270 Carboxylic acid and Dynabeads M-270 Streptavidin (Invitrogen Dynal AS, Oslo, Norway). The purified scFvs were coupled individually to the magnetic beads and used for GPS capture experiment. For Dynabeads M-270 Carboxylic acid, 25 mg scFv were coupled per 1 mg beads as previously described [11]. Briefly, 300 ml of beads (9 mg) were washed twice in 25 mM MES buffer (pH 6) with slow mixing for 10 minutes. The beads were activated with 25 mg/ml EDC (Sigma) and 47 mg/ml Sulfo-NHS (Thermo Scientific) with slow mixing for 30 minutes. After washing once with 25 mM MES and PBS respectively, scFvs were added to the beads and incubated for 45 minutes at RT with gentle mixing. The coated beads were washed twice with 50 mM Tris (pH 7.4) to block the uncoupled surface. Afterwards, the beads were washed a total of four times with 0.05% (v/v) Tween 20 in PBS (TPBS) and finally resuspended in TPBS and stored at 4 °C until further use. For Dynabeads M-270 Streptavidin, 10 mg of biotinylated scFv were coupled per 1 mg beads. Briefly, the beads were washed twice with PBS, then mixed with biotinylated scFv in PBS and incubated for 45 minutes with mixing at RT. The beads were washed once and resuspended in TPBS and stored at 4°C until further use.
GPS capture experiment
Three different capture experiments were performed, with scFv chemically coupled to beads, with biotinylated scFv coupled to
www.elsevier.com/locate/nbt 3
RESEARCH PAPER
streptavidin coated beads and with biotinylated scFvs caught onto streptavidin coated beads after capture (Fig. 1). The tryptic yeast proteome sample was defrosted and 1 mM phenylmethylsulfonyl (PMSF) were added to inhibit trypsin activity. For each capture, 35 ml beads with immobilized scFv, either with scFv chemically coupled or coupled via biotin-streptavidin, were used. The coupled beads were washed in TPBS and 35 ml of tryptic yeast sample (20 mg) was added. The mixture was incubated for 15 minutes at RT with gentle hand mixing every two minutes. After incubation, the beads were washed first with 65 ml PBS and then with 50 ml PBS and transferred to a new tube. Captured peptides were eluted by adding 8.5 ml 5% (v/v) formic acid and subsequently transferred to a new tube. The samples were dried and frozen at -20°C until analysis. Before analysis, the samples were reconstituted in 8 ml 0.1% (v/v) formic acid, 3% (v/v) acetonitrile and transferred to HPLC vials. For the in-solution capture, the capture was performed as described above with the exception that uncoupled biotinylated scFvs were mixed with the tryptic yeast sample before coupling onto beads. After 15 minutes of incubation, pre-washed uncoupled beads were added to the scFv peptide mixture and coupling of scFv onto beads were allowed to proceed for 4560 minutes. The samples were then prepared as described above.
Mass spectrometry analysis
Captured peptides were analyzed on ESI-LTQ-Orbitrap (Thermo Electron, Bremen, Germany) coupled to an Eksigent two-dimensional nano HPLC (Eksigent technologies, Dublin, CA). With the use of auto-sampler 6 ml of sample was injected. The peptides were trapped on a pre-column (PepMap 100 C18, 5 mm x 0.3 mm, 5 mm LC Packings, Amsterdam, Netherlands) and separated on a reversed-phase analytical column (10 mm fused silica emitter, 75 mm x 16 cm (PicoTip Emitter, New Objective, Inc., Woburn, MA, USA) packed in-house with Reprosil-Pur C18-QA resin (3 mm, Dr.Maisch Gmbh, Ammerbuch-Entringen, Germany)). The pep-tides were loaded onto the pre-column at a flow rate of 10 ml/min for 15 min and separated on the reversed-phase analytical column at a flow rate of 300 nl/min using a 45 min linear gradient of 335% (v/v) acetonitrile in water containing 0.1% (v/v) formic acid. Solvent A consisted of 0.1% (v/v) formic acid in water and solvent B consisted of 0.1% (v/v) formic acid in acetonitrile. The total run time including washing and equilibration of column was 120 min. The LTQ-Orbitrap was operated in a data-dependent mode to automatically switch between Orbitrap-MS and LTQ-MS/MS acquisition. Four MS/MS spectra were acquired using CID (at 35% normalized collision energy) in the LTQ and each Orbitrap-MS scan was acquired at 60 000 FWHM nominal resolution setting using the lock mass option (m/z 445.120025) for internal calibration.
Data analysis
The generated data was analyzed using the Proteios SE for generating identifications both using Mascot and X!Tandem. All data was processed in Proteios SE [24], and processed in a label free workflow [25]. Raw data files were converted to mzML and MGF using Proteowizard. Searches were performed against all Saccha-romyces cerevisiae proteins in SwissProt as of 20150506 expanded for isoforms, supplemented with an equal number of decoy proteins (reverse sequence), totaling 13 480 sequences. The following
search parameters were used: enzyme trypsin, missed cleavage 1, fixed modification carbamidomethyl (C), variable modification methionine oxidation (O). A peptide mass tolerance of 5 ppm and fragment mass tolerance of 0.4 Da was used. Peptide identifications were generated by the automated database searches in both Mascot and X!Tandem and a combination were used with a false discovery rate (FDR) of 0.01. The identified peptides for all samples were used for creating binding profiles with the use of the tool WebLogo (v. 2.8.2) [26]. The reproducibility and peptide overlap between different replicates were analyzed with the use of Venn diagrams created by BioVenn [27].
FACS analysis
The coupled magnetic beads for both coupling systems, scFvs chemically coupled to M-270 carboxylic acid magnetic beads and biotinylated scFvs coupled to M-270 streptavidin magnetic beads, were analyzed in flow cytometry experiments to evaluate if scFvs had been successfully coupled to the surface. ScFvs coupled onto magnetic surface were detected with a mouse anti-His antibody (R&D systems, Minneapolis, MN, USA). A goat anti-mouse antibody conjugated with APC fluorochrome (BD Bioscience, San Jose, CA, USA) was used for detecting the bound anti-His antibody. For the analysis, 150 mg M-270 carboxylic acid beads and 50 mg M-270 streptavidin magnetic beads carrying scFv were incubated with anti-His antibody at different concentrations for about 1 h at RT. The beads were washed twice with TPBS before addition of the anti-mouse-APC secondary antibody. To evaluate the extent of activity of the coupled scFvs, beads were incubated for 1 h at RT with biotinylated peptides complexed to Alexa 647-labelled streptavidin (Thermo Fisher Scientific) (3:1 molar ratio) or only streptavidin conjugated Alexa-647 were included in the analysis. Flow cytometry analysis was performed with BD FACSCanto II (BD Bioscience) cell analyzer.
Results
Peptide capture and identification
In the GPS platform, the combination of CIMS scFvs directed against short peptide motifs and shot-gun MS allows for investigation of complex proteomes in a discovery oriented mode. In this study, a side-by-side comparison between three different versions of the GPS platform was performed using three different CIMS scFv and by analysis of a trypsin digested yeast (S. cerevisiae) sample. The digested protein sample was exposed to either immobilized or non-immobilized CIMS scFvs and the affinity enriched tryptic peptides were identified with the use of LC-MS/MS analysis. Figure 1 outlines the three different capture types. GPS capture 1 corresponds to the previously used version of the GPS platform [11], using CIMS scFvs chemically coupled to magnetic beads. GPS capture 2 exploits site-specific biotinylated CIMS scFvs coupled to streptavidin coated magnetic beads, a method that should result in a more controlled orientation of the attached scFvs. GPS capture 3 corresponds to the platform version were uncoupled biotinylated CIMS scFvs are first incubated with the tryptic peptides prior to immobilization onto streptavidin magnetic beads (in-solution capture). The MS based peptide identification was performed in the same way in all cases.
For all three included CIMS scFvs it could be observed that capture version 2 of the GPS platform, using scFv coupled through
4 www.elsevier.com/locate/nbt
New Biotechnology• Volume 00,Number 00• December 2015
TABLE 1_
The total number of detected peptides and the number of unique detected peptides for all replicate captures for the three different scFv capture system combinations.
scFv Capture Experiment Total number Number of
version of detected unique
peptides detected
peptides
CIMS17-C08 Capture 1 1 4 3
2 18 13
3 12 8
Capture 2 1 628 335
2 846 432
3 906 461
Capture 3 1 38 28
3 15 10
CIMS17-E02 Capture 1 1 9 7
Capture 2 1 774 375
2 1008 481
3 1017 487
Capture 3 1 530 276
2 439 230
3 562 307
CIMS33-3D-F06 Capture 1 1 2 1
Capture 2 1 228 131
2 126 65
3 521 285
Capture 3 1 94 60
2 132 77
3 107 69
Negative Capture 1 1 0 0
control 2 1 1
Capture 2 1 0 0
Capture 3 1 1 1
a biotinylated tag, resulted in the highest number of both total and uniquely detected peptides (Table 1). The CIMS17-E02 antibody showed the best performance with around 400 unique identified peptides in the oriented immobilization strategy of the platform (capture version 2). One of the problems with traditional MS methods have been low reproducibility and the stochastic nature of the data sampling contributes to that the expected reproduc-ibility of identified peptides in different LC-MS/MS runs is between 35 and 50% [28]. The GPS platform with an oriented coupling strategy (capture 2) showed a good repeatability with an overall peptide overlap between replicates in several cases close to 50% (Figure S2). The chemically coupling immobilization strategy (capture 1) resulted in a substantially lower number of identified peptides for all included CIMS scFvs (Table 1). Capture version 3 of the platform also performed well for two of the included CIMS scFvs, in particular for CIMS17-E02. This demonstrates the potential for further expansion of the GPS platform approach and the
possibility of using this capture format when relevant. Capture experiments performed with beads not carrying immobilized CIMS scFvs resulted in a very low number of captured peptides (Table 1). This indicates that the platform had a very low degree of background and that the identified peptides were a result of the affinity enrichment achieved with the included CIMS scFvs. Regarding the negative control samples for the chemically coupling immobilization strategy (capture 1), only one peptide was identified for one of the replicates. This implied that even though these capture experiments resulted in a low number of identified pep-tides, the identified peptides were actual identifications and not due to the background of the system.
To ensure that the lower degree of performance of the CIMS scFvs coupled through chemical activation was not a result of an unsuccessful coupling of the antibodies to the magnetic beads, flow cytometry experiments were performed on the coupled beads included in the different capture 1 and capture 2 type experiments. With the use of an anti hexahistidine tag specific monoclonal mouse antibody, CIMS scFvs coupled to both bead systems were identified (Fig. 2). Analysis of binding to biotinylated peptide/ streptavidin conjugates demonstrated that CIMS scFvs chemically coupled to beads showed poor binding activity towards their target peptides, while the CIMS scFvs coupled through the biotinylated BAD show good binding activity (Fig. 3). In sharp contrast and as illustrated in supplementary Figure 1, ELISA data show specific binding to intended target peptide for all three CIMS scFvs used for chemical coupling (panel a, c and e) prior immobilization. This is also true for the biotinylated versions of the same CIMS scFvs (panel b, d and f). We therefore conclude that the lower performance of the chemically coupled CIMS scFvs is a result of the immobilization method used.
Binding profile and length analysis
To further assess the performance of the different GPS platform variants (Fig. 1), we also evaluated the peptide binding motifs for the CIMS scFvs included in the investigation (Fig. 4). It has previously been established that the CIMS scFvs bind peptides with a certain binding motif [14]. It was pin-pointed that the last four C-terminal amino acid residues were important for binding and that 2 to 3 positions were more conserved and acted as anchor-positions for the peptide binding. It could once again be concluded that the CIMS scFvs individually expressed specific binding patterns (Fig. 4). The established binding profiles correlated very well with the previously determined motifs. The third last and last residue (denoted 4 and 6) seemed crucial for binding and could again be identified as anchor-positions while a large variation was allowed for the remaining positions. As seen in Fig. 4, the identified binding motifs are largely consistent regardless of the immobilization strategy or for the capture format used for each respective CIMS scFv. The agreement in binding profile is kept intact although a much lower number of peptides were identified for capture version 1 of the GPS platform. Indeed the majority of the identified peptides contain the specific binding profile (Table 2). Although the predominant peptide motif was conserved between different capture approaches, it is conceivable that other parts of the binding site contribute to peptide binding beyond the core of the binding site [29,30]. The composition of the enriched peptide pool may be effected in other ways. We therefore also
www.elsevier.com/locate/nbt 5
RESEARCH PAPER
FIGURE 2
Flow cytometry detection of CIMS scFv coupled to beads using an anti-His antibody. CIMS scFv CIMS17-C08 (a,b), CIMS17-E02 (c,d) and CIMS33-3D-F06 (e,f) were bound to magnetic beads by chemical coupling (a, c, e) or via biotin-streptavidin (b, d, f).
TABLE 2
The number of detected peptides without main binding profile for the oriented capture system. Previously determined binding profiles for the included CIMS scFvs were used to evaluate the amount of unique peptides identified in this capture experiment that did not contain the same binding motifs.
scFv Selection target peptide Capture version Binding profile Experiment Frequency of peptides lacking main binding profile
CIMS17-C08 SSAYSR Capture 2 XXX F/Y/W X R All replicates 1 2 3 75/554 33/335 50/432 55/461
CIMS17-E02 SSAYSR Capture 2 X X X Y X R/K All replicates 1 2 3 135/610 55/375 98/481 96/487
CIMS33-3D-F06 LSADHR Capture 2 XXXDXR/K All replicates 1 2 3 94/299 39/131 8/65 88/285
6 www.elsevier.com/locate/nbt
New Biotechnology• Volume 00,Number 00• December 2015
FIGURE 3
Flow cytometry detection of peptide binding to scFv coupled beads. CIMS scFv CIMS17-C08 (a,b), CIMS17-E02 (c,d) and CIMS33-3D-F06 (e,f) were bound to magnetic beads by chemical coupling (a, c, e) or via biotin-streptavidin (b, d, f). Detection reagents included Streptavidin-Alexa-647 alone (top panel), an irrelevant biotinylated peptide bound to Streptavidin-Alexa-647 (mid panel) and biontinylated peptide used for CIMS scFv development bound to Streptavidin-Alexa-647 (lower panel).
investigated the lengths of the detected peptides as a surrogate marker for peptide character (Fig. 5). The low number of peptides identified after use of capture version 1 precludes efficient assessment of this parameter. Peptides collected by capture version 2 and 3 demonstrate similar length profiles (Fig. 5). The number of
peptides without the binding motif varies between the different CIMS scFvs but the number of peptides without the predominant binding motif does not exceed 30%.
Overall, the identified peptides for respective CIMS scFv antibody correlate well in both binding pattern and length to what
www.elsevier.com/locate/nbt 7
RESEARCH PAPER
FIGURE 4
Binding profiles detected for the three different scFvs in all capture formats. The binding profiles created at WebLogos include all unique peptides identified in all replicates for respective CIMS scFv and capture system.
8 www.elsevier.com/locate/nbt
previously have been established by the former existing GPS platform [14]. It can be concluded that the detected peptides do not vary much in either binding motif or length regardless of how the GPS platform is performed.
Discussion
Affinity proteomics, such as antibody microarrays, have become an important contributor to the discovery phase of biomarker identification for different complex proteomes and sample formats [31-37]. However, drawbacks involve the need of a priori known target proteins. The technology's ability to detect new targets is thereby diminished. Traditional MS techniques with classical fractionation methods have had problems with reproduc-ibility and are less convenient for analysis of large sample cohorts or of analytes present at low concentrations in complex samples [38-40]. The GPS technology combines the strength of antibody technology and MS and has shown great promise for proteomic discovery studies with regards to quantitative capability, repro-ducibility and sensitivity [13]. The methodology offers possibilities in profiling complex proteomes in a specie independent discovery-directed fashion.
In this study, we have further developed the standard GPS platform by introducing a second strategy for immobilization of the peptide-enriching affinity reagents, the CIMS scFvs, exploited in GPS. Previously, the CIMS scFvs have been immobilized with the use of chemical covalent coupling in a non-oriented fashion. Here, biotinylation of a tag on the CIMS scFvs were employed for coupling in an oriented manner to streptavidin magnetic beads. This tag is located far away from the antigen binding site of the scFvs, guaranteeing that the antigen binding site of the CIMS scFv is exposed to the surrounding when immobilized. Other platforms combining affinity enrichment and MS, such as SISCAPA and TXP, have mostly utilized monoclonal or polyclonal full-length antibodies for the enrichment target peptides [8,41-43]. However, recombinant antibody fragments, e.g. Fabs, have also been shown to be applicable affinity reagents [44]. Here protein G or specific tag-sequences have been employed to ensure the orientation of the utilized antibody or antibody fragment. In all immunoaffinity based systems the activity of the affinity probe, such as an antibody or antibody fragment, highly influences the functionality and sensitivity of the system. The activity of an antibody is closely related to its structure and the integrity of its antigen binding site. There is always a risk that the immobilization procedure of the antibody affects the structure of the antibody and thereby its activity. Covalent immobilization strategies via free amine groups on the scFv entail a coupling procedure that can affect the antigen binding site and the activity of the antibody. Such effects depend on the positions of the functional groups used for the attachment and by the sequence of the antibody at hand. The procedure can both affect the antigen binding site and denature the antibody as a result of the strain introduced from multiple attachment sites as well as by steric hindrance caused by neighboring antibodies [45]. There can also be problems with batch to batch variations between different coupling batches as a consequence of the randomness of the coupling event. When instead utilizing an oriented coupling strategy, as for example via biotin-streptavidin, the coupling procedure can be better controlled. However, the biotinylation procedure of the antibody also needs to be considered. Chemical
New Biotechnology • Volume 00, Number 00 • December 2015
FIGURE 5
Peptide lengths profiles representing unique peptides for each CIMS scFv capture system combination. The detected median length for CIMS17-C08 was 10,11, and 11 amino acids for non-oriented (capture 1), oriented (capture 2) and in-solution strategies (capture 3). For CIMS-E02 respective median lengths were 8,13, and 13 amino acids, and for CIMS33-3D-F06 the respective median lengths were 9, 12, and 13 amino acids.
labelling with biotin via primary amines (e.g. side chains of lysines) is random. It can in the same way as the non-oriented immobilization procedure affect the activity and conformation of the antibody [46,47] and is prone to batch-to-batch variability. Site-specific biotinylation is therefore preferred and may be performed with different methods [20,21,48,49]. Here we use site-specific biotinylation, via the use of a biotin acceptor domain (BAD) and biotin ligase BirA, of the CIMS scFvs. Neither the biotinylation nor the coupling procedure via streptavidin magnetic beads should affect the activity of the scFvs to a great extent. It was confirmed with ELISA (Figure S1) that the biotinylation process of the scFvs do not extensively effect their activity towards the peptide used for antibody development. The use of site-specific biotinylation of the CIMS scFvs and an oriented attachment strategy with streptavidin magnetic beads offers a high control over the immobilization procedure and likely a minimal effect on antigen binding sites.
To assess the influence of a directed and controlled immobilization method with minimal effect on the paratope of the CIMS scFv, three different GPS systems were evaluated against a tryptic yeast proteome. It was concluded that the oriented GPS platform
(capture 2) exploiting biotinylated scFv performed better compared to the non-oriented chemically coupled system (capture 1) as shown by the number of identified peptides. In our view this is the consequence of a substantially better activity of CIMS scFv coupled via biotin to the magnetic beads and not to an unsuccessful coupling in terms of problems with the actual immobilization process. Importantly, the platform version utilizing biotinylated scFv has a good reproducibility despite of the stochastic detection of peptides in the data sampling.
The CIMS scFv antibodies in the GPS platform are directed towards short tryptic motifs. Olsson et al. previously determined that certain positions within the binding motif act as anchor positions and are very important for antigen binding while a much larger variation is allowed in the remaining positions [14]. The same properties could be identified for the CIMS scFvs in this study irrespective of the capture approach. Hence, we could conclude that the capture approach does not effect the model of how the CIMS scFvs capture respective sets of target peptides but rather influences the performance of the capture.
Overall, we have enhanced our GPS platform by adding a second strategy for immobilization of the CIMS scFvs for
www.elsevier.com/locate/nbt 9
RESEARCH PAPER
improved peptide capture performance. The new approach utilizes site-directed biotinylation of the CIMS scFvs and further on immobilization to streptavidin magnetic beads. This ensures an oriented coupling procedure and a much higher activity and capacity of the immobilized CIMS scFvs and the GPS platform. The new immobilization strategy also allows for the GPS platform to be used for in-solution captures, were the CIMS scFv are not coupled to beads until after capture, which further expands the applicability of the system.
Authors contributions
AS performed all experiments and analysis of all the data, as well as drafting the manuscript. CB contributed to the overall conceptual design of the method. HP, MO and CW contributed to the conceptual design of the study and manuscript editing.
Acknowledgements
We are grateful to MSc Magdalena Godzwon for performing the flow cytometry analysis, and to Dr. Karin M. Hansson for analyzing the samples on ESI-LTQ-Orbitrap. We also gratefully acknowledge Dr. Fredrik Levander supported by BILS (Bioinformatics Infrastructure for Life Science) for all assistance with the Proteios software, and Dr. Sofia Waldemarsson for valuable discussion regarding the mass spectrometric analysis. This project was supported by grants from Swedish National Research Council (VR-NT), the foundation for Strategic Research (SSF) (Strategic Center for Translational Cancer Research - CREATE Health), and Vinnova.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.nbt.2015.12.001.
References
[1] Anderson NL, Anderson NG. The human plasma proteome: history, character, and diagnostic prospects. Mol Cell Proteomics 2002;1:845-67.
[2] Hanash S. Disease proteomics. Nature 2003;422:226-32. http://dx.doi.org/ 10.1038/nature01514.
[3] Weiß F, van den Berg BHJ, Planatscher H, Pynn CJ, Joos TO, Poetz O. Catch and measure-mass spectrometry-based immunoassays in biomarker research. Bio-chim Biophys Acta 2014;1844:927-32. http://dx.doi.org/10.1016Zj.bba-pap.2013.09.010.
[4] Paulovich AG, Whiteaker JR, Hoofnagle AN, Wang P. The interface between biomarker discovery and clinical validation: the tar pit of the protein biomarker pipeline. Proteomics Clin Appl 2008;2:1386-402. http://dx.doi.org/10.1002/ prca.200780174.
[5] Betzen C, Alhamdani MSS, Lueong S, Schröder C, Stang A, Hoheisel JD. Clinical proteomics: promises, challenges and limitations of affinity arrays. Proteomics Clin Appl 2015;9:342-7. http://dx.doi.org/10.1002/prca.201400156.
[6] Listgarten J, Emili A. Practical proteomic biomarker discovery: taking a step back to leap forward. Drug Discov Today 2005;10:1697-702. http://dx.doi.org/ 10.1016/S1359-6446(05)03645-7.
[7] Cravatt BF, Simon GM, Yates JRI. The biological impact of mass-spectrometry-based proteomics. Nature 2007;450:991-1000. http://dx.doi.org/10.1038/na-ture06525.
[8] Anderson NL, Anderson NG, Haines LR, Hardie DB, Olafson RW, Pearson TW. Mass spectrometric quantitation of peptides and proteins using Stable Isotope Standards and Capture by Anti-Peptide Antibodies (SISCAPA). J Proteome Res 2004;3:235-44.
[9] Poetz O, Hoeppe S, Templin MF, Stoll D, Joos TO. Proteome wide screening using peptide affinity capture. Proteomics 2009;9:1518-23. http://dx.doi.org/10.1002/ pmic.200800842.
[10] Whiteaker JR, Lin C, Kennedy J, Hou L, Trute M, Sokal I, et al. A targeted proteomics-based pipeline for verification of biomarkers in plasma. Nat Bio-technol 2011;29:625-34. http://dx.doi.org/10.1038/nbt.1900.
[11] Olsson N, Wingren C, Mattsson M, James P, O'Connell D, Nilsson F, et al. Proteomic analysis and discovery using affinity proteomics and mass spectrome-try. Mol Cell Proteomics 2011;10. http://dx.doi.org/10.1074/mcp.M110. 003962.
[12] Sall A, Carlsson F, Olsson N, Wingren C, Ohlin M, Persson H, et al. AFFIRM—a multiplexed immunoaffinity platform that combines recombinant antibody fragments and LC-SRM analysis. J Proteome Res 2014;13:5837-47. http:// dx.doi.org/10.1021/pr500757s.
[13] Olsson N, James P, Borrebaeck CAK, Wingren C. Quantitative proteomics targeting classes of motif-containing peptides using immunoaffinity-based mass spectrometry. Mol Cell Proteom 2012;11:342-54. http://dx.doi.org/10.1074/ mcp.M111.016238.
[14] Olsson N, Wallin S, James P, Borrebaeck CAK, Wingren C. Epitope-specificity of recombinant antibodies reveals promiscuous peptide-binding properties. Protein Sci 2012;21:1897-910. http://dx.doi.org/10.1002/pro.2173.
[15] Olsson N, Carlsson P, James P, Hansson K, Waldemarson S, Malmstrom P, et al. Grading breast cancer tissues using molecular portraits. Mol Cell Proteom 2013;12:3612-23. http://dx.doi.org/10.1074/mcp.M113.030379.
[16] Wingren C, James P, Borrebaeck CAK. Strategy for surveying the proteome using affinity proteomics and mass spectrometry. Proteomics 2009;9:1511-7. http:// dx.doi.org/10.1002/pmic.200800802.
[17] Makaraviciute A, Ramanaviciene A. Site-directed antibody immobilization techniques for immunosensors. Biosens Bioelectron 2013;50:460-71. http:// dx.doi.org/10.1016/j.bios.2013.06.060.
[18] Yang H-M, Bao R-M, Cheng Y-Z, Tang J-B. Site-specific covalent attachment of an engineered Z-domain onto a solid matrix: An efficient platform for 3D IgG immobilization. Anal Chim Acta 2015;872:1-6. http://dx.doi.org/10.1016/ j.aca.2015.03.005.
[19] Cho I-H, Paek E-H, Lee H, Kang JY, Kim TS, Paek S-H. Site-directed biotinylation of antibodies for controlled immobilization on solid surfaces. Anal Biochem 2007;365:14-23. http://dx.doi.org/10.1016/j.ab.2007.02.028.
[20] Thie H, Binius S, Schirrmann T, Hust M, Dübel S. Multimerization domains for antibody phage display and antibody production. New Biotechnol 2009;26:314-21. http://dx.doi.org/10.1016/j.nbt.2009.07.005.
[21] Cloutier SM, Couty S, Terskikh A, Marguerat L, Crivelli V, Pugnieres M, et al. Streptabody, a high avidity molecule made by tetramerization of in vivo biotinylated, phage display-selected scFv fragments on streptavidin. Mol Immunol 2000;37:1067-77.
[22] Soderlind E, Strandberg L, Jirholt P, Kobayashi N, Alexeiva V, Aberg AM, et al. Recombining germline-derived CDR sequences for creating diverse single-framework antibody libraries.Nat Biotechnol2000;18:852-6. http://dx.doi.org/10.1038/ 78458.
[23] Sall A, Sjoholm K, Waldemarson S, Happonen L, Karlsson C, Persson H, et al. Development of phage-based antibody fragment reagents for affinity enrichment of bacterial immunoglobulin G binding proteins. J Proteome Res 2015;14:4704-13. http://dx.doi.org/10.1021/acs.jproteome.5b00585.
[24] Hakkinen J, Vincic G, Mansson O, Warell K, Levander F. The proteios software environment: an extensible multiuser platform for management and analysis of proteomics data. J Proteome Res 2009;8:3037-43. http://dx.doi.org/10.1021/ pr900189c.
[25] Sandin M, Ali A, Hansson K, Mansson O, Andreasson E, Resjo S, et al. An adaptive alignment algorithm for quality-controlled label-free LC-MS. Mol Cell Proteomics 2013;12:1407-20. http://dx.doi.org/10.1074/mcp.O112.021907.
[26] Crooks GE, Hon G, Chandonia J-M, Brenner SE. WebLogo: a sequence logo generator. Genome Res 2004;14:1188-90. http://dx.doi.org/10.1101/gr.849004.
[27] Hulsen T, de Vlieg J, Alkema W. BioVenn - a web application for the comparison and visualization of biological lists using area-proportional Venn diagrams. BMC Genom 2008;9:488. http://dx.doi.org/10.1186/1471-2164-9-488.
[28] Tabb DL, Vega-Montoto L, Rudnick PA, Variyath AM, Ham A-JL, Bunk DM, et al. Repeatability and reproducibility in proteomic identifications by liquid chro-matography-tandem mass spectrometry. J Proteome Res 2010;9:761-76. http:// dx.doi.org/10.1021/pr9006365.
[29] Keitel T, Kramer A, Wessner H, Scholz C, Schneider-Mergener J, Hohne W. Crystallographic analysis of anti-p24 (HIV-1) monoclonal antibody cross-reactivity and polyspecificity. Cell 1997;91:811-20.
[30] Jirholt P, Borrebaeck CAK, Ohlin M. Antibody evolution from the centre to the periphery: applied to a human antibody fragment recognising the tumour-associated antigen mucin-1. J Mol Biol 2002;318:407-16. http://dx.doi.org/ 10.1016/S0022-2836(02)00087-6.
[31] Borrebaeck CAK, Wingren C. High-throughput proteomics using antibody microarrays: an update. Expert Rev Mol Diagn 2007;7:673-86. http:// dx.doi.org/10.1586/14737159.7.5.673.
[32] Borrebaeck CAK, Wingren C. Design of high-density antibody microarrays for disease proteomics: key technological issues. J Proteomics 2009;72:928-35. http://dx.doi.org/10.1016/j.jprot.2009.01.027.
[33] Haab BB. Applications of antibody array platforms. Curr Opin Biotechnol 2006;17:415-21. http://dx.doi.org/10.1016/j.copbio.2006.06.013.
[34] Carlsson A, Wingren C, Kristensson M, Rose C, Ferno M, Olsson H, et al. Molecular serum portraits in patients with primary breast cancer predict the development of distant metastases. Proc Natl Acad Sci USA 2011;108:14252-57. http://dx.doi.org/10.1073/pnas.1103125108.
[35] Carlsson A, Wuttge DM, Ingvarsson J, Bengtsson AA, Sturfelt G, Borrebaeck CAK, et al. Serum protein profiling of systemic lupus erythematosus and systemic sclerosis using recombinant antibody microarrays. Mol Cell Proteomics 2011;10. http://dx.doi.org/10.1074/mcp.M110.005033.
10 www.elsevier.com/locate/nbt
New Biotechnology • Volume 00, Number 00 • December 2015
[36] Sanchez-Carbayo M. Antibody microarrays as tools for biomarker discovery. Methods Mol Biol 2011;785:159-82. http://dx.doi.org/10.1007/978-1-61779-286-1_11.
[37] Dexlin-Mellby L, Sandstrom A, Antberg L, Gunnarsson J, Hansson SR, Borre-baeck CAK, et al. Design of recombinant antibody microarrays for membrane protein profiling of cell lysates and tissue extracts. Proteomics 2011;11:1550-4. http://dx.doi.org/10.1002/pmic.200900808.
[38] Lange V, Picotti P, Domon B, Aebersold R. Selected reaction monitoring for quantitative proteomics: a tutorial. Mol Syst Biol 2008;4:222. http://dx.doi.org/ 10.1038/msb.2008.61.
[39] Lange V, Malmstrom JA, Didion J, King NL, Johansson BP, Schafer J, et al. Targeted quantitative analysis of Streptococcus pyogenes virulence factors by multiple reaction monitoring. Mol Cell Proteom 2008;7:1489-500. http:// dx.doi.org/10.1074/mcp.M800032-MCP200.
[40] Picotti P, Bodenmiller B, Mueller LN, Domon B, Aebersold R. Full dynamic range proteome analysis of S. cerevisiae by targeted proteomics. Cell 2009;138:795-806. http://dx.doi.org/10.1016/jxell.2009.05.051.
[41] Whiteaker JR, Zhao L, Zhang HY, Feng L-C, Piening BD, Anderson L, et al. Antibody-based enrichment of peptides on magnetic beads for mass-spectrom-etry-based quantification of serum biomarkers. Anal Biochem 2007;362:44-54. http://dx.doi.org/10.1016/j.ab.2006.12.023.
[42] Whiteaker JR, Zhao L, Abbatiello SE, Burgess M, Kuhn E, Lin C, et al. Evaluation of large scale quantitative proteomic assay development using
peptide affinity-based mass spectrometry. Mol Cell Proteomics 2011;10. http://dx.doi.org/10.1074/mcp.M110.005645.
[43] Hoeppe S, Schreiber TD, Planatscher H, Zell A, Templin MF, Stoll D, et al. Targeting peptide termini, a novel immunoaffinity approach to reduce complexity in mass spectrometric protein identification. Mol Cell Proteomics 2011;10. http://dx.doi.org/10.1074/mcp.M110.002857.
[44] Whiteaker JR, Zhao L, Frisch C, Ylera F, Harth S, Knappik A, et al. High-affinity recombinant antibody fragments (Fabs) can be applied in peptide enrichment immuno-MRM assays. J Proteome Res 2014. http://dx.doi.org/10.1021/ pr4009404.
[45] Peluso P, Wilson DS, Do D, Tran H, Venkatasubbaiah M, Quincy D, et al. Optimizing antibody immobilization strategies for the construction of protein microarrays. Anal Biochem 2003;312:113-24.
[46] Bayer EA, Wilchek M. Protein biotinylation. Meth Enzymol 1990;184:138-60.
[47] Jokiranta TS, Meri S. Biotinylation of monoclonal antibodies prevents their ability to activate the classical pathway of complement. J Immunol 1993;151: 2124-31.
[48] Scholler N, Garvik B, Quarles T, Jiang S, Urban N. Method for generation of in vivo biotinylated recombinant antibodies by yeast mating. J Immunol Methods 2006;317:132-43. http://dx.doi.org/10.1016/j.jim.2006.10.003.
[49] Predonzani A, Arnoldi F, Lopez-Requena A, Burrone OR. In vivo site-specific biotinylation of proteins within the secretory pathway using a single vector system. BMC Biotechnol 2008;8:41. http://dx.doi.org/10.1186/1472-6750-8-41.
www.elsevier.com/locate/nbt 11
