Surface-Immobilized Peptide Aptamers as Probe Molecules for Protein

Jan 11, 2008 - Anal. Chem. , 2008, 80 (4), pp 978–983 ... Chemical Reviews 2014 114 (1), 285-366 ... The Journal of Physical Chemistry B 0 (proofing...
13 downloads 0 Views 161KB Size
Download PDF
Anal. Chem. 2008, 80, 978-983

Surface-Immobilized Peptide Aptamers as Probe Molecules for Protein Detection Steven Johnson,† David Evans,† Sophie Laurenson,‡ Debjani Paul,‡ A. Giles Davies,† Paul Ko Ferrigno,‡,§ and Christoph Wa 1 lti*,†

School of Electrical and Electronic Engineering, University of Leeds, Leeds, LS2 9JT UK, and MRC Cancer Cell Unit, Hutchison/MRC Research Centre, Hills Road, Cambridge, CB2 2XZ UK

We demonstrate the use of surface-immobilized, oriented peptide aptamers for the detection of specific target proteins from complex biological solutions. These peptide aptamers are target-specific peptides expressed within a protein scaffold engineered from the human protease inhibitor stefin A. The scaffold provides stability to the inserted peptides and increases their binding affinity owing to the resulting three-dimensional constraints. A unique cysteine residue was introduced into the protein scaffold to allow orientation-specific surface immobilization of the peptide aptamer and to ensure exposure of the binding site to the target solution. Using dual-polarization interferometry, we demonstrate a strong relationship between binding affinity and aptamer orientation and determine the affinity constant KD for the interaction between an oriented peptide aptamer STMcys+ pep9 and the target protein CDK2. Further, we demonstrate the high selectivity of the peptide aptamer STMpep9 by exposing surface-immobilized STMcys+ pep9 to a complex biological solution containing small concentrations of the target protein CDK2. A comprehensive understanding of protein pathways in cellular processes requires a thorough understanding of protein behavior and, in particular, the effect of drugs or cellular events on protein expression. Although significant advances have been made in this field over the past decade through detailed investigations of the transcriptome using oligonucleotide microarrays,1,2 there is no guarantee that any detected transcript makes a protein,20 nor can it be determined whether each protein is regulated by posttranslational modification.3,4 The ability to interrogate proteins directly in high throughput is an absolute requirement for postgenomic biology to deliver its promise.21,22,24 While traditional molecular biology techniques provide valuable information on the expression and function of proteins, these techniques cannot * To whom correspondence should be addressed. E-mail: [email protected]. † University of Leeds. ‡ Hutchison/MRC Research Centre. § Current address: Leeds Institute of Molecular Medicine, St. James’s University Hospital, Leeds LS9 7TF UK. (1) Brown, P. O.; Botstein, D. Nat. Genet. 1999, 21, 33-37. (2) Bertone, P.; Gerstein, M.; Snyder, M. Chromosome Res. 2005, 13, 259274. (3) Ramachandran, N.; Larson, D. N.; Stark, P. R. H.; Hainsworth, E.; LaBaer, J. FEBS J. 2005, 272, 5412-5425. (4) Mitchell, P. Nat. Biotechnol. 2002, 20, 225-229.

978 Analytical Chemistry, Vol. 80, No. 4, February 15, 2008

provide the massively parallel analysis necessary to map the entire proteomic space.4 Parallel sensing using arrayed systems has proved to be immensely successful in genomic research, and DNA microarrays are widely used for large-scale analysis of gene expression. However, the protein equivalent of the DNA microarray poses a significantly more difficult challenge, particularly in the identification of suitable high-affinity probe molecules, which retain their high specificity and functionality following immobilization on the arrayed sensor surface.5 The majority of current protein microarray technologies are based around antibodies whose biorecognition features have been exploited for many years in single-protein, single-analyte assays. The fabrication of complex diagnostic arrays based on antibodies has, however, proved challenging. Antibodies are typically large, multichain molecules with molecular masses of ∼150 kDa, that require post-translational modifications and disulfide bond formation for their stability. Their immobilization on surfaces leads to significant cross-reaction with proteins other than those to which they were selected to bind.6 Further, antibodies are generally selected against denatured proteins either in animals or in vitro and they predominately recognize only linear sections of peptide chains of the antigen,6 which limits their applicability where detection of native proteins is required. While significant progress has been made in selecting stable scFv antibody fragments to produce arrays,7 the biological space that any one scFv library can explore will be limited and alternatives are required. Recently, attention has turned toward reagents that may be considered as artificial antibodies, known as aptamers, made from short strands of DNA or RNA that adopt specific three-dimensional conformations allowing biorecognition of specific target proteins.8,9 While protein arrays based on DNA/RNA aptamers can borrow from the well-established immobilization protocols developed for the production of DNA microarrays, as with antibodies, selecting DNA/RNA aptamers that identify the conformation of native proteins remains problematic. However, although relatively unexplored, artificially engineered protein ligands, or peptide aptamers, are an attractive alternative biorecognition element in protein microarrays. Peptide aptamers consist of target-specific oligopep(5) Melton, L. Nature 2004, 429, 101-107. (6) Haab, P. B.; Dunham, M. J.; Brown, P. O. Genome Biol. 2001, 2, 4.1-4.13. (7) Wingren, C.; Steinhauer, C.; Ingvarsson, J.; Persson, E.; Larsson, K.; Borrebaeck, C. A. Proteomics 2005, 5, 1281-1291. (8) Kirby, R.; Cho, E. J.; Gehrke, B.; Bayer, T.; Park, Y. S.; Neikirk, D. P.; McDevitt, J. T.; Ellington, A. D. Anal. Chem. 2004, 76, 4066-4075. (9) Bock, L. C.; Griffin, L. C.; Latham, J. A.; Vermass, E. H.; Toole, J. J. Nature 1992, 355, 564-566. 10.1021/ac701688q CCC: $40.75

© 2008 American Chemical Society Published on Web 01/11/2008

Figure 1. (a) Schematic diagram of a surface-immobilized orientated STMcys+ pep9 -CDK2 complex. (b) Schematic diagram of HIS6CYS-STM fusion protein illustrating the location of the introduced cysteine residue at the amino terminus of STM.

tides inserted into a robust scaffold protein.10 The threedimensional constraint applied by the scaffold protein greatly increases the binding affinity of the inserted peptide,11 and the use of a single scaffold protein to present a wide range of different peptide recognition surfaces greatly simplifies the design of microarrays. Traditionally, the Escherichia coli protein thioredoxin A (TrxA) scaffold has been employed.12 However, many TrxAbased peptide aptamers are not expressed stably in prokaryotic cells, which will limit the applicability of these aptamers in highthroughput purification protocols. Recently, a protein scaffold, known as STM and based on a triple mutant of human stefin A, has been developed to address these problems.13 Furthermore, the STM scaffold has been engineered to give reduced interactions between human proteins and the scaffold, thus reducing crossreactivity typical with conventional protein microarray ligands.13 The STM scaffold has been mutated further to introduce a single cysteine residue (STMcys+) to allow the oriented attachment of the scaffold to a surface via the exposed sulfhydryl group.14 This cysteine residue is the only cysteine present in the scaffold and is located at the opposite side to the peptide insert region (Figure 1). While we have demonstrated previously that the cysteine residue enables direct immobilization on gold surfaces via the S-Au bond,14,15 we show here selective immobilization using a homobifunctional maleimide cross-linker for conjugation between the cysteine residue and sulfhydryl groups exposed on a thiolfunctionalized surface. This cysteine/maleimide-based immobilization chemistry is applicable to a wide variety of surfaces and thus extends the range of peptide aptamer applications, including their suitability for a number of protein array realizations. Here we demonstrate the oriented and specific immobilization of a peptide aptamer displayed by the STM scaffold with affinity (10) Crawford, M.; Woodman, R.; Ko Ferrigno, P. Briefs Funct. Genomics Proteomics 2003, 2, 72-79. (11) Ladner, R. C. Trends Biotechnol. 1995, 13, 426-430. (12) Colas, P.; Cohen, B.; Jessen, T.; Grishina, I.; McCoy, J.; Brent, R. Nature 1996, 380, 548-550. (13) Woodman, R.; Yeh, J. T.; Laurenson, S.; Ko Ferrigno, P. J. Mol. Biol. 2005, 352, 1118-1133. (14) Davis, J. J.; Tkac, J.; Laurenson, S.; Ko Ferrigno, P. Anal. Chem. 2007, 79, 1089-1096. (15) Evans, D.; Johnson, S.; Laurenson, S.; Davies, A. G.; Ko Ferrigno, P.; Wa¨lti, C. J. Biol., In press.

for cyclin-dependent kinase 2 (CDK2). We show, using dual polarization interferometry (DPI), that the affinity and specificity of the peptide aptamer is retained even following attachment to a surface, and hence, STM-based surface-immobilized peptide aptamers are viable probe molecules for protein array applications. CDK2 belongs to a group of proteins involved in the regulation of the cell cycle and, in particular, the progression of the mammalian cell from the G1 phase into the S phase.16,17,25 Deregulation of the CDK2 pathway in cancer cells is likely to require the inclusion of CDK2-specific probes in future proteinbased diagnostic arrays. The CDK2-interacting peptide aptamer (named STMpep9, where the subscript pep9 refers to the CDK2specific oligopeptide sequence) was generated by insertion of oligonucleotides encoding the previously identified CDK2interacting peptide sequence presented by thioredoxin12 into restriction sites in the open reading frame encoding the STM protein scaffold. Although we have previously characterized the surface immobilization of the empty scaffold protein,14 a detailed study of binding when the complete peptide-aptamer is immobilized on a surface, is required. MATERIALS AND METHODS Cloning of Peptide Aptamer. All oligonucleotides were purchased from Sigma-Genosys (Pampisford, UK), all enzymes from New England Biolabs (Hitchin, UK), and competent cells from Stratagene (Amsterdam, The Netherlands). Plasmid DNA and nucleotide DNA were prepared and cleaned using the appropriate Qiagen kits (Crawley, UK). A single cysteine residue was incorporated into the vector pET30a+ (Novagen, Beeston, UK) to generate the vector pET30a+CYS+ as described in ref cys+ 14. STMcys+ and STMpep9 were PCR amplified and cloned as in frame fusions to the hexahistidine tag of pET30a+CYS+ using the EcoRI and XhoI restriction sites. All constructs were sequence verified. Protein Expression and Purification. STMcys+ and cys+ were expressed as hexahistidine fusion proteins in STMpep9 BL21(DE3)pLysS E. coli cells (Stratagene). Cells were grown to mid-log phase at 37 °C and induced with 1 mM IPTG at 37 °C for 3 h. Cells were harvested by centrifugation and lysed for 20 min at room temperature with BugBuster primary amine-free Bacterial Lysis Reagent (Novagen, Beeston, UK) including Complete Protease Inhibitor Cocktail (Roche Molecular Biochemicals, Lewes, UK). The lysate was centrifuged at 4000g for 30 min and incubated with Ni-NTA resin (Qiagen, Crawley, UK) for 2 h at 4 °C. The Ni-NTA resin was washed with 30 volumes of wash buffer (50 mM sodium phosphate, 300 mM sodium chloride, 20 mM imidazole, pH 7.4). Protein was eluted from the resin with 2 volumes of elution buffer (50 mM sodium phosphate, 300 mM sodium chloride, 150 mM imidazole, pH 7.4). The eluted protein was exchanged into PBS using PD-10 columns (GE Healthcare, Little Chalfont, UK). Protein purification was analyzed by SDSPAGE and immunoblottling with anti-cystatin A monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA). Protein Extraction from Yeast Lysates. Human CDK2 in the vector pEG202 (a gift from Roger Brent (Harvard Medical School, Boston, MA)) was expressed in the yeast cells (EGY48, (16) Sherr, C. J. Appl. Phys. A: Mater. 1994, 79, 551-555. (17) O’Neill, E. M.; O’Shea, E. K. Nature 1995, 374, 121-122.

Analytical Chemistry, Vol. 80, No. 4, February 15, 2008

979

MAT R leu2::LexAop(x6)-LEU2 his3, trp1, ura3 pH18-34). Protein was extracted from 10-mL cultures grown from a single yeast colony inoculated into selective media lacking uracil and histidine. Cultures were incubated overnight at 30 °C, shaking at 225 rpm. The yeast cells were harvested by centrifugation at 2500g for 10 min and resuspended in 500 µL of cold yeast protein extraction buffer (50 mM Tris-HCl pH 7.4, 100 mM NaCl, 2 mM EDTA, 1× protease inhibitor cocktail). Approximately 300 mg of glass beads was added, and the samples were vortexed for 10 min at 4 °C. Cell debris was pelleted by centrifugation for 1 min at 16000g and the lysate transferred to fresh tubes. CDK-free lysate was obtained in a similar fashion. Baculoviral CDK2 was purchased from Cell Signaling Technologies (Danvers, MA). Dual-Polarization Interferometry. Protein immobilization and peptide aptamer-target interactions were interrogated in real time using an AnaLight Bio200 optical evanescent DPI (Farfield Sensors Ltd., Crewe, UK). This instrument has been described in detail previously.18,26 Briefly, DPI is a method for characterizing thin molecular films, based on the analysis of interference patterns resulting from coherent laser light propagating along two vertically stacked, independent optical waveguides. Interaction of the evanescent wave with a molecular layer(s) attached directly onto the exposed upper (sensing) waveguide results in a change in the relative phase of the light propagating along the two waveguides, leading to a shift in the observed interference pattern. By measuring these changes for two different optical polarizations (the transverse electric and transverse magnetic polarizations), it is possible to determine the thickness and refractive index of the adsorbed molecular film and, hence, the molecular density and mass per unit area. The technique provides a quantitative analytical measurement related directly to the structure and function of materials immobilized on the sensor surface. Each DPI sensor chip comprises two channels, i.e., two exposed waveguides, connected to a microfluidics cell consisting of a Rheodyne HPLC injector valve and an external pump (Harvard Apparatus, PHD2000). This simple fluidics system provides a controlled continuous fluid flow that can be directed across either, or both, experimental waveguide channels, together with an injection loop system for delivery of fixed fluid volumes. Prior to each experiment, each sensor chip was calibrated as follows. A phosphate-buffered saline (PBS) solution, pH 7.4, was passed over both sensor waveguides at a flow rate of 50 µL min-1 until a steady baseline was reached. This was followed by an 80% ethanol/water solution injected over both waveguides for 2 min before reverting to PBS. A 150-µL aliquot of deionized water was then injected over both waveguides for 2 min before returning again to PBS. This (18) Swann, M. J.; Peel, L. L.; Carrington, S.; Freeman, N. J. Anal. Biochem. 2004, 390, 190-198. (19) Reference deleted in proof. (20) Gygi, S. P.; Rochon, Y.; Franza, B. R.; Aebersold, R. Mol. Cell. Biol. 1999, 19, 1720-1730. (21) Mor, G.; Visintin, I.; Lai, Y.; Zhao, H.; Schwartz, P.; Rutherford, T.; Yue, L.; Bray-Ward, P.; Ward, D. C. Proc. Natl. Acad. Sci. U.S.A. 1995, 102, 77677682. (22) Michaud, G. A.; Salcius, M.; Zhou1, F.; Bangham, R.; Bonin, J.; Guo, H.; Snyder, M.; Predki1, P. F.; Schweitzer, B. I. Nat. Biotechnol. 2003, 21, 1509-1512. (23) Reference deleted in proof. (24) Emili, A. Q.; Cagney, G. Nat. Biotechnol. 2000, 18, 393-397. (25) Ladner, R. D.; Carr, S. A.; Huddleston, M. J.; McNulty, D. E.; Caradonna, S. J. J. Biol. Chem. 1996, 271, 7752-7757. (26) Karim, K.; Taylor, J. D.; Cullen, D. C.; Swann, M. J.; Freeman, N. J. Anal. Chem. 2007, 79, 3023-3031. (27) Reference deleted in proof.

980

Analytical Chemistry, Vol. 80, No. 4, February 15, 2008

procedure enabled the refractive index response of both the sensor chip and the running buffer (PBS) to be calibrated. All experiments were performed with a temperature control to (0.002 °C and with a set point of 20 °C, as controlled by the AnaLight Bio200 thermal control system. Immobilization of Peptide Aptamers. Immobilization of the cysteine-modified protein onto the silicon oxynitride waveguide surface (FB 100 sensor chips, Farfield Sensors Ltd.) was achieved using the homobifunctional, maleimide-based cross-linker 1,8-bis(maleimidoediethylene glycol) (BM(PEO)2, Pierce Biotechnology, Rockford, MA). The sensor chips were first cleaned by immersion in piranha solution (70% H2SO4, 30% H2O2) for 10 min and subsequently rinsed by sonicating in deionized water and ethanol for 10 min each. The cleaned sensor chips were then immersed in a 4% solution of mercaptotrimethoxysilane in isopropyl alcohol (IPA) for 18 h, followed by thorough rinsing in IPA, to provide a thiol functionalization. Following calibration, each thiol-modified waveguide sensor was subjected to two 150-µL injections of 20 µM BM(PEO)2 aqueous solution for 15 min before returning to PBS running buffer; the maleimide cross-linker allows conjugation between the sulfhydryl groups on the thiol-functionalized waveguide and the cysteine residue on the modified STM scaffold. Amine modification of the sensor surface was achieved by immersion of the sensor in a 4% solution of 3-(aminopropyl)triethoxysilane (Sigma) in IPA following the same piranha cleaning procedure. The sensor waveguides were subsequently exposed to protein solutions (either cysteine-modified STM scaffold (STMcys+), cyscys+ teine-modified STM containing the peptide aptamer (STMpep9 ), or nonmodified STM scaffold with neither cysteine modification nor the pep9-insert (STM)) by exposure to 150 µL of the appropriate protein at a flow rate of 10 µL min-1 before returning to PBS running buffer and increasing the flow rate to 50 µL min-1 to remove nonspecifically bound material. Upon reaching a stable state, the STM functionalized waveguides were finally exposed to 150 µL of CDK2 target solution at 10 µL min-1 before again returning to PBS running buffer at 50 µL min-1. RESULTS AND DISCUSSION Oriented Attachment of the STM Scaffold. The real-time increase in thickness and mass resulting from immobilization of the maleimide-based cross-linker BM(PEO)2 is shown in Figure 2. Following two injections, the BM(PEO)2 layer was found to be highly stable with a thickness of 1.42 nm, which compares well with the calculated size of the molecule (1.47 nm), suggesting the formation of a chemisorbed monolayer. Oriented attachment of the STM scaffold via conjugation between its cysteine residue and the BM(PEO)2 functionalized surface was demonstrated by exposing the maleimide-terminated monolayer to 150 µL of a 100 nM STMcys+ protein solution. The resulting real-time increase in thickness and mass is shown in Figure 2 and suggests a thickness of the STMcys+ adlayer of 2.7 nm. To confirm that the observed protein layer is indeed the STMcys+ protein scaffold rather than a contaminating protein that copurified with it during protein expression, the immobilized protein layer was exposed to an anti-cystatin A antibody (antiSTM), which binds to a preserved sequence of the modified STM scaffold. The anti-STM was injected over an STMcys+ adlayer at a concentration of 10 nM leading to 1.1 ng/mm2 of anti-STM bound to the STMcys+ monolayer (see Figure 3, solid line) and hence confirming the presence of the STM scaffold. Furthermore, this

Figure 2. Real-time DPI measurement of (a) thickness and (b) mass per unit area of an STMcys+ adlayer formed on a BM(PEO)2 functionalized waveguide sensor. The start and end of each injection are indicated by solid and broken arrows, respectively. The large discontinuities that occur at the start and end of each sample injection mostly reflect the transient change in refractive index of the bulk solution rather than the formation of a new layer on the waveguide surface. Steady-state, calibrated measurements of each adlayer are obtained from the plateau regions between injections (for example, marked as A).

Figure 3. Real-time DPI measurements of mass per unit area of antiSTM antibody immobilized on a nominally identical cysteinemodified, oriented STM scaffold layer, STMcys+ (solid line), and an unmodified (nonoriented) STM scaffold adlayer (broken line). The start and end of the antibody injection are indicated by a solid and broken arrow, respectively.

increase in mass is approximately three times that bound by a largely identical layer formed by physisorbing unmodifed (i.e., cysteine-free) STM scaffold proteins onto an independent waveguide surface (broken line in Figure 3). The increased binding of antiSTM to STMcys+ compared to STM results from the controlled orientation of the chemisorbed protein layer through the thiol-

Figure 4. Real-time DPI measurement of (a) thickness and (b) mass of STMcys+ and STMcys+ pep9 functionalized waveguide sensor. The start and end of the sample injections are indicated by solid and broken arrows, respectively.

maleimide bond leading to increased exposure of the anti-STM binding site. This is in good agreement with our previous studies of STM scaffold immobilization on a bare gold surface14,15 and confirms the applicability of maleimide-derivatized surfaces for specific immobilization of cysteine-modified proteins.14 STMpep9 Interactions with CDK2. To test whether the immobilized scaffold protein can present an interaction surface to recognize a specific target protein, two independent waveguides were functionalized with maleimide followed by exposure to 100 cys+ nM solutions of either the STMpep9 aptamer, which contains the CDK2-binding peptide insert identified in a screen of a TrxAconstrained library,12,13 or the non-CDK2-interacting STMcys+ protein scaffold. The real-time increase in thickness and mass occurring on the two waveguides is shown in Figure 4a and b, cys+ and respectively. While the rate of adsorption of the STMpep9 cys+ STM adlayers is largely identical, reflecting the thiol-maleimide bond formation kinetics, the final mass and thickness of cys+ the two layers is consistently different, with the STMpep9 monolayer being larger with an approximate thickness of 3.3 nm. These small differences in mass and thickness may be attributed to the cys+ increased size of STMpep9 compared to STMcys+ owing to the insertion of the peptide aptamer sequence. In order to assess the suitability of the peptide aptamer layer cys+ for the detection of specific proteins from solution, the STMpep9 cys+ and STM monolayers were simultaneously exposed to a complex protein solution generated by lysis of CDK2-expressing yeast cells. While the use of a cell lysate leads to a significant background signal resulting from nonspecific interactions, in most clinically relevant specimens, the proteins of interest are present only at very low abundance and in complex biological mixtures. Analytical Chemistry, Vol. 80, No. 4, February 15, 2008

981

Figure 5. Real-time DPI measurements showing increase in mass of material immobilized on STMcys+ (broken line) and STMcys+ pep9 (solid line) adlayers following exposure to (a) yeast cell lysate containing CDK2 and (b) yeast lysate lacking the expression of CDK2. The start and end of the CDK2 injections are indicated by a solid and broken arrow, respectivley.

Thus, the use of a lysate here is particularly relevant to future applications of peptide aptamers in clinical environments. The cys+ layer owing to the mass of material immobilized on the STMpep9 exposure to the CDK2-expressing yeast lysate is approximately double that observed on the control STMcys+ layer (see Figure 5a). The mass of material immobilized on the STMcys+ layer results from a nonspecific interaction between the STMcys+ layer and material contained within the lysate, assuming nonspecific interaccys+ tions to be similar on both the STMpep9 and STMcys+. Given the cys+ molecular masses of the STMpep9 (19.6 kDa) and CDK2 (33.4 cys+ kDa), we estimate that around a third of the STMpep9 proteins contained within the layer are bound to a CDK2 protein. We note that this may be an underestimate, as the specific high-affinity binding of CDK2 is likely to compete with, and sterically inhibit, the nonspecific binding of yeast proteins with low affinities for STM. To confirm that the observed increase in mass is related to the formation of CDK2-STMpep9 complexes, rather than to binding with other species contained within the lysate, we simultaneously cys+ exposed a pair of waveguides functionalized with STMpep9 and cys+ STM to a lysate generated from identical yeast cells, but lacking the expression of CDK2 (Figure 5b). The mass of material bound to the protein functionalized waveguide surfaces is identical on cys+ both waveguides. Given that the STMpep9 and STMcys+ layers differ only in the presence (absence) of the peptide aptamer insert, the variation in immobilized mass following exposure to the CDK2 lysate can be attributed to the specific interaction through the aptamer region, thus confirming the affinity of STMpep9 for CDK2. 982 Analytical Chemistry, Vol. 80, No. 4, February 15, 2008

Figure 6. Real-time DPI measurements showing (a) mass and (b) thickness of oriented STMcys+ pep9 layer formed on a maleimide functionalized sensor surface and nonoriented STMcys+ pep9 layer physisorbed on an amine functionalized sensor surface. (c) Real-time DPI measurements showing increase in mass immobilized on oriented and nonoriented STMcys+ pep9 layers following exposure to a 20 nM solution of CDK2 in a yeast cell lysate. The start and end of the protein injections are indicated by solid and broken arrows, respectively.

Oriented versus Nonoriented STMpep9. Unlike solutionbased systems, where the peptide aptamer can move freely in all three dimensions, thus exposing and maximizing interactions with the binding site, the reduced dimensionality associated with surface immobilization may limit access of the target protein to the binding site, therefore reducing binding efficiency. It is thus critical to orient correctly the immobilized scaffold such that the target-specific binding site is presented to the sample solution.14 To confirm the efficacy of oriented attachment through the cysteine residue, we compare the adsorption of CDK2 on both cys+ chemisorbed (oriented) STMpep9 adlayers formed on a maleimide functionalized sensor surface and physisorbed (nonoriented) cys+ STMpep9 adlayers formed on an amine functionalized surface. While the adsorption kinetics differ, with the physisorbed layer

Figure 7. Binding analysis of CDK2-STMcys+ pep9 (solid circles) and CDK2-STMcys+ (open circles) interaction. Solid line shows fit to data. Inset: Linearized form of binding data where the y-axis label M/C corresponds to the immobilized mass divided by CDK2 concentration.

forming more rapidly (Figure 6a and b), the resulting oriented and nonoriented layers are largely identical in terms of thickness and mass of immobilized material. This suggests an equivalent cys+ number of STMpep9 aptamers is contained on the two layers, therefore allowing for effective comparison of interactions occurring between the two adlayers and CDK2 target solution. Figure 6c shows that the mass of material immobilized on an oriented cys+ STMpep9 layer following exposure to a 20 nM solution of baculoviral CDK2 diluted in yeast lysate is 0.4 ng/mm2. Removing the background mass arising from nonspecific interactions with species contained within the cell lysate, obtained through similar measurements using STMcys+, we estimate that CDK2 contributes 0.33 ng/mm2 of the total immobilized mass. In contrast, there was no measurable increase of mass following exposure of cys+ nonoriented STMpep9 to an identical lysate solution (see Figure 6c). This lack of interaction was confirmed for a range of CDK2 solutions, up to a maximum concentration of 50 nM. The striking cys+ absence of interactions between the nonoriented STMpep9 and CDK2 clearly demonstrates the importance of orientation-specific immobilization. Binding Affinity of STMcys+ pep9. To study the binding affinity of the peptide aptamer to the CDK2 target, we exposed a series of cys+ identical STMpep9 layers to a yeast cell lysate doped with different concentrations of baculoviral CDK2. Figure 7 shows the mass of material deposited on the peptide aptamer adlayers as a function of CDK2 concentration, and a clear increase in mass of immobilized material with increasing CDK2 concentration is observed. In contrast, no relationship between the mass of immobilized material and CDK2 concentration was observed

following exposure of STMcys+ layers to CDK2-containing yeast lysate. Both data sets have been corrected to account for the background adsorption of CDK2-free cell lysate material. The solid line in Figure 7 represents a fit of M ) (CBmax)/(C + KD) to the data in linearized form (Scatchard plot, shown in the inset of the figure), with M the immobilized mass, C the CDK2 concentration, and Bmax the saturation mass. The fit reveals an equilibrium affinity constant KD ) 40 ( 2 nM. The extracted affinity constant is in good agreement with binding analyses of surface-immobilized CDK2-interacting peptide aptamers inserted in the TrxA scaffold where KD was found to be in the range 30-120 nM, dependent on oligopeptide sequence.12 Furthermore, this value of KD compares well to typical values of KD for surface-immobilized antibodies.14 CONCLUSIONS A comprehensive investigation of protein function and proteinprotein interactions presents a number of significant challenges, particularly in the identification of suitable high-affinity probe molecules. Here, we have demonstrated a potential solution to this problem through the use of target-specific oligopeptides presented and constrained by a robust and generic scaffold protein. These peptide aptamers have been selected from aptamer libraries using yeast-two-hybrid screening and hence bind to the native conformation of the target proteins. The STM scaffold protein has been further modified to include a single cysteine residue at the N-terminus allowing for oriented surface-immobilization, here demonstrated using maleimide functionalized surfaces. We have shown that these surface-immobilized peptide aptamers exhibit very high selectivity, enabling detection of specific target proteins from complex biological solutions. Finally, binding analyses demonstrated that the binding-affinity of correctly oriented peptide aptamers compares well with that of typical antibody-antigen interactions. ACKNOWLEDGMENT We acknowledge the support of the EPSRC (UK), the Research Councils UK Basic Technology Program, the BBSRC (UK), the University of Leeds Institute for BioNanosciences, and a grantin-aid from the Medical Research Council (UK) to the Cancer Cell Unit. S.L. gratefully recognizes the award of a scholarship from the Commonwealth Commission. Received for review August 8, 2007. Accepted October 23, 2007. AC701688Q

Analytical Chemistry, Vol. 80, No. 4, February 15, 2008

983