Surface Energy Modified Chips for Detection of Conformational States

(HRP) enzyme was shown, demonstrating that this method can be used to pattern biomolecules with their activity retained. The method presented coul...
0 downloads 11 Views 268KB Size
Download PDF
Langmuir 2006, 22, 2205-2211

2205

Surface Energy Modified Chips for Detection of Conformational States and Enzymatic Activity in Biomolecules Peter Åsberg,* K. Peter R. Nilsson, and Olle Ingana¨s Biomolecular and Organic Electronics, IFM, Linko¨ping UniVersity, S-581 83 Linko¨ping, Sweden ReceiVed October 17, 2005. In Final Form: December 6, 2005 A novel patterning method for anchoring biomolecules and noncovalent assembled conjugated polyelectrolyte (CPE)/biomolecule complexes to a chip surface is presented. The surface energy of a hydrophilic substrate is modified using an elastomeric poly(dimethylsiloxane) (PDMS) stamp, containing a relief pattern. Modification takes place on the parts where the PDMS stamp is in conformal contact with the substrate and leaves low molecular weight PDMS residues on the surface resulting in a hydrophobic modification, and then biomolecules and CPE/biomolecule complexes are then adsorbed in a specific pattern. The method constitutes a discrimination system for different conformations in biomolecules using CPEs as reporters and the PDMS modified substrates as the discriminator. Detection of different conformations in two biomacromolecules, a synthetic peptide (JR2E) and a protein (calmodulin), reported by the CPE and resolved by fluorescence was demonstrated. Also, excellent enzyme activity in patterned CPE/horseradish peroxidase (HRP) enzyme was shown, demonstrating that this method can be used to pattern biomolecules with their activity retained. The method presented could be useful in various biochip applications, such as analyzing proteins and peptides in large-scale production, in making metabolic chips, and for making multi-microarrays.

1. Introduction Biochips for studying protein recognition, conformations in biomolecules, and enzyme activity are very topical.1-3 There exist many ways to realize these, for instance ink-jet printing of biomolecules in solution, photolithographic patterning, microdispensing, soft lithography, and other methods. Some methods are based on adsorption of proteins to a surface and have been studied by protein deposition on wettability patterns.4,5 Soft lithography was introduced by Whitesides group in the early 1990s,6,7 and the initial focus was to provide a low-cost pattern transfer of alkanethiols onto gold films. Among soft lithographical methods, microcontact printing (µCP) has been the choice for transferring and patterning biomolecules onto substrates.8-13 µCP of proteins describe many possibilities, such as printing on mixed self-assembled monolayers,8 creating mixed surfaces of lipid membranes and proteins,14 printing of enzymes on gold substrates modified by µCP,,11 and even single protein printing on glass.13 When microstructured surfaces for protein confinement and * Corresponding author. E-mail: [email protected]. (1) Lueking, A.; Cahill, D. J.; Mu¨llner, S. Drug DiscoVery Today 2005, 10, 789-794. (2) Pavlickova, P.; Schneider, E. M.; Hug, H. Clin. Chim. Acta 2004, 343, 17-35. (3) Bodovitz, S.; Joos, T.; Bachmann, J. Drug DiscoVery Today 2005, 10, 283-287. (4) Sethuraman, A.; Han, M.; Kane, R. S.; Belfort, G. Langmuir 2004, 20, 7779-7788. (5) Sigal, G. B.; Mrksich, M.; Whitesides, G. M. J. Am. Chem. Soc. 1998, 120, 3464-3473. (6) Kumar, A.; Whitesides, G. M. Appl. Phys. Lett. 1993, 63, 2002-2004. (7) Kumar, A.; Biebuyck, H. A.; Whitesides, G. M. Langmuir 1994, 10, 14981511. (8) Tan, J. L.; Tien, J.; Chen, C. S. Langmuir 2002, 18, 519-523. (9) Graber, D. J.; Zieziulewicz, T. J.; Lawrence, D. A.; Shain, W.; Turner, J. N. Langmuir 2003, 19, 5431-5434. (10) Csucs, G.; Kunzler, T.; Feldman, K.; Robin, F.; Spencer, N. D. Langmuir 2003, 19, 6104-6109. (11) Wilhelm, T.; Wittstock, G. Langmuir 2002, 18, 9485-9493. (12) Bernard, A.; Fitzli, D.; Sonderegger, P.; Delamarche, E.; Michel, B.; Bosshard, H. R.; Biebuyck, H. Nat. Biotechnol. 2001, 19, 866-869. (13) Renault, J. P.; Bernard, A.; Bietsch, A.; Michel, B.; Bosshard, H. R.; Delamarche, E.; Kreiter, M.; Hecht, B.; Wild, U. P. J. Phys. Chem. B 2003, 107, 703-711. (14) Kung, L. A.; Kam, L.; Hovis, J. S.; Boxer, S. G. Langmuir 2000, 16, 6773-6776.

adsorption with µCP are compared, it is obvious that microstructuring is often much more elaborate. Biomolecules are sensitive, and so far the biological activity after printing has not been investigated thoroughly. In a recent study, printed antibodies, by µCP using patterned poly(dimethylsiloxane) (PDMS) stamps, were shown to retain much of their ability to bind the antigen.9 A recent development of µCP is patterned stamps of agarose hydrogels, where the biomolecule is stored inside the hydrogel in liquid conditions, rather then dried on a PDMS surface.15 Despite these studies, there are still some concerns, as µCP of proteins is a two step process where the protein is first adsorbed to/into a stamp and then transferred and adsorbed onto the receiving surface. Successful printing of proteins onto various substrates requires a minimum difference in contact angle between the stamp and substrate.8 If hydrophobic stamps can be used, they are preferred when printing proteins, but proteins have a tendency to both stick and denature on hydrophobic surfaces.16 Another potential concern is the surface of the PDMS stamp which is known to be very dynamic, and short oligomers diffuse from the bulk of the stamp17,18 wetting the stamp surface, due to a low glass transition temperature and high mobility of polymer chains. No systematic study of cotransfer of biomolecules and PDMS residues is available. To our knowledge, there does not exist any biochip with direct detection of conformations in biomolecules without modifying the biomolecule chemically. Also, an assay for enzymatic activity where the enzyme product can be monitored directly using fluorescence energy transfer (FRET)19 from the enzyme, close to the active site, and the product is also of interest. FRET is a good choice for this type of assay because it can detect minute amounts at extremely short distances.20-23 The present study is focused on making such biochips utilizing a novel patterning (15) Mayer, M.; Yang, J.; Gitlin, I.; Gracias, D. H.; Whitesides, G. M. Proteomics 2004, 4, 2366-2376. (16) Elwing, H. Biomaterials 1998, 19, 397-406. (17) Glasma¨star, K.; Gold, J.; Andersson, A. S.; Sutherland, D. S.; Kasemo, B. Langmuir 2003, 19, 5475-5483. (18) Wang, X. J.; O ¨ stblom, M.; Johansson, T.; Ingana¨s, O. Thin Solid Films 2004, 449, 125-132. (19) Fo¨rster, T. Faraday Soc. Discuss. 1959, 27, 1-17.

10.1021/la0527902 CCC: $33.50 © 2006 American Chemical Society Published on Web 01/12/2006

2206 Langmuir, Vol. 22, No. 5, 2006

method for biomolecules and a discrimination method for different conformations or biomolecule interactions. Candidate reporter molecules for detecting these biomolecular events are conjugated polyelectrolytes (CPEs) which offer possibilities for very sensitive measurements, which can become an important tool for genomics and proteomics. Conjugated polymers, such as polythiophene and polypyrrole, can couple analyte/receptor interactions, as well as nonspecific interactions, into observable responses.24-27 Amplification by a cooperative system response gives high sensitivity to minor perturbations of these polymers offering a key advantage compared to small molecules elements.20,21 Indeed, CPEs have been utilized to study many kinds of biomolecular events, and different biosensor devices are possible.22,28-35 An interesting class is the water soluble luminescent CPEs in which different conformations can be induced in the backbone by the act of biomolecular events.36-38 These polymers have been used to detect biomolecular interactions through optical absorption and luminescence, including DNA hybridization,36 different conformations in synthetic peptides39,40 and proteins.41,42 These CPEs are active and capable of changing their conformation on a solid support, as detected using surface plasmon resonance (SPR)43 and quartz crystal microbalance (QCM-D).44 A biochip is a preferred format for rapid screening of proteins, peptides, enzyme substrates, conformations in proteins, and other biomolecular interactions. In this study, we demonstrate three biochip variants based on the transfer of hydrophobic PDMS residues from a PDMS stamp to a hydrophilic substrate, giving a hydrophobic pattern on the hydrophilic surface.18 The PDMS(20) Marsella, M. J.; Carroll, P. J.; Swager, T. M. J. Am. Chem. Soc. 1995, 117, 9832-9841. (21) Heeger, P. S.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 1221912221. (22) Wang, J.; Wang, D. L.; Miller, E. K.; Moses, D.; Bazan, G. C.; Heeger, A. J. Macromolecules 2000, 33, 5153-5158. (23) Rasnik, I.; Myong, S.; Cheng, W.; Lohman, T. M.; Ha, T. J. Mol. Biol. 2004, 336, 395-408. (24) KorriYoussoufi, H.; Garnier, F.; Srivastava, P.; Godillot, P.; Yassar, A. J. Am. Chem. Soc. 1997, 119, 7388-7389. (25) Faı¨d, K.; Leclerc, M. J. Am. Chem. Soc. 1998, 120, 5274-5278. (26) Baek, M. G.; Stevens, R. C.; Charych, D. H. Bioconjugate Chem. 2000, 11, 777-788. (27) Kumpumbu-Kalemba, L.; Leclerc, M. Chem. Commun. 2000, 18471848. (28) Fan, C. H.; Plaxco, K. W.; Heeger, A. J. J. Am. Chem. Soc. 2002, 124, 5642-5643. (29) Wang, D. L.; Gong, X.; Heeger, P. S.; Rininsland, F.; Bazan, G. C.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 49-53. (30) Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 10954-10957. (31) Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. J. Am. Chem. Soc. 2003, 125, 896-900. (32) Liu, B.; Gaylord, B. S.; Wang, S.; Bazan, G. C. J. Am. Chem. Soc. 2003, 125, 6705-6714. (33) Kumaraswamy, S.; Bergstedt, T.; Shi, X. B.; Rininsland, F.; Kushon, S.; Xia, W. S.; Ley, K.; Achyuthan, K.; McBranch, D.; Whitten, D. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 7511-7515. (34) Xu, Q. H.; Gaylord, B. S.; Wang, S.; Bazan, G. C.; Moses, D.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 11634-11639. (35) Rininsland, F.; Xia, W. S.; Wittenburg, S.; Shi, X. B.; Stankewicz, C.; Achyuthan, K.; McBranch, D.; Whitten, D. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 15295-15300. (36) Nilsson, K. P. R.; Ingana¨s, O. Nat. Mater. 2003, 2, 419-U10. (37) Nilsson, K. P. R.; Rydberg, J.; Baltzer, L.; Ingana¨s, O. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 10170-10174. (38) Nilsson, K. P. R.; Andersson, M. R.; Ingana¨s, O. J. Phys.-Condens. Mater. 2002, 14, 10011-10020. (39) Nilsson, K. P. R.; Olsson, J. D. M.; Stabo-Eeg, F.; Lindgren, M.; Konradsson, P.; Ingana¨s, O. Macromolecules 2005, 38, 3813-6821. (40) Nilsson, K. P. R.; Rydberg, J.; Baltzer, L.; Ingana¨s, O. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 11197-11202. (41) Nilsson, K. P. R.; Ingana¨s, O. Macromolecules 2004, 37, 9109-9113. (42) Nilsson, K. P. R.; Herland, A.; Hammarstro¨m, P.; Ingana¨s, O. Biochemistry 2005, 44, 3718-3724. (43) Bjo¨rk, P.; Persson, N.-K.; Nilsson, K. P. R.; A° sberg, P.; Ingana¨s, O. Biosens. Bioelectron. 2005, 20, 1764-1771. (44) Åsberg, P.; Bjo¨rk, P.; Ho¨o¨k, F.; Ingana¨s, O. Langmuir 2005, 21, 72927298.

Åsberg et al.

modified substrates are used to control deposition and patterning of biomolecules and CPE/biomolecule complexes dissolved in a buffer solution. Sensitive molecules, such as biomolecules, should preferably be applied to the chip surface in their optimal solution. Micropatterning typically involves photolithographic techniques, not well adapted to biomolecules, and is avoided using our method. Patterning CPE/biomolecule complexes, where the CPE functions as a reporter, gives a biochip that can be used both for discrimination of conformations in the biomolecules and for screening of enzyme substrate/products. We report a method for fast evaluation of conformations in synthetic peptides and proteins decorated with conformational sensitive conjugated polymers through the optical emission from the patterned layers. Formation of enzyme formation product, from immobilized horseradish peroxidase, was detected on-line by direct excitation of the product or by FRET. 2. Experimental Section 2.1. Surface Modification of Glass Substrates using Poly(dimethylsiloxane) PDMS Stamps. PDMS modified substrates are made using a stamping procedure where a PDMS stamp creates a hydrophobic pattern on a hydrophilic substrate (step 1-3 in Figure 1 in the Supporting Information). We used glass as substrate, but other solid supports are possible. Stamps were prepared as follows: A master containing a relief pattern of SU-8 (Micro Chem Corp.) photoresist with 25 µm in height and several feature sizes, minimum width 50 µm, was prepared on a silicon wafer using standard photolithography; the pattern is later used as template for molding of PDMS. Silanization of the SU-8 structures on silicon is necessary to avoid sticking of the PDMS stamp when cured. This is done by submerging the template in a mixture of xylene (50 mL) containing 0.3 mL of dimethyldichlorosilane for 5 min and rinsed in xylene 3 times. A two component silicone elastomer, Sylgard 184 (Dow Corning Corp.), base and curing agent with mass ratio of 10:1 are mixed thoroughly in a beaker and degassed in order to avoid air bubbles. The elastomer mixture is then poured on the template and cured at 85 °C in a convection oven for 45 min; the PDMS stamp is peeled off after cooling. Surface modification is carried out by the following steps (Supporting Information, Figure 1): (1) The glass substrate is cleaned by immersion in a 5:1:1 mixture of deionized water, H2O2 (30%), and NH3 (25%) for 10 min at 85 °C and rinsed in Milli-Q water, drying (N2(g)), and oxygen plasma treatment (15 s, 150 W, 0.05 Torr). (2) A PDMS stamp is placed on the substrate. (3) After 25 min the stamp is removed from the substrate, short PDMS oligomers released from the bulk of the stamp has adhered on the surface creating a hydrophobic pattern on the hydrophilic glass substrate.17,18 2.2. Preparation of Solutions and Patterning. All buffers and polymer solutions were prepared in deionized water (Milli-Q water). The polymers used in this study were a zwitterionic polythiophene derivative poly(3-[(S)-5-amino-5-carboxyl-3- oxapentyl]-2,5-thiophenylene hydrochloride) (POWT),45 the anionic polythiophene derivative polythiophene acetic acid (PTAA),42 and the cationic poly (3-[(S)-5-amino-5-methoxycarboxyl-3-oxapentyl]-2,5-thiophenylene hydrochloride) (POMT)39 (Figures 1a, 2a, and 3). Polymer stock solutions were prepared at the day of use in deionized water at 1 mg/mL, roughly corresponding to 320-360 µM (an exact value is not possible to calculate at present due to varying chain lengths). Stock solutions at 100 mM concentration of MES-buffer at pH 5.9, Tris-HCl buffer at pH 7.8, and phosphate buffer (PB) at pH 6.8, pH 7.0 and pH 7.4. 10X PBS, phosphate-buffered saline (100 mM Na2HPO4, 100 mM NaH2PO4 and 1.5 M NaCl adjusted to pH 7.4) were prepared and diluted 10 times in Milli-Q water at the day of use. A total of 2.2 mg/mL JR2E,37 a synthetic peptide, stock was prepared in 10 mM MES-pH5.9 or PB-pH6.8. Lyophilized calmodulin (CaM) from bovine brain (Sigma-Aldrich, P2277, MW 16 680 Da) was (45) Andersson, M.; Ekeblad, P. O.; Hjertberg, T.; Wennerstro¨m, O.; Ingana¨s, O. Polym. Commun. 1991, 32, 546-548.

Surface Energy Modified Chips

Figure 1. Detection of different conformations in JR2E using POMT. (a) Fluorescence spectra (excitation at 400 nm) of POMT/JR2E at pH 5.9 (0) and POMT/JR2E at pH 6.8 (O). The chemical structure of POMT, a cationic polythiophene derivative (inset). (b,c) Typical patterning, by means of fluorescence microscopy, of (b) POMT/ JR2E at pH 5.9 and (c) POMT/JR2E at pH 6.8 on PDMS modified substrates. The scale bar corresponds to 100 µm. (d) Evaluation of the color content in panels b and c using Adobe Photoshop. diluted with 10 mM Tris-HCl pH 7.8 to a final stock concentration of 1.0 mg/mL. Horseradish peroxidase (HRP, EC: 1.11.1.7), purchased from Sigma-Aldrich (salt free powder, mol wt ∼40 000), was prepared in PB-pH7.0 at 320 µM stock concentration. Amplex UltraRed reagent (Invitrogen (A36006)) was diluted in DMSO to a stock concentration of 10 mM and stored at -20 °C until use. A 1:1 by weight PB (20 mM, pH 7.0)/glycerol solution was used to further dilute the Amplex UltraRed dye during the experiments. The following sample solutions, for patterning of PDMS modified substrates, were prepared at the time of use from the stock solutions and incubated for 5 min: POMT (0.1 mg/mL)/JR2E (0.3 mg/mL) in 10 mM MES-5.9 or 10 mM PB-6.8, POWT (0.1 mg/mL)/CaM (4 µM) complex in Tris buffer (10 mM, pH 7.8) with or without 10 mM calcium, solutions of 20 µM HRP and 5 µM PTAA (∼ 0.016 mg/mL) in 20 mM PB pH 7.0, and another containing just 20 µM HRP in 20 mM PB-pH7.0. The procedure to immobilize molecules from solution onto PDMS modified substrates can be generalized as follows (Supporting Information, Figure 1): After the 5 min incubation at room

Langmuir, Vol. 22, No. 5, 2006 2207

Figure 2. Detection of different conformations in calmodulin (CaM) using POWT. (a) Fluorescence spectra (excitation at 400 nm) of POWT (0), POWT +10 mM Ca2+ (O), POWT/CaM (4) and POWT/ CaM + 10 mM Ca2+ (]). The chemical structure of POWT, a zwitterionic polythiophene derivative (inset). (b,c) Typical patterning, by means of fluorescence microscopy, of (b) POWT/CaM and (c) POWT/CaM + 10 mM Ca2+ on PDMS modified substrates. The scale bar corresponds to 100 µm; vertical narrow stripes in the microscope images are 50 µm wide and correspond to the hydrophilic area. (d) Evaluation of the color content in panels b and c using Adobe Photoshop. temperature, a small amount (e.g., 10 µL), just to cover a small area of the pattern, of the solution is applied onto the modified surface and incubated for an appropriate time. The incubation time and temperature varies, it depends on the molecules and solvents, but 5 min at room temp is usually a good choice. Samples with HRP were incubated for 20 min. Evaporation of the liquid was avoided by placing the chip in a moist chamber. After incubation, the liquid is removed by blowing with an N2 gun, and at all times, the sample is protected from light and dust as much as possible. The immobilization step can be repeated if another molecule in solution, such as another biomolecule or CPE, is desired on the chip together with the first. After appropriate incubation this second solution is also removed by blowing with an N2 gun. 2.3. Microscopic Evaluation of the Chips and Photoluminescence Analysis in Solution. Evaluation of the chips is done with a Zeiss Axiovert 200M inverted light microscope equipped with an AxioCam HRc CCD camera and a SpectraCube SD-300 VDS

2208 Langmuir, Vol. 22, No. 5, 2006

Figure 3. Fluorescence characterization of PTAA and PTAA/HRP. Emission spectra (excitation at 400 nm) of PTAA in H2O (0), PTAA in PB-pH 7.0 (O) and PTAA/HRP in PB-pH 7.0 (4). The chemical structure of PTAA, an anionic polythiophene derivative (inset). (Applied Spectral Imaging, Inc. CA). The light source is a mercury lamp (HBO 100). Microscopic photoluminescence pictures were recorded in reflection mode using either 405 nm/30 nm, 470 nm/40 nm, or 546 nm/12 nm band-pass excitation and a 450, 515, or 590 nm long pass emission filter, respectively. The exposure time had to be selected to give the best result from each experiment. However, the exposure time was the same for each experiment within a series to allow accurate comparison between the images. The images recorded with AxioCam HRc were analyzed using Zeiss AxioVision PC software v3.1 or Adobe PhotoShop CS. By using Photoshop, we can analyze the intensity and to some extent the spectral content by evaluating the luminosity and the signal in the red, green, and blue channel in the image. To make the comparison of the color content in the region of interest, an average over at least 100 000 pixels, the color signals from each channel were divided by the luminosity and plotted in a bar graph. Images taken using the SpectraCube SD-300 VDS, Spectral Imaging v4.0 (build 0.01), was used for image acquisition and SpectraView v3.0.0.15 to analyze the images. The SpectraView system enables us to record a spectrum from each pixel in the image or an average spectrum from a region of interest. Acquisition with SpectraCube SD-300 VDS was done using the following parameters: spectral range, 450-650 nm; measurement type, SKY; spectral quality, max speed; exposure time, 200 ms; 405 nm/30 nm band-pass excitation; and a 450 nm long pass emission filter. Evaluation of HRP activity was done using Amplex UltraRed dye which is converted from nonphotoluminescent to photoluminescent Resorufin during enzyme (HRP) catalysis. Amplex UltraRed was diluted to 10 µM in PB-Glycerol containing 100 µM H2O2; the glycerol was added to reduce diffusion of Resorufin. Photoluminescence analysis in solution was performed with an ISA Jobin Yvon-Spex FluoroMax-2 fluorescence spectrometer using photoexcitation at 400 nm with the sample in 1 cm plastic cuvettes. Sample solutions from the patterning procedure were diluted in the buffer to a polymer concentration of 5 µg/mL and incubated at room temperature for 10 min before measuring fluorescence spectra in solutions.

3. Results Transfer of PDMS residues from a PDMS stamp is a fast and reliable method to modify the surface energy of a substrate.18 In principle, the surface modification is generated by placing a PDMS stamp on the substrate, and after removal of the stamp, a hydrophobic/hydrophilic PDMS pattern defined by the stamp has been created on the substrate. A sample solution with biomolecules is then applied onto the PDMS modified substrate. Depending on the hydrophobic/hydrophilic character or the conformation of the biomolecules, the biomolecules will stick

Åsberg et al.

to different parts of the surface. Hence, depending on the hydrophilic/hydrophobic character of the surface, specific patterns of biomolecules are generated. Here we have focused on adsorbing biomolecule/CPE complexes on modified glass providing an easy way to construct biomolecular arrays. Furthermore, the combination of CPE probes and PDMS modified surfaces for creating multi-arrays for the discrimination of conformational changes in biomolecules and measuring enzymatic activity in proteins is demonstrated. 3.1. Detection of Conformational Differences in a Synthetic Peptide. JR2E is a synthetic peptide37 which exists in a random coil state at pH 6.8 and as a four helix bundles at pH 5.9.39 This earlier study39 showed that the positively charged CPE, POMT (Figure 1a; inset), can be used as a conformation sensitive optical probe for discrimination between the two conformational states of peptide in solution. The fluorescence spectra of complexes between POMT and the synthetic peptide JR2E in MES pH 5.9 buffer (POMT/JR2E-pH5.9) and JR2E in PB pH 6.8 buffer (POMT/JR2E-pH6.8) are shown in Figure 1a. POMT/JR2EpH5.9 has its emission peak at 560 nm and POMT/JR2E-pH6.8 is red-shifted to 575 nm and has a lower emission yield. POMT in MES pH 5.9 or PB pH 6.8 solution has an emission peak at 563 nm (not shown). Thus, POMT is sensitive to structural changes in JR2E as well as in other peptides and can resolved by fluorescence spectroscopy (Figure 1a).37,39 POMT/JR2E complexes at pH 5.9 and 6.8 were applied on individual PDMS modified glass substrates, and microscopic images were recorded (Figure 1b,c). At pH 5.9 POMT/JR2E, where the peptide is a four helix bundle, stains the hydrophilic parts green, whereas at pH 6.8 POMT/JR2E, when the peptide is in random coil state, it stains the hydrophobic parts orange. Some aggregation is seen in the POMT/JR2E-pH6.8 samples, observed as orange dots (Figure 1c), in agreement with fluorescence data in solution (Figure 1b). These aggregates are too large to display any preference of adhesion. It might be possible to avoid such large aggregates by tuning the experimental conditions. POMT/JR2E-pH6.8 stains the hydrophobic parts of the chip, whereas POMT/JR2E-pH5.9 stains the hydrophilic areas (Figure 1b,c). To explain this difference in adhesion behavior, we need to understand how the polyelectrolyte interacts with the peptide. At pH 6.8, JR2E is negatively charged and POMT is positively charged causing the side chains of POMT to interact with the random coil peptide. When this takes place, the polyelectrolyte backbone is exposed and stretched out causing the POMT/JR2E complexes to aggregate, resulting in red shifted emission. The orange dots in Figure 1c are due to aggregation of POMT/JR2E complexes and confirm our interpretation on how these two molecules interact at this pH. When the JR2E peptide is in random coil state, it is most likely that it exposes more hydrophobic groups than as a four helix bundle. This paired with exposure of the POMT backbone is the most plausible explanation why POMT/JR2E-pH6.8 adsorbs to the hydrophobic parts. At pH 5.9, the POMT backbone might be buried in the JR2E four helix bundle as indicated by the circular dichroism and fluorescence data in an earlier study.39 The polymer is most likely sandwiched between, or resides within, the helix bundles; this supermolecule can be seen as a nanotube having a transducing polymer core with an insulating peptide cover. Therefore, the hydrophobic backbone of POMT is buried, and mainly hydrophilic groups in the POMT/JR2E complex at pH 5.9 are exposed resulting in staining of the hydrophilic parts in green. The color content in the stained areas on the chips (Figure 1d) shows that POMT/JR2E-pH6.8 contains more red color, less green color, and less blue color compared to POMT/JR2E-pH5.9, in agreement

Surface Energy Modified Chips

with the results obtained in solution. Hence, this novel technology using PDMS modified glass substrates in combination with the use of the optical changes in a CPE might offer the possibility to create multi-arrays for the discrimination of conformational changes in biomolecules. 3.2. Detection of Conformational Differences in Calmodulin. Calmodulin (CaM) is a protein with a pI of ∼4 that changes its conformation upon exposure to calcium ions (Ca2+).46,47 Upon binding of calcium, the compact calcium-free form of CaM is converted to a more open dumbbell-shaped molecule. A previous study38 has shown that a zwitterionic CPE, POWT (Figure 2a; inset), can be used as a conformation sensitive optical probe for discrimination between the two conformational states of CaM in solution. The emission spectra for POWT in Tris buffer (10 mM, pH 7.8) as well as POWT/CaM complex with or without 10 mM Ca2+ are shown in Figure 2a. Addition of Ca2+ to the buffered POWT solution does not change the emission spectra of POWT, and the emission peak remains at 565 nm in both cases, indicating a weak interaction between POWT and Ca2+ in this buffer. The addition of CaM in 1:8 (CaM:POWT) stoichiometry to the POWT solution without calcium shifts the emission spectra from 565 to 608 nm. A noncovalent complex between POWT and CaM, POWT/CaM, is formed under the proper buffer conditions. When Ca2+ is added to the POWT/CaM complex, there is again a significant change in the fluorescence spectra. Apart from blue shift of the emission peak to 601 nm, the shoulder around 540 nm is much more pronounced. These structural changes in CaM can be detected by POWT and are resolved by fluorescence spectroscopy in solution (Figure 2a), as shown in prior data.37,39 POWT/CaM and POWT/CaM-Ca2+ complexes were applied on individual PDMS modified glass substrates, and fluorescence microscopy analysis of the complexes reveals that they attach to the hydrophobic parts of the chip, but with different fluorescence colors (Figure 2b,c). POWT/CaM emits a more orange-green color with a lot of yellow dots, whereas POWT/ CaM-Ca2+ has a green color and hardly any yellow dots, in agreement with the results seen in solution (Figure 2b).38 The narrow green stripes in both images correspond to the hydrophilic area, and the green emission from these parts is from free POWT. Nonaggregated POWT molecules in Tris buffered solutions around pH 7.8 bind preferably to the hydrophilic parts and have a green fluorescence color (unpublished data). Proteins are expected to bind to the hydrophobic parts of the chip,16 and since the protein is dominating the POWT/CaM complex, this is what occurs. Evaluating the color content (Figure 2d) of the hydrophobic parts confirms that POWT/CaM is redder and less green than POWT/CaM-Ca2+, just as JR2E with different conformations (Figure 1d). The reason the adsorbed POWT/CaM complex (Figure 2b) has an orange-green color, with visible aggregation indicated by the many yellow dots, can be understood by knowing how POWT and CaM interact. CaM without calcium has a strong negative charge at pH 7.8, and the zwitterionic POWT side chains bind to CaM. The red-shifted emission for POWT/CaM suggests that the POWT backbone is stretched out in the interaction with CaM. A strong interaction between POWT side chains and CaM combined with the stretched out POWT backbone leads to exposure of the hydrophobic backbone giving aggregation due to the hydrophobic interactions, supported by fluorescence spectroscopy in solution38 and evaluation of the color content in the images. Large aggregates, yellow dots, do not display any (46) Zhang, M.; Tanaka, T.; Ikura, M. Nat. Struct. Biol. 1995, 2, 758-767. (47) Kuboniwa, H.; Tjandra, N.; Grzesiek, S.; Ren, H.; Klee, C. B.; Bax, A. Nat. Struct. Biol. 1995, 2, 768-776.

Langmuir, Vol. 22, No. 5, 2006 2209

Figure 4. Immobilization of PTAA, HRP and PTAA/HRP. (a,b) Typical patterning, by means of fluorescence microscopy (excitation at 405 nm), of (a) PTAA and (b) PTAA/HRP on PDMS modified substrates. (c,d) Immobilized and patterned HRP and HRP/PTAAcomplex, enzyme activity verified using Amplex UltraRed dye and visualized by means of fluorescence spectroscopy (excitation at 546 nm). (c) PTAA/HRP + Amplex UltraRed and (d) HRP + Amplex UltraRed. The horizontal narrow stripes in the microscope images are 100 µm wide and correspond to the hydrophilic area. The scale bar corresponds to 100 µm.

adsorption preference simply because they are too large. When 10 mM Ca2+ had been added to the sample before transfer to the chip, the adsorbed POWT/CaM-Ca2+ has a pronounced green color and hardly any aggregates (Figure 2c). In this case, the backbone in POWT is more twisted and CaM is less negatively charged due to binding two Ca2+ ions. The POWT/CaM-Ca2+ exposes a more hydrophilic surface to the surrounding and does not aggregate in aqueous solutions. Examining the absolute intensity in the images shows that bound POWT/CaM-Ca2+ has a somewhat higher intensity than POWT/CaM, also indicating separated POWT chains in the case with calcium. 3.3. Biological Activity of Patterned Proteins on PDMS Modified Substrates. Decorating and patterning an enzyme enables activity studies on adsorbed layers. We used the horseradish peroxidase (HRP) for this study. Fluorescence data in solution was obtained from PTAA (Figure 3; inset) in water, PTAA in PB-pH7, and PTAA/HRP in PB-pH7 (Figure 3). HRP was in a 4 to 1 molar excess over the PTAA in solution, ensuring that all polymer molecules were associated with HRP macromolecules. PTAA in PB is slightly red shifted, with an emission peak at 561 nm, compared to PTAA in water (emission peak at 554 nm). Luminescence from PTAA/HRP is blue shifted to 547 nm and has an increased fluorescence yield compared to both PTAA in water and PB, indicating a strong interaction between PTAA and HRP. The protein/polyelectrolyte interaction causes the backbone to become more nonplanar as well as increased separation between the individual PTAA chains, which also has been seen for PTAA/bovine insulin.42 The PTAA/HRP complex was immobilized to the PDMS modified substrates, and the adhesion and enzyme activity of the complex in the surface bound layer were studied. The chips were thoroughly rinsed in PBS and under streaming deionized water before analyzing them with fluorescence microscopy. Images of patterned PTAA and PTAA/HRP were taken using 405 nm excitation light (Figure 4a,b). In staining, the PTAA sample

2210 Langmuir, Vol. 22, No. 5, 2006

Figure 5. Time lapse study of Resorufin formation. Imaging, by means of fluorescence spectroscopy (excitation at 546 nm), of the temporal formation of Resorufin by HRP catalyzation of Amplex UltraRed. It is 3 s between the images, first image taken 1 s after addition of Amplex UltraRed. The scale bar corresponds to 100 µm.

adhered only to the hydrophilic parts of the chip, whereas the PTAA/HRP sample adhered only to the hydrophobic parts. The most likely reason for this is that the protein in PTAA/HRP dominates in the adhesion behavior and results in adsorption to the hydrophobic parts, similar to POWT/CaM (Figure 2). The polyelectrolyte/enzyme interaction is strong, since the PTAA bound to HRP could not be rinsed away despite excessive rinsing. PTAA seems to expose hydrophilic groups in PB-pH7 and therefore adheres to the hydrophilic parts (Figure 4a). HRP activity was evaluated by its ability to catalyze the reduction of H2O2 to H2O, verified using Amplex UltraRed which is converted to red fluorescent Resorufin during enzyme catalysis. Resorufin has its absorption peak at 571 nm and the emission peak is at 584 nm. A small drop (2-6 µL) of Amplex was placed on a cover slip and then pressed against the patterned HRP or HRP/PTAA on a glass slide. Glycerol (50 wt %) was used in the Amplex UltraRed solution to reduce diffusion of formed Resorufin. The microscopic fluorescence image (546 nm excitation) shows that there is a significant enzyme activity coming from the hydrophobic patches of the chip, both from PTAA/ HRP (Figure 4c) and HRP (Figure 4d) samples. A time lapse study of HRP activity performed on another patterned PTAA/ HRP sample using the same procedure as before shows the temporal increase of red fluorescent dye coming from the hydrophobic patches (Figure 5). Further evaluation of the PTAA/HRP chips was done by analyzing the spectrum from immobilized complexes using the SpectraView system. A PTAA/HRP sample, with and without Amplex UltraRed, was analyzed using the SpectraView system and 405 nm excitation light (Figure 6). Without Amplex, only the PTAA/HRP spectrum is seen from the sample with a peak at 546 nm. With Amplex, the peak at 546 nm decreased and a new peak at 586 nm appears; a HRP sample without PTAA but with Amplex did not result in any measurable emission signal using 405 nm excitation (not shown). This shows that we have FRET from the excited state in PTAA (bound to HRP in the PTAA/HRP complex) to Resorufin which then emit at 586 nm, Resorufin is not excited at 405 nm. FRET can only occur very close to the PTAA molecule;48 therefore, it was very difficult to get good pictures demonstrating this with our equipment (Supporting Information, Figure 2). A wide range of CPEs have been used as detecting elements for biological molecules in an aqueous environment, and many of these systems utilize the (48) Dung, H. T.; Knoll, L.; Welsch, D. G. Phys. ReV. A 2002, 65, art. no.043813.

Åsberg et al.

Figure 6. FRET between excited states in PTAA (PTAA/HRPcomplex) to formed Resorufin. Fluorescence from surface immobilized PTAA/HRP without Amplex UltraRed (O) and PTAA/ HRP with Amplex UltraRed (4). The spectra were recorded using the SpectraView system with excitation at 405 nm.

conditions for photoinduced charge or FRET.22,30 Here we have demonstrated FRET between a conjugated polyelectrolyte/ enzyme complex and an enzyme product.

4. Discussion and Conclusion We demonstrated a patterning and discrimination method for evaluation of biomolecules on a solid support. The immobilization method is based on modification of hydrophilic substrates, such as glass, using hydrophobic PDMS stamps containing a relief pattern. Several types of biomolecules can be immobilized and patterned on PDMS modified substrates. Noncovalent assembled CPE/biomolecule complexes were patterned on the modified substrates, and the CPEs, resolved with fluorescence, reported the biomolecular state. An important aspect with this method is that the biomolecules are dissolved in their optimal solvent when anchored to the chip surface. Anchoring was possible without engaging heating, photochemistry, covalent-coupling, dry adsorption and stamping or other protocols, generally used in typical biochip methods. The patterned substrates could differentiate between conformations in peptides and proteins. A biochip that can discriminate between different states in biomolecules has the potential to increase selectivity, speed, and simplicity of analysis which could give more efficient screening. Two variants were evaluated, Calmodulin (protein) and JR2E (synthetic peptide), both using CPEs as reporters. Using the PDMS modified substrates might provide enhanced selectivity and sensitivity to the detection of different conformations in peptides compared to solution detection, as the patterning procedure provides an extra discrimination dimension classifying the interactions as hydrophobic or hydrophilic. For the synthetic peptide, it was possible to distinguish between the two conformational states in the POMT/ JR2E complex both by color and by which area that is stained. The conformations in CaM were detected by the color on the adsorbed POWT/CaM complex to the hydrophobic parts. The color shifts were larger in the JR2E samples than in the CaM samples as seen in image analysis, probably because the differences between random coil and helix JR2E are larger than the two conformations in CaM. Surfaces can induce conformational changes in biopolymers; that is, the secondary structure of the protein fibronectin was somewhat disrupted on hydrophobic surfaces and became more extended on hydrophilic surfaces.49 (49) Baugh, L.; Vogel, V. J. Biomed. Mater. Res. A 2004, 69, 525-34.

Surface Energy Modified Chips

The hydrophobic parts of the PDMS modified substrates might therefore induce a slightly larger conformational change in POMT/ JR2E-pH6.8, resulting in an enhanced difference compared to POMT/JR2E-pH5.9 adsorbed on the hydrophilic parts. This surface induced conformation change might also be true for CaM, but since both forms of CaM bind to the hydrophobic parts, this effect is expected to be less significant then for JR2E. The conclusion from these results is that this method can provide more information than from just fluorescence spectroscopy. Another important aspect is that a chip-based method enables several samples to be spotted on a small area and evaluated at the same time which speeds up the testing. Large scale production of peptides and proteins, e.g., drugs, is associated with various problems such as, aggregation, random coil formation, and fibrillation. Standard methods, such as fluorescence spectroscopy and circular dichroism, generally require that the sample is transferred to special containers, and large volumes are needed. Smaller amounts of liquid are possible with a chip based method without loosing specificity and sensitivity. Our method permits active enzymes stained with CPEs to be immobilized and patterned, localized to the hydrophobic parts, on a substrate, thus making metabolic chips possible. An interesting application is the possibility to scan enzyme substrates and study metabolic activity. Despite the very thin PTAA/HRP

Langmuir, Vol. 22, No. 5, 2006 2211

layer and very small amount of Amplex UltraRed, the enzyme activity was excellent and easily detected. To our knowledge, no such chip method using the unique capabilities of CPEs and surface energy patterning previously existed. This could make it possible to study how long it takes for a substrate to diffuse out of the active site since it would only be possible to excite the product, by FRET, within the Fo¨rster radius (10-20 nm) and to make an assay that on-line measures the enzyme activity directly. We modeled the enzyme activity based on achieved data (Supporting Information, scetion 3). The method presented here is reliable, easy to use, and quick (results within minutes), and multimicroarrays with many different types of biomolecules should be fairly easy to accomplish. Acknowledgment. The Swedish Strategic Research Foundation has supported this work through Biomimetic Materials Science, Biomics. Supporting Information Available: Illustration of the process steps to modify a hydrophilic substrate (Figure 1). FRET between the excited state in PTAA (PTAA/HRP-complex) to Resorufin formed by HRP from Amplex UltraRed, visualized using fluorescence microscopy (Figure 2). FRET analysis (section 3). This material is available free of charge via the Internet at http://pubs.acs.org. LA0527902