NANO LETTERS
Cantilever Sensor for Nanomechanical Detection of Specific Protein Conformations
2005 Vol. 5, No. 12 2385-2388
Rupa Mukhopadhyay,*,†,‡ Vadim V. Sumbayev,†,§ Martin Lorentzen,†,‡ Jørgen Kjems,†,§ Peter A. Andreasen,†,§ and Flemming Besenbacher*,†,‡ Interdisciplinary Nanoscience Center, Department of Physics and Astronomy, and Department of Molecular Biology, UniVersity of Aarhus, 8000 Aarhus C, Denmark Received July 26, 2005; Revised Manuscript Received October 8, 2005
ABSTRACT A fast, label-free, and multiplexed method based on piezoresistive cantilevers is reported for the detection of specific protein conformations at the nanoscale level. The ligand-binding domain of the human oestrogen receptor (ERr-LBD) is used as the experimental model system, and ERr-LBD with or without oestradiol (E2) is detected using the conformation-specific peptides r/βI (Ser-Ser-Asn-His-Gln-Ser-Ser-Arg-Leu-IleGlu-Leu-Leu-Ser-Arg, which recognizes E2-bound ER) and r/βII (Ser-Ala-Pro-Arg-Ala-Thr-Ile-Ser-His-Tyr-Leu-Met-Gly-Gly, which recognizes E2-free ER). Target-specific signals are obtained in situ at protein concentrations of 2.5−20 nM. The in-build electrical readout of the piezoresistive cantilevers provides a convenient alternative to the conventional optical detection, and the presented method offers the possibility of detecting protein conformational changes using miniaturized microarrays.
Recently, it has been demonstrated that cantilever biosensors offer interesting possibilities for label-free detection of biomolecules, where the nanomechanical bending of the cantilever, induced by a differential surface stress upon the molecular binding to its functionalized surface, is registered either optically using a laser deflection system1 or electrically using piezoresistive readout.2 These techniques have been used as general methods for detecting various kinds of biomolecules based on, for example, the hybridization of complementary oligonucleotides1,2 and antibody-antigen binding.1,3,4 In these experiments, one of the binding partners was immobilized on the cantilever surface and the signal was elicited upon binding of the partner. The cantilever employed in the present study has an integrated piezoresistive element that changes resistance when the cantilever bends due to target binding. The electrical detection of DNA-DNA hybridization using piezoresistive cantilevers was demonstrated only recently.2 In the present study, the electrical readout method is used to distinguish between the conformations of oestrogen receptors (ERs) in free and oestradiol (E2)-bound form. The physiological effects of oestrogens are mediated via the ERs, ERR and ERβ, which are members of the nuclear receptor family of ligand-regulated transcription factors. When oestrogen binds to the ligand-binding domain (LBD) of ERs, a distinct and well-characterized conforma* Corresponding authors. E-mail:
[email protected];
[email protected]. † Interdisciplinary Nanoscience Center. ‡ Department of Physics and Astronomy. § Department of Molecular Biology. 10.1021/nl051449z CCC: $30.25 Published on Web 10/25/2005
© 2005 American Chemical Society
tional change occurs.5 The free and the E2-bound form of ERR-LBD were used to screen phage-displayed peptide libraries, and peptide sequences were isolated, which bound specifically to each of the two conformations, while having very low affinity to the other.6,7 The conformation-specific peptide binding to ERR is used here as the basis for sensing the two conformations. We describe a novel strategy for the use of cantilevers for following protein conformational changes based on a well-defined interaction between specific peptide sequences and specific protein conformations. This approach is different from previous articles reporting on the possible use of cantilevers for studying protein conformation. In one previous report, cantilever surface stress was suggested to change as a result of unfolding of adsorbed proteins.8 In another previous report, cantilever surface stress was suggested to change as a result of an organophosphate-induced conformational change of acetyl cholinesterase.9 It is also relevant to note that a single cantilever was used in both of these studies,8,9 and therefore no multiplexing was possible. The present report thus adds to the existing cantilever technology an exquisite control and versatility for detection of specific and well-characterized protein conformations preformed in the analyte solution by small biologically important molecules. The present experimental setup involves a two-cantilever (sensor and reference) configuration, where the peptides R/βI (Ser-Ser-Asn-His-Gln-Ser-Ser-Arg-Leu-Ile-Glu-Leu-Leu-SerArg) and R/βII (Ser-Ala-Pro-Arg-Ala-Thr-Ile-Ser-His-TyrLeu-Met-Gly-Gly) are attached on the gold-coated sides of
Figure 1. (a) Schematic drawing showing a two-cantilever configuration, the R/βI attached on one cantilever and the R/βII on the other, and the preferential binding of ERR-LBD, E2-bound or free, onto the R/βI and R/βII, respectively. (b) A cartoon showing the sensor layer on top of the surface and the blocking layer at the bottom surface of a cantilever and its bending upon target binding onto the top sensing surface.
each of the two cantilevers (Figure 1a). The silicon nitride backside of the cantilever is blocked by BSA treatment (see Supporting Information). When it is exposed to ERR-LBD in the presence of E2, the protein binding occurs preferentially onto the cantilever modified with R/βI, whereas in the absence of E2 the protein binds onto the cantilever modified with R/βII (Figure 1a). The piezoresistive cantilever, placed in a Wheatstone bridge, measures the change in resistance upon cantilever bending due to target binding. In a twocantilever sensor configuration, the preferential binding of a target probe onto one of the cantilevers can generate a positive or negative signal depending on the location of the cantilever in the bridge configuration. A distinct, positive differential signal was obtained upon the binding of E2bound ERR-LBD onto the R/βI-R/βII peptide combination (Figure 2a). The same sensing construct delivers a signal of opposite sign when exposed to ERR-LBD without E2 (Figure 2b). The reversal of the signal was interpreted as preferential binding of E2-bound LBD on R/βI and free LBD on R/βII. LBD binding reaches steady state quickly, irrespective of its solution concentration (20, 10, or 5 nM) (Figure 2c), and the steady state is maintained until the buffer replaces the sample in the flow cell. Relatively strong signals, 32.5 ( 2.5 and 27.5 ( 2.5 µV, are observed for 20 and 10 nM concentrations, respectively, from eight measurements each. The signal is reduced down to 10 ( 5 µV at 5 nM concentration (from four measurements), and 2-3 µV for concentrations less than 5 nM (data not shown). The decrease in the signal strength with the concentration indicates a dosedependent binding of LBD onto the peptide sensor layer (Figure 2c). The regeneration step using the low-pH buffer (50 mM glycine/HCl, pH 2.0) was carried out for 10 min 2386
immediately after an experiment. The chip was subsequently washed with the TBST buffer and equilibrated before another detection step was carried out. The response from a fusion protein consisting of glutathione S-transferase and ERR-LBD (GST-LBD, 64.2 kD, as compared to 34.5 kD for LBD alone) was monitored using the same peptide sensor combination R/βI-R/βII. The signals, +12.5 ((2.5) and -12.5 ((2.5) µV (from 5 measurements each), were observed for the E2-bound and E2-free forms of GST-LBD, respectively (Figure 2e and f). Independent control experiments showed that the GST alone does not bind to any of the peptides (data not shown). The specificity of the sensor was explored by monitoring the response of the E2-bound LBD in a buffer with 10 nM BSA (Figure 2d). A slightly weaker response was observed as compared to the E2-bound LBD in the absence of BSA. Importantly, a very weak response (0-2 µV) was observed when the chip was exposed only to the 10 nM BSA (Figure 2d). The signals from both LBD and GST-LBD were obtained reproducibly from the chips that were pre-exposed to BSA and generated very weak (0-2 µV) response upon BSA treatment, confirming that the signals are specific. The piezoresistive cantilever can respond to the generation of a differential stress between the top surface and the bottom surface of the cantilever when molecular adsorption occurs on the cantilever. The differential stress values and the end point deflections were calculated using methods similar to those described earlier.2 The LBD (10 nM) binding onto the cantilever generates a differential stress of 34.38 mN/m (for the mean readout 27.5 µV), whereas the stress is estimated to be 15.63 mN/m for binding of 10 nM of GST-LBD (mean readout being 12.5 µV). The corresponding end-point deflections are estimated to be 21.97 and 9.98 nm for LBD and GST-LBD, respectively. The reason for the marked difference in signal strength between the LBD and the GST-LBD is unknown at the moment. A further characterization of the observed difference between LBD and GST-LBD may lead to a better understanding of the factors determining the size of the deflection. Although there is a considerable difference in the sizes of the ligand-receptor complexes, that is, 34.8 kD for LBD and 64.5 kD for GST-LBD, the size does not seem to be decisive for the magnitude of the signal. In general, the voltage versus time curves shown in Figure 2 indicate fast uptake of the protein onto the peptide-coated surface, over the course of about 400 s, the signal depending on the absence or the presence of E2, and on which peptide is immobilized on the sensor cantilever. This association is followed by a decrease in surface stress due to sample flowing out of the fluidic chamber and the ER dissociating from the peptide. These curves do not indicate conformational change of the protein in real time because the mixture of the oestrogen receptor ligand binding domain (or ligand binding domain fused to GST) and oestradiol (E2) was injected after the mixture was incubated for at least 30 min. This means that what we measure is primarily the presence of a specific conformation, for example, E2-bound conformation, in the analyte solution. Nano Lett., Vol. 5, No. 12, 2005
Figure 2. (a and b) Nanomechanical detection of ERR-LBD (10 nM) in the E2-bound and free forms. (c) Comparative representative signals of E2-bound ERR-LBD at 20, 10, and 5 nM concentrations. (d) Target-specific response from a mixture of E2-bound ERR-LBD (10 nM) and BSA (10 nM) shown along with the control response from BSA (10 nM). (e) Response of GST-LBD (10 nM) bound to E2 and (f) ligand-free GST-LBD (10 nM). All of the signals were monitored in a continuous flow of TBST buffer (10 mM tris-HCl, pH 8.0; 150 mM NaCl; 0.05% Tween 20) at a 10 µl min-1 flow rate. The arrows indicate opening of the sample loop containing 100 µl of sample, which flows for 10 min at the given flow rate after it enters the fluidic chamber. Buffer replaces the sample in the fluidic chamber automatically at the end of this period. The peptides were dissolved in the coating buffer (100 mM Na2HPO4, 1 mM sodium ascorbate, 50 mM citric acid, pH 5.0) to the concentration of 30-50 nM prior to functionalization. The protein samples and E2 were taken in the TBST buffer. The E2 concentration is 50 times the protein concentration in all cases.
For the Wheatstone bridge configuration used in this study, a positive signal for the sensor-reference configuration or a negative signal for the reference-sensor configuration means downward bending of the sensor cantilever compared to the reference cantilever.10 The relative downward bending of the sensor cantilever indicates that compressive surface stress is generated in the molecular layer on the sensor cantilever with respect to the reference cantilever. The isoelectric points Nano Lett., Vol. 5, No. 12, 2005
of 6×His tagged LBD and GST-LBD are 6.1 and 5.6 respectively, which means that at the working pH of 8.0, both proteins would be predominantly negatively charged. The compressive stress, therefore, plausibly arises because of lateral electrostatic repulsion acting between the neighboring protein molecules confined within a monolayer that can be generated within the short time period (10 min) in which the protein molecules come in contact with the cantilever. 2387
The concern about the environmental accumulation of an increasing number of chemicals with ER ligand properties necessitates the development of biosensors for the detection of ligand-mediated ER conformational changes. Specific ER conformations induced by such chemicals can be distinguished by specific peptide recognition patterns.11 The results presented in this communication indicate that the cantilever assay is a single-step, label-free, and straightforward method for detecting protein conformational changes. Compared to ELISA, our method requires less material and a shorter data acquisition time. The direct, integrated readout makes the present cantilever sensor potentially cost-effective, userfriendly, and a candidate for miniaturized microarrays. Acknowledgment. We thank Cantion A/S for delivering the cantilever chips, Dr. M. Brown, Dr. J. A. Katzenellenbogen, and Dr. K. E. Carlson for very kindly supplying us with the reagents, and Dr. E. Laesgaard for assistance in the micromanipulator experiments. This work was supported by grant no. 2055-03-0004 from the Danish Research Agency to the iNANO Center at the University of Aarhus. Supporting Information Available: Methods for preparation of peptide and protein solutions, cantilever functionalization, and data aquisition. This material is available free of charge via the Internet at http://pubs.acs.org.
2388
References (1) Fritz, J.; Baller, M. K.; Lang, H. P.; Rothuizen, H.; Vettiger, P.; Meyer, E.; Gu¨ntherodt, H.-J.; Gerber, Ch.; Gimzewski, J. K. Science 2000, 288, 316. (2) Mukhopadhyay, R.; Loretnzen, M.; Kjems, J.; Besenbacher, F. Langmuir 2005, 21, 8400. (3) Wu, G.; Datar, R. H.; Hansen, K. M.; Thundat, T.; Cote, R. J.; Majumdar, A. Nat. Biotechnol. 2001, 19, 856. (4) Arntz, Y.; Seeling, J. D.; Lang, H. P.; Zhang, J.; Hunziker, P.; Ramseyer, J. P.; Meyer, E.; Hegner, M.; Gerber, Ch. Nanotechnology 2003, 14, 86. (5) Nilsson, S.; Makela, S.; Treuter, E.; Tujague, M.; Thomsen, J.; Andersson, G.; Enmark, E.; Pettersson, K.; Warner, M.; Gustafsson, J. A. Physiol. ReV. 2001, 81, 1535. (6) Paige, L. A.; Christensen, D. J.; Grøn, H.; Norris, J. D.; Gottlin, E. B.; Padilla, K. M.; Chang, C. Y.; Ballas, L. M.; Hamilton, P. T.; McDonnell, D. P.; Fowlkes, D. M. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 3999. (7) Tamrazi, A.; Carlson, K. E.; Katzenellenbogen, J. A. Mol. Endocrinol. 2003, 17, 2593. (8) Moulin, A. M.; O’Shea, S. J.; Welland, M. E. Ultramicroscopy 2000, 82, 23. (9) Yan, X.; Tang, Y.; Ji, H.; Lvov, Y.; Thundat, T. Instrum. Sci. Technol. 2004, 32, 175. (10) The sign of the signal upon bending of the cantilever was found using a micromanipulator. A force acting vertically was applied to the cantilever, and the resulting signal was monitored using the C-Box (Canti Lab) electronics. (11) Sumbayev, V. V.; Bonefeld-Jørgensen, E. C.; Wind, T.; Andreasen, P. A. FEBS Lett. 2005, 579, 541.
NL051449Z
Nano Lett., Vol. 5, No. 12, 2005
