Bioconjugate Chem. 2007, 18, 1705–1708
1705
Energy Transfer between CdSe/ZnS Core/Shell Quantum Dots and Fluorescent Proteins Vitor R. Hering,*,† Gary Gibson,‡ Robert I. Schumacher,† Adelaide Faljoni-Alario,† and Mário J. Politi† Departamento de Bioquímica, Instituto de Química, Universidade de São Paulo, São Paulo, SP 05513-970, Brazil, and Hewlett-Packard Laboratories, Palo Alto, California 94304. Received April 25, 2007; Revised Manuscript Received September 11, 2007
Fluorescent proteins from the green fluorescent protein (GFP) family interact strongly with CdSe/ZnS quantum dots. Photoluminescence of GFP5 is suppressed by red-emitting CdSe/ZnS quantum dots with high efficiency in a pH-dependent manner. The elevated degree of quenching, around 90%, makes it difficult to analyze the remaining signal, and it is not clear yet whether FRET is the reason behind the quenching. When the donor is a greenemitting CdSe/ZnS quantum dot and the acceptor is the HcRed1 protein, it is possible to detect quenching of the donor and sensitized emission from the acceptor. It was verified that the sensitized emission has the low anisotropy characteristic of FRET. The present characterization identifies donor–acceptor pairs formed by fluorescent proteins and CdSe/ZnS quantum dots that are suitable for the exploration of cellular events. These donor–acceptor pairs take advantage of the exceptional photochemical properties of quantum dots allied with the unique ability of fluorescent proteins to act as gene-based fluorescent probes.
INTRODUCTION CdSe/ZnS core/shell quantum dots and fluorescent proteins belonging to the green fluorescent protein (GFP) family provide components among which Förster resonance energy transfer (FRET) is possible. The emission and absorption profiles of these elements can be matched to optimize FRET, and the distances can be kept below 10 nm. The occurrence of FRET in such system suggests a great diversity of possible donor– acceptor pairs and, because both components play important roles as biological labels, potential usefulness in cell biology experiments that make use of donor–acceptor labels. CdSe/ZnS core/shell quantum dots have large absorption cross sections, high photoluminescence (PL) quantum yields (QY), size-tunable PL spectra covering the region from blue to red, exceptional photochemical stability (1, 2), and good dispersibility in biological fluids (3–6). The GFP family is composed of a group of homologous, colored, compact proteins that spontaneously generate their own fluorophore (7–9). These proteins also cover the visible spectrum and have been extensively used as gene-based probes in molecular and cell biology (10). Förster resonance energy transfer (11) between two fluorophores is typically observed over distances of less than 10 nm. Because FRET efficiency is strongly dependent on the distance separating the FRET pair and the relative orientation of the fluorophores, FRET can be used to analyze molecular interaction with high precision (11–14). In this letter, we describe some aspects of the interaction between quantum dots and fluorescent proteins, with particular emphasis on the evidence for FRET between these components. Our interest in the interaction between fluorescent proteins and semiconducting nanocrystals was piqued by the observation that GFP5 binds to 2-mercaptoethanol-covered ZnS nanocrystals, synthesized in accordance to Kosravi et al. (15). This was accomplished by cloning the gene coding for GFP5 into the * Corresponding author. E-mail:
[email protected]. Phone number: 55 11 8335 5536. † Universidade de São Paulo. ‡ Hewlett-Packard Laboratories.
Figure 1. Normalized photoluminescence (PL) spectrum of GFP5 and absorption (ABS) spectrum of red-emitting CdSe/ZnS core/shell QD solutions.
HIS-tagging vector pET28a (Qiagen) and expressing it in E. coli. GFP5 contains three mutations from wild-type GFP aimed at improving bacterial expression (16). After purification, GFP5 was poured into a suspension of ZnS nanocrystals, and it was observed that it coprecipitates with them. GFP5 bound to these nanocrystals retains its photoluminescence. Following this first evidence, a further step was given by substituting the ZnS nanocrystals by red-emitting CdSe/ZnS core/shell quantum dots coated with mercaptoundecanoic acid (MUA), henceforth denoted QD-R. The quantum dots were purchased from NNLaboratories and had a photoluminescence peak at 638 nm. Once the carboxylic end groups of MUA are exposed over the surfaces of the QDs, and given the usual pKa of carboxylic acids (∼4.5) of isolated molecules or even attached as surface groups (pKapp ∼4.5 – 6.0), negatively charged particles will predominate at pH’s g 6.0. (17). The photoluminescence of GFP5 overlaps with the absorption spectrum of the QDs, as can be seen in Figure 1. Using the measured spectral overlap and a value of 79% for the quantum yield of GFP5 (10), we calculated a Förster radius (12) of 4.2 nm for the pair GFP5-QD-R. Because the diameter of the GFP5 barrel is 2.4 nm and its height is 4.2 nm (8) and the MUA coating thickness of the QDs is around 1 nm (data
10.1021/bc700147j CCC: $37.00 2007 American Chemical Society Published on Web 09/28/2007
1706 Bioconjugate Chem., Vol. 18, No. 6, 2007
Communications
Figure 2. (A) Evolution of the photoluminescence spectra of GFP5 in GFP5-QD-R assemblies vs increasing QD-R (Q) to GFP5 (G) molar ratio. (B) Photoluminescence of GFP5 at 505 nm as a function of GFP5:QD-R mole ratio.
provided by NN-Laboratories), the central fluorophore of GFP5 may be sufficiently close to the quantum dot to allow FRET. In fact, the titration of red-emitting CdSe/ZnS quantum dots in a solution containing GFP5 results in quenching of the photoluminescence of GFP5 (see Figure 2A). The concentration of GFP5 was kept fixed at 24 nM, and the concentration of QD-R was increased from 0 to 20 nM. The GFP5-QD-R assemblies were prepared in dilute aqueous solutions where the effects of absorption of the incident light, reabsorption of PL, and aggregation of nanoparticles were negligible. The solutions also contained 5 mM NaCl and 5 mM NaH2PO4, and the pH was adjusted to 6 with NaOH. The excitation was performed at 395 nm, which coincides with the first absorption peak of GFP5 (16). There is no wavelength where GFP5 could be selectively excited, because the broad absorption of the QD-Rs covers the entire GFP5 absorption spectrum. The overlap of absorption spectra made it difficult to verify that the quenching of the PL of GFP5 resulted in any sensitized emission of the quantum dots. The stoichiometry for the GFP5-QD-R complex formation at pH 6 and room temperature were determined by the mole ratio method (18) using the quenching of the photoluminescence of GFP5 as the property to map the complexation. From the plot of photoluminescence quenching versus mole ratio (see Figure 2B), a stoichiometry of 10:1 for the GFP5-QD-R complex was established. It is important to note that the hydrodynamic radius of QD-R (nanocrystal plus MUA) is 10 nm (data provided by NN-Laboratories), which provides a surface area large enough to accommodate 10 GFP5 molecules. The formation of the GFP5-QD-R complexes is driven by electrostatic interactions, as shown by the pH dependence of the GFP5 PL quenching. At a pH of 6, 90% of the GFP5 emission is quenched at a 0.5:1 molar ratio of QD-R to GFP5. At pH 8, 75% of the GFP5 emission is quenched, while at pH 10, there is only a 10% quenching of the PL of GFP5. There are no major variations in PL intensities of GFP5 and QD-R as the pH is raised from 6 to 10, as can be seen in the control experiments demonstrating the pH dependence of PL intensities for the quantum dots and fluorescent proteins (see Supporting Information). In the Supporting Information is also presented the reversibility of the GFP5 PL quenching as the pH is raised from 6 to 10. Circular dichroism (CD) measurements verified that the suppression of the PL of GFP5 did not result from denaturing of its secondary structure (see Figure 3). At pH 6 and with a QD-R/GFP5 molar ratio of 0.4, where the PL quenching is near maximum, the secondary structure of GFP5 appears substantially unchanged when exposed to QD-Rs. To verify that FRET is occurring between the GFP5 and QD-R elements, anisotropy and nanosecond time-resolved fluorimetry measurements were undertaken. FRET would be expected to increase GFP5 anisotropy and reduce its fluorescence lifetime (12–14). Unfortunately, these measurements were
Figure 3. Circular dichroism spectra of GFP5 and GFP5 assembled to quantum dots are nearly identical, suggesting that the quantum dots do not denature the secondary structure of GFP5.
Figure 4. Normalized absorption and photoluminescence spectra of green-emitting CdSe/ZnS core/shell QDs and absorption spectrum of HcRed1 protein solutions.
inconclusive, because the strong GFP5 PL quenching, coupled with the need to keep the QD-R concentration low to avoid aggregation, resulted in poor signal-to-noise ratios. One explanation for these results is the possibility that the fraction of GFP5 complexed to the QD-Rs was completely quenched so that the remaining emission due to the uncomplexed fraction dominated the signal (19). Also, it is possible that charge transfer to the QD-R (in addition to energy transfer) contributes to the GFP5 PL quenching (20), despite the fact that GFP5 fluorophore is buried inside a proteic barrel that surrounds and protects it (7, 21). Alternative ways to determine whether the observed PL quenching is due to FRET are under investigation. One possibility would be to selectively excite the GFP5 through a chemiluminescent reaction and then observe the eventual sensitized emission from the QD-Rs. GFP has a chemiluminescent resonance energy transfer pair in ViVo, the protein aequorin (10, 22), that could be used for such finality.
Communications
Bioconjugate Chem., Vol. 18, No. 6, 2007 1707
Figure 5. (A) Evolution of the photoluminescence spectra of HcRed1 protein in QD-G-HcRed1 assemblies vs increasing HcRed1 to QD-G molar ratio. (B) Amplified view of the 590–700 nm region of part A enables us to better observe HcRed1-sensitized emission.
Figure 6. (A) pH dependency of the sensitized emission from HcRed1 complexed with green-emitting quantum dots. (B) PL of HcRed1 directly excited at 545 nm.
Alternatively, the larger two-photon cross section of GFP in comparison to the that of the QDs (23, 24) could be used to selectively excite this protein with 800 nm radiation and then try to observe the sensitized emission from the QD-Rs. The possibility of increasing the distance between GFP5 and the QD-Rs by the incorporation of an additional capping layer over the QD-Rs (25) is also being considered, aiming to reduce the efficiency of the quenching and make the remaining PL intensity of GFP5 appropriate for anisotropy and time resolved analysis. In the meantime, an inverted approach was investigated where a GFP homologous red-emitting protein, HcRed1, has been used together with a green-emitting CdSe/ZnS mercaptoundecanoic acid coated quantum dot (henceforth denoted QD-G), forming a pair where the QD-G is the donor and HcRed1 is the acceptor. The overlap between QD-Gs emission and HcRed1 absorption can be seen in Figure 4. The broad absorption spectrum of the quantum dots enabled us to find a wavelength interval, around 425 nm, where the quantum dots are selectively excited. Excited at 425 nm, the PL intensity of the quantum dots is partially suppressed by the titration of HcRed1 (see Figure 5A). The suppression is not intense as in the case of GFP5 and QDR, but the selective excitation enabled us to observe a subtle sensitized emission from HcRed1 (see Figure 5B). The concentration of QD-Gs was kept fixed at 33 nM, and the concentration of HcRed1 was augmented from 0 to 8 nM. The assemblies were prepared in dilute aqueous solutions where the effects of absorption of the incident light, reabsorption of PL, and aggregation of nanoparticles were negligible. The solutions also contained 5 mM NaCl and 5 mM NaH2PO4, and the pH was adjusted to 6 with NaOH. The formation of the HcRed1–QD-G complexes is also driven by electrostatic interactions, as shown by the pH dependence of the sensitized emission. At a pH of 6, it is possible to observe the sensitized emission at a 0.5:1 molar ratio of HcRed1 to
QD-G excited at 425 nm. At pH 10, no sensitized emission is observed (see Figure 6A) in spite of an increase in the PL intensity of HcRed1 when excited with 545 nm and no major variation in the PL intensity of the QD-G (see Supporting Information). The PL spectrum of HcRed1 excited at 545 nm has the same profile of the sensitized emission (see Figure 6B). Detection of sensitized emission anisotropy is a technique that enables the characterization of FRET with high contrast (13, 26). Sensitized emission is highly depolarized, because the orientation of emitted light is not wholly constrained by the excitation polarization (27). Due to the relatively large size and slow rotational diffusion, PL emission from fluorescent proteins is highly anisotropic (26, 27). In the case of HcRed1, the anisotropy of the PL obtained by direct excitation is nearly 0.4. The anisotropy of the sensitized emission of HcRed1, measured in the excitation interval from 400 to 470 nm, is below 0.1. The low anisotropy of the sensitized emission contrasting with the elevated anisotropy of the direct PL emission provides strong evidence that FRET occurs between QD-Gs and HcRed1. Other pairs of QD and fluorescent proteins are under present consideration. The protein HcRed1 may be substituted by other red-emitting fluorescent proteins with higher quantum yields to be the acceptors with green- or yellow-emitting quantum dot donors. Actually, HcRed1 has a quantum yield of only 0.05 (28), which may account for the low intensity of its sensitized emission. The proteins mCherry, tdTomato, and mStrawberry (29) would be attractive acceptor candidates, due to their elevated quantum yield and brightness. tdTomato (29), in particular, has a quantum yield of 0.69 and brightness superior to that of fluorescein. An optimized donor–acceptor pair would be convenient for in-depth characterization and quantification of FRET parameters and for high-contrast in ViVo investigations.
1708 Bioconjugate Chem., Vol. 18, No. 6, 2007
ACKNOWLEDGMENT The authors thank Dr. Jim Haseloff for providing the gene coding for GFP5. V.R.H. was supported by a grant from Numina Nanotecnologia Ltda. The authors wish to express their gratitude to the Brazilian funding agencies CNPQ, FAPESP, and CAPES, being also thankful for the support from the Biochemistry Department of the Chemistry Institute of the University of São Paulo, Brazil. Supporting Information Available: PL intensity variation for quantum dots and fluorescent proteins as a function of pH and reversibility of PL quenching of GFP5 by raising the pH from 6 to 10. This material is available free of charge via the Internet at http://pubs.acs.org.
LITERATURE CITED (1) Wagnier, R., Baranov, A. V., Maslov, V. G., Stsiapura, V., Artemyev, M., Pluot, M., Sukhanova, A., and Nabiev, I. (2004) Energy transfer in aqueous solutions of oppositely charged CdSe/ ZnS core/shell quantum dots and in quantum dot-nanogold assemblies. Nano Lett. 4, 451–457. (2) Dabbousi, B. O., Rodriguez-Viejo, J., Mikulec, F. V., Heine, J. R., Mattoussi, H., Ober, R., Jensen, K. F., and Bawendi, M. G. (1997) (CdSe)ZnS core-shell quantum dots: synthesis and characterization of a size series of highly luminescent nanocrystallites. J. Phys. Chem. B 101, 9463–9475. (3) Ma, J., Chen, J., Guo, J., Wang, C. C., Yang, W. L., Xu, L., and Wang, P. N. (2006) Photostability of thiol-capped CdTe quantum dots in living cells: the effect of photo-oxidation. Nanotechnology 17, 2083–2089. (4) Sukhanova, A., Devy, J., Venteo, L., Kaplan, H., Artemyev, M., Oleinikov, V., Klinov, D., Pluot, M., Cohen, J. H. M., and Nabiev, I. (2004) Biocompatible fluorescent nanocrystals for immunolabeling of membrane proteins and cells. Anal. Biochem. 324, 60–67. (5) Wenhua, L., Haiyan, X., Zhixiong, X., Zhexue, L., Jianhong, O., Xiangdong, C., and Ping, S. (2004) Exploring the mechanism of competence development in Escherichia coli using quantum dots as fluorescent probes. J. Biochem. Biophys. Methods 58, 59–66. (6) Dubertret, B., Skourides, P., Norris, D. J., Noireaux, V., Brivanlou, A. H., and Libchaber, A. (2002) In vivo imaging of quantum dots encapsulated in phospholipid micelles. Science 298, 1759–1762. (7) Wachter, R. M. (2006) The family of GFP-like proteins: structure, function, photophysics and biosensor applications. Introduction and perspective. Photochem. Photobiol. 82, 339– 344. (8) Ormo, M., Cubitt, A. B., Kallio, K., Gross, L. A., Tsien, R. Y., and Remington, S. J. (1996) Crystal structure of the Aequorea victoria green fluorescent protein. Science 273, 1392–1395. (9) Yang, F., Moss, L. G., and Phillips, G. N. (1996) The molecular structure of green fluorescent protein. Nat. Biotechnol. 14, 1246– 1251. (10) Tsien, R. Y. (1998) The green fluorescent protein. Annu. ReV. Biochem. 67, 509–544. (11) Förster, T. (1948) Zwischenmolekulare energiewanderung und fluoreszenz. Ann. Phys. 2, 55–75. (12) Lakowicz, J. R. (2006) Principles of Fluorescence Spectroscopy, 3rd ed., pp 443–475, Springer, Berlin.
Communications (13) Rizzo, M. A., Springer, G., Segawa, K., Zipfel, W. R., and Piston, D. W. (2006) Optimization of pairings and detection conditions for measurement of FRET between cyan and yellow fluorescent proteins. Microsc. Microanal. 12, 238–254. (14) Jares-Erijman, E. A., and Jovin, T. M. (2003) FRET imaging. Nat. Biotechnol. 21, 1387–1395. (15) Khosravi, A. A., Kundu, M., Kuruvilla, B. A., Shekhawat, G. S., Gupta, R. P., Sharma, A. K., Vyas, P. D., and Kulkarni, S. K. (1995) Manganese doped zinc sulphide nanoparticles by aqueous method. Appl. Phys. Lett. 67, 2506–2508. (16) Siemering, K. R., Golbik, R., Sever, R., and Haseloff, J. (1996) Mutations that suppress the thermosensitivity of green fluorescent protein. Curr. Biol. 6, 1653–1663. (17) Hiemenz, P. C. and Rajagopolan, R. (1997) Principles of Colloid and Surface Chemistry, 3rd ed., Marcel Dekker Inc., New York. (18) Meyer, A. S., and Ayres, G. H. (1957) The mole ratio method for spectrophotometric determination of complexes in solution. J. Am. Chem. Soc. 79, 49–53. (19) Song, L., Jares-Erijman, E. A., and Jovin, T. M. (2002) A photochromic acceptor as a reversible light-driven switch in fluorescence resonance energy transfer (FRET). J. Photochem. Photobiol., A 150, 177–185. (20) Nikobaakht, B., Burda, C., Braun, M., Hun, M., and El-Sayed, M. A. (2002) The quenching of CdSe quantum dots photoluminescence by gold nanoparticles in solution. Photochem. Photobiol. 75, 591–597. (21) Herberhold, H., Marchal, S., Lange, R., Scheyhing, C. H., Vogel, R. F., and Winter, R. (2003) Characterization of the pressure-induced intermediate and unfolded state of red-shifted green fluorescent protein - a static and kinetic FTIR, UV/VIS and fluorescence spectroscopy study. J. Mol. Biol. 330, 1153– 1164. (22) Shimomura, O., Musicki, B., and Kishi, Y. (1988) Semisynthetic aequorin. An improved tool for the measurement of calcium ion concentration. Biochem. J. 251, 405–410. (23) Hao, Z., Gong, H., Han, J., Zhai, Y., and Wang, Q. (2006) One- and two-photon excited fluorescence of CdSe and CdSe/ ZnS quantum dots in n-hexane. Chin. Phys. Lett. 23, 2859–2862. (24) Kirkpatrick, S. M., Naik, R. R., and Stone, M. O. (2001) Nonlinear saturation and determination of the two-photon absorption cross section of green fluorescent protein. J. Phys. Chem. B 105, 2867–2873. (25) Jafaar, S., Nam, K. T., Khademhosseini, A., Xing, J., Langer, R. S., and Belcher, A. M. (2004) Layer-by-layer surface modification and patterned electrostatic deposition of quantum dots. Nano Lett. 4, 1421–1425. (26) Rizzo, M. A., and Piston, D. W. (2004) High-contrast imaging of fluorescent protein FRET by fluorescence polarization microscopy. Biophys. J.: Biophys Lett. L14–L16. (27) Lakowicz, J. R. (2006) Principles of Fluorescence Spectroscopy, 3rd ed., pp 353–382, Springer, Berlin. (28) Wilmann, P. G., Petersen, J., Pettikiriarachchi, A., Buckle, A. M., Smith, S. C., Olsen, S., Perugini, M. A., Devenish, R. J., Prescott, M., and Rossjohn, J. (2005) The 2.1 A crystal structure of the far-red fluorescent protein HcRed: inherent conformational flexibility of the chromophore. J. Mol. Biol. 349, 223–237. (29) Shaner, N. C., Steinbach, P. A., and Tsien, R. Y. (2005) A guide to choosing fluorescent proteins. Nat. Meth. 2, 905–909. BC700147J
