pubs.acs.org/Langmuir © 2009 American Chemical Society
Developing Mixed Films of Immobilized Oligonucleotides and Quantum Dots for the Multiplexed Detection of Nucleic Acid Hybridization Using a Combination of Fluorescence Resonance Energy Transfer and Direct Excitation of Fluorescence W. Russ Algar and Ulrich J. Krull* Chemical Sensors Group, Department of Chemical and Physical Sciences, University of Toronto Mississauga, 3359 Mississauga Road, North, Mississauga, Ontario, L5L 1C6, Canada Received October 3, 2009. Revised Manuscript Received November 18, 2009 Methods have been developed for the simultaneous and selective detection of three target nucleic acid sequences based on mixed films of immobilized quantum dots (QDs) and oligonucleotide probes. CdSe/ZnS QDs were immobilized on optical fibers and conjugated with mixtures of different probe oligonucleotides. Hybridization events were detected using a combination of fluorescence from direct excitation and fluorescence sensitized by resonance energy transfer (FRET). A sandwich assay format was used to associate dye labeled reporter oligonucleotides with probe-target hybrids formed at the surface of the optical fiber. One detection channel utilized direct excitation of Pacific Blue and the two other detection channels were based on FRET. In one strategy, green emitting QDs were used as donors with Cy3 and Rhodamine Red-X acceptors. In a second strategy, green and red emitting QDs were coimmobilized and used as donors with Cy3 and Alexa Fluor 647 acceptors, respectively. Selective three-plex detection was demonstrated with both strategies. Several key design criteria that were explored to optimize the relative signal magnitude between channels included: the ratio of probe associated with direct excitation versus probes associated with FRET; the relative amounts of each FRET probe and corresponding spectral overlap; and the photoluminescence ratio between immobilized green and red emitting QDs (where applicable). Careful selection of probe sequences and lengths were important for the discrimination of single nucleotide polymorphisms in one channel without suppressing binding of target in the other two channels. This work provides a basis for the development of multiplexed biosensors that are ensemble compatible and do not require discrete sensor elements, spatial registration, sorting technology, or single molecule spectroscopy.
1. Introduction Bioanalytical and biomedical applications of luminescent semiconductor quantum dots (QDs) have been important and extremely active areas of research in recent years. The unique optical properties of QDs have enabled new developments and advancements in both biological imaging and sensing.1-7 A prevalent theme in this research has been the concept of multiplexing. The broad absorption spectra, potentially large effective Stokes’ shifts, and narrow size-tunable emission spectra of QDs have been favorable for multiplexing in several applications, including: multicolor probes as contrast agents in imaging,8,9 *Author to whom correspondence should be addressed. E-mail: ulrich.
[email protected]. (1) Sapsford, K. E.; Pons, T.; Medintz, I. L.; Mattoussi, H. Sensors 2006, 6, 925–953. (2) Medintz, I. L.; Mattoussi, H. Phys. Chem. Chem. Phys. 2009, 11, 17–45. (3) Medintz, I. L.; Mattoussi, H.; Clapp, A. R. Int. J. Nanomed. 2008, 3, 151–167. (4) Gill, R.; Zayats, M.; Willner, I. Angew. Chem., Int. Ed. 2008, 47, 7602–7625. (5) Algar, W. R.; Krull, U. J. Anal. Bioanal. Chem. 2008, 391, 1609–1618. (6) Smith, A. M.; Duan, H. W.; Mohs, A. M.; Nie, S. M. Adv. Drug. Deliver. Rev. 2008, 60, 1226–1250. (7) Michalet, X.; Pinaud, F. F.; Bentolila, L. A.; Tsay, J. M.; Doose, S.; Li, J. J.; Sundaresan, G.; Wu, A. M.; Gambhir, S. S.; Weiss, S. Science 2005, 307, 538–544. (8) Kim, B. Y. S.; Jiang, W.; Oreopoulos, J.; Yip, C. M.; Rutka, J. T.; Chan, W. C. W. Nano Lett. 2008, 8, 3887–3892. (9) Jaiswal, J. K.; Mattoussi, H.; Mauro, J. M.; Simon, S. M. Nat. Biotechnol. 2003, 21, 47–51. (10) Ghazani, A. A.; Lee, J. A.; Klostranec, J.; Xiang, Q.; Dacosta, R. S.; Wilson, B. C.; Tsao, M. S.; Chan, W. C. W. Nano Lett. 2006, 6, 2881–2886. (11) Liang, R. Q.; Li, W.; Li, Y.; Tan, C. Y.; Li, J. X.; Jin, Y. X.; Ruan, K. C. Nucleic Acids Res. 2005, 33, e17. (12) Robelek, R.; Niu, L. F.; Schmid, E. L.; Knoll, W. Anal. Chem. 2004, 76, 6160–6165.
Langmuir 2010, 26(8), 6041–6047
multicolor labels in fluorescence-based assays,10-15 and optical barcodes for identifying analytes in multiplexed assays.16-19 Another application of interest has been the development of QDs as platforms for chemosensing20-24 and biosensing25-30 via (13) Gerion, D.; Chen, F. Q.; Kannan, B.; Fu, A. H.; Parak, W. J.; Chen, D. J.; Majumdar, A.; Alivisatos, A. P. Anal. Chem. 2003, 75, 4766–4772. (14) Shingyoji, M.; Gerion, D.; Pinkel, D.; Gray, J. W.; Chen, F. Q. Talanta 2005, 67, 472–478. (15) Sun, B. Q.; Xie, W. Z.; Yi, G. S.; Chen, D. P.; Zhou, Y. X.; Cheng, J. J. Immunol. Methods 2001, 249, 85–89. (16) Klostranec, J. M.; Xiang, Q.; Farcas, G. A.; Lee, J. A.; Rhee, A.; Lafferty, E. I.; Perrault, S. D.; Kain, K. C.; Chan, W. C. W. Nano Lett. 2007, 7, 2812–2818. (17) Xu, H. X.; Sha, M. Y.; Wong, E. Y.; Uphoff, J.; Xu, Y. H.; Treadway, J. A.; Truong, A.; O’Brien, E; Asquith, S.; Stubbins, M.; Spurr, N. K.; Lai, E. H.; Mahoney, W. Nucleic Acids Res. 2003, 31, e43. (18) Wang, H. Q.; Liu, T. C.; Cao, Y. C.; Huang, Z. L.; Wang, J. H.; Li, X. Q.; Zhao, Y. D. Anal. Chim. Acta 2006, 580, 18–23. (19) Eastman, P. S.; Ruan, W. M.; Doctolero, M.; Nuttal, R.; De Feo, G.; Park, J. S.; Chu, J. S. F.; Cooke, P.; Gray, J. W.; Li, S.; Chen, F. Q. F. Nano Lett. 2006, 6, 1059–1064. (20) Wang, X.; Guo, X. Q. Analyst 2009, 134, 1248–1354. (21) Zhang, C. Y.; Johnson, L. W. Anal. Chem. 2009, 81, 3051–3055. (22) Freeman, R.; Finder, T; Bahshi, L.; Willner, I. Nano Lett. 2009, 9, 2073–2076. (23) Goldman, E. R.; Medintz, I. L.; Whitley, J. L.; Hayhurst, A.; Clapp, A. R.; Uyeda, H. T.; Deschamps, J. R.; Lassman, M. E.; Mattoussi, H. J. Am. Chem. Soc. 2005, 127, 6744–6751. (24) Snee, P. T.; Somers, R. C.; Nair, G.; Zimmer, J. P.; Bawendi, M. G.; Nocera, D. G. J. Am. Chem. Soc. 2006, 128, 13320–13321. (25) Boeneman, K.; Mei, B. C.; Dennis, A. M.; Bao, G.; Deschamps, J. R.; Mattoussi, H.; Medintz, I. L. J. Am. Chem. Soc. 2009, 131, 3828–3829. (26) Tang, B.; Cao, L. H.; Xu, K. H.; Zhuo, L. H.; Ge, J. H.; Yu, L. J. Chem.;Eur. J. 2008, 14, 3637–3644. (27) Suzuki, M.; Husimi, Y.; Komatsu, H.; Suzuki, K.; Douglas, K. T. J. Am. Chem. Soc. 2008, 130, 5720–5725. (28) Algar, W. R.; Krull, U. J. Anal. Chim. Acta 2007, 581, 193–201. (29) Zhang, C. Y.; Yeh, H. C.; Kuroki, M. T.; Wang, T. H. Nat. Mater. 2005, 4, 826–831.
Published on Web 12/09/2009
DOI: 10.1021/la903751m
6041
Article
fluorescence resonance energy transfer (FRET). In the case of the latter, QDs have been used as scaffolds for biomolecular probes. Selective binding events can mediate FRET at the surface of the QD and provide a measurable signal that can be used for biosensing. Although it is recognized that the multiplexing advantages of QDs can be applied to FRET-based sensors, multiplexed applications remain underdeveloped compared to others.2 We have recently described a pair of two-plex nucleic acid hybridization sandwich assays using immobilized QDs as donors in FRET. In one assay, two colors of immobilized QD donors were each combined with a different acceptor dye labeled reporter oligonucleotide (2D-2A).31 In a second assay, a single color of immobilized QD donor was combined with two different reporter oligonucleotides labeled with two different acceptor dyes (1D-2A).32 In both cases, FRET-sensitized acceptor photoluminescence (PL) signaled target hybridization. The FRET-based approach with QD donors enabled multiplexed analyses using a single excitation source, without the need for discrete sensor elements or spatial registration of probes- advantages amenable to miniaturized devices. The use of fluorescent acceptors with QD donors allows ratiometric measurements that are less sensitive to changes in sample conditions, variations between biosensor or assay preparations, and are less likely to give false positives or negatives. Although the multiplexing capacity of a FRET-approach is less than optical barcoding with QDs, it has the advantage of being an ensemble compatible technique. Furthermore, a solid phase approach has allowed the one-pot preparation of multiplexed sensors.31 It is also advantageous in that the immobilization of QDs allows their recovery from samples, minimizing the potential for harmful environmental or toxic effects. Immobilization also potentiates the development of a reusable biosensing device,33 and incorporation into flow cells or microfluidic platforms. While it is of interest to extend the multiplexing capacity of QD-FRET assays, it is important to do so without sacrificing any of their advantages. Moving beyond a two-plex requires the addition of another detection channel at either shorter or longer wavelengths than the existing two-plex channels. The increased spectral range may present challenges with wavelength-dependent photodetector efficiency, require shorter wavelength excitation (e.g., UV laser), alternate QD materials (e.g., CdSe is not optimal for the violet-blue or near-IR regions of the spectrum), and deconvolution of multiple overlapping spectra. A simple approach that will largely avoid these challenges is desirable. In this new work, we show that the selective three-plex detection of nucleic acids is possible by adding a direct excitation channel to a FRET-based two-plex assay utilizing immobilized QD donors. The direct excitation channel incorporated the fluorescent dye Pacific Blue (PB), which could be excited by the same violet laser source used to excite QDs. Although direct excitation was used, this approach still retained the advantages of a purely QD-FRET approach. Furthermore, the incorporation of the PB channel was compatible with QD-FRET two-plexes based on both 2D-2A and 1D-2A strategies. In these experiments, CdSe/ZnS QDs were immobilized on fused silica optical fibers using multidentate surface ligand exchange.34 As shown in Figure 1, mixtures of biotinylated oligonucleotides were coimmobilized on the immobilized QDs via a NeutrAvidin bridge. We have described (30) Medintz, I. L.; Clapp, A. R.; Mattoussi, H.; Goldman, E. R.; Fisher, B.; Mauro, J. M. Nat. Mater. 2003, 2, 630–638. (31) Algar, W. R.; Krull, U. J. Anal. Chem. 2009, 81, 4113–4120. (32) Algar, W. R.; Krull, U. J. Anal. Chem., published online November 25, 2009, http://dx.doi.org/10.1021/ac902221d. (33) Algar, W. R.; Krull, U. J. Langmuir 2009, 25, 633–638. (34) Algar, W. R.; Krull, U. J. Langmuir 2008, 24, 5514–5520.
6042 DOI: 10.1021/la903751m
Algar and Krull
Figure 1. Three-plex hybridization assays based on immobilized QDs and FRET. Mixed films of biotinylated oligonucleotide probes (P) were conjugated to QDs via a NeutrAvidin (NA) bridge. BSA was used for passivation. (a) The immobilization of green emitting QD donors (gQDs) was used to develop a three-plex assay using Cy3 and RhR FRET acceptors in combination with the direct excitation of PB. (b) The coimmobilization of gQDs and red emitting QD donors (rQDs) was used to develop a three-plex assay using Cy3 and A647 acceptors, respectively, with direct excitation of PB. The oligonucleotide sequences corresponding to the different dye labels are listed in the legend. A sandwich assay was used to create the proximity required for FRET via target (T) and reporter (R) hybridization. The arrows indicate FRET.
and characterized this chemistry previously, including aspects such as optimization of selectivity, hybridization kinetics, and response in complex matrices.31,32 A sandwich assay was used to signal hybridization through either FRET-sensitized acceptor emission or direct excitation of Pacific Blue, and avoided direct labeling of the targets. This work successfully demonstrated selective three-plex detection in 2D-2A and 1D-2A formats combined with Pacific Blue. The FRET acceptors used were Cyanine 3 (Cy3), Rhodamine Red-X (RhR), and Alexa Fluor 647 (A647). Aspects such as optimization of probe ratios, optimization of Langmuir 2010, 26(8), 6041–6047
Algar and Krull
Article
Table 1. Oligonucleotide Sequences for Probes, Targets, and Reporters.a seq. 1 probe target reporter
0
biotin-5 -TTC AGT TAA TCC TAC AAC-30 30 -AAG TCA ATT AGG ATG TTG AAA AAG ACG ACG TGG-50 Cy3-50 -TTT TTC TGC TGC ACC-30 seq. 2 (Tm = 31.4 °C)b
probe target reporter
biotin-50 -AAC AAT ATT GTC TTG ATT-30 30 -TTG TTA TAA CAG AAC TAA TCA GTA AGG ACC GTT-50 RhR-50 -AGT CAT TCC TGG CAA-30 seq. 3 (Tm = 45.3 °C)
probe target reporterc
biotin-50 -TAT GCC CGG TAA ACA GAT GAG-30 30 -ATA CGG GCC ATT TGT CTA CTC ATA ACT ACG GCT AAA-50 PB-50 -TAT TGA TGC CGA TTT-30 seq. 4 (Tm = 39.1 °C)
probe target reporter
0
biotin-5 -CTT ACT TCC ATG ATT TCT TTA ACT-30 30 -GAA TGA AGG TAC TAA AGA AAT TGA TGC GGC CCT AGG TAG-50 A647-50 -ACG CCG GGA TCC ATC-30 seq. 5 (Tm = 40.2 °C)
probe target 1BPM reporter
biotin-50 -ATT TTG TCT GAA ACC CTG T-30 30 -TAA AAC AGA CTT TGG GAC ATT CCT TTT ATT TCC T-50 30 -TAA AAC ACA CTT TGG GAC ATT CCT TTT ATT TCC T-50 Cy3-50 -AA GGA AAA TAA AGG A-30 Noncomplementary 0
3 -AGG AAA TAA AAG GAA TGT CCC AAA GTC TGT TTT A-50 a Cy3 = Cyanine 3; RhR = Rhodamine Red-X; PB = Pacific Blue; A647= Alexa Fluor 647. b Tm is the calculated melt temperature for the probe-target hybrid region.42 c The seq. 3 reporter was received as a 50 amine modified oligonucleotide from the manufacturer. It was labeled with PB as described in the Supporting Information. All other reporters were received from the manufacturer with dye modifications NC
probe length, and immobilized QD PL ratios were also addressed. As the first example of three-plex detection on the basis of QDs and FRET, this work provides an important foundation for the future development and application of multiplexed QD biosensors.
2. Experimental Section Detailed descriptions of reagents, procedures, and instrumentation used in these experiments are available in the Supporting Information. 2.1. Immobilized Quantum Dots. CdSe/ZnS QDs were obtained from the laboratory of Professor Warren C.W. Chan at the University of Toronto and were synthesized using published methods.35-37 Ligand exchange with 3-mercaptopropionic acid (MPA) was used to render the QDs water-soluble. Optical fibers (0.4 mm diameter 45 mm) were modified with multidentate surface ligands as described previously.32,34 MPA-coated QDs were immobilized on the fibers from 0.1 to 1.0 μM aqueous solutions by incubating at room temperature overnight.31-34 The immobilization of two colors of QD was from aqueous solutions containing the QDs mixed in the proportions required (35) Hines, M. A.; Guyot-Sionnest, P. J. Phys. Chem. 1996, 100, 468–471. (36) Fischer, H. C.; Liu, L. C.; Pang, K. S.; Chan, W. C. W. Adv. Funct. Mater. 2006, 16, 1299–1305. (37) Jiang, W.; Mardyani, S.; Fischer, H.; Chan, W. C. W. Chem. Mater. 2006, 18, 872–878.
Langmuir 2010, 26(8), 6041–6047
Figure 2. Normalized absorption (dashed line) and photoluminescence (PL, solid line) spectra for the QDs and dyes used in this study. The spectra are grouped into “channels” to show spectral overlap: (a) PB; (b) gQD-Cy3 FRET pair; (c) gQD-RhR FRET pair; and (d) rQD-A647 FRET pair. to give a desired PL ratio. In general, the PL ratio in the aqueous solution approximately translated to that observed on fibers after immobilization.
2.2. Interfacial Quantum Dot-Oligonucleotide Conjugates. Fibers with immobilized QDs were incubated successively in solutions of NeutrAvidin (0.2 mg/mL, borate buffer pH 8.5; 75 min), biotinylated probe oligonucleotides (1.0 μM, trisborate buffer pH 7.4; 75 min), and bovine serum albumin (BSA, 0.5 mg/mL; tris-borate buffer pH 7.4; 30 min) at room temperature. The NeutrAvidin was bound to the QDs by physisorption. The immobilization of multiple probe oligonucleotides was accomplished using solutions with a mixture of different probes. The amount of each probe was variable, but the total probe concentration was maintained at 1.0 μM. The final exposure to BSA helped block any adsorption sites, minimizing the nonspecific adsorption of oligonucleotides. The fibers were rinsed with purified water between incubations. Probe oligonucleotides are listed in Table 1. The probes were selected on the basis of having practical application: seq. 1 is complementary to a portion of the E. coli hemolysin coding gene (hly A);38 seq. 2 is complementary to a portion of the L. monocytogenes listeriolysin O coding gene (hly A);39 seq. 3 is complementary to a portion of the S. enterica invasion protein A coding gene (inv A);40 seq. 4 is complementary to a portion of the E. coli beta-glucuronidase enzyme coding gene (uid A);41 and seq. 5 is complementary to a portion of the H. sapiens survival of motor neuron protein coding gene (SMN 1).41 A sequence that is noncomplementary to all probe sequences is also listed in Table 1. However, it should be noted that any set of probe nucleotides could have been used for evaluating the multiplexing strategies presented in this work. For clarity throughout the text, each (38) Chen, S.; Zhao, S.; McDermott, P. F.; Schroeder, C. M.; White, D. G.; Meng, J. Mol. Cell. Probe 2005, 19, 195–201. (39) Amagliani, G.; Brandi, G.; Omiccioli, E.; Casiere, A.; Bruce, I. J.; Magnani, M. Food Microbiol. 2004, 21, 597–603. (40) Ikebukuro, K.; Kohiki, Y.; Sode, K. Biosens. Bioelectron. 2002, 17, 1075– 1080. (41) Watterson, J. H.; Raha, S.; Kotoris, C. C.; Wust, C. C.; Gharabaghi, F.; Jantzi, S. C.; Haynes, N. K.; Gendron, N. H.; Krull, U. J.; Mackenzie, A. E.; Piunno, P. A. E. Nucleic Acids Res. 2004, 32, e18.
DOI: 10.1021/la903751m
6043
Article
Algar and Krull
sequence is referred to by its sequence number with the corresponding acceptor dye label on the reporter oligonucleotide indicated in parentheses. 2.3. Hybridization Experiments. Fibers modified with immobilized probe oligonucleotides were incubated in solutions of target oligonucleotide (tris-borate buffer pH 7.4), rinsed with buffer, incubated in solutions of reporter oligonucleotide (trisborate buffer pH 7.4, 0.1% w/v sodium dodecyl sulfate, SDS), and rinsed again with buffer. Fibers were incubated in both target and reporter solutions for 60 min, unless otherwise noted. For hybridization experiments investigating the discrimination of single nucleotide polymorphisms, the target solutions also contained 23-25% v/v formamide. Target and reporter oligonucleotides are also listed in Table 1. 2.4. Data Collection and Analysis. PL spectra were obtained by coupling 407 nm laser light into modified optical fibers situated inside the sample compartment of a spectrofluorimeter. The raw PL spectra from modified optical fibers were background corrected and normalized to the peak QD PL. Estimates of F€ orster distances were calculated as described previously.28
3. Results and Discussion 3.1. The FRET Pairs. Green emitting QDs (gQD, peak PL at 528 nm) and red emitting QDs (rQD, peak PL at 618 nm) were used as energy donors. Both these QDs were efficiently excited by a laser with 407 nm emission. Three FRET-pairs were of interest: the gQD donor was combined with either Cy3 or RhR as an acceptor; the rQD donor was combined with A647 as an acceptor. The absorption and PL spectra of each donor and acceptor are shown in Figure 2. The spectral overlap integrals for the FRET pairs are 6.2 10-10 mol-1 cm6 for gQD-Cy3 (seq. 1 reporter); 7.9 10-10 mol-1 cm6 for gQD-Cy3 (seq. 5 reporter); 4.1 10-10 mol-1 cm6 for gQD-RhR (seq. 2 reporter); and 1.9 10-9 mol-1 cm6 for rQD-A647 (seq. 4 reporter). Note the small sequence dependence of the gQD-Cy3 spectral overlap between seq. 1 and seq. 5. Assuming a refractive index of n = 1.33, an orientation factor of κ2 = 2/3, and QD donor quantum yields between 0.2 and 1.0%, the corresponding F€orster distances calculated for the FRET pairs are: 2.3-3.0 nm for gQD-Cy3 (seq. 1 reporter); 2.4-3.2 nm for gQD-Cy3 (seq. 5 reporter); 2.2-2.8 nm for gQD-RhR; and 2.8-3.6 nm for rQD-A647. The QD quantum yield estimate is a typical range for the MPA-coated QDs. At the surface of the optical fiber, FRET pairs were formed by the interaction of immobilized QD donors and dye acceptors bound through oligonucleotide hybrids (Figure 1). These interactions did not constitute traditional FRET pairs because each acceptor was able to interact with several QD donors over a distribution of distances, and vice versa. There were no discrete donor-acceptor pairs and thus a conventional FRET efficiency cannot be reported. We have previously studied two-plex assays based on the same interfacial chemistry, and found “apparent” FRET efficiencies of ca. 15-40% for the gQD-Cy3, gQD-RhR, and rQD-A647 FRET pairs.31,32 Although these values were useful for demonstrating energy transfer, they were not realistic representations of the efficiency of transferring energy to each acceptor. In this work, energy transfer from QD donors to fluorescent dye acceptors was demonstrated by measuring the excitation spectrum of each acceptor dye as part of an individual FRET pair with a QD donor. The contribution of QD excitation to the dye excitation spectrum was readily apparent, confirming energy transfer (see Supporting Information, Figure S2). This was also used to provide a measure of the contribution of direct (42) Integrated DNA Technologies (IDT), Inc., SciTools, OligoAnalyzer 3.1. Conditions: 1 μM oligo conc.; 15 mM Naþ conc. http://www.idtdna.com/analyzer/ Applications/OligoAnalyzer/.
6044 DOI: 10.1021/la903751m
Figure 3. Response of fibers modified with different ratios of seq. 3 (PB) probe and seq. 2 (RhR) probe to samples with 150 nM seq. 3 (PB) target and 150 nM of seq. 2 (RhR) target: (a) initial immobilized QD spectrum; (b) 0:1; (c) 1:8; (d) 1:4; and (e) 1:2.
excitation of each dye in FRET experiments. In this work, direct excitation accounted for
