Homogenous rapid detection of nucleic acids using two-color quantum dots

Abstract

We report a homogenous method for rapid and sensitive detection of nucleic acids using two-color quantum dots (QDs) based on single-molecule coincidence detection. The streptavidin-coated quantum dots functioned as both a nano-scaffold and as a fluorescence pair for coincidence detection. Two biotinylated oligonucleotide probes were used to recognize and detect specific complementary target DNA through a sandwich hybridization reaction. The DNA hybrids were first caught and assembled on the surface of 605 nm-emitting QDs (605QDs) through specific streptavidin–biotin binding. The 525 nm-emitting QDs (525QDs) were then added to bind the other end of DNA hybrids. The coincidence signals were observed only when the presence of target DNA led to the formation of 605QD/DNA hybrid/525QD complexes. In comparison with a conventional QD-based assay, this assay provided high detection efficiency and short analysis time due to its high hybridization efficiency resulting from the high diffusion coefficient and no limitation of temperature treatment. This QD-based single-molecule coincidence detection offers a simple, rapid and ultra sensitive method for genomic DNA analysis in a homogenous format.

---

Introduction

Nucleic acid analysis is becoming increasingly important in the diagnosis of hereditary diseases, detection of infectious agents, forensic and paternity testing, veterinary medicine and environmental monitoring.¹ Great progress has been made in the development of hybridization assays for nucleic acid detection.^2–8 Gold nanoparticles,^2,3 bar-coded nanorods,⁴ fiber-optical DNA arrays,⁵ quantum dots-based DNA arrays,⁶ electrochemical coding⁷ and DNA-based fluorescence nanobarcodes⁸ have been developed for nucleic acid analysis. Polymerase chain reaction (PCR) enables an extremely high sensitivity for detecting low-abundance species, and has been routinely used for detection and quantification of nuclei acids;⁹ however, there are a number of complications involved in the application of real time PCR, such as contamination-induced false-positive signals.¹⁰ Single-molecule detection techniques enable the sensitive detection of low-abundance species without amplification due to the capability to detect single fluorescent molecules with a high signal-to-noise ratio.^11–13 Two-color coincident^14–18 and dual-color fluorescence cross-correlation spectroscopy (FCS)^19,20 have been developed for the homogenous sequence-specific detection of non-amplified genomic DNA and mRNA. Recently, Castro et al. demonstrated an ultra sensitive detection of DNA sequences by specific enzyme labeling without PCR amplification.²¹

Two-color organic dyes are usually used for coincidence detection and single-pair fluorescence resonance energy transfer detection (spFRET),²² however their functional limitations, such as the spectral cross-talk and non-uniform fluorophore photo bleaching rates, make subsequent quantification analysis complicated. Alternatively, quantum dots (QDs) have broad excitation and size-tunable photo luminescence spectra with narrow emission bandwidth (full-width at half-maximum of ∼25–40 nm), exceptional photochemical stability and relative high quantum yield.^23,24 QDs have been used as fluorescent markers in genomic analysis, immunoassay, fluorescence imaging and drug delivery.^25,26 Recently, they have been used as a FRET donor in biosensors to detect DNA and protein.^27–30 In this paper we report a homogenous method for rapid and sensitive detection of nucleic acids using two-color QDs based on single-molecule coincidence detection.

Experimental

1. Materials

The oligonucleotide probes and target were purchased from Integrated DNA Technol Inc. (Coralville, IA, USA) and had been purified by high-performance liquid chromatography. The sequence of oligonucleotide probes was 5′-GCA ACT AAA TTC A-3′-BioTEG (Probe-1) and BioTEG-5′-AAA GGA CCA GGC-3′ (Probe-2). The sequence of the target oligonucleotide was 5′-TGA ATT TAG TTG CGC CTG GTC CTT T-3′ (Target), which was complementary to Probe-1 and Probe-2. The sequence of the Alexa fluor 488 labeled oligonucleotide probes for fluorescence correlation spectroscopy experiment, was Alexa-5′-GCA ACT AAA TTC A-3′. Both streptavidin-coated 605QDs and 525QDs were purchased from Quantum Dot Corp. (Hayward, CA).

2. Experimental setup for single-molecule coincidence detection

The experimental setup for single-molecule coincidence detection is shown in Fig. 1A. A 488 nm-krypton/argon ion laser (Coherent Inc., Santa Clara, CA) was used as the excitation source. The laser beam was reflected by a dichroic beamsplitter (Z488RDC, Chroma Technology Corp., Rockingham, VT) and focused by an oil immersion 100×/1.30 NA objective lens (Olympus America, Inc., Melville, NY) to excite a sample in a 50 µm id capillary. The sample was passed through a laser-focused detection volume using the pressure-driven flow of a syringe pump (Harvard Apparatus, Holliston, MA). The flow rate in all measurements was kept at 1 µl min⁻¹. Fluorescence light was collected through the same objective, passed through a dichroic beamsplitter (Z488RDC, Chroma Technology Corp., Rockingham, VT) and a 50 µm pinhole (Melles Griot Co, Irvine, CA), then separated by another dichroic beamsplitter (560DCLP, Chroma Technology Corp, Rockingham, VT) and filtered through two band-pass filters (D525/40 m and D605/40 m, Chroma Technology Corp, Rockingham, VT) in the green and red channels, respectively; the light detection was accomplished by two avalanche photodiodes (Model SPCM-AQR-15, EG&G Canada, Vaudreuil, PQ, Canada). A program written in Labview (National Instruments, Austin, TX) and a digital counter (National Instruments, Austin, TX) were used to perform data acquisition and on-line data analysis.

Fig. 1 The principle of QD-based single-molecule coincidence detection. A, The experimental setup for QD-based single-molecule coincidence detection. The sample was passed through a laser-focused detection volume using the pressure-driven flow of a syringe pump. A 488 nm argon-laser was used to excite both 525QD and 605QD. The fluorescence emissions from 525QD and 605QD were separated and detected in green and red channels, separately. B, The conceptual schematic for QD-based single-molecule coincidence detection. Two biotinylated oligonucleotide probes were used to recognize and detect specific complementary target DNA through sandwich hybridization reaction. The DNA hybrids were first caught and assembled on the surface of 605QDs through specific streptavidin–biotin binding; then the second 525QDs were added to bind the other end of DNA hybrids to form the 605QD/DNA hybrid/525QD complex. C, The normalized emission spectra of 525QD and 605QD.

3. Single-molecule detection experiments

The hybridization experiments were performed in a buffered solution containing 100 mM Tris-HCl, 3 mM MgCl₂, pH 8.0. The reactions were carried out by mixing the two biotinylated probes and the target DNA at 35 °C for 30 min (the molecular ratio of the two biotinylated probes was kept at the ratio of 1 ∶ 1). After cooling to room temperature, streptavidin-coated 605QDs were added to capture the sandwiched hybrids. After incubation for 10 min, the streptavidin-coated 525QDs were added to bind the other end of DNA hybrids for another 10 min prior to the detection. Fluorescent data from both the green and red channels were collected in 1 ms intervals for a total running time of 200 s.

Fluorescence correlation spectroscopy (FCS) was used to determine the diffusion coefficient of both Alexa fluor 488-labeled oligonucleotide probes and 605QD-labeled oligonucleotide probes. The autocorrelation function G(τ) of fluorescence signal was obtained by a multiple tau digital correlator (ALV-GmbH, Langen, Germany). FCS curves were fitted using a 3D diffusion model.³¹

N denotes the number of fluorescent particles within the detection volume, τ_diff is the diffusion time of the fluorescent particle (D = w²_xy/4τ_diff, D is the diffusion coefficient of the particle), and w_xy and w_z are the radius and half height of the detection volume, respectively.

Results and discussions

Fig. 1B shows the conceptual schematic for the detection of nucleic acids using two-color QDs based on single-molecule coincidence detection. Two biotinylated oligonucleotide probes were used to recognize and detect specific complementary target DNA through a sandwich hybridization reaction. The DNA hybrids were first caught and assembled on the surface of 605QDs through specific streptavidin–biotin binding; then the second 525QDs were added to bind the other end of DNA hybrids. The coincidence signals were observed only when the presence of target DNA led to the formation of 605QD/DNA hybrid/525QD complex. Fig. 2 shows the representative fluorescence signals for both 525QD and 605QD in the absence (A and B) and in the presence of target DNA (C and D); the 525QD and 605QD signals were obtained simultaneously from two separate channels. In the absence of target DNA or in the presence of non-specific target DNA, no coincidence signals were observed (Fig. 2A and B), suggesting no false-positive coincidences due to the non-specific binding; however, in the presence of target DNA, distinct coincidence signals were observed, with each 605QD signal having a corresponding one 525QD signal (Fig. 2C and D, coincidence bursts with simultaneous fluorescence in green and red channels are marked with asterisks). Since this assay was carried out in a microfluidic flow and at a single-molecule level, the free 525QDs were far away from 605QDs, and consequently no false-positive coincidences were observed (Fig. 2A and B).

Fig. 2 Representative traces of fluorescent bursts in the absence (A and B) and in the presence of targets (C and D). A and C show the fluorescent bursts of 605QD in the red channel; B and D show the fluorescent bursts of 525QD in the green channel. Coincidence bursts with simultaneous fluorescence in green and red channels are marked with asterisks.

An important requirement of two-color coincidence detection is that the wavelengths of the two fluorescent probes have very low cross-talk to avoid false-coincidence. Since the Stokes' shift of the organic fluorophores is typically very small (20–30 nm), it is difficult to find two fluorophores that can be excited by one single-wavelength laser and meantime have no cross-talk. Selecting two fluorophores excited by two lasers with distinct wavelengths can maximize the separation between two emission wavelengths and minimizes the cross-talk; however, the confocal volumes of the two lasers cannot have perfect overlap (typically less than 30% due to the wavelength difference of two excitation lasers¹⁶), thereby decreasing the detection efficiency. QDs exhibit size-dependent tunable photoluminescence with narrow emission bandwidths; moreover, different wavelength QDs can be excited by a single-wavelength laser.^24,25 The 605QDs and 525QDs were chosen as a fluorescence pair for coincidence detection because of their good spectral resolution and no spectral overlap (Fig. 1C); no cross-talk between 605QD and 525QD emissions were observed under current experimental conditions (data not shown), making them ideal fluorescent probes for two-color coincidence detection.

To minimize the cross-linking of 605QDs, excess biotinylated oligonucleotide probes were used to saturate the binding sites of the first 605QD. Assuming three available biotin binding sites per streptavidin after conjugation to QDs, up to ∼15–30 biotinylated oligonucleotide probes or sandwiched hybrids could be captured by a single 605QD based on the assumption that each QD conjugated with 5–10 streptavidins. The multiple binding sites of streptavidin-coated QD also increase the chance of its binding to the DNA hybrids. Moreover, the long persistence length of double-stranded DNA (dsDNA) prevents the two biotinylated terminals of DNA hybrids from binding to the same 605QD (the persistence length l_p ≈ 50 nm for dsDNA,³² contour length ≈ 8.5 nm for a dsDNA 25 mer). After the DNA hybrids bound to the 605QDs, the 525QDs were added to bind the other terminal of DNA hybrids. In principle, n 525QDs linking to one 605QD through DNA hybrids will result in an n-fold increase in the green fluorescence burst of the coincidence signals, therefore amplifying the coincidence signals to distinguish them from free 525QD signals. To determine the maximum concentration of 525QDs for efficient binding to the DNA hybrid-linked 605QD, a 525QD/605QD ratio titration was performed. Fig. 3 demonstrates that the coincidence events increase with the increase of the 525QD/605QD ratio up to a ratio of about 15/1, where saturation occurs. This result indicates that the increase of the 525QD/605QD ratio improves the chance that the 525QDs will bind to the DNA hybrid/605QD complex; this is important for low-concentration target detection (one coincidence signal amounts to ∼15 copies of target DNA in this assay) and suggests that the detection efficiency of this QD-based assay may be diffusion-dependent in the solution.

Fig. 3 Variance of coincidence events as a function of the ratio of 525QD/605QD. Probe-1 concentration: 3 × 10⁻¹¹ M; Probe-2 concentration: 3 × 10⁻¹¹ M; Target concentration: 3 × 10⁻¹¹ M; 605QD concentration: 1 × 10⁻¹² M; 525QD concentration varies with the ratio of 525QD/605QD shown.

Conventional QD-based assay usually uses QD-labeled oligonucleotide probes to hybridize the DNA target;^18,33 however, the extremely low diffusion coefficients of QD-labeled oligonucleotide probes adversely influences the hybridization efficiency and results in a subsequent low detection efficiency. Fig. 4A shows the respective autocorrelation curves of Alexa fluor 488-labeled oligonucleotide probe and 605QD-labeled oligonucleotide probe and their fitted curves by FCS in the Brownian motion state. The diffusion coefficient of Alexa fluor 488-labeled oligonucleotide probe was calculated to be 8.4 × 10⁻¹¹ m² s⁻¹, while the diffusion coefficient of 605QD-labeled oligonucleotide probe was 3.8 × 10⁻¹² m² s⁻¹; there is almost one order of magnitude difference between them because the molecular weight of the QD (MW ∼ 2 × 10⁶) is much larger than that of Alexa fluor 488 (MW = 694.6). The molecular weight of biotin-TEG is 569.6; the diffusion coefficient of the biotinylated oligonucleotide probe was assumed similar to that of the Alexa fluor 488-labeled oligonucleotide probe. The difference in the diffusion coefficients between the biotinylated oligonucleotide probe and the QD-labeled oligonucleotide probe might make a difference in their access to the target DNA, and influence the subsequent hybridization efficiency, especially for low-concentration target detection. Fig. 4B shows the kinetics of DNA hybridization at 20 °C using biotinylated oligonucleotide probes and QD-labeled oligonucleotide probes separately. For biotinylated oligonucleotide probes, the coincidence events saturate within 30 min, while for QD-labeled oligonucleotide probes, the coincidence events saturate after 90 min. Even after 90 min, the coincidence events using QD-labeled oligonucleotide probes were still less than those using biotinylated oligonucleotide probes, suggesting that high hybridization efficiency may result in improved coincidence events. It is worth noticing that coincidence events with incubation at 35 °C (373 ± 19 counts) were more than that at 20 °C (150 ± 9 counts) for the same incubation time of 30 min, indicating that increasing temperature might significantly improve hybridization efficiency; this provides an alternative way to improve hybridization efficiency and the subsequent detection efficiency. However, this does not apply to QD-labeled oligonucleotide probes because QDs tend to aggregate and bind to each other at high temperature, even in the absence of target DNA (data not shown), which limits their application where temperature treatment and enzyme reaction are needed. With increasing temperature, the current assay can greatly shorten the analysis time by reducing the hybridization incubation time to as low as ∼5 min. Moreover, with the advantage of temperature treatment, the current assay can be applied to the single nucleotide polymorphisms detection assays, such as allele-specific primer extension (ASPE), single base extension assays (SBE) and oligonucleotide ligation assay (OLA).³⁴

Fig. 4 A, Autocorrelation curves (solid line: measured curve, dashed line: fitted curve) of Alexa fluor 488-labeled oligonucleotide probe and 605QD-labeled oligonucleotide probe in Brownian motion state with FCS; B, Variance of coincidence events with the hybridization time for biotinylated oligonucleotide probes (□) and QD-labeled oligonucleotide probes (●). Probe-1 concentration: 3 × 10⁻¹¹ M; Probe-2 concentration: 3 × 10⁻¹¹ M; Target concentration: 3 × 10⁻¹¹ M; 605QD concentration: 1 × 10⁻¹² M; 525QD concentration: 5 × 10⁻¹² M.

Fig. 5 shows the quantitative analysis of target DNA using two-color QDs based on single-molecule coincidence detection. The coincidence events increased with the increasing of target DNA, the detection limit is 5 × 10⁻¹⁵ M. This result demonstrates that the quantitative analysis of target DNA can be obtained by simply counting coincidence events, and thus it is possible to determine the specific target DNA concentration in complex sample without the need for DNA amplification.

Fig. 5 The variance of coincidence events with target DNA concentration. Probe-1 concentration: 3 × 10⁻¹¹ M; Probe-2 concentration: 3 × 10⁻¹¹ M; 605QD concentration: 1 × 10⁻¹² M; 525QD concentration: 5 × 10⁻¹² M. The data shows mean ± S.D.M (n = 3).

Conclusion

In this paper we have demonstrated a homogenous method for rapid and sensitive detection of nucleic acids using two-color QDs based on single-molecule coincidence detection. In comparison with conventional QD-based assay, the current assay has the advantage of high detection efficiency, short analysis time and is suitable for temperature treatment; the detection limit can reach 5 × 10⁻¹⁵ M. This assay can also be applied to the detection of single nucleotide polymorphisms, proteins and RNA.

Acknowledgements

This work was supported by NIH under award no. GM08153.

References

1. L. J. Kricka, Ann. Clin. Biochem., 2002, 39, 114–129.
2. T. A. Taton, G. Lu and C. A. Mirkin, J. Am. Chem. Soc., 2001, 123, 5164–5165.
3. Y. W. Cao, R. Jin and C. A. Mirkin, Science, 2002, 297, 1536–1540.
4. B. Reiss, L. He, I. Walton, R. Cromer, C. Keating and M. Natan, Science, 2001, 294, 137–141.
5. J. A. Ferguson, F. J. Steemers and D. R. Walt, Anal. Chem., 2000, 72, 5618–5624.
6. D. E. Gerion, F. Q. Chen, B. Kannan, A. H. Fu, W. J. Parak, D. J. Chen, A. Majumdar and A. P. Alivisatos, Anal. Chem., 2003, 75, 4766–4772.
7. J. Wang, G. D. Liu and A. Merkoci, J. Am. Chem. Soc., 2003, 125, 3214–3215.
8. Y. Li, Y. T. Hong and D. Luo, Nat. Biotechnol., 2005, 23, 885–889.
9. J. Sambrook and D. W. Russell, Molecular Cloning: A Laboratory Mannual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, 2001.
10. W. W. Wilke, R. N. Jones and L. D. Sutton, Diagn. Microbiol. Infect. Dis., 1995, 21, 181–185.
11. J. P. Knemeyer, N. Marmé and M. Sauer, Anal. Chem., 2000, 72, 3717–3724.
12. M. Eigen and R. Rigler, Proc. Natl. Acad. Sci. U. S. A., 1994, 91, 5740–5747.
13. C. Y. Zhang, S. Y. Chao and T. H. Wang, Analyst, 2005, 130, 483–488.
14. A. Castro and J. G. K. Williams, Anal. Chem., 1997, 69, 3915–3920.
15. A. Castro and R. T. Okinaka, Analyst, 2000, 125, 9–11.
16. H. T. Li, L. M. Ying, J. J. Green, S. Balasubramanian and D. Klenerman, Anal. Chem., 2003, 75, 1664–1670.
17. H. C. Yeh, E. Simone, C. Y. Zhang and T. H. Wang, 17th IEEE International Conference on Micro Electro Mechanical Systems, 2004, pp. 371–374.
18. R. L. Nolan, H. Cai, J. P. Nolan and P. M. Goodwin, Anal. Chem., 2003, 75, 6236–6243.
19. A. Camacho, K. Korn, M. Damond, J. F. Cajot, E. Litborn, B. H. Liao, P. Thyberg, H. Winter, A. Honegger, P. Gardellin and R. Rigler, J. Biotechnol., 2004, 107, 107–114.
20. R. Rigler, Z. Foldes-Papp, F. J. Meyer-Alme, C. Sammet, M. Volcker and A. Schnetz, J. Biotechnol., 1998, 63, 97–109.
21. A. Castro, D. A. R. Dalvit and L. Paz-Matos, Anal. Chem., 2004, 76, 4169–4174.
22. M. B. Wabuyele, H. Farquar, W. Stryjewski, R. P. Hammer, S. A. Soper, Y. W. Cheng and F. Barany, J. Am. Chem. Soc., 2003, 125, 6937–6945.
23. M. Bruchez, M. Moronne, P. Gin, S. Weiss and A. P. Alivisatos, Science, 1998, 281, 2013–2016.
24. W. C. W. Chan and S. M. Nie, Science, 1998, 281, 2016–2018.
25. I. L. Medintz, H. T. Uyeda, E. Goldman and H. Mattoussi, Nat. Mater., 2005, 4, 437–446.
26. A. P. Alivisatos, Nat. Biotechnol., 2004, 22, 47–52.
27. F. Patolsky, R. Gill, Y. Weizmann, T. Mokari, U. Banin and I. Willner, J. Am. Chem. Soc., 2003, 125, 13918–13919.
28. I. L. Medintz, A. R. Clapp, H. Mattoussi, E. R. Goldman, B. Fisher and J. M. Mauro, Nat. Mater., 2003, 2, 630–638.
29. I. L. Medintz, J. H. Konnert, A. R. Clapp, I. Stanish, M. E. Twigg, H. Mattoussi, J. M. Mauro and J. R. Deschamps, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 9612–9617.
30. C. Y. Zhang, H. C. Yeh, M. T. Kuroki and T. H. Wang, Nat. Mater., 2005, 4, 826–831.
31. Z. Foldes-Papp, U. Demel and G. P. Tilz, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 11509–11514.
32. M. Singh-Zocchi, S. Dixit, V. Ivanov and G. Zocchi, Proc. Natl. Acad. Sci. U. S. A., 2004, 100, 7605–7610.
33. Y. P. Ho, M. C. Kung, S. Yang and T. H. Wang, Nano Lett., 2005, 5, 1693–1697.
34. P. Y. Kwok, Single nucleotide polymorphisms methods and protocols, Human Press, Totowa, New Jersey, 2003.

---