Homogeneous Competitive Hybridization Assay Based on Two

Jan 15, 2010 - Key Laboratory of Analytical Chemistry for Biology and Medicine (Ministry of Education), College of Chemistry and Molecular Sciences, a...
2 downloads 0 Views 2MB Size
Anal. Chem. 2010, 82, 1381–1388

Homogeneous Competitive Hybridization Assay Based on Two-Photon Excitation Fluorescence Resonance Energy Transfer Lingzhi Liu,† Xiaohu Dong,† Wenlong Lian,† Xiaoniu Peng,‡ Zhihong Liu,*,† Zhike He,† and Ququan Wang‡ Key Laboratory of Analytical Chemistry for Biology and Medicine (Ministry of Education), College of Chemistry and Molecular Sciences, and Key Laboratory of Acoustic and Photonic Materials and Devices of Ministry of Education, Wuhan University, Wuhan 430072, People’s Republic of China Recently, we have successfully developed a two-photon excitation fluorescence resonance energy transfer (TPEFRET)-based homogeneous immunoassay using twophoton excitable small organic molecule as the energy donor. In the present work, the newly emerging TPEFRET technique was extended to the determination of oligonucleotide. A new TPE molecule with favorable twophoton action cross section was synthesized [2-(2,5-bis(4(dimethylamino)styryl)-1H-pyrrol-1-yl)acetic acid, abbreviated as TP-COOH], with the tagged reactive carboxyl group allowing facile conjugation with streptavidin (SA). Employing the TP-COOH molecule as energy donor and black hole quencher 1 (BHQ-1) as acceptor, a TPE-FRETbased homogeneous competitive hybridization model was constructed via a biotin-streptavidin bridge. Through the hybridization between a biotinylated single-stranded DNA (ssDNA) and a BHQ-1-linked ssDNA, and the subsequent capture of the as-formed hybrid by TP-COOH labeled SA, the donor fluorescence was quenched due to the FRET between TP-COOH and BHQ-1. Upon the competition between a target ssDNA and the quencher-linked ssDNA toward the biotinylated oligonucleotide, the donor fluorescence was recovered in a target-dependent manner. Good linearity was obtained with the target oligonucleotide ranging from 0.08 to 1.52 µM. The method was applied to spiked serum and urine samples with satisfying recoveries obtained. The results of this work verified the applicability of TPE-FRET technique in hybridization assay and confirmed the advantages of TPE-FRET in complicated matrix. The quantification and recognition of specific DNA sequences is of great importance to the development of genomics, virology, and molecular biology; hence, the establishment of highly sensitive methods with simple procedures for DNA assay has long been the focus of analytical chemistry. The hybridization technique is widely used in DNA analysis with high specificity. Many DNA * To whom correspondence should be addressed. E-mail: [email protected]. Phone: 86-27-8721-8754. Fax: 86-27-6875-4067. † College of Chemistry and Molecular Sciences. ‡ Key Laboratory of Acoustic and Photonic Materials and Devices of Ministry of Education. 10.1021/ac902467w  2010 American Chemical Society Published on Web 01/15/2010

detection systems conducted on solid-state substrates, like microarrays, are capable of providing high sensitivity due to the separation of unhybridized DNA strands from hybridized strands.1 However, the separation steps also make the assay procedure somewhat complicated, and the solid-state substrates may sometimes cause nonspecific adsorption of biomolecules on their surface.2 Therefore, homogeneous hybridization methods have attracted increasing attention and achieved remarkable advances in the study of DNA. The separation-free nature of the homogeneous assay has made the assay time shortened and the procedure simplified. What is more, detection in solution can facilitate reaction kinetics and reduce nonspecific adsorption.3 Fluorescence resonance energy transfer (FRET), which is considered as a sensitive and reliable “ruler” over distances of 10-100 Å,4 is one of the most commonly used techniques in homogeneous assays.5,6 Several FRET-based molecular probes (e.g., TaqMan probes,7,8 molecular beacons,9 and FRET probes10) have been put forward to enable separation-free detection of DNA in homogeneous solution. However, most conventionally used organic dyes emit Stokes fluorescence, i.e., the energy of the emission is typically lower than that of absorption (ultraviolet or visible light), which will inevitably cause problems in FRET-based assays for biological objectives. While the energy donor is excited with ultraviolet or visible light, autofluorescence or scattering light always arises from samples upon excitation of the energy donor.11 In addition, the energy acceptor is often coexcited with the donor because of the overlap of their excitation spectra due to the small (1) Zhang, C.-Y.; Yeh, H.-C.; Kuroki, M. T.; Wang, T.-H. Nat. Mater. 2005, 4, 826–831. (2) Taylor, S.; Smith, S.; Windle, B.; Guiseppi-Elie, A. Nucleic Acids Res. 2003, 31, e87. (3) Dubertret, B. Nat. Mater. 2005, 4, 797–798. (4) dos Remedios, C. G.; Moens, P. D. J. J. Struct. Biol. 1995, 115, 175–185. (5) Wang, S.; Mamedova, N.; Kotov, N. A.; Chen, W.; Studer, J. Nano Lett. 2002, 2, 817–822. (6) Oswald, B.; Lehmann, F.; Simon, L.; Terpetschnig, E.; Wolfbeis, O. S. Anal. Biochem. 2000, 280, 272–277. (7) Holland, P. M.; Abramson, R. D.; Watson, R.; Gelfand, D. H. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 7276–7280. (8) Livak, K. J.; Flood, S. J. A.; Marmaro, J.; Giusti, W.; Deetz, K. PCR Methods Appl. 1995, 4, 357–362. (9) Tyagi, S.; Kramer, F. R. Nat. Biotechnol. 1996, 14, 303–308. (10) Issa, N. C.; Espy, M. J.; Uhl, J. R.; Smith, T. F. J. Clin. Microbiol. 2005, 43, 1843–1845. (11) Chen, Z.; Chen, H.; Hu, H.; Yu, M.; Li, F.; Zhang, Q.; Zhou, Z.; Yi, T.; Huang, C. J. Am. Chem. Soc. 2008, 130, 3023–3029.

Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

1381

Stokes shift.12,13 Therefore, new energy donor-acceptor pairs are desired for FRET methods to resolve such problems. Although most fluorescent emitters comply with the Stokes law, a new class of fluorophores can be excited at longer wavelength (near-infrared region) to give emission at relatively shorter wavelength (visible region) through a two-photon excitation process14 or upconversion process,15,16 which is called antiStokes photoluminescence. Under the excitation with infrared light, the interference from out-of-focus autofluorescence or the scattered excitation light could be eliminated and the direct excitation of energy acceptor can be avoided, resulting in a nearzero background,17 which leads to higher signal/noise ratio and thus ensures higher sensitivity in quantitative analysis. Such merits suggest that fluorophores emitting anti-Stokes fluorescence could be promising energy donors for FRET-based bioassays. The earliest application of anti-Stokes fluorescence in homogeneous FRET bioassays was reported in 2005 by Kuningas et al. using inorganic upconverting particles (UCPs) as donors.18,19 Thereafter, several UCP-FRET methods have been developed for bioassays, including the quantitative determination of DNA using a UCPs-based DNA sensor.20 The significance of using the nearinfrared light excitable UCPs as energy donors in bioassays has been proven in these works. As a class of inorganic crystal fluorophores, UCPs exhibit outstanding photophysical properties, such as strong fluorescence intensity, sharp and symmetric emission, and good stability.17,21 Nevertheless, some negative effects of such inorganic particles, such as the requirement for surface modifications and nonspecific binding interactions, may not be ignored.10,22 Such concerns of UCPs lead naturally to the consideration of two-photon excitation (TPE), which occurs when a small organic molecule simultaneously absorbs two lower-energy photons and emits a higher-energy photon. As two types of antiStokes fluorophore which both absorb two lower-energy photons and emit at a higher-energy region, it is hard to decide which one could be better in practical applications, although this is occasionally discussed in literatures.23 What is undoubtable, the inherent advantages of a small molecule, e.g., the higher structural flexibility and the capability of avoiding nonspecific binding, make it a competent alternative (12) Schmid, J. A.; Scholze, P.; Kudlacek, O.; Freissmuth, M.; Singer, E. A.; Sitte, H. H. J. Biol. Chem. 2001, 276, 3805–3810. (13) Willard, D. M.; Carillo, L. L.; Jung, J.; Orden, A. V. Nano Lett. 2001, 1, 469–474. (14) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 2nd ed.; Kluwer Academic: New York, 1999. (15) Zijlmans, H. J. M. A. A.; Bonnet, J.; Burton, J.; Kardos, K.; Vail, T.; Niedbala, R. S.; Tanke, H. J. Anal. Biochem. 1999, 267, 30–36. (16) Niedbala, R. S.; Feindt, H.; Kardos, K.; Vail, T.; Burton, J.; Bielska, B.; Li, S.; Milunic, D.; Bourdelle, P.; Vallejo, R. Anal. Biochem. 2001, 293, 22– 30. (17) Kuningas, K.; Ukonaho, T.; Pa¨kkila¨, H.; Rantanen, T.; Rosenberg, J.; Lo ¨vgren, T.; Soukka, T. Anal. Chem. 2006, 78, 4690–4696. (18) Kuningas, K.; Rantanen, T.; Ukonaho, T.; Lo ¨vgren, T.; Soukka, T. Anal. Chem. 2005, 77, 7348–7355. (19) Wang, L.; Yan, R.; Huo, Z.; Wang, L.; Zeng, J.; Bao, J.; Wang, X.; Peng, Q.; Li, Y. Angew. Chem., Int. Ed. 2005, 44, 6054–6057. (20) Zhang, P.; Rogelj, S.; Nguyen, K.; Wheeler, D. J. Am. Chem. Soc. 2006, 128, 12410–12411. (21) Yi, G.; Lu, H.; Zhao, S.; Ge, Y.; Yang, W.; Chen, D.; Guo, L.-H. Nano Lett. 2004, 4, 2191–2196. (22) Kuningas, K.; Rantanen, T.; Karhunen, U.; Lo ¨vgren, T.; Soukka, T. Anal. Chem. 2005, 77, 2826–2834. (23) Liu, L.; Shao, M.; Dong, X.; Yu, X.; Liu, Z.; He, Z.; Wang, Q. Anal. Chem. 2008, 80, 7735–7741.

1382

Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

to UCPs. In the past few years, the special optical properties of TPE have earned increasing interest in biological applications. For example, a bioaffinity assay platform based on TPE fluorescence was built using polymer microspheres as solid-phase reaction carrier, which enables multiplexed separation-free bioaffinity assays with high sensitivity due to the absence of out-of-focus autofluorescence.24-26 TPE fluorescence was also adopted in constructing fluorescent probes for cell imaging.27,28 In comparison to the relatively more extensive application of UCP-FRET, TPE-FRET has rarely been reported. A method based on labeled DNA for determining the TPE-FRET efficiency was reported by Wahlroos et al.29 Later in 2007 Clapp et al. demonstrated FRET between quantum dots and dye acceptors driven by a two-photon process and applied it in fluorescence imaging.30 In our recent work, we have designed and synthesized some TPE molecules, with which we have developed a TPE-FRET technique with small organic molecules as energy donor, and preliminarily investigated its applicability in quantitative bioassay.23,31 The results succeeded in showing the strong ability of TPE-FRET to circumvent the problem of autofluorescence in biological samples, verifying the notable advantage of TPE-FRET over conventional OPE-FRET. Herein in the present work, with further design of the TPE molecule and construction of analytical model, we have made an attempt to set up a new TPE-FRET method for nucleic acid hybridization assay. An improved TPE molecule 2-(2,5-bis(4(dimethylamino)styryl)-1H-pyrrol-1-yl)acetic acid (written as TPCOOH for short, structure and synthesis details shown in Supporting Information Scheme S1) tagged with a reactive carboxyl group is synthesized. With a commercially available quencher BHQ-1 as acceptor, a TPE-FRET-based homogeneous competitive hybridization assay model is built up via a biotinstreptavidin bridge, with which quantitative determination of single-stranded DNA (ssDNA) is achieved. Our efforts are aimed at further completing the TPE-FRET methodology and exploring its applicability to various biological objectives. To the best of our knowledge, such a TPE-FRET-based homogeneous hybridization assay for quantitative determination of DNA has not been reported so far. EXPERIMENTAL SECTION Reagents. The donor TP-COOH was synthesized in this lab (please see the Supporting Information for details). Streptavidin (SA) was purchased from Sigma (St. Louis, MO). Fluorescein, N-hydroxysuccinimide (NHS), and N,N′-dicyclohexylcarbodiimide (DCC) were the products of Sinopharm Chemical Reagent Co. (24) Ha¨nninen, P.; Soini, A.; Meltola, N.; Soini, J.; Soukka, J.; Soini, E. Nat. Biotechnol. 2000, 18, 548–550. (25) Soini, J. T.; Soukka, J. M.; Meltola, N. J.; Soini, A. E.; Soini, E.; Ha¨nninen, P. E. Single Mol. 2000, 1, 203–206. (26) Koskinen, J. O.; Vaarno, J.; Meltola, N. J.; Soini, J. T.; Ha¨nninen, P. E.; Luotola, J.; Waris, M. E.; Soini, A. E. Anal. Biochem. 2004, 328, 210–218. (27) Kim, H. M.; Kim, B. R.; Hong, J. H.; Park, J. S.; Lee, K. J.; Cho, B. R. Angew. Chem., Int. Ed. 2007, 46, 7445–7748. (28) Dong, X.; Yang, Y.; Sun, J.; Liu, Z.; Liu, B. Chem. Commun. 2009, 3883– 3885. (29) Wahlroos, R.; Toivonen, J.; Tirri, M.; Han ¨ ninen, P. J. Fluoresc. 2006, 16, 379–386. (30) Clapp, A. R.; Pons, T.; Medintz, I. L.; Delehanty, J. B.; Melinger, J. S.; Tiefenbrunn, T.; Dawson, P. E.; Fisher, B. R.; O’Rourke, B.; Mattoussi, H. Adv. Mater. 2007, 19, 1921–1926. (31) Liu, L.; Wei, G.; Liu, Z.; He, Z.; Xiao, S.; Wang, Q. Bioconjugate Chem. 2008, 19, 574–579.

(Shanghai, China). Sephadex G-25 gel was from Amersham Pharmacia Biotech (China) Ltd., and normal rabbit serum was from Zhongshan Golden Bridge Biotechnology Co., Ltd. (China). Biotin-linked ssDNA (5′-biotin-GAG TTA GCA CCC GCA TAG TCA AGA T-3′), BHQ-1-linked ssDNA (5′-TGA CTA TGC GGG TGC TAA-BHQ-1-3′), and target ssDNA (5′-ATC TTG ACT ATG CGG GTG CTA ACT C-3′) were from Shanghai Sangon Biological Engineering Technology & Service Co., Ltd. (Shanghai, China). All the solutions were prepared in ultrapure water with a resistivity of 18.2 MΩ · cm (purified by a Milli-Q system from Millipore China Ltd.). Instruments for Fluorescence and Absorption Measurements. Two-photon excited fluorescence was measured similarly as our previous work.31 Briefly, all samples were excited at 800 nm with ∼3 ps pulse width at a repetition rate of 76 MHz. The excitation light was provided by a mode-locked Ti:Sapphire pulsed laser (Mira 900, Coherent Inc., Santa Clara, CA), and the antiStokes photoluminescence was recorded on a liquid-nitrogencooled CCD-array spectrometer (SPEC-10, Princeton Instruments, Trenton, NJ) through a monochromator (Spectrapro 2500i, Acton Research Corporation, Acton, MA). The fluorescence intensity was given in normalized form, i.e., the maximum emission of each group of data was set as 1, and others were calculated proportionally. A Tsunami-Spitfire-OPA-800C (Spectra-Physics) was used for measuring the two-photon absorption cross section. One-photon excited fluorescence was measured on a Perkin-Elmer LS 55 fluorometer (Foster City, CA) with the excitation of 420 nm, and the signal was recorded with a photomultiplier tube (PMT) detector. Absorption measurements were conducted on a UV2550 UV-vis spectrophotometer (Shimadzu Scientific Instruments Inc.). Preparation of the TP-COOH-Streptavidin (TP-SA) Conjugate. TP-COOH was first converted to succinimidyl ester to facilitate the conjugation with SA. To the solution of NHS (16 mg, 0.14 mmol) and TP-COOH (20 mg, 0.048 mmol) in 1.5 mL of anhydrous THF, DCC (21 mg, 0.1 mmol) dissolved in 1.5 mL of THF was added dropwise at 0 °C under argon atmosphere. The mixture was stirred at room temperature for 2 h. After filtration, the filtrate was evaporated and the residue was purified by column chromatography on silica gel with CH2Cl2 as eluent to give a dark-green solid of TP-COOH succinimidyl ester (TP-NHS). A one-step reaction procedure was then used for the labeling of streptavidin by the amine-reactive TP-NHS.32 SA was dissolved in 0.1 M sodium bicarbonate buffer, pH 8.3, at a concentration of 1 mg/mL. TP-NHS was dissolved in anhydrous DMF at a concentration of 5 mg/mL. With gentle stirring, a 30-fold molar excess of TP-NHS was added to 2 mL of the SA solution. After incubating for 2 h at room temperature, the mixture was centrifuged and the unconjugated label was removed by gel filtration using Sephadex G-25 equilibrated with phosphate buffer (0.01 M, pH 7.4). The purified protein fractions were further concentrated by ultrafiltration using an Amicon Ultra-4 centrifugal filter device with a MW cutoff of 50 kDa (Millipore Corp.) at 4 °C. The degree of labeling was spectrophotometrically determined. The concentration of TP-COOH in the TP-SA complex was

calculated using its molar extinction coefficient in phosphate buffer, which was experimentally obtained through a calibration curve. Since TP-COOH has a contribution to the absorbance at 280 nm, the correction factor to account for absorption of the dye at 280 nm was also estimated. Therefore, the concentration of SA was calculated from the absorbance at 280 nm after subtracting the contribution of TP-COOH, using the molar extinction coefficient of 2 × 105 M-1 cm-1.33 The dye-to-protein ratio was calculated accordingly. Building of the TPE-FRET Model. The TPE-FRET experiments were conducted through titration of the donor with acceptor. Biotinylated ssDNA and BHQ-1-ssDNA were hybridized prior to the titration. Briefly, biotin-ssDNA and BHQ1-ssDNA were prediluted to the same concentration of 2 × 10-5 M with 0.3 M phosphate-buffered saline (10 mM phosphate buffer, 0.3 M NaCl, pH 7.4). Then BHQ-1 oligonucleotide was mixed thoroughly with an equal volume of biotinylated oligonucleotide and incubated at 37 °C for 1.5 h with gentle shaking. After cooling to room temperature, TX-100 was added to the mixture to give a content of 0.1% without altering the concentration of DNA. Under room temperature, increasing amounts of the as-prepared biotin-dsDNA-BHQ-1 hybrids were added to the solution of 0.32 µM TP-SA (denoting 0.32 µM SA) in assay buffer containing 10 mM phosphate buffer, 0.3 M NaCl, and 0.1% TX-100, pH 7.4. After brief incubation, the two-photon fluorescence emission of the donor was measured under the excitation of 800 nm. The optimal molar ratio of acceptor-todonor for subsequent competitive assay was thus determined by selecting the saturation concentration of the acceptor. For comparative studies, one-photon excited (OPE) FRET experiments were performed with the same manner. TPE-FRET-Based Homogeneous Competitive Hybridization Assay for Target ssDNA. In a typical TPE-FRET-based homogeneous competitive hybridization assay for target ssDNA, 1.2 µM biotin-ssDNA and equivalent BHQ-1-ssDNA were preincubated at 37 °C for 1.5 h, then varying concentrations of target ssDNA (0, 0.04, 0.08, 0.12, 0.28, 0.48, 0.72, 0.96, 1.2, and 1.52 µM) were individually added to the above hybrids for another 1.5 h of incubation at 37 °C with continuous shaking. The thus obtained mixtures were allowed to cool down to room temperature, and then TP-SA (concentration fixed at 0.32 µM) was added to capture the biotinylated hybrids. After brief incubation, TPE fluorescence detection was performed. Identical procedures were followed for OPE experiments for comparison. The applicability of the TPE-FRET method to a real biological system was examined through detecting the recoveries of target ssDNA in spiked serum and urine samples. Normal rabbit serum was used as received without further purification, and the urine sample was obtained from a healthy female volunteer and filtered with a 0.22 µm Millipore membrane (Bedford, MA). Both serum and urine samples were 30-fold diluted with assay buffer. The experiments were also conducted with the OPE model for comparison.

(32) Hermanson, G. T. Bioconjugate Techniques; Academic Press: San Diego, CA, 1996.

(33) http://www.sigmaaldrich.com/etc/medialib/docs/Sigma/Datasheet/5/ s6402dat.Par.0001.File.tmp/s6402dat.pdf (accessed June 8, 2009).

RESULTS AND DISCUSSION Photophysical Properties of TP-COOH. As a newly prepared fluorescent label, the absorption and fluorescence properties

Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

1383

Figure 1. (A) Linear absorption spectrum of TP-COOH (1 × 10-5 M in DMF). (B) The excitation power dependence of the fluorescence intensity and the log-log plot of the fluorescence intensity vs excitation power (the inset).

of TP-COOH were characterized first. As shown in Figure 1A, the maximum absorption of TP-COOH locates at 420 nm, and no linear absorption is observed at longer wavelength (beyond 500 nm). In order to confirm the two-photon process under the excitation of near-infrared region wavelength, the excitation power dependence of the fluorescence intensity was examined (Figure 1B). As expected, the slope of the log-log plot of fluorescence intensity versus excitation power was determined as 2.0 (inset in Figure 1B), indicating that the photoluminescence of TP-COOH excited at the near-infrared region is purely the two-photoninduced fluorescence. This anti-Stokes photoluminescence comes from the combined energy of two photons that falls into the maximum absorption band of the molecule. As has been well-recognized, the two-photon action cross section is an essential attribute for two-photon excited molecules, which is defined as the product ηδ (η is the fluorescence quantum yield, while δ is the two-photon absorption cross section). This parameter would be a decisive factor to the detection sensitivity in a TPE-based assay, since it decides the ability of the fluorescent molecule to upconvert energy, i.e., to emit relatively higher-energy photons upon absorbing lower-energy photons. According to the QSAR (quantitative structure-activity relationship) studies of twophoton absorption, a larger absorption cross section requires a larger hydrophobic conjugated motif, which, unfortunately, always leads to a decrease of biocompatibility of the molecule. Therefore, as compared with our last TPE label (TP-NH2),23 the conjugation length of TP-COOH was reduced and an electron-rich heterocycle pyrrole π bridge was adopted so as to increase the biocompatibility of the label. An approximately centrosymmetric D-π-D′-π-D scaffold (Supporting Information Scheme S1) was employed instead of the symmetric D-π-A-π-D quadrupole scaffold of TP-NH2. The fluorescence quantum yield of TPCOOH is determined as 0.69 using fluorescein as the reference (please refer to the Supporting Information for details), which is quite a competitive value compared to commonly seen fluorescent labels. The δ value of TP-COOH was determined with the method of two-photon-induced fluorescence. The dependence of δ on TP excitation wavelength was obtained (Supporting Information Figure S1). The maximum TPA cross section is achieved at 680 1384

Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

nm, which is consistent with Zheng et al.’s results concerning the two-photon absorption of this kind of molecules.34 The value of 800 nm was selected as the excitation wavelength in subsequent experiments because of the highest stability of the laser source at this wavelength. The δ value at 800 nm is 107 GM, which is relatively smaller than that of TP-NH2 (which was 417 GM) because of the reduced conjugation length and the less extent of intramolecular charge transfer. Nonetheless, the product ηδ is not reduced because of the remarkably enlarged η of TPCOOH, which in our opinion could be mainly attributed to the omission of the electron-withdrawing cyano groups in TP-NH2. Spectral Characterization of the TP-SA Conjugate. A notable improvement in the molecular structure of TP-COOH is the introduction of the carboxyl group as the reaction site, which is beneficial for conjugating with biomolecules. As to the previously reported TP-NH2, the label was linked to proteins through covalent reaction with glutaraldehyde, a homobifunctional cross-linker. As can be predicted, however, the glutaraldehyde method was susceptible to protein self-conjugation. In the present work, the reactive carboxyl group is allowed to produce dye-protein conjugate without any protein polymers. With the activation of DCC, TP-COOH was easily transformed to NHS ester, which is the most typical amine reactive cross-linking reagent and is often used in one-step procedures for protein labeling. After separating the unreacted small molecules by gel filtration, pure TP-SA conjugate was acquired. The UV-vis absorption spectrum of the complex in Figure 2 (solid line) exhibits two distinct peaks around 380 and 280 nm, corresponding to the characteristic absorption of TP-COOH and SA, respectively. When combined with the absorption of pure SA (dotted line), it confirms that TP-COOH was successfully coupled to SA considering that all small molecules were exhaustively removed. On the basis of the absorption of the TP-SA complex, the labeling efficiency was obtained. The concentration of the fluorescent donor in the complex was determined using its molar extinction coefficient at 406 nm in phosphate buffer, i.e., 2.94 × 104 M-1 cm-1, which was obtained through a calibration curve (34) Zheng, S.; Beverina, L.; Barlow, S.; Zojer, E.; Fu, J.; Padilha, L. A.; Fink, C.; Kwon, O.; Yi, Y.; Shuai, Z.; Van Stryland, E. W.; Hagan, D. J.; Bre´das, J. L.; Marder, S. R. Chem. Commun. 2007, 1372–1374.

Figure 2. UV-vis absorption spectra of the TP-SA complex (solid line, 0.92 µM SA) and pure SA (dotted line, 1 µM). All the spectra were recorded in 0.01 M phosphate buffer, pH 7.4.

(Supporting Information Figure S2). With taking into account the experimentally determined correction factor of TP-COOH absorbance at 280 nm (Supporting Information Figure S2), the concentration of SA in the complex was also acquired. Accordingly, the dye-to-protein ratio was calculated. As can be predicted, reasonable increase of the dye-to-protein ratio will bring more intense signal and thus help to improve the overall sensitivity of the assay. In our experiments, with increasing the labeling ratio up to 4.0, the fluorescence of TP-SA increased and the complex kept soluble. But an excessive labeling of TP-COOH to SA caused obvious aggregation of the protein owing to the increased hydrophobicity of the conjugate. Therefore, the TP-to-SA ratio was controlled as 4.0 in subsequent assays. The fluorescence emission of pure TP-COOH (solid line) and TP-SA (dotted line) were compared with that under one-photon excitation (Figure 3). The fluorescence of pure TP-COOH exhibits distinct double-peak emission, under both TPE (Figure 3A) and OPE (Figure 3B) models. The blue-edge emission is stronger than the red-edge one under OPE, whereas the red-edge emission dominates the spectrum under the TPE model. The phenomenon might be contributed to the higher instantaneous photon flux density under two-photon excitation, which causes a higher degree of excited-state reactions (such as the formation of excited-state dimer). More interestingly, as is shown in Figure 3A, the TPE fluorescence of the TP-SA complex exhibits a new emission peak around 580 nm, which, however, is not observed in the OPE fluorescence of the same complex. In addition to the abovementioned photon flux density of TPE, the effect of protein SA on the photophysical property of the label would be another reason, since the emission appears only in the protein-dye complex. However, when the label was conjugated to another protein BSA (bovine serum albumin), the 580 nm emission did not appear under TPE (Supporting Information Figure S3). Therefore, it is rather difficult for us to give an unambiguous explanation on this emission at this stage. But anyway, such a complexity in the spectral character did not affect the subsequent hybridization assay, which was performed applying the alteration of fluorescence intensity of the donor around 493 nm (vide infra). From another point of view, however, this finding suggests that

we still need more fundamental understandings to TPE fluorescence for the sake of further developing TPE-based methods and techniques. Construction of the Competitive Hybridization Assay Model. To investigate the applicability of TPE-FRET in nucleic acid hybridization assay, a homogeneous competitive assay model using the biotin-streptavidin bridge was employed for the determination of target oligonucleotide (Figure 4). As one of the new class of high-efficiency dark quenchers, the BHQ-1 dye was selected as the energy acceptor since its absorption (maximal at 534 nm) overlaps well with the fluorescence emission of TPCOOH. A biotinylated ssDNA (5′-biotin-GAG TTA GCA CCC GCA TAG TCA AGA T-3′) was designed as the capture ssDNA, which could hybridize with the quencher-linked ssDNA (5′-TGA CTA TGC GGG TGC TAA-BHQ-1-3′) and/or the target ssDNA (5′ATC TTG ACT ATG CGG GTG CTA ACT C-3′). The sequence length was set as 25-mer and 18-mer for biotinylated oligonucleotide and BHQ-1-modified oligonucleotide, respectively, which forms an incomplete double strand after hybridization. Upon the mixture of biotin-dsDNA-BHQ-1 hybrids with TP-SA solution, the energy donor (TP-COOH) and acceptor (BHQ-1) would be in close proximity due to the specific capture of biotin by streptavidin. In this situation, FRET is supposed to occur resulting in a quenching of the donor fluorescence. Notably, the three bases at the 5′-end (GAG) in the biotinylated oligonucleotide were designed to increase the distance between the donor and the acceptor, since at very short distance the FRET theory will no longer well hold.29 For the subsequent competitive assay, the target ssDNA is to be added into the hybrids, where the quencher ssDNA and the target ssDNA will be competing toward the capture ssDNA. Since the target ssDNA has a longer sequence containing more complementary bases than the quencher ssDNA, it can form a stronger hydrogen bond with the biotinylated oligonucleotide. Thereupon, the BHQ-1-ssDNA in the hybrid would be replaced partially by the target DNA, resulting in a recovery of the donor fluorescence. With fixed concentrations of TP-SA and the quencher, the fluorescence intensity of the donor is expected to increase in a target concentration-dependent manner, which enables the quantitative assay. To be pointed out, the employment of the biotin-streptavidin bridge is also for the purpose of amplifying the FRET signal, since each streptavidin molecule contains a maximum of four biotin binding sites, which provides the donor with access to multiple acceptors. Examination on the Validity of the TPE-FRET Model. The two-photon excited FRET was confirmed with the fluorescence titration of the donor. At fixed concentration of TP-SA (0.32 µM, representing the concentration of protein SA), the TPE fluorescence intensity of the donor decreased gradually according to the increase of the quencher concentration (Figure 5A). Control experiments were performed to exclude any nonspecific interactions or collisional quenching of the fluorescence. In these experiments, the donor TP-SA was incubated with BHQ-1-labeled oligonucleotide or biotinylated ssDNA directly, and no obvious signal change was observed under such situations (data not shown). The results indicate that the quenching of the donor fluorescence is mediated by the specific binding between SA and biotin, which enables the nonradiative energy transfer between TP-COOH and BHQ-1. As the molar ratio of BHQ-1-to-SA rose Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

1385

Figure 3. Fluorescence properties of pure TP-COOH (solid line) and the TP-SA complex (dotted line). (A) Fluorescence emission spectra of pure TP-COOH (100 µM in DMF) and the TP-SA complex (0.32 µM SA) under two-photon excitation at 800 nm. (B) Fluorescence emission spectra of pure TP-COOH (1 µM in DMF) and TP-SA complex (0.16 µM SA) under one-photon excitation at 420 nm. Spectra of TP-SA complex were recorded in 10 mM phosphate buffer containing 0.3 M NaCl and 0.1% TX-100, pH 7.4. The fluorescence intensity was given in normalized form.

Figure 4. Principle of the TPE-FRET-based homogeneous competitive hybridization assay for target DNA via a biotin-streptavidin bridge, with TP-COOH as donor and BHQ-1 as acceptor.

from 1:16 to 3.75:1, i.e., the concentration of BHQ-1 varied from 0.02 to 1.2 µM, the quenching efficiency increased from 13% to 66%. But further increase of the quencher concentration did not cause further fluorescence quenching, which is probably due to the saturation of streptavidin with biotin (Figure 5B). As can be understood, to achieve the optimal sensitivity of the subsequent competitive hybridization assay which is based on the recovery of the fluorescence, a quenching efficiency as large as possible is expected from this quenching step. Meanwhile, the consumption of excessive quencher ssDNA should be avoided, since it is unfavorable for the detection sensitivity due to their competition with target ssDNA. Therefore, the optimal molar ratio of TP-SA to biotin-dsDNA-BHQ-1 hybrids was decided as 1: 3.75. As a comparison, the same fluorescence titration experiments were performed under the OPE model. The titration results of Figure 6A indicate the occurrence of OPE-FRET in the system. 1386

Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

But the saturation curve (Figure 6B) clearly shows the difference between the OPE and TPE models. Under the TPE model, the fluorescence quenching efficiency reached 66% at 3.75-fold excess of quencher. However, in the case of the OPE model, the efficiency was only 50% even at 9.2-fold excess of quencher. In our previous work,31 we investigated the background signal of avidin under one-photon and two-photon excitation and revealed the advantage of TPE in eliminating the background of biomolecules. The ability of TPE-FRET to acquire much higher signalto-noise ratio in biological samples was also verified in other papers.30 Therefore, in this OPE-FRET model, the lower quenching efficiency can also be attributed to the higher level of background signal arising from biomolecules. TPE-FRET-Based Homogeneous Competitive Hybridization Assay for Target ssDNA. At the optimized molar ratio of reagents, the determination of target ssDNA based on competitive

Figure 5. (A) Fluorescence titration of TP-SA (0.32 µM SA) with biotin-dsDNA-BHQ-1 hybrids under TPE. (B) The dependence of fluorescence quenching efficiency of the donor on the molar ratio of biotin-dsDNA-BHQ-1 hybrids and TP-SA. The fluorescence was excited at 800 nm and measured at 493 nm in assay buffer containing 10 mM phosphate buffer, 0.3 M NaCl, and 0.1% TX-100, pH 7.4.

Figure 6. (A) Fluorescence titration of TP-SA (0.16 µM SA) with biotin-dsDNA-BHQ-1 hybrids under OPE. (B) The dependence of fluorescence quenching efficiency of the donor on the molar ratio of biotin-dsDNA-BHQ-1 hybrids and TP-SA. The fluorescence was excited at 420 nm and measured at 488 nm in assay buffer containing 10 mM phosphate buffer, 0.3 M NaCl, and 0.1% TX-100, pH 7.4.

hybridization was conducted. Biotin-ssDNA (1.2 µM) was hybridized with equivalent BHQ-1-ssDNA; thereafter, varying amounts of target ssDNA were individually added to the as-formed hybrids. After sufficient time of competition reaction, 0.32 µM TP-SA was introduced to each reaction mixture initializing the biotin-SA recognition. As is shown in Figure 7A, the fluorescence intensity of the donor was gradually restored according to the increase of target ssDNA concentration. The fluorescence recovery value, ∆F (∆F ) FL - FL0, where FL0 represents the fluorescence intensity in the absence of the target and FL is the fluorescence intensity in the presence of the target), at each target concentration was calculated, respectively. The dependence of the ∆F value on the concentration of the target is illustrated in Figure 7B with the target concentration ranging from 0.08 to 1.52 µM. A good linear relationship between ∆F and the target concentration is obtained with a correlation coefficient of 0.9910, based on which a quantitative hybridization assay can be performed. The calibration curve obtained was applied to real biological systems to further illustrate the usefulness of TPE-FRET in a

complicated matrix. The recoveries of target ss-DNA in spiked serum and urine samples were detected (Table 1). It is seen that quite satisfying recoveries can be obtained within the linear range. A set of parallel detection was also performed under the OPE model with identical procedures. The results demonstrate that the analytical performance of OPE-FRET is comparable to that of TPE-FRET in buffer solution (Supporting Information Figure S4). However, the OPE-FRET model does not present rational results in serum or urine sample (Supporting Information Figure S5), where the recoveries are 150.0% (in urine) and 40.6% (in serum) calculated with the OPE-FRET calibration curve. Such results, as has been proven in our previous work, are obviously due to the interference from background signal in these complicated matrixes. CONCLUSIONS A new TPE molecule with a centrosymmetric D-π-D′-π-D scaffold is prepared, providing a favorable two-photon action cross section. Using the TPE molecule as the energy donor, a TPEFRET-based homogeneous competitive hybridization method is Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

1387

Figure 7. TPE-FRET-based homogeneous competitive hybridization assay for target ssDNA. (A) The two-photon excited fluorescence spectra of TP-SA under increasing amounts of target DNA in the presence of biotin-dsDNA-BHQ-1 hybrids. (B) The linear relationship between the increase of donor fluorescence and the target concentration within the range of 0.08-1.52 µM. The concentrations of TP-SA and biotindsDNA-BHQ-1 hybrids were fixed at 0.32 and 1.2 µM, respectively. The fluorescence was excited at 800 nm and detected at 493 nm in assay buffer containing 10 mM phosphate buffer, 0.3 M NaCl, and 0.1% TX-100, pH 7.4. Data were presented as average ( SD from three repeated measurements.

Table 1. Recoveries of ssDNA in Urine and Serum Samples under the TPE-FRET Model samples human urine rabbit serum

added (µM) found (µM) recovery (%) RSD% (n ) 3) 0.5 1.2 0.5 1.2

0.504 1.19 0.483 1.10

100.8 98.8 96.7 91.9

3.8 4.2 4.5 2.9

developed, which enables separation-free determination of ssDNA. The biotin-streptavidin bridging technique adopted in the hybridization model provides an effective approach for signal amplification and versatility due to the broad range of biotinylated reagents. Under optimized experimental conditions, a calibration curve for target ssDNA is obtained. The method is applied to spiked serum and urine samples, which yield satisfying recoveries. The conventional OPE-FRET method can afford comparable determination performance with TPE-FRET in buffer solution, but it does not work in the biological systems, which confirms the superiority of TPE-FRET in hybridization assay over the OPEFRET model when applied to complicated matrix. The results of this work suggest that the TPE-FRET technique can be readily developed for various analytical objectives, partly because of the

1388

Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

high structural flexibility of TPE molecules. The study also presents some interesting findings concerning the photophysical feature of the biomolecule conjugate under two-photon excitation, which deserves deeper understanding and needs further studies. ACKNOWLEDGMENT The authors thank the National Natural Science Foundation of China (Nos. 20675059 and 90717111), the Science Fund for Creative Research Groups (Nos. 20621502 and 20921062), and The National Key Scientific ProgramsNanoscience and Nanotechnology (No. 2006CB933103) for financial support. SUPPORTING INFORMATION AVAILABLE Details for synthesizing TP-COOH, determination of TPA cross section and fluorescence quantum yield of TP-COOH, the calibration for TP-COOH quantification, two-photon excited fluorescence spectrum of TP-COOH-BSA complex, and the results of competitive assay under OPE-FRET. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review December 21, 2009. AC902467W

October

29,

2009.

Accepted