Cation Exchange in ZnSe Nanocrystals for Signal Amplification in

Anal. Chem. 2011, 83, 402–408

Cation Exchange in ZnSe Nanocrystals for Signal Amplification in Bioassays Jingjing Yao, Samantha Schachermeyer, Yadong Yin, and Wenwan Zhong* Department of Chemistry, University of California, Riverside, California 92521-0403, United States ZnSe nanocrystals (NCs), possessing low native luminescence but high biocompatibility, were employed as labeling tags in bioassays. They were able to amplify each target recognition event thousands of times through a cationexchange reaction (CXAmp) that released over 3000 encapsulated Zn2+ from one single NC. The freed cations in turn triggered strong fluorescence from the Zn-responsive dyes. The present study demonstrated that CXAmp with ZnSe delivered superior detection performance in comparison to the conventional labeling methods. The overall fluorescence intensity of CXAmp using 5 nM ZnSe NCs was 30 times higher than that from 5 nM core-shell CdSe/ZnS quantum dots (QDs). The limit of detection (LOD) obtained with ZnSe-based CXAmp was 10-fold lower than with horseradish peroxidase (HRP) labeling, and the detection sensitivity, represented by the slope of the signalversus-concentration curve, was 20-fold higher. When applied to detect immunoglobulin E (IgE) in a sandwich format, a LOD of 1 ng/mL was achieved. The highly sensitive CXAmp also allowed detection of the total IgE content in dilute human serum, in which the abundant matrix proteins exhibited less interference and more accurate quantification could be performed. Besides high signal amplification efficiency and good biocompatibility, CXAmp with ZnSe could be easily adapted to common laboratory settings and act as a universal labeling system for reliable detection of low-abundance targets. Molecular measurements are very important to many fields, such as systems biology, health care, product quality control, environmental monitoring, biodefense, etc.1-5 However, ultrasensitive detection remains a challenge mainly due to the low tagging efficiency of reporter molecules. For example, fluorescent dyes with high quantum yields are the most common optical labels,6-8 but each reporter molecule can only be coupled with up to five dye molecules. The low labeling ratio yields weak fluorescence indiscernible from the background * To whom correspondence should be addressed. E-mail: wenwan.zhong@ ucr.edu. Fax: +1-951-827-4713. (1) Gauglitz, G. Anal. Bioanal.Chem. 2005, 381, 141–155. (2) Eubanks, L. M.; Dickerson, T. J.; Janda, K. D. Chem. Soc. Rev. 2007, 36, 458–470. (3) Tecon, R.; van der Meer, J. R. Sensors 2008, 8, 4062–4080. (4) Haab, B. B. Curr. Opin. Biotechnol. 2006, 17, 415–421. (5) Zhou, H.; Roy, S.; Schulman, H.; Natan, M. J. Trends Biotechnol. 2001, 19, S34–39.

402

Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

noise when the trace targets only capture very few reporter molecules on the array surface.9,10 Quantum dots (QDs) are 5-20 times brighter than organic dyes9,11 but have problems like high production cost, luminescent quenching after surface functionalization and bioconjugation,9,12,13 and environmental hazards imposed by the toxic element Cd.14,15 QDs made of nontoxic materials have not yet shown comparable sensing performance to those containing Cd.9,14 On the other hand, silver nanoparticles could be employed to enhance the fluorescence from dyes via surface plasmon resonance. Three to ten fold improvements have been observed.16-18 Alternatively, several techniques have been developed to amplify the detection events. The most common approach is to utilize enzymes like horseradish peroxidase (HRP) for labeling. Each enzyme can catalyze the production of a large number of colorimetric or chemiluminescent molecules,19-24 lowering the detection limits to the range of 10-100 pM. Or, nucleic acids can be amplified by polymerases before detection, but polymerase reaction is not directly applicable to non-nucleic acid molecules and suffers from severe background contamination and tedious (6) Demchenko, A. P. Anal. Biochem. 2005, 343, 1–22. (7) Resch-Genger, U.; Grabolle, M.; Cavaliere-Jaricot, S.; Nitschke, R.; Nann, T. Nat. Methods 2008, 5, 763–775. (8) Schaferling, M.; Nagl, S. Anal. Bioanal. Chem. 2006, 385, 500–517. (9) Resch-Genger, U.; Grabolle, M.; Cavaliere-Jaricot, S.; Nitschke, R.; Nann, T. Nat. Methods 2008, 5, 763–775. (10) Schaferling, M.; Nagl, S. Anal. Bioanal. Chem. 2006, 385, 500–517. (11) Ruan, G.; Agrawal, A.; Smith, A. M.; Gao, X.; Nie, S. Rev. Fluoresc. 2006, 3, 181–193. (12) Noh, M.; Kim, T.; Lee, H.; Kim, C. K.; Joo, S. W.; Lee, K. Colloids Surf., A 2010, 359, 39–44. (13) Smith, A. M.; Duan, H. W.; Mohs, A. M.; Nie, S. M. Adv. Drug Delivery Rev. 2008, 60, 1226–1240. (14) Rene-Boisneuf, L.; Scaiano, J. C. Chem. Mater. 2008, 20, 6638–6642. (15) Karabanovas, V.; Zakarevicius, E.; Sukackaite, A.; Streckyte, G.; Rotomskis, R. Photochem. Photobiol. Sci. 2008, 7, 725–729. (16) Aslan, K.; Gryczynski, I.; Malicka, J.; Matveeva, E.; Lakowicz, J. R.; Geddes, C. D. Curr. Opin. Biotechnol. 2005, 16, 55–62. (17) Deng, W.; Drozdowicz-Tomsia, K.; Jin, D. Y.; Goldys, E. M. Anal. Chem. 2009, 81, 7248–7255. (18) Deng, W.; Jin, D. Y.; Drozdowicz-Tomsia, K.; Yuan, J. L.; Goldys, E. M. Langmuir 2010, 26, 10036–10043. (19) Khatkhatay, M. I.; Desai, M. J. Immunoassay 1999, 20, 151–183. (20) Avseenko, N. V.; Morozova, T. Y.; Ataullakhanov, F. I.; Morozov, V. N. Anal. Chem. 2002, 74, 927–933. (21) Joos, T. O.; Schrenk, M.; Hopfl, P.; Kroger, K.; Chowdhury, U.; Stoll, D.; Schorner, D.; Durr, M.; Herick, K.; Rupp, S.; Sohn, K.; Hammerle, H. Electrophoresis 2000, 21, 2641–2650. (22) Servoss, S. L.; Gonzalez, R.; Varnum, S.; Zangar, R. C. Methods Mol. Biol. 2009, 520, 143–150. (23) Wiese, R.; Belosludtsev, Y.; Powdrill, T.; Thompson, P.; Hogan, M. Clin. Chem. 2001, 47, 1451–1457. (24) Woodbury, R. L.; Varnum, S. M.; Zangar, R. C. J. Proteome Res. 2002, 1, 233–237. 10.1021/ac102688s  2011 American Chemical Society Published on Web 11/30/2010

liquid handling.25-28 Besides biological enzymes, dye-doped silica nanoparticles have been produced to increase the labeling ratio by encapsulating a significant amount of dyes inside each particle.29,30 However, efforts are needed to prevent dye leakage and nonspecific binding to the silica surface, and the bulky particle size (30-60 nm) may induce steric hindrance in target binding or even block binding sites.29 Our group has developed a convenient way to use encapsulated molecules for signal amplification in in vitro assays. Our method takes advantage of the natural enclosure of thousands of ions inside each ionic nanocrystal (NC) and its high surface activity.31 The enclosed cations can be quickly released through a cationexchange (CX) reaction, and they in turn ignite thousands of metal-responsive dyes. The result is equivalent to obtaining a labeling ratio larger than 1000:1. Each target-reporter binding event can then be amplified over a thousand times. This method of signal amplification based on cation exchange in NCs (CXAmp) was first demonstrated in CdSe NCs and led to a signal 60-fold higher and a detection limit 100-fold lower than that produced by Alexa Fluor 488 labeling.31 We also successfully detected microRNA from human total RNA extract using CXAmp with CdSe, obtaining a detection limit of 35 fM.32 The gentle CX reaction involves no strong acids or corrosive oxidants, making it possible to use the metal-responsive dyes in signal production. It is also very fast, allowing in situ detection. The fluorescence reaches 80% maximum within 20 s upon mixing the NCs with the CX buffer (data not shown). Additionally, CXAmp is directly adaptable to the common bioassay formats and optical detection platforms, imposing no special requirement for instruments and laboratory settings. The most attractive feature of CXAmp in comparison with other nanostructure-based sensing techniques is its independence of the intrinsic optical properties of NCs. Thus, much greater freedom exists in the selection and modification of NCs for the pursuit of sensitive detection performance, high robustness, and excellent biocompatibility with the availability of diverse metalresponsive fluorogenic dyes.33 Since the development and application of the CdSe-based CXAmp, we have achieved significant technological advancement in the present study: using ZnSe NCs for CXAmp (Scheme 1). The toxicity concern was dramatically reduced by using the more biocompatible ZnSe NCs; the detection performance of the ZnSebased CXAmp was much superior to that of the fluorescent dyes, QDs, and HRP, and it was applied to quantify immunoglobulin E (IgE) in human serum. Moreover, a detailed study on the CXAmp (25) Fischer, N.; Tarasow, T. M.; Tok, J. B. H. Curr. Opin. Chem. Biol. 2007, 11, 316–328. (26) Gharabaghi, F.; Tellier, R.; Cheung, R.; Collins, C.; Broukhanski, G.; Drews, S. J.; Richardson, S. E. J. Clin. Virol. 2008, 42, 190–193. (27) Mahony, J. B. Clin. Microbiol. Rev. 2008, 21, 716–747. (28) Mehlmann, M.; Bonner, A. B.; Williams, J. V.; Dankbar, D. M.; Moore, C. L.; Kuchta, R. D.; Podsiad, A. B.; Tamerius, J. D.; Dawson, E. D.; Rowlen, K. L. J. Clin. Microbiol. 2007, 45, 1234–1237. (29) Wang, L.; Wang, K. M.; Santra, S.; Zhao, X. J.; Hilliard, L. R.; Smith, J. E.; Wu, J. R.; Tan, W. H. Anal. Chem. 2006, 78, 646–654. (30) Wu, J.; Ye, Z. Q.; Wang, G. L.; Jin, D. Y.; Yuan, J. L.; Guan, Y. F.; Piper, J. J. Mater. Chem. 2009, 19, 1258–1264. (31) Li, J. S.; Zhang, T. R.; Ge, H. P.; Yin, Y. D.; Zhong, W. W. Angew. Chem., Int. Ed. 2009, 48, 1588–1591. (32) Li, J. S.; Schachermeyer, S.; Wang, Y.; Yin, Y. D.; Zhong, W. W. Anal. Chem. 2009, 81, 9723–9729. (33) Domaille, D. W.; Que, E. L.; Chang, C. J. Nat. Chem. Biol. 2008, 4, 507– 507.

Scheme 1. Schematic Illustration of CXAmp with ZnSe NCsa

a

The secondary antibody was labeled with ZnSe NCs, which, after target binding, would undergo CX reaction with the added Ag+ and released Zn2+. The freed Zn2+ would then bind to the poorly luminescent FluoZin-3, and the Zn-FluoZin-3 complex emits strong fluorescence.

conditions and stability of NCs during assay and storage was conducted to confirm the robustness of our method. EXPERIMENTAL SECTION Chemicals and Materials. Chemicals used to prepare nanocrystals, solutions, and conjugations were purchased from Fisher Scientific. The amine-labeled oligonucleotide was obtained from Integrated DNA Technologies (Coralville, IA) without further purification. Human immunoglobulin G (IgG), the primary and secondary antibodies with or without HRP modification, QDot525 nm (QD), and FluoZin-3 were purchased from Invitrogen (Carlsbad, CA). Human IgE was obtained from Athens Research and Technology (Athens, GA). Lonza Biowhittaker* human serum was supplied by Fisher Scientific. Synthesis and Characterization of ZnSe NCs. The watersoluble ZnSe NCs were synthesized in aqueous solution, adapting the synthesis route previously reported for CdSe NCs.34 The detailed procedure as well as NCs characterization can be found in the Supporting Information. Bioconjugation of NCs with Antibodies or Amine-Labeled Oligonucleotides. Goat antihuman immunoglobulin G (anti-IgG) or anti-IgE was coupled to ZnSe NCs by 1-ethyl-3-(3-(dimethylamino)propyl)carbodiimide hydrochloride (EDC) and sulfo-Nhydroxysuccinimide (sulfo-NHS). The mercaptoacetic acid (MAA) molecules on the ZnSe surface provided carboxyl groups for conjugation. ZnSe powder containing approximately 1014 NCs was first suspended in 5 µL of DI water and activated by 0.4 mg of EDC and 1.1 mg of sulfo-NHS in 1 mL of linking buffer (0.1 M NaH2PO4-Na2HPO4 (pH 7.2)) for 15 min. The residual EDC was quenched by 1.2 µL of 2-mercaptoethanol before the addition of 100 µL of anti-IgG (1 mg/mL) or anti-IgE (1 mg/ mL) to prevent protein cross-linking. The mixture was gently vortexed for 3 h at room temperature (RT). Finally, the conjugated ZnSe NCs were purified with an Amicon YM-100 centrifugal filter and stored in 100 µL of 0.1 M phosphate buffer (pH 7.2). The stock was diluted a hundredfold prior to detection. Coupling of the 3′ amine-labeled 27 nt DNA (5′-AAG CTA AGACAGT15/3AmM/-3′) to ZnSe NCs was performed with the same procedure. (34) Tang, A. W.; Teng, F.; Xiong, S.; Wang, Y.; Feng, B.; Hou, Y. B. J. Electrochem. Soc. 2008, 155, K190–K194.

Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

403

Antibody conjugation onto the QDs used a one-step procedure as advised by the manufacturer. Approximately 1014 QDs were mixed with 0.2 mg of EDC and 100 µL of 1 mg/mL antibody in 1 mL of linking buffer. The mixture was vortexed at RT for 3 h, purified by the YM-100 filter, and stored in 100 µL of 0.1 M phosphate. For detection, the stock conjugated QDs were diluted 15 times. Antibody contents on the conjugated ZnSe NCs and QDs were estimated from the absorbance at 350 nm (specific for NCs) and 280 nm (containing signals from both NC and protein) obtained with the Cary 50 UV-vis spectrophotometer (Varian, CA). Both the ZnSe NCs and QDs conjugations gave out a final antibody to NC molar ratio close to 45: 1. Since the YM-100 filter could not completely remove the free antibody molecule, this number did not represent the actual number of IgG molecules per NC and was just used to estimate amounts of NCs and antibodies used in assays. Immunoassay and Fluorescence Measurement. Both the binary and sandwich assays were performed on 96-well microtiter plates. In the binary assay the wells were coated with human IgG by overnight incubation at 4 °C. The well was blocked by 1% bovine serum albumin (BSA) in 1× phosphate-buffered saline (PBS) for 30 min at RT, before being incubated with the 100 µL of anti-IgG labeled with HRP, ZnSe, fluorescein, or QDs for 1 h at 37 °C. HRP- and fluorescein-labeled anti-IgG was diluted 4000 or 40-fold, respectively, from the 1 mg/mL stock as suggested by the manufacturer. In the sandwich assay for IgE detection, the plate was coated with 100 µL of goat anti-IgE (10 µg/mL) overnight at 4 °C and blocked by 1% BSA. Then, different concentrations of IgE were added to the plate, and it was incubated at 37 °C for 1 h. After several washes with PBS, 100 µL of the ZnSe-anti-IgE (100-fold dilution from stock) in PBS/BSA (1%) was loaded into each well, and it was incubated at 37 °C for 1 h. Again, a washing step was applied to remove the residue-labeled antibodies. Absorbance signals from HRP were developed by adding 100 µL of 2 mM H2O2 and 2 mM 2,2′-azino-bis[3ethylbenziazoline-6-sulfonic acid (ABTS) to the well. The reaction lasted for 20 min at RT before the absorbance measurement was made in a plate reader. Signaling with CXAmp was done with the addition of 100 µL of 10 mM HEPES buffer containing 500 µM AgNO3 and 3 µM FluoZin-3. An excitation filter of 485 nm and an emission filter of 530/30 nm were used to detect fluorescence from FluoZin-3. The same filters were also employed to measure the signals from QDs and fluorescein with DI water as the detection solution, except that an excitation filter of 405 nm was used with the QDs. Fluorescence spectra from FluoZin-3 and CXAmp were taken on the Spex FluoroLog Tau-3 fluorescence spectrophotometer (Horiba Jobin Yvon Inc., NJ), using an excitation wavelength of 488 nm. Emission spectrum from 505 to 550 nm was collected with a slit width of 2 nm. QDs were excited at 350 nm, and the emission was recorded from 505 to 550 nm with a slit width of 2 nm. The intrinsic fluorescence emission from ZnSe NCs was excited at 350 nm and recorded from 385 to 450 nm with a slit width of 10 nm. All reported data were represented as the average and standard deviation from triplicate measurements. 404

Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

RESULTS AND DISCUSSION Maximizing the Fluorescent Signal Generated by Zn2+. Zn imposes low health risks with a tolerable upper intake level above 10 mg/day and is not regulated by the Environmental Protection Agency (EPA).35,36 Thus, Zn-based NCs are the material of choice to improve the biocompatibility of CXAmp. Since CXAmp depends on the fluorophores fluorescing upon binding to the released Zn2+, the lowest possible detection limit and the linear detection range of CXAmp are determined by the performance of the Zn2+-responsive dye. Among a large number of fluorescent Zn2+ sensors developed in the past decade,37-41 FluoZin-3 was chosen in our study. Its quantum yield increases from