Spectrally Matched Duplexed Nucleic Acid Bioassay Using Two

Oct 8, 2014 - A paper-based multiplexed resonance energy transfer nucleic acid hybridization assay using a single form of upconversion nanoparticle as...
0 downloads 0 Views 8MB Size
Download PDF
Article pubs.acs.org/ac

Spectrally Matched Duplexed Nucleic Acid Bioassay Using TwoColors from a Single Form of Upconversion Nanoparticle Feng Zhou and Ulrich J. Krull* Chemical Sensors Group, Department of Chemical and Physical Sciences, University of Toronto Mississauga, 3359 Mississauga Road, Mississauga Ontario L5L 1C6, Canada S Supporting Information *

ABSTRACT: Optical sensing can provide opportunity for simultaneous determination of multiple targets as well as for implementation of ratiometric methods that can improve accuracy and precision. Herein we report a paper-based two-color oligonucleotide detection assay with tunable sensitivity that is based on use of a single type of upconversion nanoparticle (UCNP). Water-soluble UCNPs were designed to concurrently offer green and red emission. These avidin functionalized UCNPs were adsorbed onto a cellulose support, and Cy3 was used as a green channel acceptor for Survival Motor Neuron (SMN1) target, and Cy5.5 was the red channel acceptor for the glucuronidase gene (uidA) target. Selective DNA hybridization of the labeled targets with the corresponding probe provided emission from dyes, which was the basis for concurrent quantification of both targets. The limit of detection (LOD) could be tuned by changing the relative ratio of the SMN1 and uidA probes. A higher proportion of a probe provided for a lower LOD. When the SMN1/uidA probe ratio was 1:4, the LOD for SMN1 and uidA target were 54.3 and 30.5 fmol, and when the probe ratio was 4:1, the LOD for the above targets were 22.1 and 1260 fmol, respectively. Selectivity evaluation showed that one base pair mismatched DNA for SMN1 and uidA could be discriminated in most cases. The assay showed resistance to nonspecific adsorption of interfering DNA and protein and was even functional for targets in undiluted serum. This work represents a significant step in the development of paper-based multiplexed UCNP luminescence assays.

U

respectively). These are nonradiative energy transfer processes from donor to acceptor that have found widespread use in optical assays and sensor designs.6,7 In some cases, UCNPs can participate in LRET.8,9 In a typical transduction scheme, the acceptor and donor achieve physical proximity via the specific target recognition and binding process. This approach to development of bioassays typically offers higher specificity and sensitivity when compared with assays that use labels only to mark targets.10,11 The first LRET-based detection system that made use of UCNPs was introduced in 2005.12 The method was demonstrated by means of a model assay involving avidin detection, using biotinylated UCNPs and gold nanoparticles as energy donors and quenchers, respectively. Examples of other

pconversion nanoparticles (UCNPs) are a class of luminescent nanomaterial with many intriguing features suitable for bioanalytical applications such as chemical and photostability, long lifetime, and multiple tunable emission bands.1,2 UCNPs can be excited by low energy near-infrared (NIR) radiation, offering opportunity for minimizing background interference from light scatter and autofluorescence when designing assays and sensors.3 Unlike quantum dots (QDs) and other organic dyes, UCNPs can have multiple narrow emission bands in the range from ultraviolet to NIR wavelengths. The relative intensity and wavelength of these bands can be tuned by adjusting the chemical composition and physical structure of UCNPs.4,5 The emission from UCNPs is formally described as luminescence, and the emission can be used to select an emission-absorption pair for the purpose of assay development. Perhaps the best known pairs are associated with fluorescence and luminescence resonance energy transfer (FRET and LRET, © XXXX American Chemical Society

Received: August 26, 2014 Accepted: October 8, 2014

A

dx.doi.org/10.1021/ac503207q | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Scheme 1. Duplexed Luminescence DNA Assay on Paper Using UCNPs As Donors

water, and then the nanoparticles were coated with sodium citrate to provide carboxylic groups on the UCNPs.20,21 Avidin was subsequently conjugated onto the UCNPs by the wellknown carbodiimide-succinamide (EDC/NHS) reaction. Solubilized UCNP-avidin nanoparticles were manually spotted to prepare arrays on cellulose paper. Biotinylated probes for SMN1 and uidA were then immobilized to the UCNPs by the biotin−avidin reaction. After hybridization with mixtures of Cy3-labeled SMN1 and Cy5.5-labeled uidA targets, Cy3 and Cy5.5 emission induced using a 980 nm laser was collected with an epifluorescence microscope. Quantitative duplexed analysis was demonstrated using a single type of UCNP in a relatively complex sample matrix, and analytical figures of merit were shown to be controllable by adjusting the ratios of the two different probes.

assays that have been reported include an LRET-based aptasensor for mycotoxin detection and a glucose sensor, with both of these assays using UCNPs as donors and graphene oxide as quenchers.13,14 A sandwich detection format for oligonucleotides has been reported using DNA probes conjugated to UCNPs as LRET donors and fluorescent dye as a label on DNA target.15 Emission from UCNPs has also been used to directly excite absorptive dyes by radiative energy transfer. This has been described as an inner filter effect and is also effective in development of bioassays. A recent example involves the use of two emission bands from UCNPs that could concurrently excite the enzyme cosubstrate NADH and the coenzyme FAD.16 Regardless of the mechanism of energy transfer from UCNPs to molecular dyes, it is possible to develop sensitive bioassays that can be excited by NIR and emit in the visible region of the spectrum. Paper-based analytical devices (PADs) have sparked great interest based on practical advantages of availability, low-cost, disposability, and facile surface modification.17,18 We have recently described a paper-based solid-phase method for DNA detection where UCNPs excited fluorescent dye that was attached to target DNA.3 Superior limit of detection (LOD) and dynamic range were achieved in comparison to an analogous assay using QDs as energy donors in FRET assays on paper substrates.19 This earlier work has primarily focused on transduction using only a single emission band of the UCNPs.3 The work reported herein examines the potential to concurrently use two emission bands from the same UCNPs to achieve quantitative duplexed analysis of oligonucleotide targets. The assay system has been designed using pairing of emission bands at 542 and 653 nm of green UCNPs with Cy3 and red Cy5.5 labeled DNA targets, respectively (Scheme 1). The 19 mer Survival Motor Neuron (SMN1) probe-target sequence (Cy3 labeled) has clinical relevance in the diagnostics of spinal muscular atrophy, and the 24 mer uidA probe-target sequence (Cy5.5 labeled) is commonly used as an indicator for detection of bacteria. These disparate probes have been selected to illustrate the potential for accessing diverse targets. Synthesized oleate-coated UCNPs (OA-UCNPs) were converted into ligand-free nanoparticles by treatment with acidic

■

EXPERIMENTAL SECTION Detailed description of materials and reagents, instrumentation, and experimental procedures can be found in the Supporting Information. The oligonucelotide sequences used for hybridization assays are listed in Table 1. Immobilization of Avidin-UCNPs on Paper and DNA Probe Conjugation. The stable immobilization of streptavidin coated UCNPs on cellulose paper by physical adsorption Table 1. Oligonucleotide Sequences Used in the Hybridization Assaysa SMN1 probe biotin-5′-ATT TTG TCT GAA ACC CTG T-3′ SMN1 FCT Cy3−3′-TAA AAC AGA CTT TGG GAC A-5′ SMN1 1 BPMT Cy3−3′-TAA AAC ACA CTT TGG GAC A-5′ SMN1 NCT Cy3−3′-TGT CCC AAA GTC TGT TTT A-5′ uidA probe biotin-5′-CTT ACT TCC ATG ATT TCT TTA ACT-3′ uidA FCT Cy5.5-3′-GAA TGA AGG TAC TAA AGA AAT TGA-5′ uidA 1BPMT Cy5.5-3′-GAA TGA AGG TAC TAA ATA AAT TGA-5′ uidA NCT Cy5.5-3′-TTG TTA TAA CAG AAC TAA TCA GTA-5′ Poly T (T30)-3′-TTTTTTTTTTTTTTTTTTTTTTTTTTTTTT-5′ a

FCT = fully complementary target, 1BPMT = 1 base pair mismatched target, NCT = noncomplementary target. The mismatched base in the 1 BPMT sequence is bolded and underlined.

B

dx.doi.org/10.1021/ac503207q | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 1. TEM characterization of OA-UCNPs (a) and UCNPs-avidin (b).

μM for each) were spiked with undiluted, 1:1, 1:5, and 1:10 diluted goat serum. Four replicates were determined for each sample (n = 4) unless otherwise stated.

has been previously reported.3 Briefly, 1.5 μL of solution containing avidin-UCNPs (1 mg mL−1) was pipetted as arrays of spots, where each spot was localized by printing a hydrophobic wax ring on cellulose paper (inner diameter 3 mm). The process of spotting was followed by drying under vacuum. The paper was then soaked in borate buffer and rinsed for 10 min to eliminate any weakly adsorbed avidin-UCNPs. The treated paper was then wicked on absorbent paper and dried in vacuum. For conjugation of the biotinylated oligonucleotide probe molecules, 2 μL of SMN1/uidA biotinprobe mixtures at probe ratios 1:4, 1:1, and 4:1 (total concentration of both probes, 20 μM) in borate buffer was pipetted onto the arrayed spots that contained immobilized avidin-UCNPs. The paper array was allowed to stand at room temperature for 1 h before rinsing with borate buffer containing 100 mM NaCl to remove any unconjugated probes. DNA Hybridization Assay. We have previously reported the effect of NaCl concentration on the responses associated with DNA hybridization, and 100 mM of NaCl was found to be the optimal concentration for sensitivity and selectivity.3 All hybridization assays were done using oligonucleotide target solutions prepared in borate buffers (50 mM, pH 9.25, 100 mM NaCl) unless otherwise specified. Blocking solution containing 1% BSA and 0.05% Tween 20 was pipetted onto the spots and allowed to stand for 30 min to ameliorate nonspecific binding on UCNPs prior to introduction of target DNA. Hybridization assays were done by pipetting a 1.5 μL aliquot of mixed SMN1 and uidA targets at increasing concentrations from 0 to 12.8 μM onto the paper arrays. After hybridization, the paper was rinsed in borate buffer and dried in vacuum. After drying, the paper was scanned using an epifluorescence microscope equipped with a NIR laser as the excitation source (CW 980 nm, 800 mW). For 1BPMT discrimination, different spots on the same test paper were exposed for 10 min to a mixture of FCTs of SMN1 and uidA at 4 μM for each and 1BPMT SMN1 and uidA targets in a mixture at the same concentration. After hybridization, the paper was rinsed and dried for signal collection. To enhance the stringency of hybridization, the paper was subsequently treated with 20% formamide solution for 40 min, and emission signals from the various spots were collected again. Assays associated with more complicated sample matrixes were examined to discern possible effects of nonspecific adsorption. The SMN1 and uidA FCT mixture (2.3 μM for each) in borate buffer was spiked with salmon sperm DNA (∼2.2 μM) or Poly T sequence (2.2 μM) or bovine serum albumin (BSA) (380 μM). For assays that more closely simulated clinical samples, SMN1 and uidA FCT mixtures (4

■

RESULTS AND DISCUSSION Synthesis and Characterization of NaYF4: 2% Er3+, 18% Yb3+/NaYF4 Core/Shell UCNPs. UCNPs with strong green emission were synthesized using a previously reported method.22 The resulting oleate-coated UCNPs (OA-UCNPs) were typically faceted with many having hexagonal tendency. The average diameter of nanoparticles in the ensemble was 26.1 ± 3.7 nm (Figure 1a). These nanoparticles dispersed well in hexane. To conjugate UCNPs with avidin, the hydrophobic OA-UCNPs were first modified to become water-soluble by lowering the pH to 4 using HCl as was previously reported, and subsequently they were coated with sodium citrate.21 Finally, the carboxyl groups of the citrate coated UCNPs were activated by 1-Ethyl-3-(3-(dimethylamino)propyl) carbodiimide and Nhydroxysulfo succinimide (EDC/Sulfo-NHS), and then conjugated with avidin. The TEM image (Figure 1b) of avidinUCNPs showed the surface modification did not significantly alter the morphology or size (24.5 ± 3.1 nm) of the nanoparticles. Successful conjugation of avidin was confirmed by fluorescence emission of Cy3 dye after reaction with Cy3labeled biotinylated SMN1 probe. The normalized emission spectrum of the avidin-UCNPs, the Cy3 and Cy5.5 dye-labeled targets, and the absorption spectra of Cy3 and Cy5.5 dyes are shown in Figure 2. The overlap of the green and red emissions bands of the UCNP with the Cy3 and Cy5.5 absorption envelopes indicate opportunity for energy transfer (Figure 2). Performance of the Hybridization Assay. A ratiometric approach based on the emission intensities of both UCNP (donor) and dye-labeled DNA (acceptor) was used to determine the amount of hybridized DNA. Ratiometric methods tend to improve accuracy and sensitivity of assays, and this approach has been effective for DNA hybridization assays using both QD-FRET, UCNP-LRET, and UCNPluminescence on paper substrates and solution.7,23 Previous work using paper-based DNA hybridization demonstrated fast hybridization speed, which enabled the development of signal to be complete within 2 min at room temperature and at 40 °C. The fast hybridization kinetics was ascribed to the capillary wicking effect of the fibrous paper.3 Analogous nucleic acid hybridization assays conducted in solution phase have been reported to take 30 min or longer to reach signal equilibration.24 To ensure complete reaction, a hybridization time of 10 min was used in these duplexed paperC

dx.doi.org/10.1021/ac503207q | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

various SMN1/uidA probe ratios of 1:1, 1:4, and 4:1 (Figure 4). The corresponding limit of linearity (LOL), LOD, and calibration response equation for the dynamic range are summarized in Table 2. At all ratios of SMN1/uidA probes, the luminescence signal intensity from the red uidA detection channel was significantly lower than that from the green SMN1 detection channel. It is interesting that a similar phenomenon associated with signal intensity from two channels has been reported using a mixture of two types of QDs with different emissions.25 We speculate that this could be attributed to three causes: first, the probes of different lengths were in competition for the UCNP surface during surface immobilization (24 base of uidA and 19 base of SMN1), which might affect the relative probe ratio on an UCNP surface; second, the alignment (orientation/conformation) of shorter SMN1 and longer uidA probes on UCNP might affect availability of probes during hybridization; third, the combination of spectral overlap, the intensity of excitation radiation in each channel, and the properties of the dyes were substantially different. The ratio of spectral overlap between Cy3-green and Cy5.5-orange emission of UCNP was 1.9, and greater spectral overlap usually results in greater luminescence signal. Moreover, it has been reported that Cy5.5 dye self-quenches more significantly than Cy3 when at high densities even if the quantum yield of Cy5.5 (0.23) is higher than Cy3 (0.15).26 The ratio of SMN1/uidA probes was varied with the anticipation that a relatively constant total quantity of probe would be immobilized in all cases and that the dominant probe would offer a greater dynamic range for target detection before saturation of all hybridization sites occurred. The LOL for SMN1 detection ranged from 4.8 pmol to a higher range of at least 19.2 pmol as the SMN1/uidA probe ratio was varied, and the LOL for uidA detection ranged from 0.3 pmol to a higher range of at least 19.2 pmol. The quantitative difference in LOL for the 4:1 and 1:4 ratios is not proportional to the relative concentration change for these probes when considering the reaction solutions that were used to deposit the probes. A proportional relationship was not anticipated because the length of the probes was different, which would affect both surface density and hybridization efficiency. Further appreciation of the impact of probe ratio can be derived from response behavior at low concentrations of target. The LODs were determined based on luminescence signals that were three standard deviations higher than the background signals. For SMN1 detection, the LOD decreased from 54.3 fmol to 22.1 fmol as the SMN1/uidA probe ratio increased from 1:4 to 4:1. For uidA detection, the LOD of uidA detection decreased by a much greater magnitude from 1260.0 to 30.5 fmol. The data suggests that the SMN1 probe is in excess or hybridizes more efficiently than the uidA probe. The practical outcome of the exploration of probe ratios is that the LOL and LOD both can be tuned for either SMN1 or uidA detection by adjusting the ratio of two probes and with the higher proportion of a probe resulting in a lower LOD. Single Base Pair Mismatched Target (BPMT) Discrimination Assay. The ability to discriminate a 1BPMT from FCT is indicative of high selectivity. We previously reported that the combination of ionic strength and formamide concentration can be used to enhance the stringency of nucleic acid hybridization assay.3,27 This approach was reported to be effective in the discrimination of a 1BPMT at room temperature. Borate buffer containing 100 mM NaCl was

Figure 2. Normalized absorption (dashed lines) and emission spectra (solid lines) for the UCNP-Cy3 and UCNP-Cy5.5 pairs. Pink dashed line, Cy3 absorbance spectrum; blue line, Cy3 emission spectrum; green dashed line, Cy5.5 absorbance spectrum; red line, the emission spectrum of Cy5.5; dark line, UCNP emission spectrum.

based assays. The luminescence ratios were concurrently determined in the two color channels to quantify both of the DNA targets. A control experiment was completed to investigate whether the signal originated via the LRET mechanism. Solutions containing Cy3 labeled SMN1 noncomplementary target (NCT) and Cy5.5 labeled uidA NCT (4 μM for each) were spotted onto paper arrays that were preimmobilized with avidin-UCNP, and these samples were imaged using NIR excitation. In the absence of the probes, no emission from the labeled targets was observed (Figure 3), indicating that

Figure 3. Investigation of the emission mechanism of the dye. Pseudocolor epifluorescence images collected using the green filter channel (G, 520 ± 20 nm), Cy3 yellow filter channel (Y, 585 ± 20), orange filter channel (O, 655 ± 15 nm), and Cy5.5 red filter channel (R, 725 ± 25 nm) (n = 3).

significant emission from dyes was not caused by photon emission from the UCNPs. In the presence of immobilized probes, the hybridization between UCNP-probes and dyelabeled DNA targets would result in localization of the dye near the surface of UCNPs, thus making LRET-based assay possible (Figure 4). However, it remains possible that the energy exchange is due to the inner filter effect, and lifetime and energy transfer efficiency data would be required to confirm the mechanism. Luminescence responses to increasing concentrations of SMN1 and uidA targets in the mixture were determined using D

dx.doi.org/10.1021/ac503207q | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 4. Quantitative determination of Cy3 labeled SMN1 fully complementary target (FCT) and Cy5.5 labeled uidA FCT using Cy3 and Cy5.5 emission. (a) Pseudocolor epifluorescence images collected using the green filter channel (G, 520 ± 20 nm), the Cy3 yellow filter channel (Y, 585 ± 20 nm), and the composite of the green and yellow imaging channels (M1, for SMN1 sensing); orange filter channel (O, 655 ± 15 nm), Cy5.5 red filter channel (R, 725 ± 25 nm), and the merged channel of red and orange images (M2, for uidA sensing) corresponding to SMN1 and uidA FCT mixture of concentrations ranging from 0 to 12.8 μM. (b) The response curves for the various probe ratios as determined from luminescence intensities. The images were analyzed by ImageJ using the two imaging channels for each of the FCTs. Data are expressed as mean ± standard deviation (STDEV), n = 4.

Table 2. Limit of Linearity and Detection for SMN1 and uidA Detection at Different Probe Ratios limit of linearity (pmol)

limit of detection (fmol)

SMN1/uidA probe ratios

SMN1

uidA

SMN1

1:1 1:4 4:1

>19.2 >19.2 4.8

>19.2 >19.2 0.3

34.8 54.3 22.1

uidA

R2 SMN1

52.8 30.5 1260.0

0.990 0.984 0.973

again used in this work, and formamide was used to further control the stringency of hybridization. It can be seen from Figure 5 that in the absence of formamide treatment, significant luminescence signal differences between FCT and 1BPMT for both SMN1 (FCT/ 1BPMT = 1.67) and uidA (FCT/1BPMT = 1.42) could be observed when the SMN1/uidA probe ratio was 1:1. When the SMN1/uidA probe ratios were 1:4 and 4:1, the signal ratio of FCT/1BPMT for both SMN1 and uidA ranged from 0.972 to 1.20, and no significant differences could be observed. However, after treatment of the samples with 20% formamide for 40 min, pronounced signal differences at all SMN1/uidA probe ratios were noted between FCT and 1BPMT samples for both SMN1 (FCT/1BPMT ranging from 1.60 to 2.16) and uidA targets (FCT/1BPMT ranging from 1.67 to 1.75). The exception was for uidA determination at a probe ratio of SMN1/uidA of 4:1, where the ratio of FCT/1BPMT was 0.935.

calibration equation uidA 0.972 0.991 0.877

SMN1

uidA

Y = 0.100X + 0.175 Y = 0.0394X + 0.0791 Y = 0.175X + 0.330

Y = 0.0365X + 0.0532 Y = 0.0309X + 0.0562 Y = 0.0144X + 0.0304

Hence, we conclude that this duplexed detection system can discriminate 1BPMT when suitably designed for such a purpose. Not surprisingly, limitations are associated with a probe loading that is too dilute and for a luminescent system that does not have a strong emission signal. Hybridization Assay in More Challenging Sample Matrixes. Samples containing potential interfering agents were investigated to further evaluate the selectivity of this duplexed detection system (Figure 6). The concentration of one of the dye-labeled FCTs (SMN1 or uidA) was fixed at 1 μM. This was mixed with the other target (uidA or SMN1), which served as noncomplementary material at different concentrations (0.1, 1, and 10 μM). A further set of results were collected that substituted NCT for the FCT in each permutation of concentrations. For detection of SMN1 target, the luminescence ratio reduced as the concentration of the background uidA DNA increased. For uidA target, the signals were not E

dx.doi.org/10.1021/ac503207q | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 5. Evaluation of selectivity for the detection of FCT SMN1 and uidA, and the corresponding 1BPMT and NCT luminescence ratios were determined from the fluorescent images before and after treatment with 20% (v/v) formamide. The number at the top of each histogram indicates the signal ratio between FCT and 1BPMT. Data are expressed as mean ± standard deviation (STDEV), n = 4.

significantly affected by increasing concentrations of SMN1 target. For the NCT control samples, the signal remained low at all mixture ratios. While there is indication that some small amount of the dye-labeled NCT could adsorb and contribute to background signal, overall there was a good level of selectivity demonstrated by the results.

The ability of the paper-based assays to ameliorate nonspecific adsorption of longer sequences of DNA and protein in SMN1/uidA detection was investigated. A sample of 2.3 μM SMN1 and uidA FCT was spiked in borate buffer containing various background including 2.2 μM salmon sperm DNA (2000 bp fragment), 2.2 μM polyT (T30) solution, 380 F

dx.doi.org/10.1021/ac503207q | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 6. Evaluation of selectivity when different concentrations of complementary and noncomplementary oligonucleotide targets coexisted. Data are expressed as mean ± standard deviation (STDEV), n = 4.

Figure 7. Hybridization of SMN1/uidA FCT mixtures on paper substrates in the presence of different potential interfering agents. Data are expressed as mean ± standard deviation (STDEV), n = 4.

μM BSA solution, and serum at different dilutions. The data shown in Figure 7 indicate good tolerance of the assay to matrixes containing longer DNA sequences and protein, showing no statistically significant deviations. Such tolerance has been previously noted for paper-based assays and has been in part ascribed to the surface chemistry associated with paper. For detection using samples of clinical origin, the background matrix can be far more complex than represented by solutions that contain a few spiked interferences. More representative is the target in samples of serum. SMN1 and uidA FCT (each 4 μM) were investigated in serum of different dilution ratios (no dilution, 1:1, 1:5, and 1:10 dilution). On the basis of the luminescence signals for detection of FCT by both probes, no serum dilution was required and the selectivity remained almost constant even in serum that had not been diluted.

and red emission bands of one type of UCNP, respectively. The adjustment of the relative concentration ratios of the SMN1 and uidA probes allowed for tuning of the dynamic range and the detection limit. The detection limits of the assays using UCNPs were superior to analogous QD-FRET based DNA hybridization assays on paper that made use of two different colors of QDs for two types of target DNA detection.25 Examination of selectivity demonstrated that single base mismatches within SMN1 and uidA targets could be effectively discriminated and that the assays were operational in undiluted serum. This work presents significant progress in the development of paper-based multiplexed UCNP assays with high sensitivity and selectivity. Applications of portable multiplexed oligonucleotide detection technology has potential for impact in numerous areas, such as triage for specific bacterial of viral infections in people and animals, in-field testing of pathogens in water and foods, and rapid assays of body fluids for basic information about gender, race, and physical characteristics about a suspect at a crime scene. We believe that the paper-based assay demonstrated here will eventually find use for in-field oligonucleotide detection by

■

CONCLUSION Paper-based duplexed assays for oligonucleotide determination by hybridization have been demonstrated using UCNPs for excitation of dye-labels on oligonucleotides. Cy3-labeled SMN1 target and Cy5.5-labeled uidA target were paired with the green G

dx.doi.org/10.1021/ac503207q | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

(17) Martinez, A. W.; Phillips, S. T.; Whitesides, G. M.; Carrilho, E. Anal. Chem. 2010, 82, 3−10. (18) Ge, X. X.; Asiri, A. M.; Du, D.; Wen, W.; Wang, S. F.; Lin, Y. H. TrAC, Trends Anal. Chem. 2014, 58, 31−39. (19) Noor, M. O.; Shahmuradyan, A.; Krull, U. J. Anal. Chem. 2013, 85, 1860−1867. (20) Bogdan, N.; Vetrone, F.; Ozin, G. A.; Capobianco, J. A. Nano Lett. 2011, 11, 835−840. (21) Peng, J.; Sun, Y.; Zhao, L.; Wu, Y.; Feng, W.; Gao, Y.; Li, F. Biomaterials 2013, 34, 9535−9544. (22) Carling, C. J.; Nourmohammadian, F.; Boyer, J. C.; Branda, N. R. Angew. Chem., Int. Ed. 2010, 49, 3782−3785. (23) Zhang, J.; Mi, C.; Wu, H.; Huang, H.; Mao, C.; Xu, S. Anal. Chem. 2012, 421, 673−679. (24) Yliharsila, M.; Valta, T.; Karp, M.; Hattara, L.; Harju, E.; Holsa, J.; Saviranta, P.; Waris, M.; Soukka, T. Anal. Chem. 2011, 83, 1456− 1461. (25) Noor, M. O.; Krull, U. J. Anal. Chem. 2013, 85, 7502−7511. (26) Berlier, J. E.; Rothe, A.; Buller, G.; Bradford, J.; Gray, D. R.; Filanoski, B. J.; Telford, W. G.; Yue, S.; Liu, J.; Cheung, C. Y.; Chang, W.; Hirsch, J. D.; Beechem, J. M.; Haugland, R. P.; Haugland, R. P. J. Histochem. Cytochem. 2003, 51, 1699−1712. (27) Tavares, A. J.; Noor, M. O.; Vannoy, C. H.; Algar, W. R.; Krull, U. J. Anal. Chem. 2012, 84, 312−319. (28) Noor, M. O.; Krull, U. J. Anal. Chem. 2014, DOI: 10.1021/ ac502677n.

implementation of portable signal collection devices. We recently demonstrated that a conventional digital camera can be used to collect the fluorescence signal from a paper-based detection system using QDs and a FRET-based assay.28 Interestingly, compared with the performance of the assay using hydrated paper, both the sensitivity and the LOD of the assay can be boosted 10-fold by drying the paper before signal collection. Paper-based assays are not only advantageous in terms of cost and disposability but also offer a novel avenue for nonenzymatic signal amplification, which is a significant advantage for in-field diagnostics.

■

ASSOCIATED CONTENT

S Supporting Information *

Reagents, instruments and characterization, data analysis, and upconversion nanoparticle synthesis and modification. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

* E-mail: [email protected]. Fax: 905 569 4388. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful to the Natural Sciences and Engineering Research Council of Canada for support of this research. We also thank Uvaraj Uddayasankar, Anthony Tavares, and Omair Noor from the Chemical Sensors Group at University of Toronto Mississauga for useful discussions.

■

REFERENCES

(1) Chen, G.; Qiu, H.; Prasad, P. N.; Chen, X. Chem. Rev. 2014, 114, 5161−5214. (2) Haase, M.; Schafer, H. Angew. Chem., Int. Ed. 2011, 50, 5808− 5829. (3) Zhou, F.; Noor, M. O.; Krull, U. J. Anal. Chem. 2014, 86, 2719− 2726. (4) Sun, L. D.; Wang, Y. F.; Yan, C. H. Acc. Chem. Res. 2014, 47, 1001−1009. (5) Barreto, J. A.; O’Malley, W.; Kubeil, M.; Graham, B.; Stephan, H.; Spiccia, L. Adv. Mater. 2011, 23, H18−H40. (6) Hildebrandt, N.; Wegner, K. D.; Algar, W. R. Coord. Chem. Rev. 2014, 273, 125−138. (7) Noor, M. O.; Tavares, A. J.; Krull, U. J. Anal. Chim. Acta 2013, 788, 148−157. (8) Yao, L. M.; Zhou, J.; Liu, J. L.; Feng, W.; Li, F. Y. Adv. Funct. Mater. 2012, 22, 2667−2672. (9) Ye, W. W.; Tsang, M. K.; Liu, X.; Yang, M.; Hao, J. Small 2014, 10, 2390−2397. (10) Li, H.; Wang, L. Y. Analyst 2013, 138, 1589−1595. (11) Liu, J.; Cheng, J.; Zhang, Y. Biosens. Bioelectron. 2013, 43, 252− 256. (12) 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. (13) Wu, S.; Duan, N.; Ma, X.; Xia, Y.; Wang, H.; Wang, Z.; Zhang, Q. Anal. Chem. 2012, 84, 6263−6270. (14) Zhang, C. L.; Yuan, Y. X.; Zhang, S. M.; Wang, Y. H.; Liu, Z. H. Angew. Chem., Int. Ed. 2011, 50, 6851−6854. (15) Liu, J.; Cheng, J.; Zhang, Y. Biosens. Bioelectron. 2013, 43, 252− 256. (16) Wilhelm, S.; del Barrio, M.; Heiland, J.; Himmelstoß, S. F.; Galbán, J.; Wolfbeis, O. S.; Hirsch, T. ACS Appl. Mater. Inter. 2014, 6, 15427−15433. H

dx.doi.org/10.1021/ac503207q | Anal. Chem. XXXX, XXX, XXX−XXX