Signal Enhancement in Antibody Microarrays Using Quantum Dots

Article pubs.acs.org/ac

Signal Enhancement in Antibody Microarrays Using Quantum Dots Nanocrystals: Application to Potential Alzheimer’s Disease Biomarker Screening Eden Morales-Narváez,‡,§ Helena Montón,‡,⊥ Anna Fomicheva,‡,⊥ and Arben Merkoçi*,‡,† ‡

Nanobioelectronics & Biosensors Group, Catalan Institute of Nanotechnology, 08193, Bellaterra, Barcelona, Spain Polytechnic University of Catalonia, ESAII department, 08028, Barcelona, Spain ⊥ Autonomous University of Barcelona, 08193, Bellaterra, Barcelona, Spain † Catalan Institution for Research and Advanced Studies (ICREA), 08010, Barcelona, Spain §

ABSTRACT: The performance of cadmium-selenide/zincsulfide (CdSe@ZnS) quantum dots (QDs) and the fluorescent dye Alexa 647 as reporter in an assay designed to detect apolipoprotein E (ApoE) has been compared. The assay is a sandwich immunocomplex microarray that functions via excitation by visible light. ApoE was chosen for its potential as a biomarker for Alzheimer’s disease. The two versions of the microarray (QD or Alexa 647) were assessed under the same experimental conditions and then compared to a conventional enzyme-linked immunosorbent assay (ELISA) targeting ApoE. The QDs proved to be highly effective reporters in the microarrays, although their performance strongly varied in function of the excitation wavelength. At 633 nm, the QD microarray gave a limit of detection (LOD) of ∼247 pg mL−1; however, at an excitation wavelength of 532 nm, it provided a LOD of ∼62 pg mL−1, five times more sensitive than that of the Alexa microarray (∼307 pg mL−1) and seven times more than that of the ELISA (∼470 pg mL−1). Finally, serial dilutions from a human serum sample were assayed with high sensitivity and acceptable precision and accuracy.

P

structure to function, perturbation of the structure or behavior of biomolecules must be carefully considered while designing a novel biosensing system.4 A detailed comparison between quantum dots and organic dyes has been widely covered by Resh-Genger and colleagues.8 Table 1 shows a brief comparison among the properties of organic dyes and quantum dots. Biomarkers indicate a biological status and, therefore, can reveal biological processes in normal, pathogenic, or pathological states and even during pharmacologic or therapeutic responses.21−24 Given that biomarkers can provide objective information for clinical diagnosis, new devices for biomarker screening are highly desired. For the present study, we employed Apolipoprotein E (ApoE), a potential risk factor for Alzheimer’s disease, as a model biomarker. Dementia is a cardinal health problem in developed countries: it currently affects over 25 million people worldwide and may affect over 75 million people during the next 20 years. The most frequent cause of dementia (50 to 70% of cases) is Alzheimer’s disease (AD),25 which results in a progressive loss of cognitive function and affects one in eight adults aged 65 years of age or older.26 Pathologically, AD is characterized by formation of amyloid plaques and neurofibrillary tangles in the

rotein microarray-based biosensing is an active research area with strong potential for the development of novel, multiplexed diagnostic assays.1−3 In this technology, fluorescent signals are often used to estimate analyte concentration after each assay step. The signals are provided from microarrays of labeled molecules on glass slides. Quantum dots (QDs) are semiconductor nanocrystals that offer advantageous fluorescent properties that can be exploited for optical biosensing:4−9 sizetunable emission, narrow and symmetric photoluminescence, broad and strong excitation spectra, strong luminescence, and robust photostability. The interaction of QDs in immunoassays has been widely studied by Zhu et al.10 Since QDs exhibit better photonic performance in solid phase (i.e., directly exposed to air) than in liquid phase (e.g., as an aqueous dispersion),11 they can prove highly effective as reporters in microarrays.12−14 Despite their advantages, QDs can exhibit some shortcomings; e.g., first, their heavy metal components (usually cadmium or lead) are highly toxic, but the effects are little known yet.8,15 Second, QDs present intermittence in emission or “blinking”, the causes and mechanism of which are as yet not completely understood.16−20 There is some hope that blinking can be suppressed by improved surface chemistries and addition of reducing agents.8 Third, QDs are much larger than organic fluorophores in size. Since biomolecules such as antibodies, enzymes, or aptamers are dependent on their © 2012 American Chemical Society

Received: May 20, 2012 Accepted: June 26, 2012 Published: June 26, 2012 6821

dx.doi.org/10.1021/ac301369e | Anal. Chem. 2012, 84, 6821−6827

Analytical Chemistry

Article

total ApoE and ApoE4 levels have been reported to be significantly different in AD patients compared to control patients.28 Researchers have reported using QDs as high-powered labels in chips for detection based on excitation by UV light.29,30 Here, we explore the performance of CdSe@ZnS quantum dots (QD655; maximum emission peak: ca. 655 nm) and the fluorescent dye Alexa 647 (maximum emission peak: ca. 668 nm) as reporter in a sandwich immunoassay microarray designed to detect ApoE, a potential biomarker of Alzheimer’s disease. The immunoassay in its two forms (QDs or Alexa 647) and the absorbance/emission spectra of the studied fluorophores are illustrated in Figure 1. The two versions of the microarray were compared for performance and then compared to an ApoE assay based on a conventional enzyme-linked immunosorbent assay (ELISA). The respective advantages and drawbacks of the microarrays and their potential for early diagnostic screening of real samples for Alzheimer’s disease are also discussed.

Table 1. Brief Comparison among the Properties of Organic Dyes and Quantum Dots property absorption spectra emission spectra different emission wavelengths can be efficiently excited at a single wavelength quantum yield fluorescence lifetimes toxicity

size continuity of the signal intensity brightness

organic dyes narrow and weak asymmetric and wide no

quantum dots broad and strong symmetric and narrow yes

50 to 100% 10 to 80% 1 to 10 ns 10 to 100 ns from very low to little known yet (heavy metal high; dependleakage must be prevented, ent on dye potential nanotoxicity) ∼0.5 nm ∼2 to 15 nm limited by limited by blinking effect photobleaching 1 QD ≈ 10 to 20 organic fluorophores6

■

brain, as well as neuronal loss, synaptic loss, brain atrophy, and inflammation.27 The ApoE gene provides instructions for the Apolipoprotein E yield. Apolipoprotein E is a 34 kD protein implicated in lipid transport and lipoprotein metabolism, whose levels are ultimately controlled by the gene ApoE. It is present in plasma lipoprotein particles such as chylomicrometers, very-lowdensity lipoprotein, and high-density lipoprotein. It is also found in large amounts at sites of neurological damage and appears to be involved in recycling of apoptotic remnants and of amyloidal aggregates. Apolipoprotein E is mainly produced in the liver and by macrophages but also by cell types including smooth muscle cells and neuronal cells. It exists in three isoforms: ApoE2, ApoE3, and ApoE4. Diverse pieces of evidence suggest that abnormal ApoE levels may be implicated in dementia and central nervous system diseases.25,27 In fact,

EXPERIMENTAL SECTION Materials. Glycerol, PBS, Tween 20, milk powder, and serum from human male AB plasma were obtained from SigmaAldrich (Taufkirchen, Germany). Epoxysilane glass slides were provided from Thermo Fisher Scientific (New Hampshire, USA). Assay buffer, streptavidin−HRP, monoclonal antibody, and biotinylated detection antibody against total ApoE were purchased from Mabtech (Nacka Strand, Sweden). Streptavidin−QD655 and streptavidin−Alexa 647 were obtained from Invitrogen (California, USA). PBS with 2% (v/v) glycerol was used as spotting buffer. PBS supplemented with 5% (w/v) of milk powder and 0.005% (v/v) of Tween 20 was prepared as blocking buffer. PBS supplemented with Tween 20 at 0.05% (v/v) was used as washing buffer (PBST). PBS with 0.5% (v/v) Tween 20 containing 1% of BSA fraction V (w/v) was

Figure 1. (A) Schematic of apoliprotein-E (ApoE)-screening sandwich immunocomplex microarrays using either quantum dots (QDs) or the organic dye Alexa 647 as reporter. (a) Capture antibodies are spotted onto glass slides and free sites are blocked. (b) Samples containing the analyte are incubated. (c) Biotinylated detection antibodies are incubated. (d) Bound detection antibodies are reported by a streptavidin−fluorophore complex (streptavidin−QD or streptavidin−Alexa 647). (B) Absorbance (dashed line) and emission (solid line) spectra of the studied fluorophores; in red, QD 655; in green, Alexa 647. 6822



Article

(diluted in immunobuffer at 1.5 μg mL−1) for 1 h, and then rewashed with PBST (150 μL per well, seven times). The bound detection antibodies were labeled (100 μL/well) with either quantum dot 655 (QD655; diluted in immunobuffer at 25 nM) or Alexa 647 (diluted in immunobuffer at 0.4 μg mL−1) for 30 min, depending on the studied reporter. All incubations were performed with oscillatory agitation at 300 rpm in an MTS 2/4 digital microtiter shaker (IKA; Staufen, Germany). The masked slides were washed with PBST (150 μL per well, five times), twice with PBS, and once with milli-Q water. The slides were then unmasked and subsequently dried by centrifugation (1500 rpm, 1 min). The slides were examined through a Typhoon 9410 microarray scanner (GE Healthcare; Nueva Jersey, USA) upon excitation with different lasers (457, 488, 532, or 633 nm, depending on the explored fluorophore) through an emission filter of 670 nm and a band-pass of 30 nm. Slides were quantified using GenePix (Molecular Devices; California, USA), and the signal-to-noise ratio (SNR) was measured. The fluorescence intensities were estimated by measuring the mean intensities of the assayed spots minus the local background. The intensities were normalized according to the maximum value of the respective screening. The limit of detection (LOD) of the respective calibration curve was calculated as the sum of the mean intensity among the spots (incubated with a blank) and three times the standard deviation: LOD = MI + 3SD.

employed as immunobuffer. Water used to prepare the solutions was Milli-Q. All reagents were handled according to their corresponding material safety data sheets, provided by the respective suppliers. High-Resolution Transmission Electron Microscopy (HRTEM). For HRTEM analyses, the streptavidin-QD 655 was directly diluted from stock solution in Milli-Q water to a final solution of 25 nM. A 2 μL drop was deposited onto a holey carbon layer copper grid and then air-dried. A Jeol JEM 2011 transmission electron microscope (Jeol LTD; Tokyo, Japan) operating at 200 kV was used to obtain the images for posterior analysis. The size of the QDs was determined by processing the images with Image J software (Rasband, W.S.; ImageJ; U.S. National Institutes of Health [NIH], Bethesda, Maryland, USA; http://imagej.nih.gov/ij/, 1997−2011); the data obtained were processed for statistical evaluation through Microsoft Excel (2007). Zeta-Potential Measurements. Samples of QD655 were prepared for zeta-potential evaluation by diluting them in MilliQ water to a final concentration of 12 nM. One milliliter of the solution added to a cuvette and then analyzed in a Zetasizer unit (Malvern Nanosizer ZS, Malvern Instruments Ltd., UK). The zeta-potential value of the QD655 dispersion was calculated by the instrument. Enzyme-Linked Immunosorbent Assay (ELISA). The ELISA plate microwells were coated with monoclonal antibody against total ApoE diluted in PBS at 2 μg mL−1 and incubated overnight at 4 °C. The microwells were then washed three times with PBST (200 μL/well), blocked with blocking buffer (200 μL/well) for 1 h, washed with PBST (200 μL/well, five times), and finally, incubated with different concentrations (100 μL/well) of stock solutions of ApoE2, ApoE3, and ApoE4 in assay buffer for 2 h. The microwells were washed five times with PBST (200 μL/well), incubated with biotinylated capture antibody (diluted in assay buffer at 1 μg mL−1) for 1 h, and washed again with PBST (200 μL/well, seven times). The bound detection antibodies were labeled by adding a solution of streptavidin−HRP in immunobuffer (1:1000, 100 μL). Another five rounds of washing were performed, and then, TMB (100 μL/well) was added as the chromogenic substrate. After 30 min, the reaction with TMB was stopped by adding 0.5 M H2SO4. After another 30 min, the colored solutions that had developed in the microwells were analyzed in a spectrometer at 450 nm (SpectraMax M2e; Molecular devices, Pennsylvania, USA). Fabrication of the Antibody Microarrays. The capture antibodies were generally spotted at 0.4 mg mL−1 (in spotting buffer) over epoxysilane glass slides using Microgrid II (Digilab; Massachusetts, USA). Spotting was done at room temperature and at 40 to 60% humidity. The average diameter of the printed spots was ca. 150 μm. After spotting, the slides were incubated at 4 °C overnight in a desiccated environment. The microarray slides were masked and divided by wells through a microarray cassette (Arrayit; California, USA), and each well was washed three times with PBST (150 μL) and then blocked for 30 min with blocking buffer (100 μL). Sandwich Immunoassay Microarray. Once blocked, the masked slides were washed with PBST (100 μL/well, five times) and the microarrays were incubated for 2 h with standard solutions of ApoE2, ApoE3, and ApoE4 in immunobuffer (100 μL per well) at different concentrations. The masked slides were washed with PBST (150 μL per well, five times), incubated with biotinylated capture antibody

■

RESULTS AND DISCUSSION Characterization of the QDs. The size, shape, and crystallinity of the QD655 were characterized by HRTEM. The images revealed a very homogeneous and well-dispersed solution of the QD655, which did not show any signs of aggregation (Figure 2A). The QD655 exhibited a rice-grain shape (see Figure 2a), in which the crystalline planes can be appreciated. The size distribution of QD655 is shown in Figure 2B (average size: 14 nm ±2 nm). These results are consistent with literature31 values. Since surface charge cannot be measured directly, in order to ascertain the electrical interaction forces among the dispersed particles (QD655), the zeta-potential was measured instead (estimated value: −28 mV ± 5 mV). This value indicated the high stability of the nanoparticles in an aqueous suspension, a requirement for their subsequent use in microarray immunoassays. Optimization of Fluorophore Concentration. Luo et al. compared the performance of the complexes QD−antibody and QD−streptavidin as probes for captured analytes in microplates.32 The former displayed a sensitivity of 0.21 μg L−1 and the latter of 0.06 μg L−1. The authors postulated that “the decreased detection limit is due to the aggregation of the functionalized QDs after gradually losing their colloidal stability in the PBS or the fluorescence quenching effect by conjugated biomolecules or the remaining chemicals in the solution” (after conjugation of the QDs to the antibody).32 We have employed complexes fluorophore−streptavidin as reporters of previously captured analytes. To ensure that the studied fluorophores functioned under similar experimental conditions, the concentration of each fluorophore was optimized by exploring a suitable range of concentrations and selecting the maximum fluorescence intensity (minus local background) in the range. The fluorescence intensities were scrutinized at different concentrations of QD655 or Alexa647 and at a constant concentration 6823



Article

Figure 3. Optimization of fluorophore concentration. To ensure that the studied fluorophores (QD655 or Alexa 647) would function under similar experimental conditions, the optimization was performed by exploring a suitable range of concentrations per each fluorophore and selecting the minimum value into the saturated values of such range. Fluorescence intensities were explored using a constant concentration of 10 ng mL−1 of ApoE3 in a sandwich immunocomplex. Bound biotinylated detection antibody was employed so as to be detected. Since these measurements showed a saturation at [SAv−A647] = 0.4 μg mL−1 and [SAv−QD655] = 25 nM, such concentrations were selected as optimal. The error bars were obtained from a parallel assay of 10 microspots. SAv−A647, streptavidin−Alexa 647; SAv−QD655, streptavidin−quantum dot 655. Figure 2. Characterization of the quantum dots. (A) HRTEM image of the studied quantum dots (QD655). Images show a very homogeneous and well-dispersed solution of the explored nanoparticles without any aggregation. (a) The studied nanocrystals display a rice-grain shape, in which crystalline planes can be appreciated. (B) Size distribution of the QD655 (average size: ca. 14 nm ±2 nm).

of ApoE3 (10 ng mL−1) . As described in the Experimental Section, the bound biotinylated detection antibody was employed so as to be detected upon its interaction with the streptavidin−fluorophore complex. Streptavidin−quantum dot 655 (SAv−QD655) fluorescence was analyzed from 0.7 to 50 nM, and streptavidin−Alexa 647 (SAv−A647) fluorescence was analyzed between 0.02 and 1.6 μg mL−1. Thus, the following concentrations for each complex were chosen as optimal: 25 nM for [SAv−QD655] and 0.4 μg mL−1 for [SAv−A647] (see Figure 3). Comparison of the Different ApoE-Screening Assays. ApoE Screening by ELISA. We compared our ApoE-screening microarrays with an ApoE assay done by ELISA, the gold standard of immunoassays, performed as described in the Experimental Section. ApoE2, ApoE3, and ApoE4 were mixed (in equal proportions) so as to perform total ApoE quantification at different concentrations. The ELISA gave an LOD of ca. 470 ng mL−1 and signal saturation at ca. 16 ng mL−1 (see Figure 4). ApoE Screening by Sandwich Immunocomplex Microarray. Using QDs as Reporter (at Different Excitation Wavelengths). Using the parameters given in the Experimental Section, we mixed ApoE2, ApoE3, and ApoE4 in equal proportions so as to quantify total ApoE at different concentrations in an ApoE-screening microarray containing QD reporters (maxima emission: ca. 655 nm), performed as described in the Experimental Section. The absorption spectra of the QDs were broad and strong (see Figure 1B); after the

Figure 4. ApoE screening by ELISA. ApoE2, ApoE3, and ApoE4 were mixed (in equal proportions) so as to quantify total ApoE. The error bars were obtained from three parallel assays.

assay, the formed microspots containing the immunocomplexes were inspected at different wavelengths (457, 488, 532, and 633 nm) to determine a suitable excitation wavelength for the developed system among the available excitation sources in a microarray scanner. Emission was scanned with an emission filter of 670 nm, bandpass of 30 nm, laser power of 100%, and PMT of 750 units. The signal-to-noise ratio (SNR) of the assayed spots at different excitation wavelengths were then evaluated, as were the LODs. Among the excitation wavelengths tested, 532 nm gave the highest SNR (i.e., the lowest background). The SNR lower at 457 nm was likely due to the relatively low power of the available laser (4 mW, as compared to 20 mW at 488 and 532 nm, and 10 mW at 633 nm). The background level is denoted by the limit below the respective SNR distribution (see Figure 5A). The ApoE-screening microarray showed different LODs in function of the excitation wavelength: 168 pg mL−1 at 457 nm; 76 pg mL−1 at 488 nm; 62 pg mL−1 at 532 nm; and 247 pg mL−1 at 633 nm (see Figure 6824



Article

Figure 6. Comparison of ApoE screening using the two microarrays (QD or Alexa 647 as reporter) under the same experimental conditions. ApoE2, ApoE3, and ApoE4 were mixed (in equal proportions) so as to quantify total ApoE at different concentrations. The error bars were obtained from parallel assays of 10 spots. QD655, quantum dot 655; A647, Alexa 647; Exc, excitation wavelength; LOD, limit of detection.

Figure 5. Performance of the ApoE-screening microarray using QDs as reporter at different excitation wavelengths. (A) Box plots of the signal-to-noise ratio (SNR) of the ApoE screening (from 400 to 6 ng mL−1 in 2-fold dilutions). The box plots show the median, 25th and 75th percentiles, the extreme values of the respective distribution, and the mean (+). From left to right: SNR distribution of QD655 excited at 457, 488, 532, and 633 nm and Alexa 647 excited at 633 nm. (B) Calibration curves of the ApoE screening by exciting the studied QDs at different wavelengths: 457, 488, 532, and 633 nm. ApoE2, ApoE3, and ApoE4 were mixed (in equal proportions) so as to quantify total ApoE at different concentrations. The error bars were obtained from parallel assays of ten microspots.

5B). Therefore, among the available excitation wavelengths in the scanner, 532 nm was chosen for the QDs. Using Alexa 647 as Reporter. Using the parameters given at in the Experimental Section, ApoE2, ApoE3, and ApoE4 were mixed (in equal proportions) so as to perform the total ApoE quantification at different concentrations in microarray format using Alexa 647 (A647). This dye is a well-known fluorphore with maxima excitation/emission of ca. 650/668 nm (see Figure 1B). The estimation of the SNR distribution of the ApoE screening using A647 as reporter was also performed (excitation wavelength: at 633 nm, emission filter of 670 nm, bandpass of 30 nm). In the same experimental conditions, such SNR has shown a higher background when compared with the SNR distribution of the ApoE screening using QDs at the explored wavelengths (see Figure 5A). ApoE screening reported by A647 exhibited a limit of detection of ca. 307 pg mL−1 (see Figure 6). ApoE Screening in Human Serum. Human serum from human male AB plasma was inspected in serial dilutions and assayed in a microarray as described in the Experimental Section (see Figure 7). As human serum concentration values for ApoE typically range from 16 to 169 μg mL−1,35 very small volumes (1 μL) of serum were diluted in a suitable volume of assay buffer for study. The concentration of ApoE in the assayed serum was estimated by interpolating the intensities of the ApoE screening in human serum into a calibration curve. According to the proposed microarray-QD-based system, the

Figure 7. ApoE screening in human serum using the QD microarray. Small volumes (1 μL) of serum were diluted in a suitable volume of assay buffer. The concentration of ApoE in the assayed serum was estimated by interpolating the fluorescence intensities from the screen into a calibration curve. According to the QD microarray, the original concentration of the inspected human serum was ∼110.98 ± 20.48 μg mL−1. The error bars were obtained from parallel assays of ten spots. QD655, quantum dot 655; Exc, excitation wavelength.

original concentration of ApoE in the serum sample was ∼110.98 ± 20.48 μg mL−1. Precision and Accuracy. Since antibody microarray technology lacks current validation guidelines, a standard procedure (based on the Food and Drug Administration guidelines for pharmacokinetic immunoassays) through the following acceptance criteria, accuracy (recovery) ranging from 70% to 130% and assay precision (coefficient of variation) of less than 30%, could be proposed.33,34 Precision was estimated by determining the average of the coefficients of variation across ten spots (microarrays) or three parallel assays (ELISA). The microarrays displayed a coefficient of variation of ca. 7% ± 2.3% (QD655 as reporter) or 8.4% ± 5.8% (A647 as reporter), and the ELISA displayed a coefficient of variation of ca. 6.9% ± 5.3%. To estimate the recovery of 6825



Article

Table 2. Limit of Detection (LOD), Signal Saturation, Precision, and Recovery for ApoE Screening Using the Two Microarrays and the ELISAa format

reporter

ELISA microarray

HRP A647 QD655 QD655 QD655 QD655

exc/em (nm)

SS (ng mL−1)

LOD (ng mL−1)

CV (%)

recovery (%)

633/670 633/655 532/655 488/655 457/655

∼16 ∼200 ∼200 ∼100 ∼100 ∼200

∼0.470 ∼0.307 ∼0.247 ∼0.062 ∼0.076 ∼0.168

∼6.9 ± 5.3 ∼8.4 ± 5.8

∼89 ± 6 ∼84 ± 13

∼7 ± 2.3

∼104 ± 18

a exc, excitation wavelength; em, emission wavelength; SS, signal saturation; LOD, limit of detection ; CV, coefficient of variation. Numbers after “±” denote the standard deviation.

■

FUTURE PERSPECTIVES Since incorporation of QDs into microarrays is a relatively new endeavor, various challenges in this technology remain to be addressed. For instance, many current scanners lack the tools to obtain QD fluorescent signals under suitable excitation conditions. Most of them include excitation sources for red lasers and for green lasers. Furthermore, many of these scanners cannot be configured to simultaneously select the excitation source and the filter emission. Moreover, compared to organic dyes, QDs are very expensive to purchase. However, several synthetic routes to QDs have already been published,36 and configurable scanners or other setups can be employed for QD signal acquisition in microarray platforms.37,14 Fine-tuning of microarrays for use with QDs should enable improved biosensing performance and accelerate incorporation of this technology into real-world bioanalytical scenarios for applications in diagnostics, safety, security, and environmental monitoring.

each assay, a minimum of four different quantifications of known concentrations of total ApoE (ApoE2, ApoE3, and ApoE4 mixed in equal proportions) were interpolated in the respective calibration curve. The microarrays exhibited a recovery of ca. 104.3% ± 18% (QD655 as reporter) or 84% ± 13.4% (A647 as reporter), and the ELISA exhibited a recovery of ca. 89.4% ± 5.7%. Hence, our assays meet the criteria cited in the previous paragraph. Limits of detection were estimated by interpolating the mean intensity/optical density among spots/wells incubated with a blank plus three times their standard deviation in the respective calibration curves. Signal saturation corresponds to the upper zone of the calibration curve, where the intensity (microarray) or optical density (ELISA) remains constant despite increasing amounts of analyte. Table 2 summarizes the performance of the three assays in terms of LOD, signal saturation, precision, and recovery.

■

■

CONCLUSIONS

AUTHOR INFORMATION

Corresponding Author

We have explored the biosensing performance of core−shell QDs (QD655) and a fluorescent dye (A647) while being employed in the same experimental conditions as reporters of sandwich immunocomplexes in microarray format and they are also compared with a conventional ELISA. A potential AD biomarker (ApoE) has been proposed as the model analyte. The dynamic range of ApoE screening in microarray format is broader when compared with the dynamic range of the ApoE screening in ELISA format (see table 2). Since SNR of the QDs in microarray format depends not only upon both the excitation wavelength and the background signal but also probably upon the power of the excitation source, we have observed that an excitation at 532 nm (20 mW) yields a better SNR when compared with either an excitation at 457 nm (4 mW) or an excitation at 488 nm (20 mW, displaying a higher background). Regarding the sensitivity, we have found that QDs become advantageous reporters in microarray technology even though they are not excited with an ideal wavelength (633 nm, LOD of ca. 247 pg mL−1). On the other hand, while QDs are excited at more suitable wavelengths (457, 488, and 532), we obtain up to a 7-fold enhancement in the LOD (ca. 62 pg mL−1) when compared with a conventional ELISA (ca. 470 pg mL−1) and up to a 5-fold enhancement when compared with Alexa 647 as reporter in microarray format (LOD of ca. 307 pg mL−1). Finally, very small volumes (1 μL) of human serum were assayed with high sensitivity and acceptable precision and accuracy. This approach could be extended to a multiplexed detection of ApoE isoforms and other AD related biomarkers.

*E-mail: [email protected]. Fax: +34 935868020. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS MICINN through project MAT2011-25870 and E.U. through FP7 “NADINE” project (contract number 246513) have funded this research. We acknowledge funding from CONACYT (Mexico), through a fellowship grant for E.M.-N. We thank José Ignacio Pons-Ximénez, the Genomics Service at Universitat Autònoma de Barcelona, and the Proteomics and Genomics Unit at Universidad de Alicante, for technical support with microarray technology. Finally, we thank Gregory Qushair for help with editing the manuscript.

■

REFERENCES

(1) Chen, R.; Snyder, M. J. Proteomics 2010, 73, 2147−2157. (2) Hu, S.; Xie, Z.; Qian, J.; Blackshaw, S.; Zhu, H. Wiley Interdiscip. Rev.: Syst. Biol. Med. 2011, 3, 255−268. (3) MacBeath, G. Nat. Genet. 2002, 32 (Suppl), 526−532. (4) Algar, W. R.; Krull, U. J. In Biosensing Using Nanomaterials; John Wiley & Sons, Inc.: Hoboken, NJ, 2009; pp 199−245. (5) Algar, W. R.; Tavares, A. J.; Krull, U. J. Anal. Chim. Acta 2010, 673, 1−25. (6) Jiang, W.; Singhal, A.; Fischer, H.; Mardyani, S.; Chan, W. C. W. In BioMEMS and Biomedical Nanotechnology, Vol. 3 Therapeutic Micro/ Nanotechnology; Ferrari, M., Desai, T., Bhatia, S., Eds.; Springer: New York, 2007; pp 137−156. (7) Merkoçi, A. Biosens. Bioelectron. 2010, 26, 1164−1177. 6826



Article

(36) Zhuang, Z.; Peng, Q.; Li, Y. Chem. Soc. Rev. 2011, 40, 5492− 5513. (37) Robelek, R.; Niu, L.; Schmid, E. L.; Knoll, W. Anal. Chem. 2004, 76, 6160−6165.

(8) Resch-Genger, U.; Grabolle, M.; Cavaliere-Jaricot, S.; Nitschke, R.; Nann, T. Nat. Methods 2008, 5, 763−775. (9) de la Escosura-Muñiz, A.; Parolo, C.; Merkoçi, A. Mater. Today 2010, 13, 24−34. (10) Zhu, X.; Duan, D.; Madsen, S.; Publicover, N. Anal. Bioanal. Chem. 2010, 396, 1345−1353. (11) Shi, X.; Meng, X.; Sun, L.; Liu, J.; Zheng, J.; Gai, H.; Yang, R.; Yeung, E. S. Lab Chip 2010, 10, 2844−2847. (12) Sapsford, K. E.; Spindel, S.; Jennings, T.; Tao, G.; Triulzi, R. C.; Algar, W. R.; Medintz, I. L. Sensors 2011, 11, 7879−7891. (13) Nichkova, M.; Dosev, D.; Davies, A.; Gee, S. Anal. Lett. 2007, 40, 1423−1433. (14) Rousserie, G.; Sukhanova, A.; Even-Desrumeaux, A.; Fleury, F.; Chames, P.; Baty, D.; Oleinikov, V.; Pluot, M.; Cohena, J. H. M.; Nabiev, I. Crit. Rev. Oncol. Hematol. 2010, 74, 1−15. (15) Lewinski, N.; Colvin, V.; Drezek, R. Small 2008, 4, 26−49. (16) Hohng, S.; Ha, T. Chem. Phys. Chem. 2005, 6, 956−960. (17) Medintz, I. L.; Uyeda, H. T.; Goldman, E. R.; Mattoussi, H. Nat. Mater. 2005, 4, 435−446. (18) Gomez, D. E.; Califano, M.; Mulvaney, P. Phys. Chem. Chem. Phys. 2006, 8, 4989−5011. (19) Pinaud, F.; Michalet, X.; Bentolila, L. A.; Tsay, J. M.; Doose, S.; Li, J. J.; Iyer, G.; Weiss, S. Biomaterials 2006, 27, 1679−1687. (20) Robelek, R.; Stefani, F. D.; Knoll, W. Phys. Status Solidi A 2006, 203, 3468−3475. (21) Chambers, G.; Lawrie, L.; Cash, P.; Murray, G. J. Pathol. 2000, 192, 280−288. (22) Mattson, M. P. Nature 2004, 430, 631−639. (23) Bild, A. H.; Yao, G.; Chang, J. T.; Wang, Q.; Potti, A.; Chasse, D.; Joshi, M.-B.; Harpole, D.; Lancaster, J. M.; Berchuck, A.; Olson, J. A.; Marks, J. R.; Dressman, H. K.; West, M.; Nevins, J. R. Nature 2006, 439, 353−357. (24) Alizadeh, A. A.; Eisen, M. B.; Davis, R. E.; Ma, C.; Lossos, I. S.; Rosenwald, A.; Boldrick, J. C.; Sabet, H.; Tran, T.; Yu, X.; Powell, J. I.; Yang, L.; Marti, G. E.; Moore, T.; Hudson, J.; Lu, L.; Lewis, D. B.; Tibshirani, R.; Sherlock, G.; Chan, W. C.; Greiner, T. C.; Weisenburger, D. D.; Armitage, J. O.; Warnke, R.; Levy, R.; Wilson, W.; Grever, M. R.; Byrd, J. C.; Botstein, D.; Brown, P. O.; Staudt, L. M. Nature 2000, 403, 503−511. (25) Takeda, M.; Martínez, R.; Kudo, T.; Tanaka, T.; Okochi, M.; Tagami, S.; Morihara, T.; Hashimoto, R.; Cacabelos, R. Psychiatry Clin. Neurosci. 2010, 64, 592−607. (26) Ray, S.; Britschgi, M.; Herbert, C.; Takeda-Uchimura, Y.; Boxer, A.; Blennow, K.; Friedman, L. F.; Galasko, D. R.; Jutel, M.; Karydas, A.; Kaye, J. A.; Leszek, J.; Miller, B. L.; Minthon, L.; Quinn, J. F.; Rabinovici, G. D.; Robinson, W. H.; Sabbagh, M. N.; So, Y. T.; Sparks, D. L.; Tabaton, M.; Tinklenberg, J.; Yesavage, J. A.; Tibshirani, R.; Wyss-Coray, T. Nat. Med. 2007, 13, 1359−1362. (27) Kim, J.; Basak, J. M.; Holtzman, D. M. Neuron 2009, 63, 287− 303. (28) Gupta, V. B.; Laws, S. M.; Villemagne, V. L.; Ames, D.; Bush, A. I.; Ellis, K. A.; Lui, J. K.; Masters, C.; Rowe, C. C.; Szoeke, C.; Taddei, K.; Martins, R. N. Neurology 2011, 76, 1091−1098. (29) Geho, D.; Lahar, N.; Gurnani, P.; Huebschman, M.; Herrmann, P.; Espina, V.; Shi, A.; Wulfkuhle, J.; Garner, H.; Petricoin, E.; Liotta, L. A.; Rosenblatt, K. P. Bioconjugate Chem. 2005, 16, 559−566. (30) Hu, M.; Yan, J.; He, Y.; Lu, H.; Weng, L.; Song, S.; Fan, C.; Wang, L. ACS Nano 2009, 4, 488−494. (31) Montón, H.; Nogués, C.; Rossinyol, E.; Castell, O.; Roldán, M. J. Nanobiotechnol. 2009, 7, 4. (32) Luo, Y.; Zhang, B.; Chen, M.; Jiang, T.; Zhou, D.; Huang, J.; Fu, W. J. Transl. Med. 2012, 10, 24. (33) Findlay, J. W. A.; Smith, W. C.; Lee, J. W.; Nordblom, G. D.; Das, I. J. Pharm. Biomed. Anal. 2000, 21, 1249−1273. (34) Urbanowska, T.; Mangialaio, S.; Zickler, C.; Cheevapruk, S.; Hasler, P.; Regenass, S.; Legay, F. J. Immunol. Methods 2006, 316, 1−7. (35) Vincent-Viry, M.; Schiele, F.; Gueguen, R.; Bohnet, K.; Visvikis, S.; Siest, G. Clin. Chem. 1998, 44, 957−965. 6827


Signal Enhancement in Antibody Microarrays Using Quantum Dots

Recommend Documents