ANALYTICAL SCIENCES MARCH 2016, VOL. 32 307

Introduction

Alzheimer’s disease (AD) is the most common age-related neurodegenerative disorder, and one of the most devastating diagnoses that patients can receive.1 AD is characterized by senile or neuritic plaques and neurofibrillary tangles in the medial temporal lobe structures and cortical areas of the brain, together with degeneration of the neurons and synapses.2 The amyloid beta (Aβ) protein is a 39- to 43-amino acid polypeptide that is the primary constituent of senile plaques and cerebrovascular deposits in AD.3 This amphiphilic peptide spontaneously forms fibrillar aggregates called amyloid fibrils in aqueous solutions at or below physiological pH. Amyloid fibrils are typically rigid and exhibit a cross-β structure, in which the β-strands are aligned perpendicular to the long axis of the fibril. Aβ fibrils and intermediate oligomers are cytotoxic, and their physiological presence and tissue deposition are associated with AD. Therefore, detection of these Aβ aggregates is important for early recognition of the disease.4

The amyloid fibrils were mainly detected with the help of amyloid specific probes, such as thioflavin T (ThT) and Congo red.4,5 However, it has some drawbacks. First, sub-μM amounts of amyloid fibrils are necessary for the detection. It was also reported that in the presence of exogenous additives such as polyphenols, the strong absorptive and fluorescent properties of these compounds were found to significantly bias the ThT fluorescence.6 Enzyme-linked immunosorbent assay (ELISA) using anti-Aβ antibody can be used for sensitive and specific Aβ detection.7 However, it cannot distinguish aggregates from

monomer molecules. Thus, it is necessary to develop a sensitive and specific amyloid aggregates detection method.

A previous study reported that Aβ fibrils can be detected by the precipitation of gold nanoparticles (AuNPs (100 nm in diameter)) modified with Aβ antibodies (Ab-AuNPs).8 Red-colored precipitates can be seen by the naked-eye with the addition of Aβ fibrils at the μM range into Ab-AuNP solution. In this work, in an attempt to achieve higher sensitivity of Aβ fibrils detection, dark field microscopy (DFM), which can detect scattered light from individual single metal nanostructures,9,10 was utilized. Since Aβ fibrils-induced AuNP aggregates can be detected at the single particle level using DFM, it is expected that sensitive detection of Aβ fibrils can be achieved. Previously, we showed that DNA could be detected sensitively by observing DNA-induced non-crosslinking AuNP aggregates with DFM.11 In this study, we demonstrated that the limit of detection (LOD) of Aβ fibrils using this method reached as low as the pM level. It is expected that this AuNP-based immune sensing method could be developed with higher sensitivity than conventional detection methods.

Experimental

Reagents and chemicalsAβ1-42 peptide was purchased from Peptide Institute (Osaka,

Japan). AuNP (40 nm in diameter) was obtained from British Biocell International (Cardiff, UK). This particle size is suitable for the detection of AuNP aggregates with DFM.11 Mouse monoclonal antibody to Aβ (6E10) and HRP conjugated anti-mouse immunoglobulin G (IgG) were purchased from Abcam (Cambridge, UK) and R&D Systems (Minneapolis, MN, USA), respectively. The enhanced chemiluminescent (ECL) detection kit was purchased from GE Healthcare (Little Chalfont, UK). PBS was from Takara (Osaka, Japan). Bovine serum albumin (BSA) was obtained from Sigma Chemical Company (St. Louis, MO, USA).

2016 © The Japan Society for Analytical Chemistry

† To whom correspondence should be addressed.E-mail: zako@riken.jpT. Z. present address: Department of Chemistry and Biology, Graduate School of Science and Engineering, Ehime University, 2-5 Bunkyo, Matsuyama, Ehime 790–8577, Japan.

Dark Field Microscopic Sensitive Detection of Amyloid Fibrils Using Gold Nanoparticles Modified with Antibody

Tong BU,*,** Tamotsu ZAKO,**† and Mizuo MAEDA*,**

* Department of Advanced Materials Science, School of Frontier Sciences, The University of Tokyo, 5-1-5 Kashiwanoha, Kashiwa, Chiba 277–8561, Japan

** Bioengineering Laboratory, RIKEN Institute, 2-1 Hirosawa, Wako, Saitama 351–0198, Japan

Dark field microscopy (DFM) was employed to detect amyloid β (Aβ) fibrils-induced gold nanoparticle (AuNP) aggregation at the single-particle level, with a detection limit of 40 pM fibrils. The sensitivity of this method is higher than that of the current fibril-specific detection method using probe dye, such as thioflavin T, for which sub-μM level of fibrils are necessary. This study further proved the potential application of DFM in the analytical methods based on AuNP aggregation.

Keywords Dark field microscopy, gold nanoparticle, single molecule analysis, amyloid β fibrils, sensitive detection

(Received December 9, 2015; Accepted January 8, 2016; Published March 10, 2016)

Notes

308 ANALYTICAL SCIENCES MARCH 2016, VOL. 32

Preparation of Aβ samplesSeed-free Aβ monomers were prepared as previously

described.12,13 Briefly, Aβ peptide was dissolved in a 0.1% ammonia solution at 200 μM, followed by centrifuging at 100000 rpm for 3 h at 4°C to obtain a seed-free Aβ solution (Aβ monomer). The supernatant was collected and stored as aliquots at –80°C. Before usage, the aliquots were thawed, diluted with PBS and incubated at 25 μM and 37°C for 24 h to obtain Aβ fibrils. The concentration of the Aβ samples was measured using the BCA protein assay kit (Thermo Scientific, Roskilde, Denmark).

Immobilization of Aβ antibodies on AuNPAntibody was immobilized on AuNP as described with slight

modifications.8 A  100-μL AuNP solution was centrifuged at 6000 rpm for 10 min. The supernatant was removed and AuNP precipitates were suspended into 100 μL of 10-times diluted PBS including 50 nM Aβ antibody (6E10), followed by incubation for 1 h at room temperature. After incubation, 100 μL of 4-times diluted PBS containing BSA (5 wt%) was added to the sample tube, followed by incubation for 30 min at room temperature. BSA was added to stabilize the AuNP.8 The sample tube was then centrifuged at 6000 rpm for 5 min, and the supernatant was removed. The resulting antibody-modified AuNP (Ab-AuNP) precipitates were re-suspended in 100 μL of 1 × PBS and used for further studies. Dot blot assay was performed to quantify the amount of Aβ antibody immobilized on the surface of AuNPs (see Supporting Information for details).

Color change and UV-Vis spectrometer for observation of AuNP aggregates

As primary experiments, 5 μM Aβ fibrils and Aβ monomer was added to 30 pM Ab-AuNP, followed by incubation for 1 h at RT. The photoimages of tubes were taken by a digital camera. The UV-Vis spectra were taken by Cary 50 UV-VIS spectrometer (Varian, Palo Alto, USA). DFM images were taken as described below.

Dark field microscopy analysisScheme for detection of Aβ fibrils using DFM is shown in

Fig. 1. Various concentrations (from 0 to 5 μM) of Aβ fibrils were added to the 30 pM Ab-AuNP colloid in PBS buffer, and incubated for 1 h at room temperature. Ab-AuNP without addition of Aβ sample was used as a negative control sample (background). The DFM analysis was carried out as previously described.11 The AuNP samples (6 μL) were deposited onto

glass slides (high density amine coated slides, Matsunami Glass Ind., Osaka, Japan) and covered with cover glass. DFM images were taken using a BX53 microscope (Olympus, Tokyo, Japan) equipped with UDCW dark field condenser, UPlanFLN 60× objective lens and DP73 CCD camera (Olympus). Images were taken using CellSens Standard software Ver. 1.6 (Olympus). AuNPs were identified as bright orange-colored spots in DFM images and the scattered light intensity of each spot was quantified to obtain the intensity histogram.

To determine the detection limit of DFM, the ratio of AuNP aggregates at each target Aβ fibrils concentration was obtained from the histogram. The brighter spots that have intensities higher than 45 a.u. were defined as aggregates, since 90% of the  dispersed negative control sample to which no Aβ fibrils was added, showed an intensity lower than 45 a.u. as described in the results section. The LOD was evaluated by the 3σ criterion method, where σ denotes the standard deviation of zero-concentration background data (n = 6) as previously described.11,14 Average values from three histograms obtained by three different glass slides were used. The data points of the calibration curve were fitted with a four-parameter logistic function: y = d + (a – d)/[1 +(x/c)b], where a, b, c and d are fitting parameters. These parameters were optimized by nonlinear least-square regression weighted by the reciprocal of the square of standard deviation of each datum point. The LOD was calculated from the fitting equation as the concentration that gives higher value than 3σ line (= background (zero-concentration) + 3σ).

Results and Discussion

Antibody immobilization on AuNPIn this study, Ab-AuNPs were used for detection of Aβ fibrils.

The Ab-AuNP aggregates were formed due to binding of Ab-AuNP to Aβ fibrils that possess multiple antibody binding sites. Aβ antibody was immobilized on the AuNP surface via non-covalent interaction. The amount of immobilized antibody was estimated with dot blot assay (Fig. S1, Supporting Information), and was calculated to be 128 molecules per particle.

Comparison of DFM with color change and UV-VIS spectroscopy in the detection of Aβ fibrils

First, in order to confirm the advantage of DFM for the sensitive detection of Aβ fibrils, the Ab-AuNP sample incubated with 5 μM Aβ fibrils was observed by the naked eye, UV-VIS spectroscopy and DFM (Fig. 2). Aβ monomer was used as a

Fig. 1　Detection of Aβ fibrils with DFM using antibody-modified gold nanoparticles (Ab-AuNPs). Aβ fibrils were added into the gold nanoparticles modified with Aβ antibody (Ab-AuNP) solution. Ab-AuNP aggregates induced by Aβ fibrils were observed with dark field microscopy (DFM) at the single particle level. The scattered light intensity of each spot in the DFM image was quantified.

ANALYTICAL SCIENCES MARCH 2016, VOL. 32 309

negative control. As shown in the figure, no obvious color change can be observed by the naked eye for the Ab-AuNP sample incubated with 5 μM Aβ fibrils (Fig. 2a). The UV-VIS spectrum also did not show a visible change with the addition of 5 μM Aβ fibrils (Fig. 2b). On the contrary, the DFM image of Ab-AuNP incubated with 5 μM fibrils showed brighter spots than samples without Aβ samples, due to aggregation of AuNPs caused by Aβ fibrils. Importantly, Ab-AuNP incubated with Aβ monomer showed a similar DFM image as the Ab-AuNP only sample, supporting Aβ fibrils-specific detection. These results indicate that DFM would be able to detect Aβ fibrils-induced AuNP aggregates with higher sensitivity than color change and UV-VIS spectroscopy.

LOD of DFM method for the detection of Aβ fibrilsTo determine the detection limit of DFM, various

concentrations of Aβ fibrils (from 0 to 5 μM) were added to Ab-AuNP solution, and Ab-AuNP aggregates were analyzed at the single-particle level. The intensities of the spots in the DFM images were used to obtain corresponding histograms (Fig. 3a). As shown in the figure, larger AuNP aggregates with higher intensities were observed when an increased amount of Aβ fibrils was added.

The detection limit of the Aβ fibrils concentration was estimated from the ratio of aggregates at each concentration.

As shown in Fig. 3b, the ratio of aggregates increased at higher concentrations of Aβ fibrils. The LOD was evaluated by the 3σ criterion method, and was determined to be as low as 40 pM (corresponding to 0.24 femtomol in a 6-μL sample volume).

This detection limit is better than the fibril detection method using ThT dye where sub-μM amounts of samples are necessary.5 Another advantage of this method is that fibrils can be distinguished from the monomer. Although Aβ samples can be sensitively detected with ELISA with a detection limit in the pM range, it cannot distinguish Aβ aggregates from Aβ monomer. Thus, our method is advantageous for specific sensitive detection of Aβ aggregates. Since AuNP aggregates are formed by binding of the antibody immobilized on the AuNP surface to Aβ fibrils, removal of factors that inhibit the interaction would be necessary for application to biological samples such as blood and cerebrospinal fluid. Since it was shown that pretreatments of serum including removal of excess albumin protein and inactivation of proteases enhanced immunoassay sensitivity,15 such treatments would also be important for application of our method to biological samples.

Conclusion

DFM was employed to detect Aβ fibrils-induced AuNP

Fig. 2　Comparison of color change, UV-VIS spectroscopy and DFM in the detection of Aβ fibrils. Aβ fibrils or Aβ monomer (5 μM) was added to 30 pM Ab-AuNP, followed by incubation for 1 h at room temperature. Samples were estimated with the naked eye (a), UV-VIS spectroscopy (b) and DFM (c). The scale bar = 20 μm.

310 ANALYTICAL SCIENCES MARCH 2016, VOL. 32

aggregation at the single-particle level. The detection sensitivity of Aβ fibrils reached as low as the pM level. This study further proved the potential application of DFM in the analytical methods based on AuNP aggregations.

Acknowledgements

We thank Dr. Kazuo Hosokawa (Bioengineering Lab., RIKEN) for help with LOD calculation. T. B. is a RIKEN International Program Associate. This work was financially supported by RIKEN (Molecular Systems Research) and JSPS KAKENHI (Grant Nos. 24570143 and 25286015).

Supporting Information

Dot blot assay for estimation of the amount of Aβ antibody immobilized on surface of AuNPs. This material is available free of charge on the Web at http://www.jsac.or.jp/analsci/.

References

1. F. Chiti and C. M. Dobson, Annu. Rev. Biochem., 2006, 75, 333.

2. K. Blennow, M. J. de Leon, and H. Zetterberg, Lancet, 2006, 368, 387.

3. B. A. Yankner, L. K. Duffy, and D. A. Kirschner, Science,

Fig. 3　Detection of Aβ fibrils with DFM. (a) Intensity histogram of AuNP aggregates observed by DFM caused by adding different concentrations of the Aβ fibrils ranging from 0 to 5 μM (number of spots >200). Ab-AuNP without addition of Aβ sample was used as a negative control sample (background). (b) Ratio of aggregated AuNPs at various concentrations of Aβ fibrils. Ab-AuNP without addition of Aβ fibrils was used as a negative control sample (triangle, background). The error bars represent standard deviations (SD) of three different histograms. The “3σ line”, indicating background (zero-concentration, aggregate ratio = 0.076 ± 0.028, n = 6) + 3SD, was used for calculation of the LOD.

ANALYTICAL SCIENCES MARCH 2016, VOL. 32 311

1990, 250, 279. 4. T. Zako and M. Maeda, Biomater. Sci., 2014, 2, 951. 5. R. Eisert, L. Felau, and L. R. Brown, Anal. Biochem., 2006,

353, 144. 6. S. A. Hudson, H. Ecroyd, T. W. Kee, and J. A. Carver,

FEBS J., 2009, 276, 5960. 7. S. A. Gravina, L. Ho, C. B. Eckman, K. E. Long, L. Otvos,

Jr., L. H. Younkin, N. Suzuki, and S. G. Younkin, J. Biol. Chem., 1995, 270, 7013.

8. M. Sakono, T. Zako, and M. Maeda, Anal. Sci., 2012, 28, 73.

9. C. L. Nehl, N. K. Grady, G. P. Goodrich, F. Tam, N. J. Halas, and J. H. Hafner, Nano Lett., 2004, 4, 2355.

10. J. A. Fan, K. Bao, J. B. Lassiter, J. Bao, N. J. Halas, P.

Nordlander, and F. Capasso, Nano Lett., 2012, 12, 2817. 11. T. Bu, T. Zako, M. Fujita, and M. Maeda, Chem. Commun.,

2013, 49, 7531. 12. K. M. Sörgjerd, T. Zako, M. Sakono, P. C. Stirling, M. R.

Leroux, T. Saito, P. Nilsson, M. Sekimoto, T. C. Saido, and M. Maeda, Biochemistry, 2013, 52, 3532.

13. N. Yamamoto, E. Matsubara, S. Maeda, H. Minagawa, A. Takashima, W. Maruyama, M. Michikawa, and K. Yanagisawa, J. Biol. Chem., 2007, 282, 2646.

14. H. Arata, H. Komatsu, K. Hosokawa, and M. Maeda, PLOS ONE, 2012, 7, e48329.

15. M. Ihara, A. Yoshikawa, Y. Wu, H. Takahashi, K. Mawatari, K. Shimura, K. Sato, T. Kitamori, and H. Ueda, Lab. Chip., 2010, 10, 92.