Surface-enhanced Raman scattering based lateral flow immunochromatographic assay for sensitive influenza detection

Abstract

A sensitive surface-enhanced Raman scattering-based lateral flow immunochromatography test system (SERS-LFIA) was developed using multi-branched gold nanostars (AuNS) coated with a Raman active molecule, 4-aminothiophenol (ATP), as a signal reporter. The unique characteristic of the AuNS having multi-arms and surface roughness features allows high SERS performance in the system to be achieved via the increased bioconjugation sites and hot spot regions. To demonstrate the system performance, influenza A nucleoprotein was used as the target molecule. After a facile one-step synthesis, AuNS was tagged with the ATP molecule and conjugated to the antibody specific to influenza A nucleoprotein, and then used as both the SERS signal reporter and the detection probe in the system. It was demonstrated that the lower limit of detection using visual and SERS detection was 67 and 6.7 ng mL⁻¹, respectively. With the SERS detection signal, the detection sensitivity was significantly improved approximately 37 and 300 fold, as compared with fluorescence-based and conventional LFIA, respectively. The integrated SERS-LFIA platform reported here demonstrated an ultrahigh detection sensitivity, which could potentially facilitate a sensitive point-of-care test for diagnosis in many diseases.

---

Introduction

Over the past few decades, there has been an increasing demand for the development of point-of-care testing devices with high sensitivity and specificity for diagnostics purposes and other applications. Among various point-of-care testing platforms, a membrane-based lateral flow immunochromatographic assay (LFIA) has emerged as a popular technology and a powerful tool for rapid screening due to its simplicity and flexibility for various applications. To date, most LFIA test interpretation is based upon color visualization using colloidal gold nanoparticles as reporters. However, the low analytical sensitivity of the system often limits its use in further applications requiring a high sensitive detection. Whilst, there are attempts using other enhanced sensitivity systems; for example, enzyme-amplified signal system,^1,2 fluorescence-based signal enhancement system,^3,4 magnetic nanoparticle-based system,^5–7 and near-infrared dye,⁸ as reporters, these systems may either suffer from optical interferences or require additional steps for signal measurement.

Recently, surface-enhanced Raman scattering (SERS) technique has received much attention in various analytical applications. Taking advantages that it provides ultrasensitive detection in a single molecular level and enable the detection through unique Raman fingerprints, SERS based LFIA for target analyte detection has been reported.^9–11 The combination of SERS and LFIA shows promises in overcoming low sensitivity problem of conventional LFIA and lack of precision of signal measurement from optical or fluorescence signal reader.¹² The technique exploits the enhancement of Raman scattering by molecular absorption of Raman-active reporter on specific metallic surfaces such as silver (Ag) and gold (Au).^13,14 Spherical gold and silver nanoparticles are often employed as materials used for the enhancement of SERS.^15,16 Whilst the extinction coefficient of the surface plasmon bands produced by silver nanoparticles is four-time greater than that of gold nanoparticles, silver nanoparticles suffer from low stability under biological conditions. It was reported that, anisotropic geometries of gold nanoparticles like triangles, cubes, rods, and multi-branches such as stars or urchin provide advantages in producing concentrated electromagnetic field at the tips of anisotropic particles, leading to stronger SERS activity relative to spherical ones.¹⁷ Amongst them, AuNS was dominant due to their tunable surface plasmon and multiple sharp branches.¹⁸ AuNS was applied to various substrates in order to produce the SERS active substrates for biosensing applications, for instance; paper-based SERS active substrate for detection of the oxidation products of apomorphine;¹⁹ gold nanostars immobilized on a gold substrate for detection of 4-mercaptobenzoic acid.²⁰ With the application of AuNS in the LFIA system, previous works using urchin-like gold nanoparticles as the reporter in LFIA were reported. However, the system only reported the application in visual detection without SERS signal measurement.^21,22

Herein, we report a novel SERS-LFIA integration system using AuNS coated with a Raman active molecule as a SERS signal reporter for influenza A nucleoprotein detection. The system could provide both qualitative and quantitative analysis of target analyte. With the combination of LFIA and SERS techniques, ultrasensitive detection and simultaneous separation of the target analyte from biological matrices could be achieved.

Experimental

Materials and instruments

Gold(III) chloride trihydrate (HAuCl₄·3H₂O), trisodium citrate, hydroquinone, 4-aminothiophenol (ATP), bovine serum albumin (BSA), boric acid, sodium tetraborate, sodium azide and sodium chloride (NaCl) were purchased from Sigma-Aldrich (St. Louis, MO). Hydrochloric acid and ethanol were purchased from Carlo Erba. Nitric acid (HNO₃) was purchased from J.T. Baker. All reagent was used as received without further purification. All aqueous solutions were prepared with ultrapure water from a Milli-Q system (Millipore). All other chemical reagents used in this study were of molecular grade and used without further purification. All glassware was cleaned with aqua regia before experiments.

Monoclonal antibody specific to influenza A nucleoprotein (mAb) was purchased from Innova Biotechnology (Bangkok, Thailand). Goat anti-mouse immunoglobulin (IgG) antibody was supplied by KPL (MD, USA). Recombinant nucleoprotein (rNP) used in this study was generated and characterized as described previously.²³ Allantoic fluid infected with influenza A (H1N1)pdm09 virus was prepared according to a method described earlier.²⁴ The LFIA components including a sample pad, a conjugate pad, a nitrocellulose membrane and an absorbent pad were supplied from GE Healthcare (Buckinghamshire, UK). The CM4000 Guillotine cutting module and the dispensing platform were obtained from BioDot (Irvine, USA).

Preparation of gold nanostar (AuNS)

Spherical gold nanoparticles with a diameter of 12 nm were produced by the standard Frens method using HAuCl₄·3H₂O and sodium citrate as reducing agent.²⁵ The resulting particles were used as seeds in the synthesis process of AuNS produced by a modified protocol reported by Perrault and Chan.²⁶ Briefly, an aliquot of 400 μL of a seed solution, described above, was diluted in a 50 mL freshly prepared ultrafiltered water, followed by the addition of 1 mL of a 10 mg mL⁻¹ gold(III) chloride trihydrate solution under magnetic stirring at room temperature. Then, 1 mL of 11 mg mL⁻¹ hydroquinone solution (in water) was fast added, resulting in the appearance of dark greenish-blue color indicating the growth of AuNS. After 5 minutes of preparation, particles were then stabilized by addition of a 300 μL of 10 mg mL⁻¹ trisodium citrate solution. Then, the AuNS was washed by centrifugation at 7500 rpm for 15 minutes and resuspended in a 1% sodium citrate solution.

Preparation of SERS tags (AuNS–ATP) and antibody conjugation (AuNS–ATP–mAb NP)

Preparation of the SERS tags and the antibody conjugation are depicted in Fig. 1A. The SERS tags were prepared by attaching a Raman reporter molecule, 4-aminothiophenol (ATP), on the AuNS surface. Briefly, 1 mL of AuNS colloid was centrifuged at 10 000 rpm for 5 minutes and resuspended in 1 mL of ethanol. Then, 5 μL of ethanoic solution of 0.1 M ATP was added to the AuNS suspension with gentle mix at room temperature for 1 h. The ATP coated AuNS were separated from solution by centrifugation at 10 000 rpm and redispersed in 900 μL of 20 mM borate buffer pH 9 for further antibody conjugation.

Fig. 1 Schematic illustration of the SERS-LFIA test system, including (A) the preparation of the AuNS–ATP–mAb SERS tags, (B) the assembly of the LFIA test strip, and the principle of the SERS-LFIA test system in (C) the presence and (D) absence of influenza A nucleoprotein.

The mAb specific to nucleoprotein (NP) was conjugated to the AuNS–ATP in a 20 mM borate buffer (pH 9) using a physical adsorption method, as described previously.²⁷ Briefly, 200 μL of 0.1 mg mL⁻¹ of the mAb in 20 mM borate buffer (pH 9) was mixed with 900 μL of AuNS–ATP (optical density, O.D. at 598 nm = 1). This was followed by the addition of 100 μL of 20 mM borate buffer, pH 9, and 15 min incubation step. Then, 100 μL of 10 mg mL⁻¹ BSA in 20 mM borate buffer (pH 9) was added to the solution as a blocking solution and incubated for 15 min. The AuNS–ATP–mAb NP conjugates were then collected by centrifugation at 10 000 rpm for 5 min at 4 °C and resuspended in 45 μL of 50 mM Tris–HCl buffer, pH 8.2 containing 1% (w/v) BSA and 1% (w/v) NaN₃.

Characterization of the AuNS and the AuNS conjugates

The optical response of nanoparticle was measured using a microplate spectrophotometer (BioTek Instrument, USA). The morphology and structure the particles were determined through with transmission electron microscope operated at 200 kV (TEM, JEM 2100, JEOL Ltd, Japan). Images of SERS tag in the paper fiber pores of the LFIA membrane was obtained by using field-emission scanning electron microscope (FE-SEM, SU5000, Hitachi, Japan). An uncoated specimen was characterized at 10 kV using backscattered-electron (BSE) mode.²⁸ Hydrodynamic diameter (number weighted) and zeta potential during the encoding process were investigated by dynamic light scattering using a Zeta Nanosizer (Malvern Instruments, UK). SERS signals of the nanoparticles in solution phase were measured using a portable Raman spectroscopy (AvaRaman, Avantes, Netherlands) with a 25 mW laser at 785 nm, and the signal acquisition time of 10 s.

Fabrication of SERS-LFIA test system

The LFIA test strip was composed of 4 components including a sample pad, a conjugate pad, a fluid-flow nitrocellulose membrane, and an absorbent pad (Fig. 1B). Firstly, the fluid-flow membrane was prepared by immobilizing the mAb (1 mg mL⁻¹) and secondary antibody (1 mg mL⁻¹) at the area defined as the test (T) and control (C) lines, respectively. The conjugate pad was prepared by dispensing a desired volume of the AuNS–ATP–mAb NP conjugates. After dried at 25 °C, all components of the test strip were assembled on a plastic backing card and a cassette, respectively.

Assay procedure and SERS-LFIA measurement method

The assay was performed by applying a 100 μL sample liquid on a sample pad of the test strip. After 15 minutes, the signal visualization was performed. The presence of both T and C lines indicates a positive result, whilst the presence of the C line only indicates the negative result. For SERS-LFIA measurement method, SERS spectral measurements of the T and C lines were performed using a Raman spectrometer (NT-MDT, Russia) equipped with an inverted confocal microscope (Olympus IX71, Olympus, USA). The system was connected with a CCD detector cooled at −60 °C. A He–Ne laser operating at λ = 633 nm was used as an excitation source with a power of 3.3 mW. The laser light was coupled through 100× objective lens with a laser spot size at 333 nm. The system was calibrated with Raman signal from a standard silicon wafer at 520 cm⁻¹. The SERS spectra were obtained in a range of 500–2000 cm⁻¹ with the exposure time of 10 s and 12 accumulations.

Performance evaluation and specificity

To evaluate the sensitivity of the test system, various concentrations of the rNP (0–6700 ng mL⁻¹) was used as the antigen. The volume of 100 μL of the prepared rNP samples were applied directly to the sample pad of the test strip and the signal was measured as described previously. The lower limit of detection was defined as the lowest concentration of the rNP producing positive T line signal by both visual and SERS detection. Borate buffer (20 mM, pH 9) was used as a negative control. Other proteins including amicase, avidin, BSA, lysozyme, tryptone, and other closely related viruses such as influenza B, and canine distemper virus (CDV) (1250 ng mL⁻¹) were used to evaluate the specificity of the system. To assess the potential application of the system in detection the nucleoprotein in biological matrices, allantoic fluid containing the influenza A (H1N1)pdm09 virus at 5.6 × 10³ TCID₅₀ mL⁻¹ (50% tissue culture infectious dose) was used. A volume of 100 μL of the prepared allantoic fluid was applied directly to the test and processes in the same manner, as described previously.

Results and discussion

Principle of the SERS-LFIA test system

The principle of the SERS-LFIA test system for influenza nucleoprotein detection is illustrated in Fig. 1C and D. The test system is relied upon the combination of the principle of chromatography in separating the target analyte from sample matrices and the high sensitivity of SERS technique. When the sample containing the target protein was dispensed on the test system, the sample liquid moved along to the conjugate pad, where the specific recognition between the nucleoprotein and the conjugates was taken place. The nucleoprotein–conjugate complex continued to migrate along the test system through capillary action and captured by the immobilized the mAb on the T line, forming the sandwich immunocomplex accumulated at the T line. The excess conjugates continued to migrate along the membrane and were captured by the secondary antibody on the C line. After the assay was completed, SERS spectral measurements of the T and C lines were performed using a Raman spectrometer (Fig. 1C). In the absence of the nucleoprotein, only the signal from the C line was observed, validating the test system (Fig. 1D). Quantitative analysis was performed using the Raman intensity obtained from T line.

Characterization of AuNS and SERS tags

AuNS was synthesized successfully by an aqueous synthesis through a reduction of gold salt via a seed-mediated growth approach using hydroquinone as a reducing agent. Hydroquinone was selected because it can induce an epitaxial growth of branches on gold surface and provide spherical shape of AuNS with short branches.¹⁷ This configuration of AuNS could reduce hindrance effect occurred from antibody–antigen binding and allow AuNS immunocomplex flew along the LFIA membrane, while remained electromagnetic field enhancement property for SERS applications at the tip of branched particles. In addition, based on high surface to volume ratio of their nanosizes, AuNS could be a good candidate as reporter for SERS-LFIA platform. The TEM image of multi-branched gold nanostar with average crystal size of 85 nm was shown in Fig. 2A. The inset in Fig. 2A showed the photograph of the greenish-blue colored of AuNS. The hydrodynamic diameter and zeta potential of AuNS after surface modification were demonstrated in Fig. 2B. The zeta potential in Fig. 2B demonstrated the highly negative charge on AuNS surface due to citrate coverage. After coating the AuNS surface with a Raman reporter (ATP) through thiol–gold interaction, the hydrodynamic size was slightly changed. However, the surface charge of the SERS tags (AuNS–ATP) became much less negative due to the replacement of citrate by ATP molecules. In this study, mAb was immobilized on the SERS tags by physisorption and the successful conjugation was confirmed by significant change in hydrodynamic size as seen in Fig. 2B. The as-prepared AuNS showed the maximum surface plasmon absorption peaks at 598 nm, while the SERS tags and conjugates showed slightly plasmon shift to 608, and 616 nm, respectively (Fig. 2C). It is worth noting that the plasmon absorption of the conjugates at 616 nm overlapping with the laser excitation wavelength at 633 nm can facilitate the maximum SERS enhancement by resonance effect. In order to prove that the particles can be used as a SERS reporter in the SERS-LFIA test system, the solutions of AuNS, AuNS–ATP, and AuNS–ATP–mAb was subjected to Raman spectroscopy as demonstrated in Fig. 2D. The Raman spectrum of AuNS–ATP was dominated with the A1 vibrational modes (in-plane, in-phase modes) at 1086 (ν(CS)) and 1588 (ν(CC)) cm⁻¹ corresponding to the Raman spectra of ATP with slight shift.²⁹ After antibody conjugation, the AuNS–ATP–mAb SERS tags still showed characteristic peaks of ATP with relatively strong signal. The overall characterization result demonstrated a successful conjugation of ATP molecules and antibodies onto the surface of AuNS. Thus, the SERS tag conjugates can be further used as a reporter for the SERS-LFIA system.

Fig. 2 (A) TEM image of the AuNS with the scale bar of 100 nm. Inset demonstrated a photograph of AuNS solution. (B) A table indicated hydrodynamic size and zeta potential, (C) surface plasmon absorption, and (D) Raman spectra of individual step of the SERS tags assembly. (i) AuNS; (ii) 4-ATP coated on AuNS (AuNS–ATP); (iii) AuNS–ATP–mAb SERS tags.

Detection of target analyte using the SERS-LFIA test system

Herein, the SERS-LFIA platform for analysis of nucleoprotein of influenza A was demonstrated. By applying a sample containing the target analyte, the rNP was reacted with AuNS–ATP–mAb SERS tags and captured by the immobilized mAb at the T line forming the sandwich mAb–rNP–conjugate complex accumulated at the test zone. Fig. 3 illustrated SEM images obtained for AuNS immunocomplex in the test area of LFIA using backscattered-electron detector. At the presence of rNP target antigen (670 ng mL⁻¹), AuNSs were observed as small bright spots in the paper fiber pores of the membrane due to the accumulation of the sandwich mAb–nucleoprotein–conjugate complexes at the T line. This observation was corresponding to the dark grey color visualized on the T line, as well as the SERS activity of the substrate (Fig. 3A). In the absence of rNP target antigen (negative test), no color band was appeared at the T line and negligible bright spot was observed in the paper fiber from SEM image. In this case, no SERS activity was detected on the substrate either. All these observations demonstrated that rNP target antigen could be qualitatively detected by naked eyes and quantitatively detected by monitoring the characteristic SERS peak of the SERS tag.

Fig. 3 Typical photographic image of SERS spectra and SEM images of SERS-LFIA test strip (T line area) in (A) the presence and (B) absence of the rNP (670 ng per test).

To evaluate the detection sensitivity of SERS-LFIA system, various concentrations of rNP solution ranging from 0–6700 ng mL⁻¹ were applied as samples. A photograph of SERS-LFIA strips loading different concentration of rNP were displayed in Fig. 4A. The LOD of colorimetric detection was demonstrated to be 67 ng mL⁻¹ as shown in Fig. 4A. Quantitative analysis was also conducted by measuring the Raman intensity at the test zones and the Raman spectra were shown in Fig. 4B. The relative Raman intensity was then plotted as a function of rNP concentration in Fig. 4C. The results showed a continuous increase in Raman shifts at 1086 and 1588 cm⁻¹ which are characteristic peaks of the SERS tags as the concentration of the rNP was raised due to the more sandwich immunocomplex were formed at the substrate. The limit of detection (LOD) of the system using SERS-based measurement was calculated to be 6.7 ng mL⁻¹ (0.67 ng of rNP per test) based upon the determination of standard deviation (signal to noise ratio (S/N) = 3). The sensitivity of this system was compared to previous studies of rNP detection using different platforms. For instance, a sandwich ELISA provided limit of detection at 4 μg mL⁻¹,³⁰ whereas electrochemical immunoassay reported LOD at 10 ng mL⁻¹.³¹ In addition, the fluorescence-based LFIA which was developed for sensitivity enhancement showed detection limit of 250 ng mL⁻¹ (ref. 24) whereas dual-layered and double-targeted nanogold based LFIA provided LOD at 4.7 × 10¹ TCID₅₀ mL⁻¹ in the influenza A (H1N1) infected MDCK cells.²⁷ Compared to the SERS-based half sandwich assay for the detection of influenza A (H5N1) virus like particle (VLP) which reported the LOD of 30 ng mL⁻¹,³² more sensitive and quantitative detection of rNP can be achieved using the SERS-LFIA system in this work. The comparative table of sensitive influenza detection with other strategies as listed in Table 1. The high sensitivity of this system can be attributed to high sensitivity of the SERS based technique especially when the Raman signal was enhanced by AuNS containing multi-branch feature which can generate multiple hot spots on their tips. With this SERS-LFIA system, highly sensitive quantification of target analytes is possible.

Fig. 4 Sensitivity evaluation using various concentration of rNP from 0–6700 ng mL⁻¹. (A) Signal visualization of the SERS-LFIA test strips with the limit of detection (LOD) of 67 ng mL⁻¹. (B) SERS spectra obtained from the T line after applying different concentrations of the rNP. The limit of detection (LOD) was 6.7 ng mL⁻¹ by determining the Raman shifts at 1086 and 1588 cm⁻¹. (C) Quantitative analysis of the SERS-LFIA for rNP detection by calculating relative Raman intensity at 1086 and 1588 cm⁻¹. Error bars represent standard deviations from three measurements.

Table 1 Comparison of the efficiency of some methods for sensitive influenza A detection

Detection method	Signal reporter	LOD	Target/sample type	Ref.
Sandwich ELISA	Enzymatic reaction	4 μg mL^−1	rNP and influenza A (H1N1) infected allantoic fluid	30
Electrochemical ELISA immunoassay	Enzymatic reaction and redox reaction of 4-aminophenol	10 ng mL^−1	rNP	31
Fluorescence-based LFIA	Silica dye doped nanoparticles	250 ng mL^−1	rNP and influenza A (H1N1) infected allantoic fluid	24
Dual combine antigen LFIA	15 and 40 nm spherical gold nanoparticles	4.7 × 10^1 TCID50 mL^−1	Influenza A (H1N1) infected MDCK cells	27
SERS based half-sandwich assay	40 nm spherical gold nanoparticles	30 ng mL^−1	Influenza A (H5N1) virus like particle (VLP)	32
SERS-LFIA	Gold nanostar	6.7 ng mL^−1	rNP and influenza A (H1N1) infected allantoic fluid	This work

---

The specificity of the SERS-LFIA was assessed using the target rNP (670 ng mL⁻¹) compared to other proteins, such as amicase, avidin, BSA, lysozyme, tryptone. Moreover, other closely related viruses such as influenza B, and canine distemper virus (CDV) are also included (1250 ng mL⁻¹). Visual and SERS signal detection were displayed in Fig. 5. As shown in Fig. 5A, two colored lines appeared in the presence of rNP (positive), whereas only one colored band was observed at the T line for other protein samples (negative). Based on normalized Raman intensity at the T lines (Fig. 5B), strong SERS signal at 1086 and 1588 cm⁻¹ was observed for the positive sample, whereas other proteins and viruses showed no significant Raman shift at background corresponding to the visual detection. The results indicated high specificity of the SERS-LFIA system against other proteins and viruses, which could interfere the detection of the target analyte in complex systems.

Fig. 5 Specificity of the SERS-LFIA. (A) The strip was successfully applied to detect rNP (670 ng mL⁻¹) as demonstrated by (B) strong Raman intensity while negative results was obtained from other proteins and viruses (1250 ng mL⁻¹).

Evaluation of SERS-LFIA test system using biological fluid

One of the challenges in developing the test system for rapid target detection is sample matrix interference. In order to evaluate the potential feasibility of the system for the detection of influenza nucleoprotein in complex biological matrices, infected allantoic fluid containing the influenza A (H1N1)pdm09 virus at 5.6 × 10³ TCID₅₀ mL⁻¹ was used as a sample. Under the optimized condition, it was demonstrated that the system was able to detect influenza A antigen in the allantoic fluid containing the virus by both visual and SERS detection (Fig. 6). Fig. 6A demonstrated a photograph of positive and negative strips and Fig. 6B showed the Raman spectra at the test zones. The Raman shifts at 1086 and 1588 cm⁻¹ were observed at the T line when the allantoic fluid containing the virus was applied as a sample. In contrast, no significant Raman shift was observed from the T line when only running buffer was used as a sample. In both cases, the Raman peaks at the C line at 1086 and 1588 cm⁻¹ can be observed, indicating the validity of the test system.

Fig. 6 Performance of SERS-LFIA in detecting nucleoprotein in the allantoic fluid infected with influenza A virus. (A) Photograph of the SERS-LFIA strip after applying infected fluid as a positive test and buffer as a negative test. (B) Raman spectra at the test lines of SERS-LFIA strip.

Conclusions

In the present study, a novel AuNS-based SERS-LFIA was developed for a sensitive detection of influenza A nucleoprotein. AuNS was used as a SERS detection probe in the LFIA platform instead of spherical gold nanoparticles in order to overcome the limitation of system in term of low sensitivity and quantitative analysis. The integrated system reported here demonstrated an ultrahigh detection sensitivity as compared with conventional methods. Together with portable Raman spectrometer, it is expected that our technique could potentially facilitate a sensitive point-of-care test for diagnosis in many diseases in the field.

Acknowledgements

The authors would like acknowledge the financial support from the National Nanotechnology Center (NANOTEC), National Science and Technology Development Agency (NSTDA), Thailand.

References

1. I.-L. Cho, A. Bhunia and J. Irudayaraj, Int. J. Food Microbiol., 2015, 206, 60–66.
2. X. Gao, L.-P. Xu, T. Wu, Y. Wen, X. Ma and X. Zhang, Talanta, 2016, 146, 648–654.
3. G. J. Worsley, S. L. Attree, J. E. Noble and A. M. Horgan, Biosens. Bioelectron., 2012, 34, 215–220.
4. L.-W. Song, Y.-B. Wang, L.-L. Fang, Y. Wu, L. Yang, J.-Y. Chen, S.-X. Ge, J. Zhang, Y.-Z. Xiong, X.-M. Deng, X.-P. Min, J. Zhang, P.-J. Chen, Q. Yuan and N.-S. Xia, Anal. Chem., 2015, 87, 5173–5180.
5. Q. Xu, H. Xu, H. Gu, J. Li, Y. Wang and M. Wei, Mater. Sci. Eng., C, 2009, 29, 702–707.
6. D.-B. Wang, B. Tian, Z.-P. Zhang, J.-Y. Deng, Z.-Q. Cui, R.-F. Yang, X.-Y. Wang, H.-P. Wei and X.-E. Zhang, Biosens. Bioelectron., 2013, 42, 661–667.
7. C. Liu, Q. Jia, C. Yang, R. Qiao, L. Jing, L. Wang, C. Xu and M. Gao, Anal. Chem., 2011, 83, 6778–6784.
8. C. Swanson and A. D'Andrea, Clin. Chem., 2013, 59, 641–648.
9. J. Baniukevic, I. H. Boyaci, A. G. Bozkurt, U. Tamer, A. Ramanavicius and A. Ramanaviciene, Biosens. Bioelectron., 2013, 43, 281–288.
10. Y. Wang, B. Yan and L. Chen, Chem. Rev., 2013, 113, 1391–1428.
11. J. Hwang, S. Lee and J. Choo, Nanoscale, 2016, 8, 11418–11425.
12. Y. Li, Y. Wang, J. Wang, Z. Tang, J. G. Pounds and Y. Lin, Anal. Chem., 2010, 82, 7008–7014.
13. X. Dou, T. Takama, Y. Yamaguchi and H. Yamamoto, Anal. Chem., 1997, 69, 1492–1495.
14. I. H. Cho, M. Das, P. Bhandari and J. Irudayaraj, Sens. Actuators, B, 2015, 213, 209–214.
15. X. Fu, Z. Cheng, J. Yu, P. Choo, L. Chen and J. Choo, Biosens. Bioelectron., 2016, 78, 530–537.
16. Q. Fu, H. L. Liu, Z. Wu, A. Liu, C. Yao, X. Li, W. Xiao, S. Yu, Z. Luo and Y. Tang, J. Nanobiotechnol., 2015, 13, 1–9.
17. J. Li, J. Wu, X. Zhang, Y. Li, D. Zhou, H. Sun, H. Zhang and B. Yang, J. Phys. Chem. C, 2011, 115, 3630–3637.
18. A. M. Fales, H. Yuan and T. Vo-Dinh, J. Phys. Chem. C, 2014, 118, 3708–3715.
19. D. Mehn, C. Morasso, R. Vanna, M. Bedoni, D. Prosperi and F. Gramatica, Vib. Spectrosc., 2013, 68, 45–50.
20. A. S. D. S. Indrasekara, S. Meyers, S. Shubeita, L. C. Feldman, T. Gustafsson and L. Fabris, Nanoscale, 2014, 6, 8891–8899.
21. W. Ren, Z. Huang, Y. Xu, Y. Li, Y. Ji and B. Su, Anal. Bioanal. Chem., 2015, 407, 7341–7348.
22. Y. Ji, M. Ren, Y. Li, Z. Huang, M. Shu, H. Yang, Y. Xiong and Y. Xu, Talanta, 2015, 142, 206–212.
23. K. Watcharatanyatip, S. Boonmoh, K. Chaichoun, T. Songserm, M. Woratanti and T. Dharakul, J. Virol. Methods, 2010, 163, 238–243.
24. S. Bamrungsap, C. Apiwat, W. Chantima, T. Dharakul and N. Wiriyachaiporn, Microchim. Acta, 2014, 181, 223–230.
25. G. Frens, Nature, 1973, 241, 20–22.
26. S. D. Perrault and W. C. W. Chan, J. Am. Chem. Soc., 2009, 131, 17042–17044.
27. N. Wiriyachaiporn, W. Maneeprakorn, C. Apiwat and T. Dharakul, Microchim. Acta, 2015, 182, 85–93.
28. U. Heinzmann, A. Reininger, R. Autrata and H. Hofler, Scanning, 1994, 16, 241–245.
29. J. Zheng, Y. Zhou, X. Li, Y. Ji, T. Lu and R. Gu, Langmuir, 2003, 19, 632–636.
30. P. Jian-umpunkul, C. Thepthai, N. Apiwat, W. Chantima, K. Poomputsa, N. Wiriyachaiporn and T. Dharakul, J. Virol. Methods, 2012, 185, 24–31.
31. K. Ohtsuka, H. Endo, K. Morimoto, B. N. Vuong, H. Ogawa, K. Imai and S. Takenaka, Anal. Sci., 2008, 24, 1619–1622.
32. Y. J. Lin, C.-Y. Wu, T. Li, P.-W. Hsiao and D.-K. Chang, J. Biosens. Bioelectron., 2014, 5 DOI:10.4172/2155-6210.1000150.

---

Footnote

† These authors contributed equally to this work.

---