As featured in: Showcasing research from Prof. Horn-Jiunn Sheen's laboratory, Institute of Applied Mechanics, National Taiwan University, Taiwan and Prof. Pei-Kuen Wei's laboratory, Research Center for Applied Sciences, Academia Sinica, Taiwan.

LMP1 gene detection using a capped gold nanowire array surface plasmon resonance sensor in a microfl uidic chip

A new detection device by using SPR nanowire array chip and a microfl uidics system was developed. A simple, low-cost and reproducible SPR nanowire chip with a visible light source displayed real-time detection capability. The test results demonstrated that the device has high sensitivity of detecting amplifi ed LMP1 DNA products without labeling or DNA modifi cation. Furthermore, this device can detect DNA products with lower amplifi cation cycles than agarose gel electrophoresis.

Registered charity number: 207890

rsc.li/analyst

See Yu-Jui Fan, Horn-Jiunn Sheen et al. , Analyst , 2020, 145 , 52.

Analystrsc.li/analyst

Volume 145Number 17 January 2020Pages 1-298

ISSN 0003-2654

PAPER Zhiping Zhang et al. Metal salt assisted electrospray ionization mass spectrometry for the soft ionization of GAP polymers in negative ion mode

Analyst

PAPER

Cite this: Analyst, 2020, 145, 52

Received 27th July 2019,Accepted 3rd November 2019

DOI: 10.1039/c9an01419e

rsc.li/analyst

LMP1 gene detection using a capped goldnanowire array surface plasmon resonancesensor in a microfluidic chip

Chih-Shen Chuang, a,b Chieh-Ying Wu,a Po-Han Juan,a Nai-Cheng Hou,a

Yu-Jui Fan,*b,c,d Pei-Kuen Weie and Horn-Jiunn Sheen *a

Surface plasmon resonance (SPR) nanowire array chips with a microfluidic system are an effective detec-

tion method for a rapid test device. This study investigated a capped gold nanowire array and a microflui-

dic test platform to provide a fundamental understanding of the kinetic binding of SPR nanowires and the

surface gold refractive index. The device sensitivity of the SPR nanowire array was 485 nm RIU−1 and the

detection limit was 4.1 × 10−5 RIU. Moreover, a kinetic binding analysis also indicated that a peak shift

resulted from a specific hybridization of the target molecule with the immobilized probe on the gold

nanostructures. The peak shift (red-shift) value of latent membrane protein 1 (LMP1) DNA was 2.21 nm.

The results demonstrated that this new method had high sensitivity to detect amplified DNA products

without labeling or complex sample treatment. The SPR nanowire chip can detect the PCR products at

lower cycle numbers compared to gel electrophoresis due to probe and DNA specificity. Furthermore,

the mechanisms of SPR nanowire array fabrication and the detection of the LMP1 gene were studied. The

findings can assist in improving the biosensing of DNA-amplified products and in developing rapid detec-

tion devices with a small-footprint nanostructured SPR chip.

Introduction

Nasopharyngeal carcinoma (NPC) is one of the epithelialcarcinomas characterized by the additional co-expression oflatent membrane proteins (LMPs, i.e., LMP1, LMP-2A, andLMP-2B).1,2 Standard methods used to detect LMPs or theEpstein–Barr virus (EBV) are polymerase chain reaction (PCR)and real-time (rt) PCR.3–7 A rapid, label-free and highly sensi-tive detection method for LMPs would provide a more efficientand economical way for the diagnosis of NPC. Lab-on-a-chip(LOC) with a microfluidic system and surface plasmon reso-nance (SPR) are two effective methods for the development ofnext-generation testing devices.8 Recently, cost-effectiveness

and measurement sensitivity have been the two significantchallenges for the development of next-generation testingdevices.8 On the other hand, the cost and handling time ofsample pretreatment are also a challenge for rapid detectiondevices. Kranzfelder et al.9 pointed out that NucleoSpin®Tissue XS kit, DNeasy® Blood and Tissue kit, andQuickExtrat™ DNA extraction solution can provide goodresults in isolating DNA. The extraction time of QuickExtrat™DNA extraction solution is less than 5 min, and this kit is easyto handle.9 Therefore, the measurement sensitivity and limitare the important challenges for a rapid detection device.

SPR is a resonant oscillation of electrons at the metal anddielectric interface induced by an incident electromagneticwave. In the past two decades, SPR sensing was found to haveadvantages of real-time and label-free measurements of bio-molecular interactions on metallic surfaces, especially goldsurfaces.7,10 The widely used method utilizes an optical prismwhich can induce SPR on a flat gold surface, known as theKretschmann configuration. The prism coupling method canalso be used to excite metallic nanostructures.11–14 However,the prism coupling method needs a precise control of the inci-dent angle and is too complicated to be integrated with chip-based microfluidic devices.

Recently, transparent types of SPR sensors using extraordi-nary optical transmission (EOT) in a periodic gold nanohole

aInstitute of Applied Mechanics, National Taiwan University, 1 Roosevelt Rd.,

Section 4, Taipei 10617, Taiwan. E-mail: sheenh@ntu.edu.tw;

Fax: +886-2-3366-5633; Tel: +886-2-3366-5632bSchool of Biomedical Engineering, Taipei Medical University, 250 Wuxing St.,

Taipei 11031, Taiwan. E-mail: ray.yj.fan@tmu.edu.tw;

Tel: +886-2-2736-1661, ext.7722cInternational PhD Program for Biomedical Engineering, Taipei Medical University,

250 Wuxing St., Taipei 11031, TaiwandGraduate Institute of Biomedical Optomechatronics, Taipei Medical University,

250 Wuxing St., Taipei 11031, TaiwaneResearch Center for Applied Sciences, Academia Sinica, 128 Academia Road,

Section 2, Nankang, Taipei 11529, Taiwan

52 | Analyst, 2020, 145, 52–60 This journal is © The Royal Society of Chemistry 2020

array or nanoslit to generate SPR have been used for biosen-sing applications. This method takes advantage of normalincident light and chip-based configuration without the use ofan optical prism. Biosensors based on extraordinary trans-missions of periodic nanohole arrays and nanoslit arrays ingold have been proposed15–20 and used in medical diagnostics,environmental monitoring, and food safety.21–23 Differentfrom localized surface plasmons (LSPs), which are directlyexcited by the incident light into metallic nanoparticles, theSPR in a metallic hole array or nanoslit array is based on thecoupling of LSPs in nanoapertures and Bloch-wave surfaceplasmon polariton (BW-SPP) on a periodic metallic surface.The coupling often forms the Fano-like resonance and canenhance the detection of biomolecules due to its sharper reso-nance profile.23 Periodic gold-nanostructure sensors takeadvantage of a chip-based and straightforward optical setup.They are easily incorporated into microfluidic devices. The bio-molecular interaction causes a change of the refractive indexon the metal surface, which can be measured by recording theresonance change in the transmission spectrum.24,25

Compared to the reflective-type SPR sensors, transparent-typeSPR sensors benefit from the normal incidence that is capableof being integrated with multiplex microfluidic devices. Theyalso provide a feasible way to achieve chip-based, high-throughput, label-free detection of modern DNA and proteinmicroarrays.

In recent years, several researchers have developed SPRmicrofluidic devices for microRNA,26–28 protein,29,30 methyl-ation DNA,31,32 and PCR detection.33,34 However, the SPR fabri-cation process with the Fano-like resonance and the detectionof cancer DNA are rarely elucidated. Furthermore, the inte-gration of SPR with LOC-based microfluidic devices for detect-ing different amplification cycles of cancer DNA remainsunclear. DNAs are small biomolecules; therefore the change ofthe refractive index by a small amount of DNA is very small.The non-specific binding of other substrate biomoleculeswould lead to a false-positive result. PCR amplification oftarget DNA can significantly enrich the amount of DNA andthus enhance the specificity of the detection. In this study, thecapped gold nanowire array-based SPR and a microfluidic testplatform were evaluated. Furthermore, PCR amplification pro-ducts of LMP1 DNA were investigated to provide a fundamentalunderstanding of the kinetic binding analysis of the SPR nano-wire array and the surface gold refractive index. The microflui-dic SPR platform is a label-free biosensor with high detectionsensitivity. Furthermore, low-cost nanostructures can beachieved by using an automatic nanoimprint technique,35

which makes the SPR-microfluidic devices feasible for variousrapid detection applications.

ExperimentalFabrication of a capped gold nanowire SPR chip

An automatic hot-embossing nanoimprinting machine wasused in this study to fabricate the capped gold nanostructure.

This fabrication method followed the method used by Leeet al.35 and Lee et al.36 Fig. 1(a) and (b) show the fabricationprocess. The nanowire structure was fabricated using a metalmold for rapid replication. A periodic nanowire structure wasdefined and fabricated on a silicon wafer by E-beam lithogra-phy and reactive ion etching. Then, the structure was trans-ferred onto a nickel–cobalt alloy mold using the electrodeposi-tion method. For the rapid hot-embossing method, the Ni–Comold is essential. It has low adhesion to plastic films and canwithstand the hitting force during the rapid stamping process.The nanowire structures were produced on a polycarbonate(PC) film using an automatic hot-embossing nanoimprintingmachine. After imprinting the nanowire structure into aplastic film, the metal mold was peeled off from the replicatedplastic film without lowering the temperature of the mold. Thetime needed for a single stamping process was only 5 s.35 Thenanowire structure can be rapidly replicated, and the mold canbe used repeatedly. Polycarbonate (PC) was adopted as the sub-strate material of our sensor owing to its low cost and excellentthermal properties for nanoimprinting fabrication. The nano-structure on the Ni–Co mold was transferred onto a PC film byhot-embossing nanoimprinting lithography, at a temperatureof 165 °C and a working pressure of 130 psi. This temperatureis much higher than the Tg of PC (135 °C), to rapidly heat thesurface of the plastic film. Gold was deposited onto the nano-wire structure of PC using a DC sputter, the deposition timewas 70 s, and the thickness of gold was 50 nm.

In this study, the nanowire array SPR is based on the coup-ling of LSPs in nanoapertures and Bloch-wave surface plasmonpolariton (BW-SPP) on a periodic metallic surface. There weresharp Fano resonances in the spectra. The coupling oftenforms the Fano-like resonance, as shown in Fig. 1(c). The wave-length sensitivity is comparable with most metallic nano-structure-based sensors using SPR or localized surfaceplasmon resonance (LSPR). It is noted that the Fano resonancehad a very sharp resonance slope. Therefore, the intensity sen-sitivity (intensity change at a fixed wavelength under a refrac-tive index change) will be much higher compared to conven-tional SPR sensors. We have checked the resonance quality ofthe SPR chip. The chips were well reproduced as shown inrevised Fig. 1(c). The variations of the resonance wavelengthare within 1% for different chips (N = 6). The experimentalresults of functional reproducibility indicated that theSPR nanowire chip has stable production and functionalreproducibility.

A scanning electron microscope (Hitachi S-4800, Japan) wasused to evaluate the nanowire width and interval. The voltageof the scanning electron microscope was 5 kV. The SEM imageof the SPR chip is shown in Fig. 1(d). The experimental resultsindicated that the interval of the nanowire was 500 nm, andthe nanowire width was 97.5 nm. The surface sensitivity of thegold nanostructure array is optimized using the structureswith a 500 nm interval.19 On the other hand, the width of thenanostructure plays an essential role in the resonance line-width.18 The SPR of the fabricated capped gold nanowire arraywith an interval of 500 nm was manifested as a transmission

Analyst Paper

This journal is © The Royal Society of Chemistry 2020 Analyst, 2020, 145, 52–60 | 53

spectrum in a wavelength range of 670–690 nm when distilledwater was in contact with the gold film.

Fabrication of the integrated microfluidic chip

In this experiment, a microfluidic device combined with thetransmission (t)-SPR detection technique was used to detectthe LMP1 gene. The disposable microfluidic device is a single-layered structure and was fabricated using polydimethyl-siloxane (PDMS), as shown in Fig. 2. The microfluidic channel

was integrated with the gold capped nanowire array as the bio-sensor. The Fano resonance is a result of two-mode couplingbetween the LSP resonance in gold nanowires and the BW-SPPmode on the periodic nanowire structure. The incident lightwas normally incident on the sample surface. The trans-mission light was filtered with a linear polarizer with its polar-ization direction perpendicular to the nanowires because theSPR can only be excited by the transverse magnetic (TM) wave.The TM light was collected through a collimator and passed

Fig. 1 (a) The gold coating nanowire structure of a surface plasmon resonance (SPR) chip fabricated with a nanoimprinting technique, (b) the hot-melting machine for the rapid fabrication on the PC film, (c) the functional reproducibility of the Fano resonance and the nanowire structure, and (d)the scanning electron microscopy (SEM) image of the surface plasmon resonance (SPR) chip using periodic nanowires. The nanowire width is97.5 nm and the period is 500 nm.

Paper Analyst

54 | Analyst, 2020, 145, 52–60 This journal is © The Royal Society of Chemistry 2020

through a 230 μm optical fiber, which was connected to aspectrometer (model: USB 4000, Ocean Optics, FL, USA) andthe optical resolution was 0.3 nm by using a low-cost device.The optical setup and the structure of the microfluidic chipare shown in Fig. 2(a) and (b), respectively. Furthermore, theSPR gold surface needs to be modified to ensure good bindingto the probe. In this study, cysteamine was used to modify theSPR gold surface. Immerse 100 µM cysteamine on SPR surfacefor 2 h. The melting temperature (Tm) of the probe used in thisstudy was 64 °C. This process is shown in Fig. 2(c). The probewas used to react with DNA, and then the SPR chip could beeasily detected.

Optical setup of the nanowire SPR chip and the test platformassembly

All transmission SPR measurements were carried out by usingthe custom-built transmission SPR system. The spectralmeasurement parameters were controlled using a custom-designed LabVIEW program (National Instrument, USA) pro-vided by the manufacturer, and each spectrum recorded con-sisted of an average of 100 individual spectra (integration time:

6000 ms), which were further processed into transmissionspectra using

T ð%Þ ¼ ðSλ � DλÞ=ðRλ � DλÞ � 100%; ð1Þ

where Sλ, Dλ, and Rλ are the sample spectrum, dark spectrum,and light source reference spectrum, respectively. The custom-designed LabVIEW software is packaged in an executableapplication. To analyze the Fano peak spectral changes overtime, the software identifies the Fano peak of the experimentby the spectral centroid method. This method uses theweighted mean approach to calculate the central position ofthe spectral area under the Fano resonance as indicated ineqn (2):

λCM ¼P

IðλiÞλiPIðλiÞ ð2Þ

where λi is the digital wavelength with a pixel resolution of0.3 nm for the USB spectrometer, and I(λi) is the intensityat different digital wavelengths. In our experiment, the Fano

Fig. 2 (a) Schematic diagrams of the DNA detection setup, (b) the layered structure of the microfluidic chip integrated with a gold capped nanowirearray and (c) the schematic of the surface plasmon resonance (SPR) chip with surface modification with cysteamine and linked with a probe andlatent membrane protein 1 (LMP1) DNA.

Analyst Paper

This journal is © The Royal Society of Chemistry 2020 Analyst, 2020, 145, 52–60 | 55

resonance is around 680 nm. Therefore, λi is selected from650 nm to 710 nm to cover the resonance profile.

SPR and sensitivity measurement

A TM-polarized wave in the capped gold nanowire array gener-ated sharp and asymmetric Fano resonances in the trans-mission spectra. The position of the SPR spectrum shifts dueto environmental changes around the gold nanostructures.Raw data were obtained and processed using a home-writtenMATLAB® program. The resonance wavelength of the SPRspectrum was identified by applying the spectral centroidmethod. The sensor graph represents the change of resonancewavelength against time.

The sensitivity test of the SPR nanowire was verified by therefractive index unit (RIU) sensitivity test. Refractive index sen-sitivities were measured by injecting purified water mixed withvarious ratios of glycerin into the microfluidic device. Therefractive indexes of the mixtures (from 0% to 20% glycerin)ranged from 1.334 to 1.344.

Polymerase chain reaction analysis

Conventional thermal-cycler PCR was also used to amplify theLMP1 DNA templates with a PCR machine (Analytik Jena,Germany). The LMP1 DNA, primer sequences, and probesequences are given in Table 1. The LMP1 DNA, primersequence and probe sequence design were those used by Ryanet al.4 and Guo et al.37 and also modified by us. The PCRmixture contained 25 µL of Taq master mix, 1 µL of forwardand reverse primers, 1 µL of a DNA template, and 22 µL ofnuclease-free water. Amplification reactions were performedfor 30 thermal cycles. The temperature conditions required forconventional PCR were set at 95 °C (30 s) for denaturation,60 °C (30 s) for annealing, and 72 °C (50 s) for extension.Moreover, the initial denaturation and final extension wereperformed at 95 °C for 5 min and 72 °C for 10 min, respect-ively. Furthermore, the 4 different PCR thermal cycles (5, 10,20, and 30 cycles) for LMP1 DNA amplification were also evalu-ated in this study. Agarose gel electrophoresis based analysiswas also employed to confirm the DNA amplicon results. A

SPR chip was then used to analyze each of the different PCRproducts and the results were also compared with gel electro-phoresis to investigate the limitation of the different tests.

Results and discussionThe sensitivity of the capped gold nanowire array

The refractive index unit (RIU) sensitivity test was used toevaluate the sensitivity of the SPR chip by injecting purifiedwater, mixed with various ratios of glycerin, into the microflui-dic device. The refractive index of the mixtures (from 0% to20% glycerin) ranged from 1.334 to 1.344. Fig. 3(a) shows theresonance wavelengths of different glycerin percentages. Theexperimental results indicated that the resonance wavelengthof 0% glycerin was 674 nm. The resonance wavelengths of 5%,10%, 15%, and 20% glycerin were 676, 679, 682, and 686 nm,respectively. The noise level determined the detection limit ofthe chip and the system in the measurement. Fig. 3(b) showsthe wavelength sensitivity of the 500 nm period SPR nanowire

Table 1 Latent membrane protein 1 (LMP1) DNA template, primersequences, and probe sequences

DNA template: (311 bp)CCATGACCCGCTGCCTCATAACCCTAGCGACTCTGCTGGAAATGATGGAGGCCCTCCAAAATTGACGGAAGAGGTTGAAAACAAAGGAGGTGACCGGGGCCCGCCTTCGATGACAGACGGTGGCGGCGGTCATCCACACCTTCCTACACTGCTTTTGGGTACTTCTGGTTCCGGTGGAGATGATGACGACCCCCACGGCCCAGTTCAGCTAAGCTACTATGACTAACCTTTCTTTACTTCTAGGCATTACCATGTCATAGGCTTGCCTGACTGACTCTCCCTCCATTTACTGGGAATGCCTTAGCTAATCA

Primer F (21 bp)AGC GAC TCT GCT GGA AAT GAT

Primer R (21 bp)TGA TTA GCT AAG GCA TTC CCA

Probe (29 bp)GTC ATA GTA GCT TAG CTG AAC TGG GCC GT

Fig. 3 Surface plasmon resonance (SPR) sensor sensitivity test using aglycerol/water mixture. (a) The red-shift of the resonance wavelengthwas induced by nine different concentrations of glycerol. (b) The wave-length sensitivity of the SPR chip. The deviations were below 0.1%.

Paper Analyst

56 | Analyst, 2020, 145, 52–60 This journal is © The Royal Society of Chemistry 2020

array. The peak wavelength was 674 nm, and the wavelengthsensitivity was 485 nm RIU−1. With this resolution, the currentnanostructure and simple system achieved a detection limit of4.1 × 10−5 RIU. Furthermore, the wavelength at 674 nm wasvisible and was easily covered by the LED light source.Cennamo et al.38 pointed out that the SPR sensor showed highsensitivity for measuring low concentration glycerin and thesensor could be used in biochemical applications.

In contrast, the deviations of each wavelength are alsoshown in Fig. 3(b). The deviations were obtained by measuringthe sample 10 times (N = 10). From the experimental results,the deviations of Δλ/λ0 were found to be below 0.1%.Compared to other SPR sensors, the current SPR nanowirechip demonstrated cost-effectiveness and good reproducibilityof production and measurement. The SPR nanowire chip witha visible light source can be used to develop a portable real-time detection device for biological binding reactions.Furthermore, the SPR nanowire chip can be integrated withthe on-chip PCR device in further studies.

Kinetic binding analysis of the SPR nanowire chip

Two different kinetic binding analyses were employed to evalu-ate the wavelength change of the probe and the LMP1 DNA pro-ducts. Fig. 4(a) shows the red-shift spectra with normalizedwavelengths of two different kinetic binding analyses. Thecurve fitting results showed two peaks at different wavelengths.Moreover, the spectral centroid method gave two centralpoints in the range from 650 nm to 710 nm. The results indi-cated that the LMP1-amplified DNA was detected by the SPRnanowire chip. The experimental results depicted that the SPRchip could be used to detect different DNA-based moleculeswithout any labeling. The current results were in good agree-ment with those of Mousavi et al.27 and Lee et al.35 Fig. 4(b)shows the stability of the resonance wavelength in pure wateras a function of time. The resonance wavelength was obtainedby the spectral centroid method. To compare the performanceof the analytic method, we also calculated the peak resonancewavelength using a curve fitting method. It can be seen thatthe wavelength resolution is greatly improved using the spec-tral centroid method. The standard deviation of the wave-length measurement is 0.02 nm.

Furthermore, the real-time progression reaction of the probeand LMP1 DNA was first detected in this study. Fig. 4(c) showsthe kinetic binding analysis of the entire test. The SPR chip wasfirst treated with cysteamine to form thiol bonding. After that,the spectrum of the SPR chip was stabilized by using water.After 2 min, the probe was injected into the microchannel for10 min. After 10 min of immersion, the excess probe waswashed away for 2 min. After that, the amplified LMP1 DNA pro-ducts were first denatured, and then linked to the probe at theprobe Tm (64 °C). After another 5 min of immersion, the excessDNA which had not been hybridized was washed away. InFig. 4(c), the spectrum exhibited a sharp peak shift at 1140 swhen the amplified DNA products reacted with the probe.Moreover, compared to the spectrum of the probe, the differ-ence was a red-shift caused by the DNA product.

Fig. 4 (a) The red-shift of the probe and the amplified DNA product inthe normalized intensity curve using the spectral centroid method andthe curve fitting method; (b) the stability tests of the resonance wave-length using the spectral centroid method and the curve fitting method;and (c) the sensor graph of SPR responses of an LMP1 DNA test with thereal-time progression reaction.

Analyst Paper

This journal is © The Royal Society of Chemistry 2020 Analyst, 2020, 145, 52–60 | 57

For the single-stranded DNA (ss-DNA) molecule detection,owing to the low molecular weight and low wavelength sensi-tivity of the LSPR, the DNA detection limit and specificity haveto be improved. A widely used method to improve the speci-ficity is to use the sandwich-type (capture DNA probe–targetDNA–reporter DNA probe) method. To enhance the detectionlimit, the reporter DNAs are usually conjugated with goldnanoparticles.39–41 On the other hand, methylation of DNAcan also be used to improve the microfluidic-based SPRdetection.31,32 In this study, we used the low-cost hot-emboss-ing nanoimprint method to fabricate a 97 nm wide, 500 nmperiod nanowire array on the plastic substrate. After coating a50 nm gold film on the nanowire array, we can generate theFano-like resonance on the gold surface. Such Fano resonanceis the coupling between the LSPR in the nanowires and thepropagation SPR on the continuous gold surface. It has a nar-rower bandwidth and higher intensity sensitivity.26,27 From theexperimental results, the immobilized probe on the cappedgold nanowire array and the target molecule leads to a red-shift in the SPR resonance wavelength. Therefore, the reporterDNA probe is not required, and the unmethylated DNA can bedetected during the test.

LMP1 DNA detection by the SPR nanowire chip

As mentioned earlier, the SPR chip combined with a micro-channel can be used to detect DNA molecules without thereporter DNA probe. Therefore, the LMP1 DNA amplified byusing the conventional PCR machine can also be used tounderstand the wavelength change mechanism and the detec-tion limit of the SPR nanowire chip. To evaluate the limitationof the different LMP1 DNA amplification products, agarose gelelectrophoresis based analysis was employed in this study.Agarose gel electrophoresis of 4 different PCR thermal cycles(5, 10, 20, and 30 cycles) of the LMP1 DNA amplification isshown in Fig. 5(a). Lane M was the DNA size marker, lane 1was the negative control, and lanes 2–5 denoted the LMP1 DNAamplification cycle numbers of 5, 10, 20, and 30, respectively.The experimental results indicated that the fluorescence bandwas not clearly visible at the cycle numbers of 5 and 10(lanes 2 and 3). However, the fluorescence band can be easilyobserved at the PCR cycle numbers of 20 and 30 (lanes 4and 5).

The kinetic binding analysis of 4 different PCR thermalcycles of the LMP1 DNA amplification is shown in Fig. 5(b). Inthe figure, the red-shift of 4 different PCR amplified productswas observed by using the spectral centroid method. Theresults indicated that the lower cycle numbers of the PCR pro-ducts could be detected by the SPR nanowire chip. In Fig. 5(c),the experimental results demonstrated that the red-shift valueswere 0.67, 0.92, 1.22 and 2.21 nm for the PCR cycle numbersof 5, 10, 20, and 30, respectively. The red-shift values showed alinear increase from 5 to 30 cycles.

Furthermore, the SPR nanowire chip combined with amicrofluidic device may ensure the detection of the PCR pro-ducts at lower cycle numbers, i.e., 5 and 10, which could notproduce a visible brightness band by the agarose gel electro-

phoresis method. The observed results, Fig. 5(a) and (c), werein good agreement with Nguyen et al.,33 Trinh et al.,34 andZhang et al.42 In this study, the variations of the resonancewavelengths are within 1% for different chips. Moreover, weused the same primer ratio for the PCR process. However,

Fig. 5 (a) The gel fluorescence image of the PCR results in the electro-phoresis analysis of the latent membrane protein 1 (LMP1) DNA amplifi-cation product. Lane M shows the DNA size marker. Lane 1 shows thenegative control. Lanes 2, 3, 4, and 5 denote the 5, 10, 20, and 30 LMP1DNA amplification cycle numbers, respectively; (b) the kinetic bindinganalysis of the different LMP1 DNA amplification cycle numbers usingthe spectral centroid method; and (c) the red-shift values of thedifferent LMP1 DNA amplification cycle numbers and the probe.

Paper Analyst

58 | Analyst, 2020, 145, 52–60 This journal is © The Royal Society of Chemistry 2020

Nagai et al.43 pointed out that the accurate value for the DNAconcentration should be obtained from the optimal primerratio. Zhang et al.42 also reported the high sensitivity of theDNA based on the powerful amplification and the excessprimer removal. Because of the variation of the amplifiedDNAs, the measured SPR shift has a significant standarddeviation.

Gold nanowire arrays have been applied to detect bio-molecular interactions by using localized surface plasmonresonance (LSPR) in the nanowires. In this work, we appliedthe PCR method instead of the sandwich-type measurement.PCR enhances explicitly the number of DNA molecules for thetarget DNA. Therefore, it does not need the reporter DNAprobes and the conjugation of gold nanoparticles. Using thePCR product and the proposed chip modified with the captureDNA probe, we can achieve DNA detection in a lower DNA copynumber.

Commercial PCR machines only amplify specific DNAs. Fordetection, the widely used method is electrophoresis analysis.It needs additional fluorescent labeling to detect the DNAs.The electrophoresis process takes time and requires the prepa-ration of running gels. In this work, the detection is performedby using the low-cost gold-coated nanowire chip. The chip wasfirst modified with a capture DNA probe. When the amplifiedDNAs are captured explicitly by the DNA probe, the change inthe surface refractive index results in a significant change inthe SPR peak and this can be detected using a spectrometer.Since there is no requirement for DNA labeling, the detectionis much simpler and quicker.

As shown in Fig. 5, the measurement sensitivity is betterthan the fluorescence electrophoresis method. Because thesensing chip has a low price and the measurement system issimple, the chip can also be used to develop a new portabledevice in the future. However, sample pretreatment isanother crucial issue. There are some commercial kits forDNA sample pretreatment, which can provide good results inisolating the DNA. The extraction time of QuickExtrat™ DNAextraction solution is less than 5 min and it is easy tohandle.9 After the sample treatment, the extracted DNA canbe individually amplified using the PCR machine and thendetected using the gold nanowire array chip modified withthe capture DNA probe. It is noted that the specific DNAamplification can also be performed by using the PCR micro-fluidic chip.44

In other words, this work is promising for biosensing appli-cations using the LMP1 gene. In comparison with the SPRnanowire chip with the microfluidic device, the red shift canbe easily observed at the PCR cycle numbers of 5 to 30. SPRnanowire chip can detect lower cycle numbers of the DNAamplification due to probe and DNA specificity. To summarizethis study, the SPR nanowire chip coated with the probe had alower detection limit compared to agarose gel electrophoresis.Furthermore, our gold nanowire array sensing chip can beintegrated into the PCR chip. The detection would be moreconvenient by directly injecting the extracted DNA into the on-chip PCR/SPR device in the future.

Conclusions

In this study, LMP1 gene detection was achieved by using amicrofluidic device with a SPR chip consisting of a gold-capped nanowire array. A transmission-type SPR sensing plat-form with high sensitivity and low background noise wasestablished to measure the SPR wavelength by the spectral cen-troid method. The SPR nanostructure with a 500 nm periodcan be used to detect DNA and genes without labeling. Withthe PCR amplification, the specificity and detection signalscan be significantly improved. The current SPR nanowire chipis proved to ensure the detection of the PCR products at lowercycle numbers. In contrast, using the agarose gel electrophor-esis method, the visible brightness band could not be pro-duced. The SPR nanowire chip can detect lower cycle numbersof the DNA amplification due to probe and DNA specificity.The detection systems used in this study have the advantagesof cost-effectiveness and good measurement and productionrepeatability. In summary, the results can assist in improvingthe biosensing of DNA-amplified products and in developingrapid detection devices with small-footprint nanostructuredSPR chips. Meanwhile, the SPR chip can be integrated into theonline PCR device platform in the future.

Conflicts of interest

The authors declare no conflicts of interest.

Acknowledgements

Financial support (MOST 106-2221-E-002-127, 107-2221-E-002-131 and 107-2622-E-002-016-CC2) from the Ministry of Scienceand Technology of Taiwan is acknowledged.

Notes and references

1 M. L. Gulley and W. Tang, J. Mol. Diagn., 2008, 10, 279–292.2 T. Riedel, C. Rodriguez-Emmenegger, A. de los Santos

Pereira, A. Bědajánková, P. Jinoch, P. M. Boltovets andE. Brynda, Biosens. Bioelectron., 2014, 55, 278–284.

3 X. Liu, J. Tang, M. Wang, Q. Ma and Y. Wang, Int. J. Mol.Sci., 2013, 14, 23922–23940.

4 J. L. Ryan, H. Fan, S. L. Glaser, S. A. Schichman, N. Raab-Traub and M. L. Gulley, J. Mol. Diagn., 2004, 6, 378–385.

5 D. E. Tsai, L. Douglas, C. Andreadis, D. T. Vogl, S. Arnoldi,R. Kotloff, J. Svoboda, R. D. Bloom, K. M. Olthoff,S. C. Brozena, S. J. Schuster, E. A. Stadtmauer,E. S. Robertson, M. A. Wasil and V. N. Ahya,Am. J. Transplant., 2008, 8, 1016–1024.

6 M. N. Islam, S. Yadav, M. H. Haque, A. Munaz, F. Islam,M. S. A. Hossain, V. Gopalan, A. K. Lam and N. T. Nguyen,Biosens. Bioelectron., 2017, 92, 668–678.

7 H. Nagai, K. Tomioka and S. Okumura, Appl. Biochem.Biotechnol., 2019, 187, 323–337.

Analyst Paper

This journal is © The Royal Society of Chemistry 2020 Analyst, 2020, 145, 52–60 | 59

8 S. K. Vashist, P. B. Luppa, L. Y. Yeo, A. Ozcan andJ. H. T. Luong, Trends Biotechnol., 2015, 33, 692–705.

9 P. Kranzfelder, T. Ekrem and E. Stur, Mol. Ecol. Resour.,2016, 16, 353–363.

10 S. A. Maier, Plasmonics: Fundamentals and Applications,Springer Science, New York, 2007, pp. 188–192.

11 M. E. Stewart, C. R. Anderton, L. B. Thompson, J. Maria,S. K. Gray, J. A. Rogers and R. G. Nuzzo, Chem. Rev., 2008,108, 494–521.

12 H. J. Lezec, A. Degiron, E. Devaux, R. A. Linke, L. Martin-Moreno, F. J. Garcia-Vidal and T. W. Ebbesen, Science,2002, 297, 820–822.

13 J. N. Anker, W. P. Hall, O. Lyandres, N. C. Shah, J. Zhao andR. P. Van Duyne, Nat. Mater., 2008, 7, 442–453.

14 A. G. Brolo, R. Gordon, B. Leathem and K. L. Kavanagh,Langmuir, 2004, 20, 4813–4815.

15 K. A. Tetz, L. Pang and Y. Fainman, Opt. Lett., 2006, 31,1528–1530.

16 J. C. Yang, J. Ji, J. M. Hogle and D. N. Larson, Nano Lett.,2008, 8, 2718–2724.

17 K. L. Lee, C. W. Lee, W. S. Wang and P. K. Wei, J. Biomed.Opt., 2007, 12, 044023.

18 K. L. Lee, S. H. Wu and P. K. Wei, Opt. Express, 2009, 17,23104.

19 K. L. Lee and P. K. Wei, Appl. Phys. Lett., 2011, 99, 083108.20 D. Zhang, Y. Yan, Q. Li, T. Yu, W. Cheng, L. Wong, H. Ju

and S. Ding, J. Biotechnol., 2012, 160, 123–128.21 O. Tokel, U. H. Yildiz, F. Inci, N. G. Durmus, O. O. Ekiz,

B. Turker, C. Cetin, S. Rao, K. Sridhar, N. Natarajan,H. Shafiee, A. Dana and U. Demirci, Sci. Rep., 2015, 5, 9152.

22 H. Shi, K. Nie, B. Dong, M. Long, H. Xu and Z. Liu, Chem.Eng. J., 2019, 361, 635–650.

23 K. L. Lee, C. C. Chang, M. L. You, M. Y. Pan and P. K. Wei,Sci. Rep., 2018, 8, 9762.

24 J. Sambles, G. Bradbery and F. Yang, Contemp. Phys., 1991,32, 173–183.

25 J. Homola, S. S. Yee and G. Gauglitz, Sens. Actuators, B,1999, 54, 3–15.

26 M. Z. Mousavi, H. Y. Chen, S. H. Wu, S. W. Peng, K. L. Lee,P. K. Wei and J. Y. Cheng, Analyst, 2013, 138, 2740–2748.

27 M. Z. Mousavi, H. Y. Chen, K. L. Lee, H. Lin, H. H. Chen,Y. F. Lin, C. S. Wong, H. F. Lin, P. K. Wei and J. Y. Cheng,Analyst, 2015, 140, 4097–4104.

28 W. K. Yeung, H. Y. Chen, J. J. Sun, T. H. Hsieh,Z. M. Mansoureh, H. H. Chen, K. L. Lee, H. Lin, P. K. Weiand J. Y. Cheng, Analyst, 2018, 143, 4715–4722.

29 K. L. Lee, M. L. You, C. H. Tsai, E. H. Lin, S. Y. Hsieh,M. H. Ho, J. C. Hsu and P. K. Wei, Biosens. Bioelectron.,2016, 75, 88–95.

30 J. Burger, C. Ruth, J. Woehrle, P. A. Meyer, N. B. Ammar,N. Kilb, T. Brandstetter, F. Proll, G. Proll, G. Urban andG. Roth, SLAS Technol., 2017, 22, 437–446.

31 A. De, W. Sparreboom, A. Berg and E. T. Carlen,Biomicrofluidics, 2014, 8, 054119.

32 R. Kurita, H. Yanagisawa, K. Yoshioka and O. Niwa, Anal.Chem., 2015, 87, 11581–11586.

33 T. T. Nguyen, K. T. L. Trinh, W. J. Yoon and N. Y. Lee, Sens.Actuators, B, 2017, 242, 1–8.

34 K. T. L. Trinh, Y. Zhang and N. Y. Lee, Anal. Chim. Acta,2018, 1040, 63–73.

35 K. L. Lee, T. Y. Wu, H. Y. Hsu, S. Y. Yang and P. K. Wei,Sensors, 2017, 17, 1548.

36 K. L. Lee, H. Y. Hsu, M. L. You, C. C. Chang, M. Y. Pan,X. Shi, K. Ueno, H. Misawa and P. K. Wei, Sci. Rep., 2017, 7,44104.

37 Q. Guo, I. W. K. Tham, S. Lin, Y. Su, Z. Chen, J. Lin, L. Han,Q. Lin, J. Pan and J. J. Lu, Head Neck Oncol., 2012, 4, 81–87.

38 N. Cennamo, L. Zeni, E. Catalano, F. Arcadio andA. Minardo, Appl. Sci., 2018, 8, 1172.

39 X. Hong and E. A. H. Hall, Analyst, 2012, 137, 4712–4719.40 Q. Wang, R. Liu, X. Yang, K. Wang, J. Zhu, L. He and Q. Li,

Sens. Actuators, B, 2016, 223, 613–620.41 N. Yaakov, Y. Chaikin, E. Wexselblatt, Y. Tor, A. Vaskevich

and I. Rubinstein, Chem. – Eur. J., 2017, 23, 10148–10155.42 D. Zhang, Y. Yan, Q. Li, T. Yu, L. Wang, H. Ju and S. Ding,

J. Biotechnol., 2012, 160, 123–128.43 H. Nagai, K. Tomioka and S. Okumura, Appl. Biochem.

Biotechnol., 2019, 187, 323–337.44 C. Zhang, J. Xu, W. Ma and W. Zheng, Biotechnol. Adv.,

2006, 24, 243–284.

Paper Analyst

60 | Analyst, 2020, 145, 52–60 This journal is © The Royal Society of Chemistry 2020