Biosensors and Bioelectronics 63 (2015) 196–203

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

Rapid and amplification-free detection of fish pathogens by utilizing a molecular beacon-based microfluidic system$ Yi-Chih Su a, Chih-Hung Wang a, Wen-Hsin Chang a, Tzong-Yueh Chen d,n, Gwo-Bin Lee a,b,c,nn a

Department of Power Mechanical Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan Institute of Biomedical Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan c Institute of NanoEngineering and Microsystems, National Tsing Hua University, Hsinchu 30013, Taiwan d Institute of Biotechnology, National Cheng Kung University, Tainan 70101, Taiwan b

art ic l e i nf o

a b s t r a c t

Article history: Received 7 May 2014 Received in revised form 11 July 2014 Accepted 11 July 2014 Available online 24 July 2014

Nervous necrosis virus (NNV) and iridovirus are highly infectious pathogens that can cause lethal diseases in various species of fish. These infectious diseases have no effective treatments and the mortality rate is over 80%, which could cause dramatic economic losses in the aquaculture industry. Conventional diagnostic methods of NNV or iridovirus infected fishes, such as virus culture, enzymelinked immunosorbent assays and nucleic acid assays usually require time-consuming and complex procedures performed by specialized technicians with delicate laboratory facilities. Rapid, simple, accurate and on-site detection of NNV and iridovirus infections would enable timely preventive measures such as immediate sacrifice of infected fishes, and is therefore critically needed for the aquaculture industry. In this study, a microfluidic-based assay that employ magnetic beads conjugated with viral deoxyribonucleic acid (DNA) capturing probes and fluorescent DNA molecular beacons were developed to rapidly detect NNV and iridovirus. Importantly, this new assay was realized in an integrated microfluidic system with a custom-made control system. With this approach, direct and automated NNV and iridovirus detection from infected fishes can be achieved in less than 30 min. Therefore, this molecular-beacon based microfluidic system presents a potentially promising tool for rapid diagnosis of fish pathogens in the field in the future. & Elsevier B.V. All rights reserved.

Keywords: Nervous necrosis virus Iridovirus Probe conjugated magnetic beads Molecular beacon Microfluidics

1. Introduction There are more than 50 different species of groupers inhabiting around Taiwan and grouper farming represents a 7.7-billion industry in Taiwan (Kuo et al., 2011). For this industry, viral infection poses a significant threat and causes heavy economic losses. In particular, nervous necrosis virus (NNV) and iridovirus are two strains of highly contagious virus that cause high mortality rates among grouper larvae and juveniles (Harikrishnan et al., 2010). Furthermore, because there are no effective treatments for

☆ The preliminary results in this paper have been presented at the 16th International Conference on Miniaturized Systems for Chemistry and Life Sciences, mTAS 2012 Conference, OKINAWA, October 28–November 1, 2012. n Corresponding author. Tel.: þ 886 6 2757575x65622x610; fax: þ886 6 2766505. nn Corresponding author at: Department of Power Mechanical Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan. Tel.: þ 886 3 5715131x33765; fax: þ 886 3 5722840. E-mail addresses: [email protected] (T.-Y. Chen), [email protected] (G.-B. Lee).

http://dx.doi.org/10.1016/j.bios.2014.07.035 0956-5663/& Elsevier B.V. All rights reserved.

diseases caused by NNV and iridovirus, fishes infected with NNV or iridovirus are immediately sacrificed in order to prevent further transmission (Kuo et al., 2012a, 2012b). Simple, rapid and accurate methods for detecting NNV and iridovirus directly from fishes are therefore crucial for monitoring the infected fishes and minimizing economic losses. Up to date, conventional methods for detecting these viruses, including histopathology (Tanaka et al., 2004), virology (Fukuda et al., 1996; Chou et al., 1998), immunohistochemistry (Shi et al., 2003) or molecular diagnosis (Kurita et al., 1998; Dalla Valle et al., 2005) have been employed. Unfortunately, these methods suffer from several drawbacks, such as complex and time-consuming procedures and requirements of large amount of samples/reagents, bulky instruments and well-trained technicians, which have significantly limited their effectiveness as simple, rapid and accurate detection methods. Micro-total-analysis-systems (μ-TAS), also called integrated microfluidic systems or lab-on-a-chip (LOC), are systems that can perform complex processes, such as sample pretreatment, transportation, mixing, separation, reaction and detection within a single miniaturized chip, and have recently been actively explored

Y.-C. Su et al. / Biosensors and Bioelectronics 63 (2015) 196–203

Nomenclature ANOVA BHQ1 DEV DNA EV FRET H1 H3 IB LOC

analysis of variance Black Hole Quencher-1 dengue virus type 2 deoxyribonucleic acid enterovirus type 71 fluorescence resonance energy transfer influenza A/H1N1 virus influenza A/H3N2 virus influenza B virus lab-on-a-chip

in the research community. Such μ-TAS integrates various microfluidic modules, such as microvalves, micropumps, micromixers, microfilters, microheaters, microsensors, and microreactors and achieves chemical and biomedical analysis with lower sample or reagent consumption, shorter reaction times and higher throughput (Hessel et al., 2005; Thorsen et al., 2002). Furthermore, these systems could perform the entire process in an automatic format. Therefore, μ-TAS may offer a robust and effective platform for detection of fish viruses that may replace traditional complicated diagnostics and potentially save considerable analysis time and costs. In order to achieve simple and rapid detection, detecting the viral nucleic acids via an amplification-free method should be employed. Molecular beacons (Tyagi and Kramer, 1996) serve this purpose well because they produce an easily detectable signal directly upon hybridization with the target nucleic acid sequence. The structure of molecular beacons includes a loop that is designed to be complementary to the target sequence and a selfcomplementary stem to hold with a fluorophore and a quencher in proximity. It has been widely used in chemistry, biology, biotechnology and medical sciences as bio-molecular recognition probes due to their ease of synthesis, unique functionality, molecular specificity and structural tolerance to various modifications (Tan et al., 2004; Whitcombe et al., 1999; Wong and Medrano, 2005; Mercier-Delarue et al., 2014) for diagnostic assays. Furthermore, when employed in concert with a second sequence-specific deoxyribonucleic acid (DNA) capturing probe that is conjugated to magnetic beads (Wang et al., 2011), the specificity of molecular beacon-based assays could be further improved. The magnetic probes not only enhance the specificity for detecting the target sequence, but also help purify the target viral DNA or ribonucleic acid (RNA) from debris from the fish sample (Chang et al., 2013a). Therefore, the combination of molecular beacons and sequencespecific magnetic capture probes performed on an integrated microfluidic system may present an amplification-free process that decreases the detection time. This study therefore reports the development of an integrated microfluidic chip that contains some micro-modules, including reagent and washing buffer chambers, normally-closed valves and micropumps. The specific probe conjugated magnetic beads and molecular beacons were used to purify samples and hybridize with the target nucleic acid sequences for detection, respectively. In addition, a custom-made control system including a temperature control module, a microfluidic control module and an optical detection module was used to carry out the entire diagnosis process automatically (Chang et al., 2013b). As a demonstration, this microfluidic platform could carry out the detection of NNV and iridovirus simply and directly from fish samples.

MCP NNV PCR PDMS RNA RT-PCR S/N ratio VIB a.u. ddH2O FAM μ-TAS

197

major capsid protein nervous necrosis virus polymerase chain reaction polydimethylsiloxane ribonucleic acid reverse transcription polymerase chain reaction signal-to-noise ratio Vibrio sp. arbitrary unit double-distilled water 6-carboxyfluorescein micro-total-analysis-systems

2. Materials and methods 2.1. Viral strain The NNV-infected, iridovirus-infected fish and Vibrio sp. (VIB) samples were from the Institute of Biotechnology, National Cheng Kung University, Taiwan. In this study, inactivated influenza A/ H1N1 virus (H1), influenza A/H3N2 virus (H3), influenza B virus (IB), dengue virus type 2 (DEV), and enterovirus type 71 (EV), were prepared and generously provided by Dr. Chih-Peng Chang from the Department of Microbiology and Immunology, National Cheng Kung University, Taiwan. 2.2. Experimental procedures The working principle of the integrated microfluidic system for the rapid diagnosis of NNV and iridovirus developed in this work is shown in Fig. 1. First, virus-infected fishes were grinded by a bio-vortex homogenizer (No. 1083-MC, Bio Spec Products Inc., USA). Subsequently, the 10 mL of grinded fish suspension was incubated with 5 mL of specific molecular beacons and 5 mL of specific probe-conjugated magnetic beads at 95 °C for 10 min, to lyse viral particles and denature the stem-loop structure of the molecular beacons and any secondary structure of viral DNA and RNA. Then single-stranded molecular beacons were allowed to hybridize to the complementary sequence of target viral DNA or RNA at 75 °C for 5–15 min. After that, the temperature was cooled to 55 °C for the probe-conjugated magnetic beads to anneal to the target viral sequence. The beacon–probe–target complex was collected by an external magnet that unbound beacons and debris from fish samples were washed away by using double-distilled water (ddH2O, pH 7.0) for three times. In is study, the fluorescence donor and the quencher of the molecular beacons may contact to cause direct energy transfer and then dissipate with heat energy by FRET (fluorescence resonance energy transfer). When the molecular beacon was denatured as single-strand molecular beacon by heating treatment, the ssDNA was hybridized to the target region of the tested nucleic acids. The fluorescence donor then provided a stronger signal that resulted from the donor since the quencher was located far apart for a long distance. Finally, the fluorophore of the molecular beacon was excited by a laser (473 nm) and the intensity of fluorescence signal was measured by a custom-developed software (PCR_MFD ver 1.0, Mirle Co., Taiwan) to determine the diagnostic results. Note that, with the exception of grinding the fish sample, the whole assay was carried out in an integrated microfluidic system in a rapid and automated fashion by a developed custom-made control system (Chang et al., 2013b). Note that NNV is a RNA-virus and iridovirus is a DNAvirus. A specific NNV RNA2 probe and an iridovirus major capsid

198

Y.-C. Su et al. / Biosensors and Bioelectronics 63 (2015) 196–203

Table 1 The sequences of designed primers, probes and molecular beacons. Primer

Sequence 5′-3′

NNV RNA2 RNA2 RNA2 RNA2

ATAGCAATCGCACCCTCCTT TGGCGTTGCCTACTACTCTG amide-TTTTTTTTTTTGACTCGGAAAACTAACCGG 6FAM-CCACCCAAGGAGGGTGCGATGGCGAGGG- BHQ1

Reverse Forward probe beacon

Iridiovirus MCP Reverse MCP Forward MCP probe MCP beacon

GGCATTGTCCGACACCTTAC GTATGTCAACAACAAGGT amide-TTTTTTTTTTGGGTTGATATTGCCCATGTC 6FAM-CGCGATCAGCGACACCTCGGACAGGGATCGCG- BHQ1

probe-conjugated magnetic beads were suspended into 1000 mL of ddH2O (pH 7.0) and stored at 4 °C until use. Note that the pH values of 7.0–7.5 were normally used for most biological experiments such that the stability of enzyme activity could be maintained (Darby et al., 2002). In this study, the zeta potential of the probe-conjugated beads was measured to be about 2.7 mV. The modified magnetic beads were reported to be stable such that they could be used for diagnostic applications from microscopic observation (Klamp et al., 2013; Hung et al., 2014). 2.4. Molecular beacons, PCR primers and operating conditions Fig. 1. The experimental procedure for the rapid detection of NNV and iridovirus. (A) Molecular beacons, virus and probe conjugated magnetic beads were loaded into the reaction chamber. (B) The temperature was increased to 95 °C to lyse the virus and denature the beacon. (C) Next, the temperature was decreased to 75 °C to allow the beacon to anneal to the target RNA or DNA. (D) The temperature was lowered to 55 °C for the probe to anneal to the target sequence, followed by a washing step to remove unbound material. The fluorescent signal of the sample was then measured.

protein (MCP) gene probe were used to capture the released DNA and RNA from virus particles. 2.3. Probe-conjugated magnetic beads The probe-conjugated magnetic beads can eliminate debris that may significantly interfere with the fluorescence signal and concentrate tested nucleic acid. The nucleotide sequences of specific NNV RNA2 probe and the iridovirus MCP gene probe were synthesized and modified with a C6 amide group on the 5' terminus (from Sigma Corp., USA). The amide-modified nucleotide probes were conjugated with a carboxylated group on the surface of the magnetic beads (ؼ 4.5 mm (micrometer), 4  107 beads/mL, MAGBEAD AGT-003-05, Applied Gene Technologies, USA). The carboxylation started with a mixture of 100 mL magnetic beads and 900 mL of ddH2O collected by a magnetic rack (DynaMagTM-2, Invitrogen, USA) for 3 min at room temperature. Then, the supernatant was discarded and dissolved in 1000 mL of ddH2O. Next, the magnetic beads were pelleted and the supernatant was removed again. Following, 950 mL of ddH2O, 30 mL of 100 mM amine-modified probes and 20 mL of 120 mg/mL 1-ethy1-3-(3-dimethylaminopropyl) (EDAC, Invitrogen Co., USA) were added. The mixture was rotated on a wheeling rotator at 20 rpm for 18 h at room temperature and kept in the dark. Next, the mixture was washed twice with 1 mL of 0.02% Tween-20 (Sigma Co., USA), followed by two washes of 1 mL 0.1% sodium dodecyl sulfate (Merck Co., USA) to remove the unbound probes. To block unbound coupling sites on the magnetic beads, the beads were incubated in 1000 mL of 1 M ethanolamine (Sigma Co., USA) at room temperature for one hour. One last washing process was performed and the

The sequences of NNV and iridovirus molecular beacons, NNV RNA2 and iridovirus MCP sequences, and forward and reverse primers that confirmed by PCR were designed by Vector NTI 8 software (InforMax Inc., USA). The primer sequences for each virus and molecular beacons are listed in Table 1. The designed molecular beacons and primers were synthesized and purified by high performance liquid chromatography (Sigma Co., USA). The fluorophore and quencher of the molecular beacons were 6FAM (6-carboxyfluorescein) and BHQ1 (Black Hole Quencher-1), respectively. Polymerase chain reaction (PCR) (for iridovirus MCP) and reverse transcription PCR (RT-PCR) (for NNV RNA2) were employed to confirm the specific probe capturing or molecular beacon based detection. For PCR amplification, the 5 mL of tested samples and magnetic beads were added to 3 mL of 10  buffer, 1 mL of 10 mM dNTP, 0.25 mL of 25 mM MgCl2, 0.5 mL of forward primer and 0.5 mL of reverse primer, 0.5 mL of Superthermo Gold Taq DNA polymerase and 19.25 mL of ddH2O for PCR mixture. The optimal PCR and RT-PCR operating conditions were determined by changing temperatures, denaturation time, amplification cycles and final extension steps. Briefly, the initial denaturation step of PCR was performed at 95 °C for 10 min. The next thermocycling steps were performed at 95 °C for 30 s, 58 °C for 30 s and 72 °C for 30 s. Finally, an extension step was carried out at 72 °C for 5 min to ensure that the remaining single-stranded DNA was fully extended. The RT-PCR was used for nucleic acid amplification of the RNA-based virus by using the MMLV reverse transcriptase (Promega, USA). An additional reverse-transcription step was performed at 42 °C for 45 min prior to the PCR processes. 2.5. Chip design, fabrication and processes performed on the chip The microfluidic chip was fabricated by using a standard soft lithography microfabrication process (Lin et al., 2002) using polymethymethacrylate (PMMA) master molds created by a computer-numerical-control machining process (CNC EGX-400, Roland Inc., Japan). Briefly, a 0.5-mm drill bit was used to fabricate microfluidic structures on by using the CNC machining process. Second, PDMS was poured (mixed with a ratio of 10:1 for Sylgard

Y.-C. Su et al. / Biosensors and Bioelectronics 63 (2015) 196–203

199

channel layer for liquid transport. Transported reagent volume could be readily adjusted by changing the negative gauge pressure applied to the designed suction-type micropump (Weng et al., 2011). The length, the width and the depth of the air layer were 5.9, 7.4, and 1.0 cm, respectively. In addition, the length, the width and the depth of the liquid channel layers were 5.9, 7.4, and 0.2 cm, respectively. A glass layer was used as a bottom substrate for sealing the microchannels and chambers. This integrated microfluidic chip contained four identical sample processing and detection units for performing four parallel assays for simultaneously detecting both NNV and iridovirus. Furthermore, the chip also contained two identical sets of chambers for performing the negative control and the positive control for each virus. Each sample processing and detection unit included washing buffer and reaction chambers, suction-type micropumps, normally-closed valves and a waste unit designed to perform the entire process for virus detection. 2.6. Molecular beacon analysis and operating conditions The stem-loop structure was denatured and then hybridized to the target gene region for molecular beacon based detection. To initiate the molecular beacon analytic process by the developed chip, the 5 mL of specific molecular beacons, 10 mL of ground fish samples and 5 mL of specific probe-conjugated magnetic beads were loaded into the reaction chambers, while ddH2O (pH 7.0) was loaded into the wash buffer chambers. The microfluidic chip was then mounted in the custom-developed control system. The temperatures under the reaction chambers were sequentially changed to achieve thermal lysis of virus, denaturation of viral DNA or RNA, molecular beacon hybridization and capture probe hybridization. After these sample processing steps, an external magnet was placed under the reaction chambers to capture the magnetic beads, while the micropump pumped ddH2O from wash buffer chambers through reaction chambers to waste outlets to wash away any unbound materials and debris from fish samples. The hybridization temperature and time were further optimized to determine the optimal condition for the molecular beacon assay. Finally, the fluorescent signals were detected in the reaction chambers.

3. Results and discussion 3.1. Optimal hybridization temperature of molecular beacons

Fig. 2. (A) A top view, (B) an exploded view, and (C) a photograph of the developed molecular-beacon-based microfluidic chip. The chip consisted of normally-closed valves, washing buffer chambers, reaction chambers, waste units, and suction-type micropumps. The size of the chip was measured to be 7.4 (L)  5.9 (W) cm2.

184A and Sylgard 184B, Sil-More Industrial Ltd., USA) into the PMMA mold to fabricate the air chamber layer and the liquid channel layer. Next, the microstructure was replicated when placed into an oven at 80 °C for 3 h. Finally, oxygen plasma was applied to the PDMS and glass substrates for surface treatment and then the two PDMS layers and glass substrate were bonded together. The microfluidic chip consisted of two PDMS layers and one glass layer, as shown in Fig. 2. One of the PDMS layers was an air chamber layer that could be driven by applying a negative gauge pressure (vacuum) to lift up the membrane-type microvalves and micropumps. The other PDMS layer was a liquid

The optimal hybridization temperature for the NNV RNA2 and iridovirus MCP molecular beacons were first explored. The molecular beacons and their specific target RNA or DNA were first denatured at 95 °C, subsequently allowed to hybridize to each other at 70, 75, 80, and 85 °C for 15 min and the resulting fluorescence signals were measured. In parallel, the beacons were also incubated with negative control samples that contained no tested virus to determine the background fluorescence signal. Then, the signal-to-noise (S/N) ratio was determined by taking the ratio of the fluorescence signals between the target-containing samples and the negative control. Note that all results shown here were obtained from three consecutive measurements. As shown in Fig. 3(A), the measured S/N ratios for NNV at 70, 75, 80 and 85 °C were measured to be 2.773, 3.145, 2.555 and 2.417, respectively. Similarly, the S/N ratios for iridovirus at 70, 75, 80 and 85 °C were measured to be 2.260, 3.283, 2.406 and 1.990, respectively. This test revealed that the optimal hybridization temperature for both molecular beacons were determined to be 75 °C, as the S/N ratio for both beacons at 75 °C peaked at 43-fold, as shown in Fig. 3(B). It should be noted that at 80 and 85 °C, although the sample

200

Y.-C. Su et al. / Biosensors and Bioelectronics 63 (2015) 196–203

Fig. 3. The optimal hybridization conditions for NNV RNA2 and iridovirus MCP molecular beacons. The hybridization temperature optimization for (A) NNV RNA2 and (B) iridovirus MCP molecular beacons. The results indicated that the optimal hybridization temperatures of NNV RNA and iridovirus MCP molecular beacons were 75 °C, which expressed the highest S/N ratio. Note that the values and the error bars in this figure represent the mean and the standard deviation of three consecutive measurements. The hybridization time optimization of the (C) NNV RNA2 and (D) Iridovirus MCP molecular beacons at 75 °C. The results indicated that the optimal hybridization times for NNV RNA2 and Iridovirus MCP molecular beacons were 5 and 15 min, respectively. The values and the error bars in this figure represent the mean and the standard deviation of three consecutive measurements.

fluorescence signals increased, the background signals also significantly increased, which resulted in decreased S/N ratios at these two higher temperatures. The decrease in S/N ratios was presumably due to the fact that these temperatures were above the melting temperature of the molecular beacons. In other words, at these higher temperatures, a significant portion of probes remained open, which facilitated the hybridization to targets and yielded greater signals but resulted in higher background signals.

complete diagnosis time. On the other hand, Fig. 3(D) showed that the measured S/N ratios for iridovirus at 75 °C for 5, 10 and 15 min were 1.994, 2.065 and 2.724, respectively, which suggested that the optimal hybridization time was 15 min. Note that the slower hybridization of the iridovirus molecular beacon might be caused by its higher melting temperature (around 85.7 °C) and may be further optimized. 3.3. Molecular beacon specificity and sensitivity

3.2. Optimal hybridization time of molecular beacons The hybridization time of the molecular beacons was also optimized. The molecular beacons and their specific target RNA or DNA were first denatured at 95 °C, subsequently allowed to hybridize to each other at optimized temperature 75 °C for 5, 10 and 15 min and the resulting fluorescence signals were measured. In parallel, the beacons were also incubated with negative control samples that contained no target to determine the background fluorescence signal. Note that all results shown here were obtained from three consecutive measurements. As shown in Fig. 3(C), the measured S/N ratios for NNV at 75 °C for 5, 10 and 15 min were measured to be 3.578, 2.535 and 3.589, respectively. The results indicated that different hybridization times produced similar S/N ratios for the NNV RNA2 molecular beacon. As a result, the optimal hybridization time of the NNV RNA2 molecular beacon used in this study was determined to be 5 min in order to decrease the

The molecular beacons demonstrated high specificity for their specific targets, which is important for preventing unnecessary sacrifice of fishes and consequent economic losses. The specificity of the NNV RNA2 and iridovirus MCP molecular beacons was demonstrated by testing against the RNA or DNA of various kinds of viruses and bacteria, including influenza A H1NI (H1), influenza A H3N2 (H3), influenza B virus (IB), dengue virus (DEV), enterovirus (EV), and Vibrio sp. (VIB). Of note, Vibrio sp. is a common strain of bacteria found in fishes, which therefore represents a particularly useful test for beacon specificity. Unless otherwise stated, all subsequently molecular beacon hybridization conditions were 75 °C for 5 or 15 min. Note that all results shown here were obtained from three consecutive measurements. As shown in Fig. 4, both molecular beacons displayed high specificity as they produced robust fluorescence signals only when the specific targets were used. When non-targets were used, the fluorescence

Y.-C. Su et al. / Biosensors and Bioelectronics 63 (2015) 196–203

Fluorescence signal (a.u.)

A

1200 1000 800 600 400 200 0 N

Fluorescence signal (a.u.)

B

P

Irido

H1

H3

IB

EV

DEV

VIB

H3

IB

EV

DEV

Sample 1200 1000 800 600 400 200

201

negative control versus 5, 10, 20, 40 ng/mL of iridovirus plasmid DNA were calculated to be 0.228, 0.005, 0.016 and 0.007, respectively (Fig. 5(B)). The results showed that the limit of detection of MCP was measured as 10 ng/mL. the LOD equation of MCP was determined based on the relation between the fluorescence intensity (Y) and the concentration of the tested nucleic acids (ng/mL) (X), which could be represented as Y ¼19.3X 410.0 (R2 ¼0.9254). The detectable nucleic acid concentrations of NNV and MCP from tested infected fishes were then measured to be 22.3373.94 and 37.17 76.50 ng/mL, respectively. Note that these two values are sensitive enough for practical applications. This value is comparable to the previous study using a non-amplification molecular beacon assay (Zhang et al., 2005). Similarly, the PCR process was carried out to detect iridovirus non-infected and iridovirus infected fish after the capturing the nucleic acids from the fish samples. Note that this PCR results were just used for comparison. Fig. 5(C) shows the electropherogram results of RT-PCR, which used specific primers to amplify a specific NNV RNA fragment (Wang et al., 2011). Similarly, Fig. 5(D) shows the electropherogram results of PCR, which used specific primers to amplify a specific iridovirus MCP gene (Lien, et al., 2009). The results show that the data from the traditional PCR and RT-PCR processes were consistent with ones from the developed microfluidic system, which verifies its accuracy. 3.4. Diagnosis of fish sample

0 N

P

NNV

VIB

H1

Sample Fig. 4. The specificity test for (A) NNV RNA2 and (B) iridovirus MCP molecular beacons using various strains of virus and bacteria. The results indicated statistically significant differences between the positive control and the no-target groups. Here, N is the negative control using ddH2O. P is the positive control. Irido, H1, H3, IB, EV, DEV and VIB indicate the samples containing Iridovirus plasmid DNA, influenza A/H1N1 viral RNA, influenza A/H3N2 viral RNA, influenza B virus viral RNA, enterovirus viral RNA, dengue virus viral RNA and bacterial DNA of Vibrio, respectively. The values and the error bars in this figure represent the mean and the standard deviation of three consecutive measurements.

intensities were similar to the background signals. The results were further analyzed by one-way ANOVA (analysis of variance) test, which is a test for analyzing the difference between groups. The specificity results of the NNV RNA2 beacon by ANOVA analysis indicated statistically significant differences between the NNV RNA2 positive control plasmid DNA and the other groups (F (8,36) ¼ 7.942, p ¼0.003) (Fig. 4(A)). Furthermore, the results of the ANOVA test indicated statistically significant differences between the MCP positive control plasmid DNA and the no-target groups (F(8,36) ¼4.428, p¼ 0.019). Therefore, the developed assay is confirmed to have high specificity for detection. The detection limits of the NNV RNA2 and iridovirus molecular beacons were evaluated and shown in Fig. 5. Here, NNV RNA2 or iridovirus MCP molecular beacons were allowed to hybridize to their respective target plasmid DNA at different concentrations (5, 10, 20, 40 ng/μL). Note that all results shown here were obtained from three consecutive measurements. After the fluorescence signal from each sample was acquired, student's two-tail t test was performed to determine the statistical difference between each target sample and the no-target, negative control sample. The p value of the negative control versus 5, 10, 20, 40 ng/mL of NNV plasmid DNA were found to be 0.258, 0,018, 0.005 and 0.001, respectively (Fig. 5(A)). The results indicated that the limit of detection of NNV was determined to be 20 ng/mL. Moreover, the LOD equation of NNV was determined based on the relation between the fluorescence intensity (Y) and the concentration of the tested nucleic acids (ng/mL) (X), which could be represented as Y¼ 65.5X  105.0 (R2 ¼0.9897). Similarly, the p value of the

Once the optimal hybridization conditions were determined, fishes that are either healthy (non-infected) or infected with NNV or iridovirus were diagnosed with the microfluidic system to assess the applicability of this assay for clinical diagnostics. In order to ensure the accuracy of the developed microfluidic system, a RT-PCR was also carried out for comparison to detect NNV noninfected and NNV infected fish. The fluorescent intensities of the negative control, positive control, infected and non-infected NNV were 320.842 739.352, 913.472 727.672, 1357.708 7152.867 and 355.786 7 45.660, (arbitrary unit, a.u.), respectively. Similarly, Fig. 5(F) shows the fluorescent intensity of the negative and positive controls, iridovirus infected and non-infected samples were measured to be 395.372 750.108, 879.0757 92,372, 1127.4457125,989 and 349.672 751.577 (a.u.), respectively. Both of the analytic results of NNV and iridovirus showed a dramatic difference of the fluorescence intensity between the infected sample and the negative control. The detectable nucleic acid concentrations of NNV and MCP from tested infected fishes were then measured to be 22.3373.94 and 37.17 76.50 ng/mL, respectively. Note that these two values are sensitive enough for practical applications. Furthermore, there is no dramatically difference between non-infected sample and the negative control by using the student's two-tail t test assay. Therefore, the developed assay is capable of rapidly and automatically detecting NNV and iridovirus in infected biological samples.

4. Conclusions In this study, an integrated microfluidic chip was developed to perform rapid, automatic and amplification-free detection of NNV and iridovirus in fish by utilizing molecular beacons and nucleic acid sequence-specific captured probes conjugated to magnetic beads. Compared with the developed microfluidic system, the conventional methods may provide comparable throughputs and similar sensitivity. However, the developed microfluidic system integrated four steps in one chip, including sample pretreatment for tested nucleic purification/concentration, nucleic acid denaturation, beacon hybridization and fluorescent detection.

202

Y.-C. Su et al. / Biosensors and Bioelectronics 63 (2015) 196–203

A

B

C

D

E

F

1600

1600

1200

Fluorescence signal (a.u.)

Fluorescence signal (a.u.)

1400

1000 800 600 400 200

1400 1200 1000 800 600 400 200 0

0 N

P

S1

S2

Sample

N

P

S1

S2

Sample

Fig. 5. The fluorescent signal intensity of (A) NNV RNA2 molecular beacon, and (B) iridovirus MCP molecular beacon. (C) The gel electrophoresis of the RT-PCR results. Lane L is the 100-bp DNA ladder, N is negative control uses ddH2O, P is positive control that contains NNV RNA2 plasmid DNA, and S1 and S2 contain NNV infected and non-infected samples, respectively. (D) The gel electrophoresis of the PCR results. Lane L is the 100-bp DNA ladder, N is negative control uses ddH2O, P is positive control that contains Iridovirus MCP plasmid DNA, and S1 and S2 contain Iridovirus infected and non-infected samples, respectively. The detection limits for (E) NNV RNA2 and (F) iridovirus MCP molecular beacons. The values and the error bars in this figure represent the mean and the standard deviation of three consecutive measurements.

Moreover, the entire process was automatically performed such that artificial errors among those steps and whole analytic operation could be dramatically reduced. The entire diagnostic process, including sample pre-treatment, hybridization and optical detection, was automatic performed by using a custom-made control system and could be completed in less than 30 min, which is much

faster when compared to the traditional nucleic acid amplification process. Both molecular beacons have had high specificity to influenza A/H1NI virus, influenza A/H3N2 virus, influenza B virus, dengue virus, enterovirus and Vibrio sp. Most importantly, the developed microfluidic system was able to diagnose the two strains of virus directly from fish samples. Therefore, this

Y.-C. Su et al. / Biosensors and Bioelectronics 63 (2015) 196–203

developed integrated microfluidic system may be a promising tool for the detection of fish pathogens in the future.

Acknowledgements The authors gratefully acknowledge the financial support provided to this study by the National Science Council in Taiwan (NSC102-2218-E-007-001). The authors would like to thank Dr. Kuangwen Hsieh for editing and providing valuable feedback. The authors also thank Dr. Chih-Peng Chang for providing some virus samples.

References Chang, W.H., Yang, S.Y., Lin, C.L., Wang, C.H., Li, P.C., Chen, T.Y., Jan, F.J., Lee, G.B., 2013a. Nanomed. Nanotechnol. Biol. Med. 9, 1274–1282. Chang, W.H., Yang, S.Y., Wang, C.H., Tsai, M.A., Wang, P.C., Chen, T.Y., Chen, S.C., Lee, G.B., 2013b. Sens. Actuators B: Chem. 180, 96–106. Chou, H.Y., Hsu, C.C., Peng, T.Y., 1998. Fish Pathol. 33, 201–206. Dalla Valle, L., Toffolo, V., Lamprecht, M., Maltese, C., Bovo, G., Belvedere, P., 2005. Vet. Microbiol. 110, 167–179. Darby, R.A., Sollogoub, M., McKeen, C., Brown, L., Risitano, A., Brown, N., Barton, C., Brown, T., Fox, K.R., 2002. Nucleic Acids Res. 30 (9), e39. Fukuda, Y., Nguyen, H.D., Furuhashi, M., Nakai, T., 1996. Fish Pathol. 31, 165–170. Harikrishnan, R., Balasundaram, C., Heo, M.S., 2010. Aquaculture 309, 1–14.

203

Hessel, V., Lowe, H., Schonfeld, 2005. Chem. Eng. Sci. 60, 2479–2501. Hung, L.Y., Chang, J.C., Tsai, Y.C., Huang, C.C., Chang, C.P., Yeh, C.S., Lee, G.B., 2014. Nanomedicine 10, 819–829. Klamp, T., Camps, M., Nieto, B., Guasch, F., Ranasinghe, R.T., Wiedemann, J., Petrášek, Z., Schwille, P., Klenerman, D., 2013. Sci. Rep. 3, 1852–1858. Kuo, H.C., Wang, T.Y., Chen, P.P., Chen, Y.M., Chuang, H.C., Chen, T.Y., 2011. J. Clin. Microbiol. 49, 1090–1096. Kuo, H.C., Wang, T.Y., Hsu, H.H., Lee, S.H., Chen, Y.M., Tsai, T.J., Ou, M.C., Ku, H.T., Lee, G.B., Chen, T.Y., 2012a. PLoS One 7, e42203. Kuo, H.C., Wang, T.Y., Hsu, H.H., Chen, P.P., Lee, S.H., Chen, Y.M., Tsai, T.J., Wang, C.K., Ku, H.T., Lee, G.B., Chen, T.Y., 2012b. PLoS One 7, e36183. Kurita, J., Nakajima, K., Hirono, I., Aoki, T., 1998. Fish Pathol. 33, 17–23. Lien, K.Y., Lee, S.H., Tsai, T.J., Chen, T.Y., Lee, G.B., 2009. Microfluid. Nanofluid. 7, 795–806. Lin, C.H., Lee, G.B., Chang, B.W., Chang, G.L., 2002. J. Micromech. Microeng. 12, 590–597. Mercier-Delarue, S., Vary, M., Plantier, J.C., Maillard, T., Adjout, Z., de Olivera, F., Schnepf, N., Maylin, S., Simon, F., Delaugerre, C., 2014. J. Clin. Microbiol. 52, 52–56. Shi, C.Y., Wei, Q., Gin, K.Y.H., Lam, T.J., 2003. J. Virol. Methods 107, 147–154. Tan, W.H., Wang, K.M., Drake, T.J., 2004. Curr. Opin. Chem. Biol. 8, 547–553. Tanaka, S., Takagi, M., Miyazaki, T., 2004. J. Fish Dis. 27, 385–399. Tyagi, S., Kramer, F.R., 1996. Nat. Biotechnol. 14, 303–308. Thorsen, T., Maerkl, S.J., Quake, S.R., 2002. Science 298, 580–584. Wang, C.H., Lien, K.Y., Wang, T.Y., Chen, T.Y., Lee, G.B., 2011. Biosens. Bioelectron. 26, 2045–2052. Weng, C.H., Lien, K.Y., Yang, S.Y., Lee, G.B., 2011. Microfluid. Nanofluid. 10, 301–310. Whitcombe, D., Theaker, J., Guy, S.P., Brown, T., Little, S., 1999. Nat. Biotechnol. 17, 804–807. Wong, M.L., Medrano, J.F., 2005. Biotechniques 39, 75–85. Zhang, C.Y., Chao, S.Y., Wang, T.H., 2005. Analyst 130, 483–488.