Rapid detection of multiple respiratory viruses based on microfluidic isothermal amplification and a real-time colorimetric method

Ruili Wang§ abc, Rongtao Zhao§ a, Yang Li a, Wen Kong a, Xudong Guo a, Yi Yang a, Feng Wu a, Wanying Liu a, Hongbin Song *a and Rongzhang Hao *a
aInstitute of Disease Control and Prevention, PLA, Beijing 100071, China. E-mail: hongbinsong@263.net; hrongzhang@163.com
bBeijing National Laboratory for Molecular Sciences (BNLMS), CAS Key Laboratory of Molecular Recognition and Function, CAS Research/Education Center for Excellence in Molecular Sciences, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China
cUniversity of Chinese Academy of Sciences, Beijing 100049, China

Received 13th August 2018 , Accepted 12th October 2018

First published on 15th October 2018


Respiratory viruses are major threats causing development of acute respiratory tract infections, which are common causes of illness and death throughout the world. Here, an integrated microsystem based on real-time colorimetry was developed for diagnosing multiple respiratory viruses. The microsystem employed magnetic beads for nucleic acid extraction and an eight-channel microfluidic array chip integrated with a loop-mediated isothermal amplification system for point-of-care screening of respiratory viruses. The overall detection process (including sample collection, nucleic acid extraction, sample loading, real-time detection, and signal output) could be completed within 1 h. Our results show that the developed method could specifically recognize influenza A virus subtypes (H1N1, H3N2, H5N1, and H7N9), influenza B virus, and human adenoviruses. The results obtained with 109 clinical samples indicate that the developed method has high specificity (100%, confidence interval 94.9–100.0) and sensitivity (96%, confidence interval 78.1–99.9). The integration of magnetic bead-based pre-treatment techniques and microfluidic isothermal amplification provides an effective solution for rapidly detecting etiological agents of respiratory diseases. The strategy of using a closed chip system and real-time colorimetry reduced aerosol contamination and ensured the accuracy of the results. The developed method provides an effective alternative for rapid point-of-care screening for viruses that cause respiratory disease syndromes and further aids in accurate and timely detection to control and prevent the spread of respiratory diseases caused by such pathogens.


Introduction

Respiratory diseases (especially acute respiratory tract infections) are still major public health threats in developing and developed countries, and cause considerable morbidity and fatality (particularly among infants and elderly people) as well as considerable economic losses.1,2 Respiratory pathogens include viruses, bacteria, mycoplasma, chlamydia, and other microorganisms, among which viruses account for approximately 90% of all respiratory diseases.3 The main viruses causing respiratory diseases include influenza A virus (FluA), influenza B virus (FluB), human adenovirus (HAdV), human rhinovirus (HRV), and respiratory syncytial virus (RSV).4,5 Influenza virus outbreaks have occurred every year since the 21st century; FluA outbreaks include the emergence of a mutated H5N1 strain in 1997, the global influenza pandemic and severe pneumonia caused by the mutated H1N1 swine flu in 2009, and the influenza outbreak caused by the mutant H7N9 avian influenza in 2013.4,6–9 Because of concerns regarding the disease severity during new or sudden virus emergence, it is paramount to establish rapid and accurate diagnostic technology to screen for respiratory disease syndromes and to enhance the surveillance for respiratory diseases. Rapid detection of respiratory viruses would be beneficial for taking suitable and timely measures to control and prevent the spread of respiratory infectious diseases caused by such pathogens.10,11

Respiratory syndromes share morbidity characteristics with some other non-viral infections, such as coughing and fever caused by bacteria, and distinguishing viral infections from bacterial infections provides a valuable reference for further symptomatic treatment and to avoid the abuse of antibiotics. Nonetheless, the detection of a single sample is time-consuming, whereas multi-channel screening can quickly detect multiple samples, by quickly distinguishing between viral and non-viral infections and saving time for further symptomatic treatment. The development of methods for simultaneous detection of multiple pathogens can effectively shorten the time required for respiratory syndrome screening.

The conventional diagnosis of respiratory viruses is mainly based on cell culture, serological, immunological, and molecular biological methods.12,13 To date, conventional culture methods are still being used to detect respiratory viruses. The results of culture methods are usually accurate; however, the operating procedures are complicated and time-consuming.14 Other detection methods (such as immunological methods) also enable rapid detection capability, but their accuracy and sensitivity are much less than those of nucleic acid detection, making it difficult to meet the requirements of point-of-care tests (POCTs) for respiratory viruses.15–17 In recent years, molecular diagnostic methods, especially polymerase chain reaction (PCR), have been widely used to detect respiratory viruses and other pathogens.18–20 Based on the specific and rapid detection of target sequences of interest, real-time PCR is a popular choice for the molecular diagnosis of pathogens.18,21 However, real-time PCR relies upon thermal cycling instruments and fluorescence detection, which significantly hinders its applications in POCT situations for detecting respiratory viruses. Therefore, the development of new methods and detection devices for POCTs is needed. The loop-mediated isothermal amplification (LAMP) method has attracted widespread attention as an alternative to real-time PCR.22–25 The critical difference is that isothermal amplification does not involve thermal cycling and the polymerase can amplify the target sequence at a constant temperature.22 The only equipment needed for isothermal amplification is a regular laboratory water bath or a heat block that provides a constant temperature of 60–65 °C, which eliminates the requirement for a sophisticated instrument. In addition, isothermal amplification offers the advantages of simple operation and a significantly shortened amplification time.24 Thus, the LAMP method has advantages for rapid, on-the-spot detection of respiratory viruses.

The LAMP method has been used to detect a variety of pathogens, such as influenza virus, adenovirus, RSV, severe acute respiratory syndrome, and Middle East respiratory syndrome coronavirus.24,26–29 To date, LAMP has also been applied for the detection of other samples.30 Although LAMP technology has advantages for rapid detection, it is also susceptible to aerosol contamination, which causes false positive results and is problematic for POCTS. There is an urgent need to develop a sealed instrument for isothermal amplification, and for rapid detection on-the-spot. It is also necessary to develop a suitable portable instrument for rapid detection of multiple respiratory viruses.

In recent years, the LAMP method has often been combined with microfluidic systems for rapid and reliable detection of respiratory viruses, which has facilitated the use of portable detection platforms and thus enabled on-the-spot treatment of epidemics.31,32 Previous results demonstrated the feasibility of using a microfluidic LAMP array for detecting respiratory viruses. Previous reports have shown success using the integrated microfluidic LAMP method for on-the-spot pathogen diagnostics.10,33 Although sample preparation was integrated into the chip, the design and complexity of operation make it difficult to handle multiple samples.33 Some researchers have reported the use of integrated lab-on-a-disc systems with colorimetric detection for rapid molecular diagnosis of pathogens.10,33–35 To date, the formation of LAMP reaction products has been detected based on the formation of a white precipitate, or with indicators such as calcein, SYBR green, and hydroxynaphthol blue (HNB),22,23,36–38 which is convenient for the result readout especially by the naked eye. However, it can be difficult to interpret the test results correctly when little color change occurs after the reaction is complete, making subjective interpretations with the naked eye inevitable, which can lead to misinterpretation of the results. Thus, it is essential to develop a platform for automatic monitoring of LAMP,39 and the LAMP can be recorded in real-time or determined at end-point.40,41 Real-time LAMP could provide abundant information for the amplification, which would help improve the accuracy and quantification of the detection. End-point LAMP is cost-effective and convenient, but less accurate. Besides, the LAMP reaction is susceptible to aerosol contamination, which may cause false positive results. Moreover, integrated detection equipment with suitable capability should be developed to meet the requirement for the rapid detection of respiratory viruses in the field. However, until now, the development of a portable detection platform is still in its infancy. Few studies in previous reports have realized the rapid POCT detection of multiple respiratory viruses at the same time. In this study, we developed a LAMP-integrated microfluidic chip system for multiplexed respiratory virus assays (LMCS-MRVAs). The developed chip was constructed with eight channels and allowed rapid detection on a single chip. The chip system was closed, which reduced aerosol contamination during LAMP. The real-time colorimetric detection instrument developed in this study could draw amplification curves based on the color of the chip microchambers, which changed from violet to sky blue when amplification occurred. The real-time colorimetric design strategy increased the accuracy of the test results. We developed monodisperse magnetic nanoparticles for rapid nucleic acid extraction from samples. We simplified the chip design and used low-cost polymer materials to fabricate the chip, which can enable the cost-effective mass production of the chip. The developed LMCS-MRVA could detect four chips at the same time, which allowed detection of multiple respiratory viruses simultaneously and screening for multiple respiratory syndromes.

Experimental

Fig. 1 shows the details of the whole detection process. The whole detection process includes sample collection, nucleic acid extraction, sample loading, real-time detection, and signal output. The whole detection process was completed within 1 h, and the integrated system developed in this study was suitable for the rapid detection of clinical respiratory viruses.
image file: c8lc00841h-f1.tif
Fig. 1 Schematic illustration of the strategy used to detect multiple respiratory viruses with a LAMP-integrated microfluidic chip system for multiplexed respiratory virus assays (LMCS-MRVAs).

Materials and reagents

All the chemicals were obtained from commercial sources and used without further purification. Hydroxynaphthol blue and betaine were purchased from Sigma-Aldrich, and ThermoPol reaction buffer (20 mM Tris-HCl, 2 mM MgSO4, 10 mM KCl, 10 mm (NH4)2SO4 and 0.1% Triton X-100; pH 8.8), MgSO4, and BST 2.0 WarmStart® DNA polymerase were purchased from New England BioLabs. dNTPs and AMV reverse transcriptase were purchased from Promega. Monodisperse magnetic beads were prepared in our laboratory. GelRed nucleic acid gel stain was purchased from Biotium. A DNA marker was purchased from Transgen. 3-(2-Aminoethylamino)propyldimethoxymethylsilane (AEAPS) was purchased from Dow Corning. All the solutions were prepared with DNase/RNase-free distilled water. The NanoDrop spectrophotometer was from Thermo Scientific. The PCR thermal cycler (Veriti) was from Applied Biosystems. The iQ5 real-time PCR detection system was from Bio-Rad, and the S-4800 scanning electron microscope was from Hitachi.

All the DNA oligonucleotides used as LAMP primers and PCR primers in this work were supplied by Sangon Biotech (Shanghai, China), and their sequences are shown in the ESI Tables S1 and S2. In Table S1, F3 and B3 represent outer primers. FIP and BIP represent inner primers. LF and LB represent loop primers. Six LAMP primers were used in this study. In Table S2, F and R represent PCR primers, which were used for comparison with the established microfluidic system and LAMP methods.

Preparation of samples

Viral lysates from throat swabs provided by the Chinese Center for Disease Control and Prevention (Beijing, China) were used as H5N1 and H7N9 FluA RNA templates. Clinical FluA (H1N1 and H3N2), FluB, and HAdV samples were provided by the Beijing Center for Disease Prevention and Control (Beijing, China). The identities of those viruses were confirmed by real-time PCR according to the World Health Organization guidelines.42 Routine laboratory methods were used to propagate the influenza viruses. The viruses were cultured in chicken embryo cells, as described in the World Health Organization guidelines. The nucleic acids from FluA (H1N1 and H3N2), FluB, and HAdV were isolated from the infected chicken embryo cells, using the homemade monodisperse magnetic beads. The procedure took approximately 30 min to complete.

Fabrication of monodisperse magnetic beads

Monodisperse Fe3O4 nanoparticles. A suspension of sodium acetate (2.0 g, 14.7 mmol) and iron(III) chloride hexahydrate (0.25 g, 0.9 mmol) in ethylene glycol solution (50 mL) was stirred at room temperature for 1 h. Subsequently, the reaction mixture was transferred to a reactor and stirred for 10 h at 100 °C.
Monodisperse magnetic beads. A suspension of Fe3O4 nanoparticles (100 mg, 62.6 mmol) was prepared in deionized water (20 ml), after which anhydrous ethanol (80 ml) and 28 wt% NH3·H2O (1 ml) were added successively into the reaction mixture. Next, ethyl silicate (1 ml) was added dropwise to the reaction mixture and stirred at room temperature for 6 h. After performing the reaction, Fe3O4@SiO2 core/shell nanoparticles were prepared and treated with the silane coupling agent 3-(2-aminoethylamino)propyldimethoxymethylsilane (to obtain monodisperse magnetic beads) and dried in 60 °C. The monodisperse Fe3O4 nanoparticles and monodisperse magnetic beads were characterized by field emission scanning electron microscopy (FESEM).
Real-time detection microfluidic chip and instrument. The LMCS-MRVA system examined in this study included a disposable eight-channel microfluidic chip and a real-time detection instrument. The instrument was capable of isothermal incubation and real-time colorimetric detection, and could simultaneously detect four chips. The fabricated microfluidic chip and real-time detection instrument are shown in Fig. 2A and B. The chip was produced using an internal injection-molding method using polycarbonate. The microfluidic chip contained eight independent reaction microchambers (2 × 1 × 1 mm), with a corresponding reaction volume of approximately 2 μl. The microchip microchannels had a width of about 500 μm and a depth of 200 μm, and flowed through the injection hole, microchambers, and connecting channel. The microfluidic chip was sealed with a polymer film with a thickness of 200 μm, which could be appropriately pressurized to ensure that the chip was sealed with no bubbles. After the nucleic acids were extracted from the samples, the nucleic acids were mixed with reaction solution, and the mixture was added to the microchambers of the microfluidic chip using a micropipette. Each solution was then diverted into the eight microchambers along the different channels and sealed with a film to ensure that the LAMP was carried out within a sealed chip. As LAMP proceeded, the color of the reaction system gradually changed from violet to sky blue, and the detection instrument developed in this study drew a real-time amplification curve based on the colorimetric change, as shown in Fig. 2C. The measurement system that we developed was hand-held, with a size of 41 × 31 × 12 cm and a weight of 2.7 kg. The schematic drawing of the measurement system is shown in Fig. S1. The instrument was mainly composed of the heating module, the detection module, the control and processing display module and the outer cover. The heating module consists of a heating film and a heating machine. The detection module is composed of a high-definition camera, a light source and a fixed structure. The control and processing display module is composed of a circuit board, an ARM board and a screen. The instrument used a double-row array of white LEDs as the light source and the detector was a high-definition color camera. During the LAMP, the RGB color values in each reaction microchamber were obtained by an image recognition algorithm, and the blue component was taken as the real-time curve value to draw the real-time LAMP curve. The time required for the real-time LAMP signal to exceed the threshold signal was used for quantitation, similar to the threshold cycle number (Ct) in real-time PCR. Based on the experimental results, the negative signal intensity was less than 5, so the positive signal intensity was set to 15, i.e., three times the signal-to-noise ratio. To define the start time of amplification, we drew a straight line parallel to the abscissa with a signal value of 15 on the ordinate, and added a perpendicular line from the abscissa with the point intersecting the amplification curve. The intersection with the abscissa represented the amplification start time. With this detection system, the eight-channel microfluidic chip was disposable, and the real-time detection instrument can be reused.
image file: c8lc00841h-f2.tif
Fig. 2 The developed microfluidic chip, real-time detection instrument, and representative real-time amplification curves. (A) Photograph of a microfluidic chip. (B) Real-time detection instrument. (C) Real-time amplification curves.

Establishment of the LMCS-MRVA method

Design of LAMP primers. Based on the information from the Global Initiative on Sharing All Influenza Data (https://www.gisaid.org/), the hemagglutinin gene sequences of human FluA (H1N1, H3N2, H5N1, and H7N9) and FluB, as well as the hexon gene sequence of HAdV, were searched using BLAST, and the conserved sequences were analyzed using Mega software. The primers were designed using Primer Explorer V5 (https://primerexplorer.jp/lampv5/index.html, Renzo Chemical Co., Ltd., Japan) based on the conserved sequences. The primers included two outer primers (F3 and B3), two inner primers (FIP and BIP), and two loop primers (LF and LB). The primers could specifically identify eight target sequences of the gene of interest. The primers were synthesized by Sangon Biotech. See Table S1 for details.
LAMP-integrated microfluidic chip system. All the solutions were prepared using ultrapure DNase/RNase-free distilled water. With each respiratory virus, the LAMP primers were used at a concentration of 1.6 μM each for FIP and BIP, 0.2 μM each for F3 and B3, and 0.8 μM each for LF and LB. Random primers were used as a negative control. Each corresponding LAMP primer mixture and negative control were embedded in the eight different microchambers of the microfluidic chip in advance (1: H1N1 primers, 2: H3N2 primers, 3: H5N1 primers, 4: H7N9 primers, 5: FluB primers, 6: HAdV primers, 7: HAdV primers, and 8: negative control) and then dried before sealing. Seventy microliters of the whole reaction system contained 5 μl of the nucleic acid sample to be tested extracted with the homemade magnetic beads. The reaction mixture also included 7.5 μl ThermoPol buffer, 4.5 μl MgSO4 (120 mM), 10.5 μl dNTPs (10 mM), 3 μl AMV reverse transcriptase (1000 units), 4 μl betaine (250 mM), 1 μl HNB (20 mM), 3 μl Bst 2.0 WarmStart® DNA polymerase (8000 units), and 31.5 μl DNase/RNase-free distilled water. The reaction system was mixed and injected into the injection hole of the chip to the separate microchambers through the eight microchannels. Then, the injection port and channels were re-sealed. Finally, the chips were inserted into the detection instrument. The reaction temperature was preset to 65 °C for 60 min for isothermal amplification and 80 °C for 10 min to terminate the reaction. HNB served as a metal ion indicator in the system. Initially, Mg2+ binding to HNB caused the color of the reaction system to be violet. As the reaction proceeded, Mg2+ reacted with precipitated pyrophosphate ions to form the magnesium pyrophosphate precipitate, such that HNB lost the bound Mg2+, making the system color become sky blue (Fig. S2). In the absence of LAMP, the system maintained a violet color. In order to reduce the influence of bubbles on the LAMP, we treated the reagents with an ultrasonic procedure before amplification, and compared every microchamber with the negative control microchamber to give the amplification curve. Moreover, we used the developed LMCS-MRVA system to detect those six kinds of respiratory virus, and used the gel electrophoresis assay to compare the detection results (Fig. S3).
Specificity and sensitivity assays for the LMCS-MRVA system. To evaluate the specificity of the fabricated LMCS-MRVA system, various LAMP primers were embedded in advance as described in the LAMP-integrating microfluidic chip system, and FluA (H1N1, H3N2, H5N1, and H7N9), FluB, HAdV (HAdV-3 and HAdV-7) and negative control (nucleic acids extracted from other non-respiratory virus samples) were used to test the established method.

To evaluate the sensitivity of the fabricated LMCS-MRVA system, 10 ng μl−1 nucleic acid solution was tested at various concentrations (with dilution ratios of 1, 1/10, 1/100, 1/1000, 1/10[thin space (1/6-em)]000, 1/100[thin space (1/6-em)]000, and 1/1[thin space (1/6-em)]000[thin space (1/6-em)]000). After dilution, 2 μl of the different target nucleic acid templates were embedded in the 8 different microchambers of the microfluidic chip in advance. The reactions were mixed, and the system was operated as described in the LAMP-integrated microfluidic chip system section above.

Clinical sample evaluation

The clinical samples were preserved in the laboratory after treating an outbreak of respiratory-like cases. The number of samples was 109. The nucleic acids were extracted from the samples and then tested with the LMCS-MRVA system and by real-time PCR. The detection results were analyzed using the MedCalc software “diagnostic test evaluation calculator” to evaluate the specificity and sensitivity, and the 95% confidence intervals (https://www.medcalc.org/calc/diagnostic_test.php).

Results and discussion

Characterization and nucleic acid extraction efficiency of monodisperse magnetic beads

To quickly extract nucleic acids from multiple samples to be tested, we developed our own monodisperse magnetic beads. The FESEM results of the monodisperse Fe3O4 nanoparticles showed that the monodisperse Fe3O4 nanoparticles had a relatively uniform size and shape (with a large specific surface area) and good dispersion in a certain medium, which can effectively shorten the sedimentation time of the material (Fig. 3A). The above advantages are required for nucleic acid extraction with the magnetic beads. The roughness of the bead surface was determined by the degree of nano-monodispersity, which was favorable for dispersing the materials in solution. The particle size was approximately 300 nm, and no agglomeration phenomenon was found. The FESEM results showed that the silicon matrix layer evenly coated the surface of the magnetic nanoparticles and formed the core–shell magnetic beads, which had smooth surfaces and showed good dispersion (Fig. 3B). The particle size was approximately 320 nm, which indicates that the coated silicon matrix shell was about 20 nm. The silane coupling agent AEAPS did not change the morphology of the Fe3O4@SiO2 core/shell nanoparticles after sialylation. Our homemade magnetic beads were compared with commercial magnetic beads in the extraction efficiency of nucleic acid. The results in Fig. 3C show that the nucleic acid concentration extracted by the commercial magnetic beads from companies a, b, and c and our homemade magnetic beads was 46.83 ng μl−1, 21.28 ng μl−1, 5.76 ng μl−1 and 91.22 ng μl−1, respectively, and the value of OD260/280 was 1.95, 1.87, 1.70, and 2.06, respectively. The results demonstrated that the homemade magnetic beads had an improved quality compared to the commercial polydisperse magnetic beads (Fig. 3C). Those results indicate that the homemade monodisperse magnetic beads could be used to extract nucleic acid from the samples. In addition, the homemade magnetic beads could be used for the extraction of nucleic acid without an instrument, making them suitable for detecting samples in POCT settings.
image file: c8lc00841h-f3.tif
Fig. 3 Monodisperse Fe3O4 nanoparticles, monodisperse magnetic beads, and comparison of the results with commercially available magnetic beads. (A) FESEM images of the monodisperse Fe3O4 nanoparticles. (B) SEM images of the monodisperse magnetic beads. (C) Comparison of the results with the commercially available magnetic beads. The results shown in a–c represent the commercially available polydisperse magnetic beads. Black columns represent nucleic acid concentrations, and grey columns represent OD 260/280 values. The data shown represent the monodisperse magnetic beads used in this study. Error bars represent the standard deviation of three independent experiments.

Specificity and sensitivity of the LMCS-MRVA system

To verify the detection performance of the LMCS-MRVA system, we carried out the specificity and sensitivity experiments. For the specificity and sensitivity testing of the LMCS-MRVA system, we used the microfluidic chips with the embedded LAMP primers and negative control, as described in the Experimental section.

The specificity test results of the LMCS-MRVA using 10 ng μl−1 virus solution are shown in Fig. 4. We use a microfluidic chip embedded with the LAMP primers. When the corresponding test sample was added, only the corresponding well had an amplification signal, and the remaining wells showed no amplification. For example, when H1N1 was detected, only the corresponding microchamber 1, in which the H1N1 LAMP primer was embedded showed an amplification curve, and the rest of the wells showed no amplification signal. The results showed that the LMCS-MRVA had high detection specificity for each of the six kinds of respiratory viruses studied. The good specificity benefits from the channel design of the chip. The elongated channel operates as a liquid seal and at the same time as a link between other connections, so that the reaction holes are independent and do not interfere with each other, thus, providing reliable results for rapid detection of multiple respiratory viruses.


image file: c8lc00841h-f4.tif
Fig. 4 Results of the specific evaluation of the different virus samples by the array chips embedded with the different primers. Abscissa: 1, H1N1 primer; 2, H3N2 primer; 3, H5N1 primer; 4, H7N9 primer; 5, Flu B primer; 6, HAdV primer; 7, HAdV primer; 8, negative control. The specificity of the test results is shown for (A) H1N1, (B) H3N2, (C) H5N1, (D) H7N9, (E) Flu B, and (F) HAdV. Error bars represent the standard deviation of three independent experiments.

The sensitivity test results of the LMCS-MRVA system obtained using 10-fold serial dilutions of virus solution in DNase/RNase-free distilled water (diluted from 1 × 10−1 to 1 × 10−5 ng μl−1) are shown in Fig. 5. The time required for detection (i.e., the initiation time) gradually increased with decreasing target concentrations, and the detection began within 30 min. The expanded views around the cross points with a threshold value of 15 are inserted in each amplification curve graph. The LMCS-MRVA detected all the six kinds of respiratory viruses at a concentration of 10–100 fg μl−1 (Fig. 5). The limit of detection (LOD) for the LMCS-MRVA system was 10 to 100 times better than that of PCR (Fig. S4), which is close to real-time PCR (Fig. S5).


image file: c8lc00841h-f5.tif
Fig. 5 Sensitivity and standard curves for the LMCS-MRVA system. Sensitivity test results of the LMCS-MRVA using 10-fold serial dilutions of virus stocks in PBS (diluted from 1 × 10−1 to 1 × 10−5 ng μl−1) as determined using amplification plots (A1, B1, C1, D1, E1, and F1) and standard curves (A2, B2, C2, D2, E2, and F2). The error bars represent the standard deviation of three independent experiments.

To quantify the target nucleic acid concentrations, we established a linear equation describing the relationship between the microfluidic LAMP threshold value and the target concentration. The concentration of virus per test and the threshold time showed a good linear relationship, with a linear correlation coefficient of 0.97–0.99. We established a linear curve equation, which was used to calculate the LOD. Taking FluA (H1N1) as an example, the resulting linear equation was Y = −5.203[thin space (1/6-em)]log[thin space (1/6-em)]X + 34.448 with a correlation coefficient of 0.9894. The LOD of FluA (H1N1) was calculated as 4 fg μl−1 using the equation LOD = 3σ/slope (where σ is the relative standard deviation over the negative response, n = 3).43,44 In addition, the LODs for FluA (H3N2, H5N1, and H7N9) were approximately 4 fg μl−1, 4 fg μl−1, and 2 fg μl−1, respectively, whereas those for FluB and HAdV were about 3 fg μl−1 and 4 fg μl−1, respectively. The above data show that we could use the LMCS-MRVA system for quantitative detection of all the six kinds of respiratory viruses studied; thus, the quantitative LAMP data were obtained, and the system is not limited to qualitative studies.

Clinical sample evaluation results

To further investigate the performance of the LMCS-MRVA system, 109 clinical throat swab samples stored in our laboratory were tested and these results were compared with those of the real-time PCR method (Table 1). It is shown that the detection result of the LMCS-MRVA method for one sample did not match with that of real-time PCR. The real-time PCR detection results in a high Ct value (>35) and bad exponential amplification curve, which indicated weakly positive, and we added a column to describe the result. Anyway, the LMCS-MRVA system presents a high sensitivity (>96%) and specificity (100%). Statistical analysis was conducted using the online “diagnostic test evaluation calculator.”45 The results of the LMCS-MRVA system were comparable to the real-time PCR results, indicating that the LMCS-MRVA system had high accuracy and met the rapid detection requirement for clinical respiratory virus samples. The LMCS-MRVA system only required 1 h to complete the detection, which is faster than real-time PCR (3–4 h); thus, the LMCS-MRVA system is suitable for rapid detection of multiple respiratory viruses on-the-spot.
Table 1 Detection results for the LMCS-MRVA system and real-time PCR
Clinical sample (sample number) Test results
LMCS-MRVA Real-time PCR Sensitivity (95% CIa) Specificity (95% CI)
Positive Weakly positive Positive Weakly positive
a CI: confidence interval.
FluA H1N1 (32) 32 0 32 0 100% (89.1–100.0) 100% (95.3–100.0)
FluA H3N2 (15) 15 0 15 0 100% (78.2–100.0) 100% (96.2–100.0)
FluB (23) 22 0 22 1 96% (78.1–99.9) 100% (95.8–100.0)
HAdV (39) 39 0 39 0 100% (91.0–100.0) 100% (94.9–100.0)
Total (109) 108 0 108 1 96% (78.1–99.9) 100% (94.9–100.0)
Negative control
Positive control + +


Conclusions

In summary, we developed the LAMP-integrated microfluidic chip system for multiplexed respiratory virus assays (LMCS-MRVAs), which incorporates magnetic beads for pre-treating samples, an eight-channel microfluidic array chip integrated with the LAMP method for isothermal amplification, and real-time colorimetric detection for point-of-care testing of multiple respiratory viruses, enabling quick diagnosis. The detection system could prevent typical problems associated with the LAMP by fabricating relatively independent and totally sealed reaction microchamber to avoid aerosol contamination, reagent evaporation, and cross-contamination during LAMP reactions. Using the LMCS-MRVA showed that multiplex respiratory virus detection could be accomplished with an LOD of 2–4 fg μl−1 from the throat swab samples with high specificity (100%, confidence interval 94.9–100.0) and sensitivity (96%, confidence interval 78.1–99.9) within 1 h, which is much quicker than commercial methods (3–4 h). The results showed that our LMCS-MRVA system is suitable for rapid detection with high sensitivity, as well as low cost. The high performance of the LMCS-MRVA system provides great potential as a user-friendly method that can be applied for use in resource-limited environments. Based on this detection strategy, we could develop more multiple microfluidic chips to screen for other respiratory viruses through integrating different LAMP primers.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

This work was funded by the Mega-projects of Science and Technology Research (grant numbers 2018ZX10711001003002 and 2017ZX10304403) and the National Natural Science Foundation of China (grant number 81871738). Supporting work in the manufacture of the measurement device by Guohao Zhang and Guojun Wu from Baicare Biotech (Beijing) Co., Ltd. is greatly appreciated.

Notes and references

  1. A. D. Lopez, C. D. Mathers, M. Ezzati, D. T. Jamison and C. J. L. Murray, Lancet, 2006, 367, 1747–1757 CrossRef.
  2. A. R. Falsey, P. A. Hennessey, M. A. Formica, C. Cox and E. E. Walsh, N. Engl. J. Med., 2005, 352, 1749–1759 CrossRef CAS PubMed.
  3. D. M. Morens, G. K. Folkers and A. S. Fauci, Nature, 2004, 430, 242 CrossRef CAS PubMed.
  4. D. J. Jamieson, M. A. Honein, S. A. Rasmussen, J. L. Williams, D. L. Swerdlow, M. S. Biggerstaff, S. Lindstrom, J. K. Louie, C. M. Christ, S. R. Bohm, V. P. Fonseca, K. A. Ritger, D. J. Kuhles, P. Eggers, H. Bruce, H. A. Davidson, E. Lutterloh, M. L. Harris, C. Burke, N. Cocoros, L. Finelli, K. F. MacFarlane, B. Shu and S. J. Olsen, Lancet, 2009, 374, 451–458 CrossRef.
  5. W. W. Thompson, D. K. Shay and E. Weintraub, et al. , JAMA, 2003, 289, 179–186 CrossRef PubMed.
  6. P. K. S. Chan, Clin. Infect. Dis., 2002, 34, S58–S64 CrossRef PubMed.
  7. R. Gao, B. Cao, Y. Hu, Z. Feng, D. Wang, W. Hu, J. Chen, Z. Jie, H. Qiu, K. Xu, X. Xu, H. Lu, W. Zhu, Z. Gao, N. Xiang, Y. Shen, Z. He, Y. Gu, Z. Zhang, Y. Yang, X. Zhao, L. Zhou, X. Li, S. Zou, Y. Zhang, X. Li, L. Yang, J. Guo, J. Dong, Q. Li, L. Dong, Y. Zhu, T. Bai, S. Wang, P. Hao, W. Yang, Y. Zhang, J. Han, H. Yu, D. Li, G. F. Gao, G. Wu, Y. Wang, Z. Yuan and Y. Shu, N. Engl. J. Med., 2013, 368, 1888–1897 CrossRef CAS PubMed.
  8. S. Dong, R. Zhao, J. Zhu, X. Lu, Y. Li, S. Qiu, L. Jia, X. Jiao, S. Song, C. Fan, R. Hao and H. Song, ACS Appl. Mater. Interfaces, 2015, 7, 8834–8842 CrossRef CAS PubMed.
  9. R. Hao, Y. Zhang, P. Li, Y. Wang, S. Qiu, Z. Li, L. Wang, Z. Wu, R. Lin, N. Liu, G. Yang, C. Yang, J. D. Klena and H. Song, Clin. Infect. Dis., 2013, 57, 776–778 CrossRef PubMed.
  10. B. H. Park, S. J. Oh, J. H. Jung, G. Choi, J. H. Seo, D. H. Kim, E. Y. Lee and T. S. Seo, Biosens. Bioelectron., 2017, 91, 334–340 CrossRef CAS PubMed.
  11. P. Craw and W. Balachandran, Lab Chip, 2012, 12, 2469–2486 RSC.
  12. C. H. Krause, K. Eastick and M. M. Ogilvie, J. Clin. Virol., 2006, 37, 184–189 CrossRef CAS PubMed.
  13. F. François, V. Astrid, C. N. Delphine, S. Sandrine, D. Julia, L. Loïc, G. Stéphanie, P. Joëlle, E. Philippe and B. Jacques, J. Med. Virol., 2006, 78, 1498–1504 CrossRef PubMed.
  14. A. J. Eisfeld, G. Neumann and Y. Kawaoka, Nat. Protoc., 2014, 9, 2663 CrossRef CAS PubMed.
  15. A. W. Chung, M. P. Kumar, K. B. Arnold, W. H. Yu, M. K. Schoen, L. J. Dunphy, T. J. Suscovich, N. Frahm, C. Linde, A. E. Mahan, M. Hoffner, H. Streeck, M. E. Ackerman, M. J. McElrath, H. Schuitemaker, M. G. Pau, L. R. Baden, J. H. Kim, N. L. Michael, D. H. Barouch, D. A. Lauffenburger and G. Alter, Cell, 2015, 163, 988–998 CrossRef CAS PubMed.
  16. R. M. Zellweger and S. Shresta, Front. Immunol., 2014, 5, 151 Search PubMed.
  17. R. R. Thangavel and N. M. Bouvier, J. Immunol. Methods, 2014, 410, 60–79 CrossRef CAS PubMed.
  18. K. J. Henrickson, Pediatr. Infect. Dis. J., 2004, 23, S6–S10 CrossRef PubMed.
  19. E. Spackman, D. A. Senne, T. J. Myers, L. L. Bulaga, L. P. Garber, M. L. Perdue, K. Lohman, L. T. Daum and D. L. Suarez, J. Clin. Microbiol., 2002, 40, 3256–3260 CrossRef CAS PubMed.
  20. M. Munch, L. P. Nielsen, K. J. Handberg and P. H. Jørgensen, Arch. Virol., 2001, 146, 87–97 CrossRef CAS PubMed.
  21. L. J. R. van Elden, M. Nijhuis, P. Schipper, R. Schuurman and A. M. van Loon, J. Clin. Microbiol., 2001, 39, 196–200 CrossRef CAS PubMed.
  22. T. Notomi, H. Okayama, H. Masubuchi, T. Yonekawa, K. Watanabe, N. Amino and T. Hase, Nucleic Acids Res., 2000, 28, E63 CrossRef CAS PubMed.
  23. N. Tomita, Y. Mori, H. Kanda and T. Notomi, Nat. Protoc., 2008, 3, 877 CrossRef CAS PubMed.
  24. Y. Mori and T. Notomi, J. Infect. Chemother., 2009, 15, 62–69 CrossRef CAS PubMed.
  25. K. Nagamine, T. Hase and T. Notomi, Mol. Cell. Probes, 2002, 16, 223–229 CrossRef CAS PubMed.
  26. P. G. Ziros, P. A. Kokkinos, A. Allard and A. Vantarakis, Food Environ. Virol., 2015, 7, 276–285 CrossRef CAS PubMed.
  27. H. Zhou, M. Zhao, X. Li, D. Zhang, S. Zhou, C. Chen, Z. Feng and X. Ma, Arch. Virol., 2017, 162, 1311–1318 CrossRef CAS PubMed.
  28. L. L. M. Poon, C. S. W. Leung, M. Tashiro, K. H. Chan, B. W. Y. Wong, K. Y. Yuen, Y. Guan and J. S. M. Peiris, Clin. Chem., 2004, 50, 1050–1052 CrossRef CAS PubMed.
  29. S. H. Lee, Y. H. Baek, Y. H. Kim, Y. K. Choi, M. S. Song and J. Y. Ahn, Front. Microbiol., 2017, 7, 2166 Search PubMed.
  30. N. Shao, J. Chen, J. Hu, R. Li, D. Zhang, S. Guo, J. Hui, P. Liu, L. Yang and S. C. Tao, Lab Chip, 2017, 17, 521–529 RSC.
  31. J. Hiltunen, C. Liedert, M. Hiltunen, O. H. Huttunen, J. Hiitola Keinänen, S. Aikio, M. Harjanne, M. Kurkinen, L. Hakalahti and L. P. Lee, Lab Chip, 2018, 18, 1552–1559 RSC.
  32. Y. D. Ma, K. Luo, W. H. Chang and G. B. Lee, Lab Chip, 2018, 18, 296–303 RSC.
  33. S. J. Oh, B. H. Park, G. Choi, J. H. Seo, J. H. Jung, J. S. Choi, D. H. Kim and T. S. Seo, Lab Chip, 2016, 16, 1917–1926 RSC.
  34. J. F. C. Loo, C. C. H. Leung, H. C. Kwok, S. Y. Wu, I. L. G. Law, M. L. Chin, M. Hui, S. K. Kong and H. P. Ho, Procedia Technol., 2017, 27, 224–225 CrossRef.
  35. F. Stumpf, F. Schwemmer, T. Hutzenlaub, D. Baumann, O. Strohmeier, G. Dingemanns, G. Simons, C. Sager, L. Plobner, F. von Stetten, R. Zengerle and D. Mark, Lab Chip, 2016, 16, 199–207 RSC.
  36. X. Fang, Y. Liu, J. Kong and X. Jiang, Anal. Chem., 2010, 82, 3002–3006 CrossRef CAS PubMed.
  37. M. Goto, E. Honda, A. Ogura, A. Nomoto and K.-I. Hanaki, BioTechniques, 2009, 46, 167–172 CrossRef CAS PubMed.
  38. D. Yuan, J. Kong, X. Li, X. Fang and Q. Chen, Sci. Rep., 2018, 8, 8682 CrossRef PubMed.
  39. A. Qin, L. T. Fu, J. K. F. Wong, L. Y. Chau, S. P. Yip and T. M. H. Lee, ACS Appl. Mater. Interfaces, 2017, 9, 10472–10480 CrossRef CAS PubMed.
  40. T. Wang, J. P. Devadhasan, D. Y. Lee and S. Kim, Anal. Sci., 2016, 32, 653–658 CrossRef CAS PubMed.
  41. A. Sayad, F. Ibrahim, S. Mukim Uddin, J. Cho, M. Madou and K. L. Thong, Biosens. Bioelectron., 2018, 100, 96–104 CrossRef CAS PubMed.
  42. http://www.who.int/influenza/gisrs_laboratory/molecular_diagnosis/en/, 2017.
  43. S. Liu, N. Xu, C. Tan, W. Fang, Y. Tan and Y. Jiang, Anal. Chim. Acta, 2018, 1018, 86–93 CrossRef CAS PubMed.
  44. D. Bai, D. Ji, J. Shang, Y. Hu, J. Gao, Z. Lin, J. Ge and Z. Li, Sens. Actuators, B, 2017, 252, 1146–1152 CrossRef CAS.
  45. C. h. Gao, J. Y. Wang, F. Shi, D. Steverding, X. Wang, Y. t. Yang and X. N. Zhou, Parasites Vectors, 2018, 11, 311 CrossRef PubMed.

Footnotes

Authors' contributions: Rongzhang Hao, Ruili Wang and Hongbin Song designed the research plan; Ruili Wang, Yang Li, Wen Kong and Xudong Guo performed the experiments; Ruili Wang, Rongtao Zhao, Yi Yang, Feng Wu and Wanying Liu wrote the paper.
Electronic supplementary information (ESI) available. See DOI: 10.1039/c8lc00841h
§ These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2018