Facile, Rapid, and Low-Cost Detection for Influenza Viruses and Respiratory Syncytial Virus Based on a Catalytic DNA Assembly Circuit.

Influenza viruses and respiratory syncytial virus (RSV) have contributed to severe respiratory infections, causing huge economic and healthcare burdens. To achieve rapid and precise detection of influenza viruses and RSV, we proposed a catalytic hairpin assembly (CHA) combined with the lateral flow immunoassay (CHA-LFIA) detection method. The presence of the target RNA triggers the initiation of CHA circuits. H1/H2 complexes, the amplified signal products, which were labeled with digoxin and biotin, were detected with a highly sensitive lateral flow immunoassay system. The sensitivity of the CHA-LFIA system to influenza A and B viruses and RSV reached up to 1, 1, and 5 pM, respectively. In addition, this method exhibited excellent capability for differentiating between target RNA and base-mismatched RNA. The results demonstrated that an enzyme-free, rapid, highly sensitive, and specific method had been developed to detect influenza A and B viruses and RSV.

Introduction

Influenza viruses and respiratory syncytial virus (RSV) infections are responsible for significant global morbidity, mortality, and healthcare burden each year.[1−3] In particular, influenza viruses cause millions of deaths worldwide each year, and RSV is responsible for 200,000 deaths in a single year, posing a clinical and economic burden.[4−7] Moreover, the symptoms of influenza viruses and RSV infections are cough, rhinorrhea, and fever, which are difficult to distinguish from those of other respiratory virus infections.[8−10] Therefore, a precise and rapid diagnostic method is of utmost importance. The traditional viral culture is the gold standard for respiratory viral infection diagnosis; however, it is time-consuming, with the viral detection timeframe ranging from 2 to 14 days.[11−13] The transition from a conventional viral culture to a reverse-transcription polymerase chain reaction (RT-PCR) is a major leap. However, labor-intensive procedures along with sophisticated equipment have greatly limited its application in point-of-care analysis.[14−16] Therefore, it is essential to develop a sensitive, specific, and convenient strategy to detect influenza viruses and RSV.

In recent years, nucleic acid isothermal amplification techniques have garnered extensive attention because of their simplicity, rapid detection, and high sensitivity and specificity.[17] The isothermal nucleic amplification method can be performed at specific temperatures in a constant-temperature water bath pot to overcome the process of repeated heating and cooling.[18,19] Catalytic hairpin assembly (CHA) is an isothermal nucleic amplification method that has promising applications. CHA, first proposed by Pierce et al.,[20] is an enzyme-free isothermal amplification that relies on a thermodynamically driven entropy gain process to achieve exponential signal amplification.[21−25] Circuits in CHA provide rapid and efficient amplification with minimal background and fast turnover rates.[26] Furthermore, CHA can be adopted in various analytical formats, such as colorimetry, surface plasmon resonance, electrochemistry, and Raman spectroscopy.[27−30] However, these strategies are complex, time-consuming, involve expensive equipment, and rely on nanoparticles.

In this study, a facile and sensitive detection strategy for influenza A virus (Flu A), influenza B virus (Flu B), and RSV was designed based on a catalytic hairpin assembly and the lateral flow immunoassay (CHA–LFIA) system. As shown in Scheme , the CHA–LFIA detection method consists of two parts: (1) signal amplification by CHA and (2) fluorescence signal detection by LFIA. The mixture of rationally designed probes Hairpin 1 (H1), Hairpin 2 (H2), and target RNA was incubated in a constant-temperature water bath pot for 15 min to produce large amounts of H1/H2 complexes. The fluorescence signals were read between 10 and 30 min after the addition of the CHA reaction solution. Moreover, the CHA–LFIA detection system realizes quantitative detection. The fluorescence and target RNA concentrations exhibited a good linear relationship (R2 = 0.9915 for Flu A, R2 = 0.9830 for Flu B, and R2 = 0.9828 for RSV). In addition, this method exhibits excellent capability for differentiating between the target RNA and base-mismatched RNA. In brief, the CHA–LFIA system has the advantages of easy performance, rapid detection, high accuracy, sensitivity, and specificity.

Scheme 1

Schematic of the CHA–LFIA System for Influenza Viruses and RSV Detection. (A) Mechanism of the CHA Reaction. (B) Principle Applied by the LFIA Strip to Detect H1/H2 Complexes Labeled with Digoxin and Biotin.

Experimental Section

Chemicals and Materials

The oligonucleotides listed in Table S1 were synthesized and purified by Sangon Biotech Co., Ltd. (Shanghai, China). Native-PAGE gel and DNA loading buffer (6×) were purchased from Sangon Biotech Co., Ltd. (Shanghai, China). Ultrapure water was obtained from a Millipore water purification system.

Preparation of the CHA Probes

Genome sequences of influenza A and B viruses and RSV were obtained from NCBI Gene Bank. Highly conserved regions of viruses were identified by multiple sequence alignments, and hairpin probes were designed based on genome conserved regions.

Hairpin probes were designed based on NUPACK,[31] and the ideal structures of H1 and H2 were analyzed by mFold,[32] as shown in Figure S1. H1 and H2 were diluted in a 1× TNaK buffer (20 × 10–3 M Tris, pH = 7.5; 140 × 10–3 M NaCl; 5 × 10–3 M KCl). For the pretreatment of hairpin probes, H1 and H2 were separately heated at 95 °C for 5 min and slowly cooled down to room temperature.

Fluorescence Signal Monitoring of the CHA Reaction

The CHA reaction mixture containing H1 (300 nM), H2 (300 nM), and target RNA (300 nM) to a total volume of 30 μL was monitored using a CFX96 real-time system at intervals of 30 s. H2 was labeled with a fluorescent reporter (FAM) and quencher dyes (BHQ1) at opposite ends. As the CHA reaction progressed, a number of H2 molecules opened and generated remarkable fluorescence signals. For comparison, a mixture of H1 (300 nM), H2 (300 nM), and TNaK buffer was measured without the addition of the target RNA.

Gel Electrophoresis Analysis

Native-PAGE electrophoresis was performed to detect the CHA reaction. (Lane 1–7: H1, H2, target RNA, H1 + H2, H1 + target RNA, H2 + target RNA, and H1 + H2 + target RNA, respectively). The mixture was reacted at 37 °C in a water bath pot for 15 min. Thereafter, 15 μL of the mixture was loaded onto a gel for electrophoresis analysis that was performed at 120 V for 1 h in a 1× TAE buffer. The gels were thereafter stained with 1% ethidium bromide for 10 min and imaged under ultraviolet irradiation.

Optimization of Assay Conditions

To improve the sensitivity of the detection system, several parameters, including the reaction temperature, ratio of H1 and H2, LFIA strip detection time, and concentrations of H1 and H2, were systematically optimized. The fluorescence signal (S) was defined by the fluorescence value obtained by the mixture of H1, H2, and target RNA. In the absence of the target RNA, the fluorescence signal produced by spontaneous H1 and H2 binding served as the background signal (N). The signal-to-noise ratio (S/N) was used as the evaluation criterion of condition optimization. First, we optimized the reaction temperature of CHA. Briefly, H1, H2, and target RNA were diluted in TNaK buffer, and the concentrations of H1, H2, and target RNA were held at 300 nM. A reaction mixture of H1, H2, and target RNA was incubated at varying temperatures (25, 30, 35, 40, 45, and 50 °C) in 300 μL. Second, the ratio of H1 and H2 was optimized at the optimum temperature. H1 (100, 200, 300, and 400 nM) and H2 (100 nM) were mixed with the target RNA (100 nM), which were diluted in 300 μL of TNaK buffer for 15 min incubation. Then, an experiment on the LFIA strip detection time was performed. The mixture of H1, H2, and target RNA was incubated at the optimum temperature, and 75 μL of the mixture was dropped on the immunoassay strip. Thereafter, fluorescence signals were detected every 5 min. During the optimal concentrations of H1 and H2, target RNA was fixed at 1 nM. The concentrations of H1 were 2, 4, 40, and 200 nm, and the corresponding concentrations of H2 were 500 pm and 1, 10, and 50 nm based on the optimum ratio of H1 and H2 for Flu A. Similarly, optimal concentrations of H1 and H2 for Flu B and RSV are illustrated in Figure S2.

Sensitivity of CHA–LFIA

Under optimized experimental conditions, H1 and H2 detected the target RNA of varying concentrations in a virus preservation solution at 1 nM and 800, 600, 500, 400, 200, 100, 50, 10, 5, 1, and 0 pM, similar to the aforementioned procedure.

Specificity of CHA–LFIA

To evaluate the selectivity of the CHA–LFIA system, target RNA, single-base-mismatched RNA (SM), and double-base-mismatched RNA (DM) were detected using a procedure similar to that described previously.

Results and Discussion

Principle of CHA–LFIA for Influenza Viruses and RSV Detection

The principle of CHA–LFIA is shown in Scheme . CHA is composed of three elements: target RNA and a pair of complementary hairpin probes H1 and H2. H1 and H2 were labeled with biotin and digoxin at the 5′ terminus, respectively, to facilitate follow-up immunoassay strip testing. Without the target RNA, spontaneous hybridization between H1 and H2 hardly occurs because the complementary domain is kinetically blocked by hairpin stems.[33] However, when the target RNA is present, the CHA reaction is triggered via a toehold-mediated strand displacement reaction. In particular, the target RNA hybridizes with the toehold of H1 and forms the H1/target RNA complex. Similarly, the exposed H1 initiates assembly with H2 and forms the H1/H2 complex. Gradually, the target RNA is displaced through strand displacement and further participates in the cycle of hairpin opening and assembly. Consequently, one copy of the target RNA can generate numerous H1/H2 complexes that produce significant signal amplification.

Another part of the CHA–LFIA system is the lateral flow immunoassay strip. The construction of the test strip was based on our previous study.[34] Briefly, the strip is composed of a sample pad, conjugation pad, nitrocellulose membrane with a test line (T), a control line (C), and an absorbent pad. The conjugation pad was prepared with polyethylene (PE) nanoparticles labeled with streptavidin (SA) and fluorophore Alexa Fluor 647. H1/H2 complexes labeled with digoxin and biotin gradually flowed and were captured by a prefixed digoxin antibody. Thereafter, PE nanoparticles labeled with SA and fluorophore Alexa Fluor 647 were combined with the captured H1/H2 complexes to form a sandwich hybridization complex. Finally, fluorescence signals were detected using a fluorescence immunochromatographic quantitative analyzer.

Feasibility of the CHA Reaction

Real-time fluorescence signals were monitored to validate the feasibility of CHA. When the reaction solution contained the target RNA, it triggered the recycling of the CHA reaction. H2 was labeled with FAM and BHQ1 at opposite ends. With the hairpin opening of H2, fluorescence signals were detected by fluorescence resonance energy transfer. As the reaction proceeded, the fluorescence signals rapidly increased. Owing to the substrate concentration consumption, the reaction velocity became slower and reached the platform period. Figure A–C shows the changes in fluorescence signals as a function of the reaction time for Flu A (A), Flu B (B), and RSV(C), respectively. In the presence of the target RNA, fluorescence signals grew significantly. However, fluorescence signals barely changed in the absence of the target RNA.

Figure 1

Feasibility of the CHA reaction. (A–C) Real-time fluorescence signals monitoring of the CHA reaction for Flu A (A), Flu (B), and RSV (C). (D–F) Native polyacrylamide gel electrophoresis of the CHA reaction for Flu A (D), Flu (E), and RSV (F). Lane 1: H1, Lane 2: H2, Lane 3: target RNA, Lane 4: H1 + H2, Lane 5: H1 + target RNA, Lane 6: H2 + target RNA, and Lane 7: H1 + H2 + target RNA.

Native polyacrylamide gel electrophoresis was also performed to reveal the feasibility of CHA via electrophoretic bands. As shown in Figure D–F, Lanes 1 (H1), 2 (H2), and 3 (target RNA) exhibited a clear electrophoresis band. When the target RNA was absent (Lane 4), most H1 and H2 could coexist. Research showed that there was approximately 3.5% of H1 and H2 spontaneously hybridized without an initiator.[20] When samples contained H1 and H2 and target RNA, an obvious band (Lane 7) appeared representing the product H1/H2 complexes.

Optimization of Assay Conditions

As shown in Figure A–C, 20 cycles reached the platform period, which was approximately 15 min. Thus, 15 min is sufficient for the CHA reaction. Optimization of assay conditions for Flu A is shown in Figure . The fluorescence signal values are shown in Figure A at varying temperatures. The signal-to-noise ratio (S/N) of fluorescence (Figure D) showed that the optimum temperature was 35 °C; Figure E shows that the optimal ratio of H1/H2 was 4:1 considering the S/N and cost. With the increasing concentration of H1 and H2, both the fluorescence and background signals increased (Figure C). When the concentration of H2 was 1 nM, S/N reached a maximum (Figure F). The optimized conditions for flu B and RSV are shown in Figures S2 and S3. Optimum temperatures were 25 and 30 °C for Flu B and RSV, respectively; the optimal ratios of H1/H2 were 1:1 and 1:3 for Flu B and RSV, respectively; the optimum concentration of H2 was 1 and 3 nM for Flu B and RSV, respectively. The fluorescence value fluctuated substantially during the first 10 min, and there was nearly no clear change from 10 to 30 min (Figure S4). Thus, an immunoassay strip was detected between 10 and 30 min after dropping the CHA reaction solution.

Figure 2

Optimization of assay conditions for Flu A. (A, D) Reaction temperature and S/N of CHA; (B, E) optimization of the ratio of H1/H2 and S/N; (C, F) concentration and S/N of H1 and H2. Data are represented as the means ± standard deviation (SD) (n = 3).

Sensitivity of the CHA–LFIA

Owing to the complex biological environment of oropharyngeal buffer, there are numerous factors that may influence the outcome of the investigation. To make the detection environment closer to clinical detection, H1, H2, and target RNA of subsequent experiments were studied in an oropharyngeal virus preservation solution.

As shown in Figure A–C, the fluorescence value decreased gradually as the concentration of target RNA decreased. Figure D–F shows an excellent linear relationship between the fluorescence value (Y) and target RNA concentration (X), and the linear equations were expressed as follows: Y = 6.696*X + 1606 (R2 = 0.9915), Y = 11.93*X + 1138 (R2 = 0.9830), and Y = 12.06*X + 1339 (R2 = 0.9828) for Flu A, Flu B, and RSV, respectively. The cutoff values (the mean value plus 3 × the standard deviation for negative samples)[35] were 1613, 352, and 885 for Flu A, Flu B, and RSV, respectively. The CHA–LFIA system exhibited a wide linear range with limits of detection of 1, 1, and 5 pM, respectively.

Figure 3

Sensitivity of the CHA–LFIA. (A–C) Fluorescence value for varying concentrations of target RNA in the virus preservation solution at 1 nM and 800, 600, 500, 400, 200, 100, 50, 10, 5, 1, and 0 pM for Flu A (A), Flu B (B), and RSV (C). (D–F) Calibration curve of the fluorescence value against varying concentrations of the target RNA for Flu A (D), Flu B (E), and RSV (F). Data are represented as the means ± SD (n = 3).

Specificity of the CHA–LFIA

To evaluate the selectivity of the CHA–LFIA system, target RNA, SM, and DM were detected under similar conditions. As shown in Figure , in comparison with the target RNA, the fluorescence value of mismatched RNA was significantly lower, demonstrating that this CHA–LFIA detection method had excellent selectivity and high discrimination capability.

Figure 4

Selective investigation of the target RNA, SM, and DM with a concentration of 1 nM for Flu A (A), Flu B (B), and RSV (C). Data are represented as the means ± SD (n = 3).

Conclusions

The CHA–LFIA detection system required only 30 min to obtain results and had excellent detection performance, high sensitivity to detect targets as low as 1 pM, and excellent specificity to identify the target. In addition, this system did not require complicated and expensive detection equipment, only a water bath pot and a portable immunofluorescence detection analyzer that facilitated point-of-care diagnostic testing. The CHA–LFIA detection system has an exponential signal amplification capability. In comparison with the CHA with fluorescence, as shown in Figure S4, the minimum detection was 1 nM for influenza A virus. Fluorescence signals were further amplified by approximately 3 orders of magnitude via a lateral flow immunoassay strip that is an innovation from the traditional fluorescence signal detection of CHA.

In conclusion, we developed a facile, rapid, sensitive, and convenient strategy for detecting influenza A virus, influenza B virus, and RSV by coupling the CHA and lateral flow immunoassay strip. By designing the probe precisely and optimizing experimental conditions, the proposed CHA–LFIA system exhibited promising performance in detecting influenza A virus, influenza B virus, and RSV with a minimum detection limit of 1, 1, and 5 pM, respectively. Owing to its high sensitivity, low cost, and rapid detection, the proposed CHA–LFIA system provides a new avenue for influenza viruses and RSV detection in clinical fields.