A novel multiplex isothermal amplification method for rapid detection and identification of viruses

A rapid multiplex isothermal amplification assay has been developed for detection and identification of multiple blood-borne viruses that infect millions of people world-wide. These infections may lead to chronic diseases or death if not diagnosed and treated in a timely manner. Sets of virus-specific oligonucleotides and oligofluorophores were designed and used in a reverse-transcription loop-mediated multiplexed isothermal amplification reaction for detection and gel electrophoretic identification of human Immunodeficiency virus (HIV), hepatitis-B virus (HBV), hepatitis-C virus (HCV), hepatitis-E virus (HEV), dengue virus (DENV), and West Nile (WNV) virus infection in blood plasma. Amplification was catalyzed with two thermostable enzymes for 30–60 minutes under isothermal condition, utilizing a simple digital heat source. Electrophoretic analysis of amplified products demonstrated simultaneous detection of 6 viruses that were distinctly identified by unique ladder-like banding patterns. Naked-eye fluorescent visualization of amplicons revealed intensely fluorescing products that indicated positive detection. The test demonstrated a 97% sensitivity and a 100% specificity, with no cross-reaction with other viruses observed. This portable detection tool may have clinical and field utility in the developing and developed world settings. This may enable rapid diagnosis and identification of viruses for targeted therapeutic intervention and prevention of disease transmission.

Scientific RepoRts | 5:17925 | DOI: 10.1038/srep17925 multiplexing capability to simultaneously detect different types of viruses [13][14][15][16][17] . A recent study reported a combined multiplex RT-LAMP and immunochromatographic system for colorimetric detection, but this method was confined to the detection and subtyping of only influenza A virus 18 . Other LAMP systems have relied on the use of SYBR Green, GelGreen, calcein or hydroxy naphthol blue for quantitation or visual detection of amplified products, but the dyes are target-independent and non-specific 15,17,[19][20][21] .
This work describes the development of the first multiplex fluorogenic loop-mediated isothermal amplification assay for simultaneous detection and identification of 6 different viral pathogens in human plasma. This assay is based on fluorogenic hybridization and autocycling strand-displacement DNA synthesis that is loop-mediated under isothermal amplification conditions and produce repeats of nucleic acid amplicons as ladder-like banding patterns [13][14][15][16][17]22 . The method utilizes 3 pairs of virus-specific primers (including target-specific fluorogenic oligonucleotides) that target 8 different sequences within the genomic region of interest (Table 1; Fig. 1). The specific fluorogenic oligonucleotides hybridize only to the target-sequence of the amplified DNA, resulting in emission of fluorescence that is visible to the naked-eye under UV-illumination and measureable with a fluorospectrophotometer. The amplification process is accomplished within 30-60 minutes, utilizing a portable digital heating source. Here, we present a simple diagnostic approach that have clinical application for rapid detection and simultaneous identification of multiple viral pathogens.

Results
Detection of viral nucleic acid. We applied standards of viral nucleic acids to isothermal multiplexed amplification reaction in order to evaluate the ability of the assay to detect multiple viral targets. Electrophoretic analysis of reaction products revealed amplification of nucleic acids of HIV, HBV, HCV, HEV, DENV and WNV. Positive amplification was demonstrated by the presence of specific ladder-like banding patterns. No banding-pattern was observed for the no-template control (NTC), parvovirus (PV), and cytomegalovirus (CMV) which served as negative controls ( Fig. 2A; see Supplementary Figs S1-S4 online).
Fluorescence detection. All virus-specific fluorogenic oligonucleotides were used in a multiplex amplification reaction mixture to enable visual detection of amplified viral nucleic acids and facilitate measurement of fluorescence intensities. The reaction products were visualized under UV-illumination and revealed an intense fluorescence in tubes with amplified products (HIV, HBV, HCV, HEV, DENV, and WNV), but no glow was observed in the negative control tubes (Fig. 2B). Spectrofluorometric measurement of the same products revealed higher relative fluorescence units (RFU) in the amplified target as compared with the negative controls ( Fig. 2B).
Analytical sensitivity. Dilutions of all target nucleic acids, including HBV-DNA standards ranging from 10 4 -25 IU/reaction and HCV-RNA standards ranging from 10 3 -10 IU/reaction (IU/rxn) were used in amplification reaction to investigate the analytical sensitivity of the assay. Results of electrophoretic gel analysis revealed detection of 50 IU/rxn of HBV-DNA and 10 2 IU/rxn of HCV-RNA (Fig. 3, Panel-1A and Panel-2A); evaluation of assay sensitivity using the other targets revealed a detection limit detection of approximately 10 2 -50 viral particles per reaction. Spectrofluorometric measurement revealed higher RFU of HBV-DNA serial dilutions 50 IU/rxn as compared with the negative controls (Fig. 3, Panel-1B and 1C; Table 2A). Interestingly, the 25 IU HBV-DNA dilution revealed higher RFU than the controls, but reacted negative on gel electrophoresis. Similarly, the 10 2 and 10 IU/rxn of HCV-RNA showed higher RFUs as compared with the negative controls, but the 10 IU/rxn dilution showed no banding pattern on agarose gel electrophoresis (Fig. 3, Panel-2B and 2C; Table 2B).
Analytical specificity. Primer sets and fluorogenic oligonucleotides were tested against nucleic acids of HCV, HIV, HBV, DENV, HEV, WNV, CMV, and PV in the multiplex isothermal amplification assay in order to investigate the analytical specificity. Electrophoretic analysis of reaction products demonstrated amplification of the targeted viral pathogens by their specific oligonucleotide-sets and produced distinctive ladder-like banding patterns for each detected virus ( Fluorimetric analysis of reaction products. Spectrofluorometric measurement of reaction products was performed in order to quantify the fluorescence emitted by the amplified targets as compared with the reaction negative controls. Analysis of the results revealed higher RFU values for the detected targets, but lower RUFs for the negative controls (Figs 2B, 3B,C-all panels, and 4B-all panels). In Fig. 5, the two specific fluorogenic oligonucleotides that were employed for detection of an HIV-HCV co-infection successfully hybridized with their targets. This In step 1, in the absence of hybridization to amplified product (target sequence), the reporter remains quenched, with no fluorescence emission due to the close proximity of the quencher to the fluorophore. In the presence of amplified product in steps 2 and 3, the LRp hybridizes to the specific target between the R1c and R2 segments of the gene sequence, thus fluorophore begins to fluoresce as it separates from the quencher. From step 3 to step 4 as Bst DNA Polymerase catalyzes the reaction, quenching is disrupted and fluorescence is increased due to the increased distance (separation) between the reporter and quencher when the fluorooligonucleotide hybridizes and also due to fluorescence resonance energy transfer (FRET). The emitted fluorescence which corresponds to the amount of amplified nucleic acid is measured by spectrofluorometer in relative fluorescence units (RFU) and visualized by the naked-eye under UVillumination.
was demonstrated by the higher RFUs of the HCV and HIV duplicate samples, while lower RFUs were observed for the non-target nucleic acids.
Test of clinical specimens. Nucleic acids were extracted from 89 clinical plasma specimens, including HIV, HBV, HCV, WNV, CMV, DENV, PV, and healthy human plasma (see Supplementary Table S1 online). The DNA and RNA extracts were tested to validate the clinical performance of the multiplex isothermal assay. Results of amplification showed that the assay detected 36 out of 37 infected clinical specimens, thereby demonstrating a 97% diagnostic sensitivity. The results further revealed that all 52 assay-controlled plasma specimens reacted negative, thus demonstrating a 100% diagnostic specificity (Table 3).

Discussion
We have developed a rapid fluorogenic multiplex isothermal amplification assay that simultaneously detects and identifies different viral pathogens. The assay utilized a reaction master mixture that simultaneously amplified both DNA and RNA, and specifically amplified RNA in a one-step single-tube procedure. In addition, this method provides three end-point read-outs, utilizing distinctive gel-electrophoretic banding patterns for viral differentiation, measurement of fluorescence intensities, and naked-eye fluorescence visualization for rapid detection (Figs 2-5). Furthermore, the assay demonstrated an analytical sensitivity of 10 2 IU/rxn using RNA and 50 IU/rxn utilizing DNA. The assay also revealed a specificity characterized by amplification of only the viruses that were targeted for detection. For example, the specific-primer sets only detected HIV, HBV, HCV, HEV, DENV, and WNV, but reacted negative to PV and CMV (Figs 2 and 4). For assay validation, nucleic acids extracts from clinical donor plasma specimens were used in amplification reaction (see Supplementary Table S1 online). As shown in Table 3, the assay revealed a plausible diagnostic sensitivity of 97% and a 100% diagnostic specificity when compared with the FDA-approved Procleix test. These diagnostic characteristics point to the potential applicability of the assay for blood screening and clinical point-of-care diagnostic use (Table 3). Detection systems such as quantitative reverse-transcription polymerase chain reaction (qRT-PCR) employ a single pair of primer and a probe for detection of targets. In contrast, we used 3 pairs of pathogen-specific primers (including an oligofluorophore) that target 8 different sequences on the genomic region of interest (Fig. 1A). In this study, we designed target-specific Loop Reverse oligonucleotides (LRp) of each primer set as fluorogenic probes covalently linked with either 6-FAM, Tet, or Texas-Red as reporter on the 5′ end and Black Hole Quenchers (BHQ) on the 3′ end (Fig. 1B). As illustrated in Fig. 1, the fluorogenic oligonucleotides hybridize only to the virus-specific sequence located between the R1c and R2 segments of the targeted gene. During strand displacement DNA synthesis by Bst DNA Polymerase, the reporter-fluorophore at the 5′ end of the fluorogenic oligonucleotide is separated from the quencher at the 3′ end by the increase in distance upon hybridization. An intense fluorescence is emitted and measured by a spectrofluorometer or visualized by the naked-eye (Figs 1B, 2B, 4B-all panels, and 5). The biomolecular mechanism used in this study ensured amplification efficiency and detection specificity.
Each virus of interest was detected on agarose gel due to the presence of all the primer-sets in the multiplex master reaction-mixture. The distinguishing ladder-like banding-patterns generated by each primer-set then enabled pathogen identification (Figs 2A and 4A-all panels; see Supplementary Figs S1-S4 online). The differences in banding patterns are identifiable by the group-formation of the bands relative to the molecular marker, the size of the bands within the patterns, and position or laddering-shift of the detected targets as illustrated by the color lines and dots (Figs 2A and 4; see Supplementary Figs S1-S4 online). Like products of other amplification platforms, gel electrophoresis of RT-LAMP products does risk cross-contamination by amplified DNA. Precautions to avoid this and the use of negative (no-template) control assays are essential to identify contamination problems and monitor assay integrity. Furthermore, upon post-amplification UV-visualization of reaction tubes, only the amplified targets produced intense fluorescence (Fig. 4B-all panels and 5). The results of this biomolecular process suggest that the oligofluorophores hybridized only to the specific viral nucleic acids without cross-reacting with nucleic acids of the non-target viruses. This approach obviated the post-amplification addition of DNA intercalating dyes such as SYBR Green and GelGreen or the use of metal indicator, calcein which are non-specific indicators of amplified DNA 15,17,[19][20][21] . On the contrary, the fluorogenic oligonucleotides used in our study are specific; hybridization to their targets were confirmed by the higher RFUs and increased fluorescent intensity of the detected targets as compared with the non-target viruses. (Fig. 4B-all panels and 5).    Fluorophores have been traditionally used in immunohistochemistry for identification of cell structures and in qPCR applications for detection and quantitation of pathogen load 22,23 . The properties of these fluorogenic dyes to absorb and emit light or transfer energy were employed by this study for detection and measurement of the fluorescence of amplified viral nucleic acid. In this study, we successfully utilized three different fluorophores (Tet, 6-FAM, and Texas-Red) for detection. Furthermore, we observed that amplified nucleic acid demonstrated higher RFUs that indicated positive detection, while the non-target viral nucleic acids and negative controls revealed lower RFUs. For example, the results of the assay sensitivity test showed a visual and quantitative increase in fluorescence intensity of the HBV-DNA and HCV-RNA dilutions as compared with the negative control samples ( Fig. 3; Table 2). Also, viruses that were detected by their specific oligofluorophores produced higher RFU values, while the non-targeted viruses and negative controls revealed lower RFU values (Fig. 4B-all panels). In this study, we have consistently observed that the increase fluorescence intensity corresponds to amplification of the target nucleic acid. Therefore, we propose that this biomolecular process could potentially be used for rapid empirical quantitation of amplified nucleic acid as this could correspond to pathogen load in an infected specimen. However, additional studies are warranted.
In view of the prevalence of co-infection of the hepatitis viruses and HIV particularly among injection drug users (IUDs) 5 , we simulated a test for detecting a co-infection of this kind. When two specific oligofluorophores targeting HCV-RNA and HIV-RNA were simultaneously utilized in the amplification reaction mixture, the results revealed detection of only the two target-viruses as observed in Fig. 5. The specificity of this co-detection was confirmed by the intense fluorescence of the amplified targets. These amplified specimens also produced higher RFUs as compared with the RFUs of the controls and non-target samples (Fig. 5). Although the detection characteristics of this assay have proven relevant to clinical applications, we believe that additional optimization and improvements could further enhance its capabilities. Efforts are underway in this endeavor, including plans for a large-scale field trial.
The data reported in this study have demonstrated the potential application of this assay for simultaneous detection, identification, and measurement of amplified products of multiple viral targets. The assay revealed a plausible diagnostic sensitivity and specificity. Taken together, these results suggest that the assay could be useful  Table 1 for detection of only hepatitis C virus (HCV) and human Immunodeficiency virus (HIV) nucleic acid extracts. Fluorospectrophotometric results demonstrate brighter fluorescence and higher RFUs in the HCV samples (lanes 5 and 6) and the HIV (lanes 9 and 10) as compared with the no-template controls (NTCs) in lanes 1 and 2, hepatitis B virus (HBV) in lanes 3 and 4, and hepatitis E virus (HEV) in lanes 7 and 8; UV = ultraviolet naked-eye visualization.  for point-of-care diagnostics, donor blood screening, and investigating the epidemiological trend of viral pathogens in resource-limited communities of the developed and developing world settings.

Methods
Nucleic acid standards. DNA  Oligonucleotides and probes design. Several full-length sequences of various viral pathogens (e.g. HIV, HBV, HCV, HEV, WNV, and DENV) were obtained from the GenBank database and analyzed using CLUSTALW-2.
Conserved genetic regions and selected sequences were targeted for primer design (Table 1) were also added to the reaction mixture. Viral RNA and DNA template-volume of 5 μ L of both target and assay-controlled viruses were added to the amplification reactions. A no-template (water) control and nucleic acid extracted from normal human plasma were included in amplification runs as controls. Reaction reagents were prepared in a PCR work-station (Plas-Lab, MI, USA), with precautionary measures observed to prevent cross-contamination. Amplification-reactions were conducted under isothermal condition at 60 °C for 60 minutes on a portable heating device. Reactions were terminated by placing reaction tubes on ice.
Fluorescence detection and spectrofluorometric analysis of products. In order to investigate as to whether the virus-specific fluorogenic oligonucleotides hybridized to their targets, fluorescence intensities in reaction tubes were evaluated with the naked-eye using a handheld UV-transilluminator at 302 nm. Reaction tubes were photographed using a BlackBerry Z30 smartphone camera (Research-In-Motion, Canada) and the iPad Air tablet camera (Apple, Inc., USA). Additionally, fluorimetric evaluation of fluorescence intensity was performed by measuring 1.5 μ L of target and control amplification products on a NanoDrop-3300 spectrofluorometer (ThermoScientific, DE, USA).

Electrophoretic evaluation of amplification products.
To confirm amplification and evaluate the different ladder-like banding patterns of the targets, 5 μ L of (RT)-LAMP reaction products were electrophoresed on 2.8% agarose gels. Gels were prepared in 1x Tris-borate EDTA buffer (TBE) and contained 10x GelRed DNA intercalating dye (Phenix Research, Candler, NC). Electrophoretic pattern differences were observed and identified by the location and formation of banding-patterns in relation to the molecular marker, size of the bands within the patterns, spacing of each group of patterns as well as the laddering-shift of the patterns of detected targets. Gels were visualized under UV-transilluminator at 302 nm and photographed with the GBOX gel documentation system (Syngene, Frederick, MD, USA).
Assay specificity and sensitivity. In order to investigate the specificity of the assay, primer sets and fluorogenic oligonucleotides were tested against nucleic acids of HCV, HIV, HBV, DENV, HEV, WNV, CMV, and PV in the multiplex isothermal amplification assay. In order to determine the assay's limit of detection, dilutions of nucleic acids of various targets were evaluated. Sensitivity tests representative of DNA and RNA viruses were performed by testing dilutions of HBV DNA standards ranging from 10 4 -25 IU/rxn and HCV RNA standards ranging from 10 3 -10 IU/rxn.

Test of clinical plasma specimens.
A total of 89 clinical donor specimens that were randomly pre-selected with the FDA-approved Procleix test 23 (and packet insert) were used to investigate the clinical utility of the isothermal multiplex assay in comparison to the Procleix test: 52 non-target control specimens, including healthy human plasma and clinical donor plasma specimens infected by cytomegalovirus (CMV) plasma and parvovirus (PV) were tested; 37 targeted clinical specimens, including donor plasma from patients infected by HIV, WNV, DENV, HBV, and HCV were also tested (see Supplementary Table S1 online).