Rapid Diagnostic Tests for the Detection of the Four Dengue Virus Serotypes in Clinically Relevant Matrices

Abstract

The efficient and accurate diagnosis of dengue, a major mosquito-borne disease, is of primary importance for clinical care, surveillance, and outbreak control. The identification of specific dengue virus serotype 1 (DENV-1) to DENV-4 can help in understanding the transmission dynamics and spread of dengue disease. The four rapid low-resource serotype-specific dengue tests use a simple sample preparation reagent followed by reverse transcription-isothermal recombinase polymerase amplification (RT-RPA) combined with lateral flow detection (LFD) technology. Results are obtained directly from clinical sample matrices in 35 min, requiring only a heating block and pipettes for liquid handling. In addition, we demonstrate that the rapid sample preparation step inactivates DENV, improving laboratory safety. Human plasma and serum were spiked with DENV, and DENV was detected with analytical sensitivities of 333 to 22,500 median tissue culture infectious doses (TCID₅₀)/mL. The analytical sensitivities in blood were 94,000 to 333,000 TCID₅₀/mL. Analytical specificity testing confirmed that each test could detect multiple serotype-specific strains but did not respond to strains of other serotypes, closely related flaviviruses, or chikungunya virus. Clinical testing on 80 human serum samples demonstrated test specificities of between 94 and 100%, with a DENV-2 test sensitivity of 100%, detecting down to 0.004 PFU/μL, similar to the sensitivity of the PCR test; the other DENV tests detected down to 0.03 to 10.9 PFU/μL. Collectively, our data suggest that some of our rapid dengue serotyping tests provide a potential alternative to conventional labor-intensive RT-quantitative PCR (RT-qPCR) detection, which requires expensive thermal cycling instrumentation, technical expertise, and prolonged testing times. Our tests provide performance and speed without compromising specificity in human plasma and serum and could become promising tools for the detection of high DENV loads in resource-limited settings. IMPORTANCE The efficient and accurate diagnosis of dengue, a major mosquito-borne disease, is of primary importance for clinical care, surveillance, and outbreak control. This study describes the evaluation of four rapid low-resource serotype-specific dengue tests for the detection of specific DENV serotypes in clinical sample matrices. The tests use a simple sample preparation reagent followed by reverse transcription-isothermal recombinase polymerase amplification (RT-RPA) combined with lateral flow detection (LFD) technology. These tests have several advantages compared to RT-qPCR detection, such as a simple workflow, rapid sample processing and turnaround times (35 min from sample preparation to detection), minimal equipment needs, and improved laboratory safety through the inactivation of the virus during the sample preparation step. The low-resource formats of these rapid dengue serotyping tests have the potential to support effective dengue disease surveillance and enhance the diagnostic testing capacity in resource-limited countries with both endemic dengue and intense coronavirus disease 2019 (COVID-19) transmission.

Keywords: NS5 gene; dengue virus; isothermal amplification; lateral flow detection; rapid molecular assays; rapid sample preparation; recombinase polymerase amplification.

FIG 1

Analytical sensitivities and serotype specificities of the four serotype-specific DENV RT-RPA-LFD assays using kit-purified DENV isolate RNA. (A) Sensitivity testing used purified RNA from a DENV-1 isolate (ET00.243). (B) Serotype specificity was trialed with purified RNA from DENV-1 (ET00.243), DENV-2 (New Guinea C), DENV-3 (ET00.209), and DENV-4 (ET00.288) isolates at 10⁵ copies/μL as determined by universal dengue RT-qPCR. (A and B, left) Photographs of lateral flow strips with control bands (all samples) and test bands (positive samples). Nuclease-free water was used as a no-template control (NTC). (A and B, right) Normalized pixel densities (black values) from the test displayed. (C) Heat maps summarizing serotype-specific DENV RT-RPA-LFD assay results for the detection of purified DENV-1, -2, -3, and -4 RNAs, with the corresponding isolates displaying sensitivity (left) and serotype specificity (right) data. (D) Heat map summarizing RT-RPA-LFD test results for the detection of DENV-1, -2, -3, and -4, testing RNA isolated from chikungunya virus (CKGV), Japanese encephalitis virus (JEV), West Nile virus subtype Kunjin (WNV_KUN), Murray Valley encephalitis virus (MVEV), yellow fever virus (YFV), and Zika virus (ZIKV), with DENV-1, -2, -3, and -4 RNA transcripts as the positive controls at 10³ copies/μL in the corresponding assays. Nuclease-free water was used as NTC for all testing.

FIG 2

Detection and inactivation of cultured DENV-1 to -4 using TNA-Cifer reagent E. (A and B) TNA-Cifer reagent E (A) or cell culture medium (B) was added to dilutions of a DENV-1 isolate (ET00.243) in a 5:1 ratio (sample to TNA-Cifer reagent E or cell culture medium), incubated for 10 min at room temperature, and tested for detection with the rapid DENV-1 serotyping test. (A and B, left) Sample preparation diagrams; (middle) photographs of lateral flow strips with control bands (all samples) and test bands (positive samples). (A and B, right) Normalized pixel densities (black values) from the test displayed; (bottom) heat map summarizing RT-RPA-LFD test results. Cell culture medium was used as the no-template control (NTC). (C) Workflow of inactivation testing performed with TNA-Cifer reagent E and DENV-1, including virus stock incubation (1), serial dilution (2), infection plate setup (3), and a TCID₅₀ ELISA with fixed cells (4). OD, optical density. (D) TNA-Cifer reagent E inactivation of DENV-1 (ET00.243), DENV-2 (New Guinea C), DENV-3 (ET00.209), and DENV-4 (ET00.288) after a 10-min incubation at room temperature at a 5:1 ratio. Averages ± standard deviations are shown (n = 3). (E) Virus titer after DENV-1 (ET00.243) was incubated with TNA-Cifer reagent E at a 5:1 ratio (i.e., 100 μL of virus in cell culture medium added to 20 μL of TNA-Cifer reagent E) for 0, 0.167, 0.5, 1, 2, 5, and 10 min. This experiment was performed independently three times, with one sample taken per time point; the average results ± standard deviations are displayed.

FIG 3

Optimized rapid DENV-1 serotyping test detection of DENV-1 in human blood, plasma, and serum spiked with a DENV-1 isolate. (Left box) Cartoon displaying the simplified procedure, with the time consumed (left); sensitivity testing using human K₃ EDTA blood, plasma, or serum containing dilutions of cultured DENV-1 (ET00.243) (second from the left); sample preparation (sample prep) and dilution ratios (middle); a photograph of lateral flow strips with control bands (all samples) and test bands (positive samples) (second from the right); and normalized pixel densities (black values) from the test displayed (right). (Right box) Comparative RT-qPCR cycle threshold (C_T) values and copy numbers (copies per microliter) (left) and cartoon illustrating the simplified methodology with the time consumed (right). (Top) Nuclease-free water was used as the no-template control (NTC), and transcribed RNA of DENV-1 was used as the positive control (PTC) at 10⁶ copies/μL. Blood, plasma, and serum without the spiked DENV-1 isolate were used as negative controls (no virus). ND, not detected.

FIG 4

DENV serotype test performance characteristics in relation to DENV serotyping RT-qPCR as a reference. Test performances for each serotype are presented as 2-by-2 tables. As this was a stratified sample set, positive predictive values (PPV) and negative predictive values (NPV) were calculated based on an assumed dengue prevalence of 40%. (Raw data are available in Table S3 in the supplemental material.)

FIG 5

Workflow for rapid dengue serotyping tests. Shown is the basic workflow for rapid dengue serotyping tests, including sample processing, dilution, amplification, and lateral flow detection steps.