Rapid and Visual Detection of Muscovy Duck Reovirus Using a Reverse Transcription Recombinase‐Aided Amplification Assay in a Lyophilized Format for Point‐of‐Care Applications

TL;DR

A new rapid and portable test for detecting Muscovy duck reovirus was developed, enabling on-site diagnosis without needing advanced lab equipment.

Contribution

The first lyophilized RT-RAA assay for MDRV detection, enabling point-of-care testing without cold-chain dependence.

Findings

The assay detected MDRV with a limit of 1.03 × 10¹ copies/μL within 20 minutes.

The lyophilized format showed 95.8% sensitivity and 100% specificity compared to RT-qPCR.

A portable blue light imager allowed visual detection without specialized instruments.

Abstract

Muscovy duck reovirus (MDRV) causes substantial economic losses in the waterfowl industry, necessitating rapid and field‐deployable diagnostic tools. In this study, we developed, for the first time, a lyophilized reagent‐based reverse transcription recombinase‐aided amplification (RT‐RAA) assay targeting the σC gene of MDRV for point‐of‐care testing (POCT). The assay was optimized to operate at 39°C with a probe concentration of 100 nM. It achieved a detection limit of 1.03 × 101 copies/μL within 20 min, which is comparable to RT‐quantitative polymerase chain reaction (qPCR) and superior to existing methods. It exhibited high specificity for MDRV with no cross‐reactivity against other common waterfowl pathogens. Particularly, the assay was successfully lyophilized into a ready‐to‐use format, representing the first reported RT‐RAA‐based method for MDRV detection in such a format. The lyophilized reagents retained the key analytical performance characteristics without cold‐chain dependance. When evaluated on 42 clinical samples, the lyophilized RT‐RAA demonstrated 95.8% diagnostic sensitivity and 100% specificity compared to RT‐qPCR, with almost perfect agreement (κ = 0.952). A visual readout using a portable blue light imager further enabled naked‐eye interpretation without instrumentation. This robust, rapid, and field‐deployable RT‐RAA assay provides a practical solution for on‐site MDRV surveillance, particularly in low‐resource settings.

Figures

Schematic workflow for rapid detection of MDRV using the RT‐RAA assay. The entire process, from sample to result, can be completed within 25 min. Key steps include: (1) Nucleic acid release via rapid lysis of tissue samples; (2) Isothermal amplification of target RNA by RT‐RAA at 39°C for 20 min, which can be performed using a portable fluorescence detector or a real‐time PCR system; (3) Result interpretation via real‐time fluorescence curves or visual assessment using a portable blue‐light imager. This schematic was created with BioRender (https://www.biorender.com).

Tables

Table 1. Primer and probe sequences used for RT‐RAA and RT‐qPCR assays.

Methods	Primers/probes	Sequences (5′ to 3′)	Source
Real‐time RT‐RAA	F478‐505a	TTTCATCACTTTGGACAACGGACAACAT	Designed in this study
	F480‐507	TCATCACTTTGGACAACGGACAACATTG
	F466‐493	CTGAACCTGCGGTTTCATCACTTTGGAC
	R594‐621	ACTAGGAGTCAACGGGGAGGTATCGTAC
	R598‐625	CAGGACTAGGAGTCAACGGGGAGGTATC
	R557‐585	AATGAGACGAGACGTGTCCAACACCAAGT
	F478‐505	TTTCATCACTTTGGACAACGGACAACAT
	F479‐506	TTCATCACTTTGGACAACGGACAACATT
	F481‐508	CATCACTTTGGACAACGGACAACATTGT
	F482‐509	ATCACTTTGGACAACGGACAACATTGTT
	R596‐623	ACGATACCTCCCCGTTGACTCCTAGTCC
	R597‐624	CGATACCTCCCCGTTGACTCCTAGTCCT
	R599‐626	ATACCTCCCCGTTGACTCCTAGTCCTGC
	R600‐627	TACCTCCCCGTTGACTCCTAGTCCTGCG
	Exo‐probeb	TTCTATCCTCTCCTAACGTTGTTACTG/i6FAM‐dT/THF/iBHQ1‐dT/TCCCAGGCTCCAG [C3‐spacer]
RT‐qPCR	F	CCACCACCGCATATGATGTTG	[15]
	R	CATAACTGCTACCAACGGTGAA
	TaqMan‐probe	FAM‐GGGAAAGCCTGCTAAG‐MGB

Table 2. Comparison of the established method with previous methods for MDRV detections.

Analytical method	Sensitivity	Reaction temperature	Suitable for on‐site use?	Detection time	References
Virus isolation	—	—	No	7 days	[8]
Indirect IFA	—	—	No	2 h	[10]
ELISA	—	—	No	~ 2.5 h	[11]
RT‐PCR	1 fg viral RNA	Thermal cycling	No	3 h	[12]
Nested RT‐PCR	0.1 pg viral RNA	Thermal cycling	No	~ 7 h	[13]
qPCR	2.83 × 101 copies/μL	Thermal cycling	No	1 h	[14]
RT‐RAA	1.03 × 101 copies/μL	Constant 39°C	Yes	20 min	This study

Table 3. Diagnostic performance of RT‐RAA assay compared to RT‐qPCR for MDRV detection.

Assay			RT‐qPCR				κ	p‐Value
				Positive	Negative	Total
RT‐RAA (Real‐time)			Positive	23	0	23	0.952	<0.001
			Negative	1	18	19	—	—
			Total	24	18	42	—	—
RT‐RAA (Visual)			Positive	20	0	20	0.814	<0.001
			Negative	4	18	22	—	—
			Total	24	18	42	—	—

Full text

1. Introduction

Muscovy duck reovirus (MDRV), a member of the genus Orthoreovirus in the family Reoviridae, possesses a genome comprising 10 double‐stranded RNA (dsRNA) segments [1]. These segments are categorized by size into large (L1, L2, and L3), medium (M1, M2, and M3), and small (S1, S2, S3, and S4) based on their electrophoretic mobility [1, 2]. The viral genome encodes at least 11 primary translation products: the three λ proteins (λA, λB, and λC) are encoded by the L1–L3 segments; the three μ proteins (μA, μB, and μNS) by M1–M3; and the three σ proteins (σA, σB, and σNS) by S1–S3, respectively [1, 2]. Notably, the S4 segment features a partially overlapping gene arrangement and encodes two proteins, p10.8 and σC [1, 3]. The σC protein is a homo‐trimer that exhibits cell‐binding activity and induces type‐specific neutralizing antibodies [4, 5].

The MDRV infection was first reported in South Africa in 1950; however, the virus was not isolated until 1972 in France [6, 7]. MDRV primarily infects Muscovy ducklings within 40 days of age, causing a disease characterized by leg weakness, growth retardation, and watery diarrhea. Necropsy reveals distinctive gray‐white necrotic foci in the liver and spleen, indicative of extensive tissue damage [8]. Notably, MDRV demonstrates a pronounced affinity for immune organs, resulting in the disruption of both cellular and humoral immune responses and ultimately leading to immunosuppression [9]. Since its emergence in China in 1997, MDRV has become endemic in duck‐producing regions, causing widespread outbreaks and considerable economic losses in the waterfowl industry [8].

Current diagnostic approaches for MDRV include virus isolation [8], serological assays such as the indirect immunofluorescence assay (IFA) [10] and the indirect enzyme linked immunosorbent assay (ELISA) [11], as well as molecular techniques including conventional and real‐time reverse transcription polymerase chain reaction (RT‐PCR) [12–14]. Although these methods are reliable, they often necessitate specialized equipment, trained personnel, and extended processing times, thereby limiting their applicability in field settings. Furthermore, virus isolation is a time‐intensive process, and serological assays are unable to detect early infections due to delayed antibody responses. These limitations highlight the critical need for a rapid, sensitive, and field‐deployable diagnostic tool to facilitate timely surveillance and effective outbreak control.

Recombinase‐aided amplification (RAA) is an innovative isothermal nucleic acid amplification technology well‐suited for point‐of‐care testing (POCT). RAA operates at a constant low temperature, typically ranging from 37 to 42°C, facilitating amplification within 10–20 min without necessitating thermal cycling. This method demonstrates a high tolerance to inhibitors present in crude samples and eliminates the requirement for cold‐chain storage of reagents, thereby rendering it particularly advantageous for deployment in remote areas [15, 16]. Owing to these benefits, RAA has been effectively utilized for the detection of a diverse range of pathogens. For instance, in viral detection, it has been successfully applied to identify bovine coronavirus and avian influenza virus subtype H7, demonstrating high sensitivity and specificity in clinical samples [16, 17]. In the bacterial domain, RAA combined with lateral flow dipstick has facilitated the rapid and simultaneous detection of major foodborne pathogens such as Escherichia coli O157, Salmonella, and Shigella in meat products [18]. For fungal pathogens, an RAA assay targeting Candida auris exhibited a detection limit of 10 copies/μL, outperforming real‐time PCR in some clinical sample types [19]. Furthermore, RAA has also been adapted for parasitic detection, as evidenced by the visual identification of anisakid nematodes in fish samples using RPA with SYBR Green I [20]. These applications highlight the versatility, robustness, and field‐deployability of RAA technology across a broad spectrum of infectious agents.

In this study, we developed a real‐time reverse transcription RAA (RT‐RAA) assay targeting the σC gene of MDRV for rapid and specific detection. The assay was systematically optimized and rigorously evaluated for its sensitivity, specificity, repeatability, and reproducibility. Notably, to facilitate its application in field settings and enhance storage stability, we established, to our knowledge, the first lyophilized reagent format for MDRV detection based on RT‐RAA. This formulation was integrated with a simplified RNA release protocol and supports dual‐mode readout: visual readout using a portable blue‐light imager or real‐time fluorescence monitoring using a compact handheld device (Figure 1), allowing for result interpretation without sophisticated instrumentation. The development of this lyophilized RT‐RAA assay combines high analytical performance with operational simplicity and improved field adaptability, presenting a significant advance toward practical, point‐of‐care diagnostics for MDRV surveillance, particularly in resource‐constrained regions.

Schematic workflow for rapid detection of MDRV using the RT‐RAA assay. The entire process, from sample to result, can be completed within 25 min. Key steps include: (1) Nucleic acid release via rapid lysis of tissue samples; (2) Isothermal amplification of target RNA by RT‐RAA at 39°C for 20 min, which can be performed using a portable fluorescence detector or a real‐time PCR system; (3) Result interpretation via real‐time fluorescence curves or visual assessment using a portable blue‐light imager. This schematic was created with BioRender (https://www.biorender.com).

2. Materials and Methods

2.1. Viral Strains

The following viral strains were used in this study: MDRV (GenBank accession no. AY580159.1), novel duck reovirus (NDRV, GenBank accession no. GQ888710.1), Muscovy duck origin goose parvovirus (MDGPV, GenBank accession no. OQ301813.1), Muscovy duck parvovirus (MDPV, GenBank accession no. KU844281.1), duck plague virus (DPV), and duck adenovirus serotype B2 (DAdV B2, GenBank accession no. PP763217.1). All viral strains were isolated and stored at the Institute of Animal Husbandry and Veterinary Medicine, Fujian Academy of Agricultural Sciences. The egg drop syndrome virus (EDSV) reference strain was provided by the China Institute of Veterinary Drug Control.

2.2. Preparation of RNA Molecular Standard

The σC gene (835 bp) of MDRV strain MW9710 was amplified by PCR using the viral genome as template. The resulting PCR product was cloned into the pGEM‐T Easy Vector System (Promega, USA) via TA ligation and subsequently transformed into E. coli DH5α competent cells (Sangon Biotech Co., Ltd., Shanghai, China). Positive recombinant clones were selected and confirmed by Sanger sequencing.

RNA transcripts were synthesized in vitro using the SP6/T7 RNA Transcription Kit (Thermo Fisher Scientific, USA) and purified. Concentration was quantified using the Quant‐iT RiboGreen RNA Assay Kit (Thermo Fisher Scientific, USA) according to the manufacturer’s instructions.

RNA copy number was calculated using the following formula:

$$[eqn]$$

The final RNA standard had a concentration of 1.03 × 10^10^ copies/μL. A 10‐fold serial dilution series ranging from 1.03 × 10^7^ to 1.03 × 10^0^ copies/μL was prepared and stored at –80°C for subsequent use.

2.3. Primers and Probes Design for MDRV Detection

Primers and an exo‐probe for RT‐RAA were designed based on the σC gene sequence of MDRV MW9710 (GenBank accession no. AY580159.1), following RAA design guidelines. Three primer sets and one exo‐probe were designed and evaluated for specificity using Primer‐BLAST (NCBI) to ensure no cross‐reactivity with nontarget sequences. All oligonucleotides were synthesized by Sangon Biotech (Shanghai, China); sequences and positions are listed in Table 1.

2.4. Screening of Optimal Primer Combinations

To identify the optimal primer set, all possible combinations of three forward and three reverse primers were evaluated for specificity and amplification efficiency. Reactions were performed using an RT‐RAA exo kit (Hangzhou ZC Bio‐Sci & Tech Co., Ltd., China) according to the manufacturer’s protocol with minor adjustments. Each lyophilized pellet was rehydrated with 45.5 µL of a master mix containing 25 µL of buffer A, 15.9 µL of nuclease‐free water, 2 µL of each primer (10 µM), and 0.6 µL of probe (10 μM). The solution was divided equally into two tubes, and 1 µL of RNA standard (1.03 × 10^7^ copies/μL) was added to each. To initiate amplification, 1.25 µL of 280 mM magnesium acetate was applied to the inner cap of each tube, followed by sealing, brief vortexing, and centrifugation. Nuclease‐free water was used as a negative control (NC).

Amplification was carried out at 39°C for 20 min in a Roche LightCycler 96 system, with fluorescence acquisition in the FAM channel every 30 s. Primer pairs showing early amplification and high fluorescence signals were considered candidates for optimal efficiency. Specificity was further verified by basic RT‐RAA followed by agarose gel electrophoresis (see Supplementary Information). The final primer pair was selected based on the presence of a strong, specific band, indicating high amplification yield and specificity.

2.5. Optimization of Real‐Time RT‐RAA Reaction Conditions

To improve the sensitivity of the assay, a systematic optimization of key reaction parameters, including incubation temperature and probe concentration was performed. The optimization was performed using the selected primer pair and an RNA standard (1.03 × 10^7^ copies/μL), evaluating performance across a range of temperatures (37–40°C) and exo‐probe concentrations (80, 100, and 120 nM).

2.6. Evaluation of Sensitivity and Specificity

Analytical sensitivity of the RT‐RAA assay was determined using 10‐fold serial dilutions of RNA standard ranging from 1.03 × 10^4^ to 1.03 × 10^0^ copies/μL. Nuclease‐free water served as a NC. The detection limit of the RT‐RAA assay was compared with a reference RT‐quantitative PCR (qPCR) method.

For specificity testing, nucleic acids were extracted from a variety of common duck pathogens, including NDRV, MDGPV, MDPV, DPV, DAdV B2, and EDSV, were tested using the optimized RT‐RAA conditions. MDRV was used as a positive control, and nuclease‐free water as a NC. All viral nucleic acids were extracted using the EasyPure Viral DNA/RNA Kit (TransGen Biotechnology Co., Ltd., Beijing, China) according to the manufacturer’s instructions.

2.7. Analysis of Repeatability and Reproducibility

Intra‐assay repeatability was determined by performing triplicate reactions within a single experimental run, whereas inter‐assay reproducibility was evaluated across three independent assays conducted on different days by different operators. Two RNA standard concentrations (1.03 × 10^7^ and 1.03 × 10^3^ copies/μL) were examined under both conditions, with nuclease‐free water as a NC. The coefficient of variation (CV) of the cycle threshold values was calculated for each concentration to quantify intra‐assay and inter‐assay variability.

2.8. Preparation of Ready‐to‐Use Lyophilized RT‐RAA Reagents

To facilitate on‐site deployment and enable refrigeration‐free storage and transport, a two‐component lyophilized RT‐RAA system was developed. The system consists of the following lyophilized reagents: (1) Reagent A: a lyophilized enzyme matrix containing primers and an exo‐probe. The formulation for each reaction contains 1 μL each of forward and reverse primers (10 μM), 0.3 μL of exo‐probe (10 μM), and 19.7 μL of an enzyme solution comprising recombinase, DNA polymerase, single‐stranded DNA‐binding protein (SSB), MLV reverse transcriptase, 7% (w/v) PEG 35,000, and 5% (w/v) trehalose. (2) Reagent B: a lyophilized mixture of isothermal amplification buffer and magnesium acetat. Each tube contains 12.5 μL of isothermal amplification buffer and 1.25 μL of 280 mM magnesium acetate.

Prior to lyophilization, all components of Reagent A and Reagent B were individually mixed and aliquoted into PCR strips (one reaction per tube). The lyophilization procedure included pre‐freezing at –80°C for 3 h, followed by a two‐step drying process in a freeze‐dryer: primary drying at –35°C and 10 mbar for 6 h, and secondary drying at 25°C under deep vacuum for an additional 6 h, with relative humidity maintained below 30%. Finally, the tubes were vacuum‐sealed in aluminum foil pouches and stored at –20°C until use.

2.9. Field Applicability Assessment of Ready‐to‐Use Lyophilized RT‐RAA Reagents

The clinical performance and field applicability of the lyophilized RT‐RAA assay were evaluated using 42 clinical samples collected from diseased Muscovy ducks across six farms in Putian and Zhangzhou, Fujian Province. Cardiac tissues were harvested within 2 h postmortem to preserve nucleic acid integrity and immediately immersed in DNA/RNA Shield (New England Biolabs, China) to prevent degradation. Samples were transported on dry ice and processed within 6 h of collection.

For POCT application, a rapid lysis protocol was employed to release target nucleic acids directly from tissue homogenates. Briefly, 2 µL of homogenate supernatant was mixed with 50 µL of lysis buffer and incubated at room temperature for 5 min to obtain a crude lysate. The lyophilized Reagent A and Reagent B were each rehydrated with 21 μL of nuclease‐free water. Subsequently, 4 μL of the crude lysate was added to the reconstituted reaction mixture. After gentle vortexing and brief centrifugation, amplification was carried out in a portable isothermal fluorescent detector (NAT‐08A; Genuine Biological Technology, Hangzhou, China). Reactions were conducted at 39°C for 20 min with fluorescence signals monitored in the FAM channel every 30 s. Nuclease‐free water served as a NC in each run. A sample was deemed positive if its amplification curve crossed the threshold defined by the NC. All results were visually confirmed using a TGreen portable blue‐light imager (Tiangen Biotech, Beijing, China). For comparative analysis, all clinical samples were tested in parallel using a previously established RT‐qPCR assay.

2.10. Statistical Analysis

Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software Inc., San Diego, CA, USA). Cohen’s kappa coefficient (κ) was calculated to evaluate the agreement between the results of the lyophilized RT‐RAA assay and RT‐qPCR. Diagnostic sensitivity, specificity, and predictive values with their 95% confidence intervals were calculated using the Clopper‐Pearson exact method. A κ value less than 0.4 was interpreted as indicating poor agreement, while a κ value of 0.75 or greater was considered to represent good agreement.

3. Results and Discussion

3.1. Screening of Primers for the Real Time RT‐RAA Assay

The optimal primer pair for the real‐time RT‐RAA assay was determined through a systematic screening process. Three forward primers (F478–505, F480–507, and F466–493) and three reverse primers (R594–621, R598–625, and R557–585) were designed and cross‐combined, resulting in nine candidate pairs (Figure 2a). Initial screening was performed according to the TwistAmp recommendation: all forward primers were tested against one randomly selected reverse primer. The forward primer F480–507 showed the shortest time to threshold (Tt) and was therefore selected for subsequent screening against all three reverse primers (Figure 2b). Among these, the combination of F480–507 and R598–625 exhibited the fastest amplification (Figure 2c) and was identified as the top candidate in the primary screen. Agarose gel electrophoresis confirmed that this pair produced a single, bright band of the expected size (146 bp) without non‐specific amplification or primer dimers (Figure S1), corroborating the fluorescence‐based results.

Figure 2. Primer design and screening for the real‐time RT‐RAA assay. (a) Schematic of the primary primer screening strategy. Numerals in primer names indicate binding positions within the σC gene of the MDRV MW9710 strain (GenBank accession no. AY580159.1). (b) Screening of three forward primers paired with reverse primer R594–621. (c) Screening of three reverse primers paired with the selected forward primer F480–507. (d) Strategy for secondary screening. The positions of the top primers (F480–507 and R598–625) were incrementally shifted by 1–2 nucleotides toward either the 5′ or 3′ end, while maintaining constant primer length, generating four forward and four reverse primer variants. (e) Evaluation of four forward primer variants with fixed reverse primer R596–623. (f) Testing of four reverse primer variants with the optimal forward primer F479–506. (g) Performance comparison of the two final candidate pairs using a low‐concentration RNA template (1.03 × 10^3^ copies/μL). Unless specified, all reactions used 1 μL of RNA standard (1.03 × 10^7^ copies/μL) at 39°C. Nuclease‐free water served as the negative control (NC). Fluorescence was monitored in the FAM channel for 20 min.(a)(b)(c)(d)(e)(f)(g)

To further enhance performance, a secondary screening was conducted by shifting the positions of the selected forward and reverse primers by one or two nucleotides toward the 5′ or 3′ end while maintaining primer length. This generated four forward and four reverse variants (Figure 2d). Screening of the forward variants identified F479–506 as the most efficient (Figure 2e), while evaluation of the reverse variants revealed R600–627 as the top performer (Figure 2f).

To validate these findings under more challenging conditions, the leading primer pairs—F479–506/R597–624 and F479–506/R600–627—were compared using a low‐concentration RNA standard (1.03 × 10^3^ copies/μL). The pair F479–506/R600–627 achieved a shorter Tt (Figure 2g), confirming its superior kinetics and sensitivity at low template levels. Based on these consistent results across screening stages and template concentrations, F479–506 and R600–627 were selected as the final primer pair for all subsequent assays.

3.2. Optimization of Reaction Conditions for the Real‐Time RT‐RAA Assay

To achieve optimal performance, two critical parameters—probe concentration and reaction temperature—were systematically optimized utilizing the F479–506 and R600–627 primer pair. The probe concentration was first evaluated at 80, 100, and 120 nM, with the reaction temperature held constant at 39°C. As shown in Figure 3a, a concentration of 100 nM produced the strongest fluorescence signal and the shortest time to threshold, whereas both lower (80 nM) and higher (120 nM) concentrations led to diminished signals or delayed amplification. Therefore, 100 nM was selected as the optimal probe concentration for all subsequent experiments.

Figure 3. Optimization of RT‐RAA assay conditions. (a) Probe concentration was evaluated at 80, 100, and 120 nM. A concentration of 100 nM (red curve) yielded the earliest amplification time and strongest fluorescence signal was selected as optimal. (b) Reaction temperature was tested from 37 to 40°C. The most rapid and robust amplification was achieved at 39°C (red curve), establishing it as the optimal temperature. All reactions used the F479–506/R600–627 primer pair and 1 μL of RNA standard (1.03 × 10^7^ copies/μL). Nuclease‐free water served as the negative control (NC).(a)(b)

Next, the influence of temperature on amplification efficiency was examined at 37, 38, 39, and 40°C using the optimized probe concentration. As illustrated in Figure 3b, 39°C (red curve) yielded the highest amplification efficiency, marked by the earliest signal onset and the highest fluorescence gain. By contrast, reactions performed at lower temperatures exhibited delayed amplification kinetics, suggesting suboptimal recombinase activity. Based on these results, 39°C was identified as the ideal reaction temperature and adopted for all further RT‐RAA assays.

3.3. Sensitivity and Specificity Evaluation of the RT‐RAA Assay

The analytical sensitivity of the optimized RT‐RAA assay was evaluated using a 10‐fold serial dilution of standard RNA template, ranging from 1.03 × 10^4^ to 1.03 × 10^0^ copies/μL. As depicted in Figure 4a, distinct amplification signals were observed from 1.03 × 10^4^ to 1.03 × 10^1^ copies/μL, while the lowest concentration (1.03 × 10^0^ copies/μL) produced no detectable signal. Thus, the limit of detection (LoD) for the liquid‐format assay was determined to be 1.03 × 10^1^ copies/μL. This sensitivity is comparable to that of a reference RT‐qPCR assay (Figure S2) and surpasses other reported methods for MDRV detection (Table 2).

Figure 4. Evaluation of sensitivity and specificity of the RT‐RAA assay for MDRV detection. (a) Analytical sensitivity. Ten‐fold serial dilutions of MDRV RNA standard (ranging from 1.03 × 10^4^ to 1.03 × 10^0^ copies/μL) were tested, with 1 μL of RNA standard used as template per reaction. The limit of detection was determined to be 1.03 × 10^1^ copies/μL. No amplification was observed in the nuclease‐free water negative control (NC). (b) Specificity analysis by real‐time fluorescence. The assay was challenged with nucleic acids from Muscovy duck reovirus (MDRV) and non‐target pathogens including novel duck reovirus (NDRV), Muscovy duck origin goose parvovirus (MDGPV), Muscovy duck parvovirus (MDPV), duck plague virus (DPV), duck adenovirus serotype B2 (DAdV B2), and egg drop syndrome virus (EDSV). All templates were normalized to the same concentration. Specific amplification was exclusively observed in the MDRV sample, with no cross‐reactivity detected in non‐target viruses or the negative control. (c) Visual specificity under a portable blue‐light imager (excitation at 480 nm). 1–8 correspond to: (1) MDRV, (2) NDRV, (3) MDGPV, (4) MDPV, (5) DPV, (6) DAdV‐B2, (7) EDSV, and (8) nuclease‐free water (NC). Bright green fluorescence was observed solely in the MDRV sample, confirming high specificity.(a)(b)(c)

To facilitate field deployment, the assay components were lyophilized into a ready‐to‐use format. Upon rehydration and testing under identical conditions, the lyophilized assay achieved the same LoD (1.03 × 10^1^ copies/μL) with no significant delay in amplification kinetics (Figure S3a), indicating that the lyophilization process did not compromise the assay’s intrinsic sensitivity. The maintained performance is largely attributable to the use of cryoprotectants such as trehalose, which protects biomacromolecules during dehydration by substituting for water and forming hydrogen bonds, thereby preventing structural denaturation and inactivation [21].

The compatibility of the lyophilized assay with POCT was further verified on a portable isothermal amplification device. As illustrated in Figure S4a, the assay consistently detected MDRV RNA at concentrations as low as 1.03 × 10^1^ copies/μL, producing clear amplification curves within 20 min without false positives in NCs. This confirms that the reagents can be reliably rehydrated with water without loss of detection performance, even on miniaturized instrumentation.

Assay specificity was rigorously evaluated against a panel of common waterfowl pathogens, including NDRV, MDGPV, MDPV, DPV, DAdV B2, and EDSV. Across all formats—liquid, lyophilized, and portable—amplification was observed exclusively in MDRV‐positive samples, with no cross‐reactivity detected (Figure 4b and Figures S3b and S4b). These results confirm that neither lyophilization nor portable instrumentation compromised the high specificity of the assay.

In addition to real‐time fluorescence detection, amplification products were visually assessed using a portable blue‐light imager. Only the MDRV sample exhibited bright green fluorescence; nontarget viruses and NCs showed no signal (Figure 4c). This visual readout provides an easily interpretable result suitable for nonexpert users, reinforcing the assay’s suitability for field use. Similar endpoint detection strategies have been successfully implemented in lyophilized LAMP assays [22–25].

Our findings are consistent with studies on other lyophilized isothermal assays. Prado et al. [22] reported that RT‐LAMP reagents retained diagnostic performance after 28 days at ambient temperature, while Matl et al. [25] demonstrated that lyophilized assays perform robustly in less developed regions without cold chain dependance. These precedents highlight the potential of lyophilization in enhancing the feasibility of field‐deployable diagnostic tools.

In summary, this study establishes a lyophilized RT‐RAA assay that exhibits high sensitivity and exceptional specificity for MDRV, with no cross‐reactivity detected against related avian pathogens. The lyophilized format preserves the fundamental analytical performance of the assay, enabling instrument‐free detection suited for field use. Future research aimed at thoroughly assessing the long‐term stability of this format will be essential to determine its overall viability in resource‐limited environments.

3.4. Analytical Repeatability and Reproducibility of the RT‐RAA Assay

The reliability of the developed RT‐RAA assay was rigorously assessed by evaluating its repeatability (intra‐assay precision) and reproducibility (inter‐assay precision). Intra‐batch and inter‐batch experiments were performed using RNA standards at two concentrations (1.03 × 10^7^ and 1.03 × 10^3^ copie/μL). For the liquid‐format assay, the CVs ranged from 2.36% to 4.09% for repeatability and from 1.76% to 4.48% for reproducibility (Figure 5c). All CV values were well below the 5% threshold generally accepted for molecular diagnostics, demonstrating a level of precision comparable to or exceeding that of other RAA‐based virus detection methods [26, 27]. This high level of precision reflects the robustness of the RT‐RAA assay.

Figure 5. Assessment of the repeatability and reproducibility of the RT‐RAA assay. (a) Repeatability (intra‐assay precision) and (b) reproducibility (inter‐assay precision) were assessed using RNA standards at high (1.03 × 10^7^) and low (1.03 × 10^3^) copies/μL. (c) Summary of the results. Data points show mean cycle threshold values from three independent assays, with error bars indicating standard deviation (SD). The consistently low coefficients of variation (CV) affirm the high precision and reliability of the assay.(a)(b)(c)

A similar precision evaluation was performed using the lyophilized RT‐RAA formulation. As summarized in Figure S5, both intra‐ and inter‐assay CVs for high‐ and low‐concentration RNA standards consistently remained below 5%, ranging from 1.36% to 4.09%. These results confirm that the freeze‐drying process did not adversely affect assay precision and that the dried reagents maintain high consistency across batches and experimental runs.

Collectively, these data demonstrate that both the liquid and lyophilized formats of the RT‐RAA assay exhibit excellent repeatability and reproducibility. The robust precision profiles further support the reliability of the assay in both laboratory and potential point‐of‐care settings.

3.5. Field Applicability Assessment of the Lyophilized RT‐RAA Assay

The clinical performance of the lyophilized RT‐RAA assay was evaluated using 42 tissue samples collected from diseased Muscovy ducks exhibiting characteristic MDRV infection symptoms, including gray‐white necrotic foci on the surfaces of the spleen and liver (Figure S6). All samples were tested under simulated field conditions using the freeze‐dried reagents. Compared with RT‐qPCR as the reference method, the real‐time fluorescence‐based RT‐RAA assay demonstrated excellent diagnostic agreement (Table 3). Of the 42 clinical samples, the RT‐RAA assay identified 23 positives and 19 negatives, while RT‐qPCR detected 24 positives and 18 negatives. The kappa value between the two methods was 0.952 (p < 0.001), indicating almost perfect concordance.

The fluorescence‐based RT‐RAA assay showed a sensitivity of 95.8% (23/24; 95% CI: 79.8%–99.9%) and a specificity of 100% (18/18; 95% CI: 81.5%–100%). The positive and negative predictive values were 100% (95% CI: 85.2%–100%) and 94.7% (95% CI: 74.0%–99.9%), respectively, with an overall agreement of 97.6% (41/42). The single discordant sample, which was positive by RT‐qPCR (Ct = 37.8) but negative by RT‐RAA, also tested negative in virus isolation. This discrepancy may be attributed to stochastic variation at the very low viral load (near the common detection limit of both assays), or to the presence of substances in the crude lysate that differentially inhibited the RT‐RAA amplification.

When the same amplification products were visually assessed under blue light, 20 of the 24 RT‐qPCR‐positive samples showed distinct green fluorescence, corresponding to a visual detection sensitivity of 83.3%. No false‐positive signals were observed, confirming 100% specificity in visual readout. The four missed samples by visual readout all had high Ct values (>35) in RT‐qPCR, indicating low viral loads near the detection limit. This is consistent with previous reports that visual interpretation exhibits slightly reduced sensitivity compared to instrument‐based fluorescence detection, particularly near the detection threshold [26]. The lower sensitivity of the visual readout is likely attributable to the inherent limitations of human visual perception in distinguishing weak positive fluorescence from the background when the amplicon yield is low. Despite the lower sensitivity of the visual readout, the absence of false positives and the operational simplicity of the lyophilized format make the assay highly suitable for field deployment. Future studies should therefore focus on enhancing the sensitivity of the visual readout. A promising strategy would be the development of low‐cost, smartphone‐based imaging systems equipped with optimized optical filters to better capture weak fluorescence signals. Integrating such devices with automated image processing algorithms for background subtraction and signal quantification should significantly lower the detection threshold for samples with low viral loads, thereby bridging the sensitivity gap with instrumental fluorescence readouts [28, 29]. The stability and portability of lyophilized reagents further support their use in resource‐limited settings [22, 30, 31], where rapid screening and ease of use are often prioritized over maximal analytical sensitivity.

In summary, these findings confirm that the RT‐RAA assay is highly accurate and reliable for MDRV detection, showing performance comparable to the reference RT‐qPCR method (Table 2). This robust performance is consistently achieved with both instrument‐based fluorescence and visual readout formats.

4. Conclusion

In summary, this study successfully developed a lyophilized real‐time RT‐RAA assay for the rapid and on‐site detection of MDRV. The assay achieved a LoD of 1.03 × 10^1^ copies/μL within 20 min, exhibiting sensitivity comparable to that of RT‐qPCR through the analysis of serial dilutions of standard RNA, while offering a significantly reduced time to result. The assay displayed high specificity and was successfully lyophilized into a ready‐to‐use format, representing the first report of a lyophilized RT‐RAA method for MDRV detection. Clinical validation further confirmed its excellent agreement with the reference RT‐qPCR method. This robust, portable, and user‐friendly assay provides a powerful tool for field‐deployable diagnosis and active surveillance of MDRV, particularly in underdeveloped areas. The successful lyophilization presents a crucial step towards instrument‐free testing, pending further validation of long‐term reagent stability.

Funding

This research was funded by the Natural Science Foundation of Fujian Province (Grant 2025J011234), the Fujian Public Welfare Project (Grant No. 2023R1024002, 2024R1025003, and 2025R1024001) and the “5511″ Collaborative Innovation Project of Fujian Academy of Agricultural Sciences, China (Grant No. XTCXGC2021018 and XTCXGC2021012).

Disclosure

All authors contributed to the article and approved the submitted version.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting Information

Additional supporting information can be found online in the Supporting Information section.

Supporting information

Supporting Information Figure S1: Screening of primer pairs for the RT‐RAA assay. Combinations of three forward and three reverse primers were evaluated by agarose gel electrophoresis to identify the optimal pair for specific amplification of MDRV. Figure S2: Analytical sensitivity of the RT‐qPCR assay. Amplification curves and corresponding standard curve generated from serial dilutions of MDRV RNA standard, demonstrating a detection limit of 1.03 × 10^1^ copies/μL. Figure S3: Analytical performance of the lyophilized RT‐RAA assay upon rehydration. Sensitivity and specificity testing against MDRV RNA and non‐target pathogens, showing a detection limit of 1.03 × 10^1^ copies/μL and no cross‐reactivity. Figure S4: Performance of the RT‐RAA assay on a portable isothermal fluorescent detector. Sensitivity and specificity results confirming consistent detection of MDRV without interference from related avian viruses. Figure S5: Repeatability and reproducibility of the lyophilized RT‐RAA assay. Intra‐ and inter‐assay precision data at high and low RNA concentrations, with coefficients of variation below 5%, indicating robust assay performance post‐lyophilization. Figure S6: Gross pathological lesions in Muscovy ducks naturally infected with MDRV. Images of spleen and liver showing characteristic gray‐white necrotic foci indicative of infection.