High-sensitivity multiplex real-time RT-PCR for universal detection and virulence typing of Newcastle disease virus in poultry field samples

Abstract

Newcastle disease virus (NDV) remains a major global threat to poultry production despite widespread vaccination, driven by the emergence of genetically diverse and virulent Class II strains. We developed and validated a one-step multiplex real-time RT-PCR assay enabling simultaneous universal detection of all Class II NDV genotypes and differentiation between virulent (velogenic/mesogenic) and non-virulent (lentogenic/asymptomatic) pathotypes in a single reaction. The assay showed high analytical sensitivity, detecting as few as 10^1.0 EID₅₀ /0.1 mL for HN target and 10^2.0EID₅₀/0.1 mL for F target, with limits of 10 and 100 copies/tube using plasmid standards, and exhibited no cross-reactivity with non-target avian viruses. When evaluated against 97 field isolates and seven reference strains collected over a 75-year period (1946-2021) from multiple continents, the assay achieved 100% concordance with four established molecular assays. All 80 isolates with multi-basic F protein cleavage site motifs (¹¹²RRQKRF¹¹⁷, ¹¹²RRRKRF¹¹⁷, ¹¹²KRRKRF¹¹⁷, and ¹¹²RRQRRF¹¹⁷) were positive for both targets, whereas those with monobasic motifs (¹¹²GKQGRL¹¹⁷ and ¹¹²GRQGRL¹¹⁷) were detected only by the universal HN assay, consistent with a non-virulent phenotype.

The assay detected NDV lineages spanning >75 years of virus evolution, from historical strains (Herts 33/56, Kr-KJW/49) to recent isolates (UPM/1051/2018, UPM111), capturing up to 23.7% nucleotide divergence in complete HN genes and 13.4% in F genes. This platform integrates broad-spectrum detection with precise virulence differentiation, offering a rapid, sensitive, and reliable tool for Class II NDV surveillance, outbreak response, and vaccine strain monitoring.

Introduction

Newcastle disease (ND) is a major transboundary avian disease that causes high morbidity and mortality in both domestic and wild birds, resulting in severe economic losses for the poultry industry.

The etiologic agent, avian paramyxovirus type I (APMV-1), commonly referred to as Newcastle disease virus (NDV; genus Avulavirus, family Paramyxoviridae), possesses a non-segmented, single-stranded, negative-sense RNA genome approximately 15.2 kb in length. This genome encodes six structural proteins, nucleocapsid (NP), phosphoprotein (P), matrix (M), fusion (F), hemagglutinin–neuraminidase (HN), and large polymerase (L), which coordinate virion assembly, host cell entry, and genome replication (Gao et al., 2019). Among them, the F and HN glycoproteins mediate viral attachment and membrane fusion, while NP, P, and L proteins form the ribonucleoprotein complex essential for transcription and replication (Shabbir et al., 2024).

Phylogenetic analyses have classified NDV strains into two major classes: Class I and II (Dimitrov et al., 2019; Zereen et al., 2025; Zhang et al., 2025). Class I strains, typically of low virulence, circulate predominantly in wild waterfowl, shorebirds, and live‑bird markets (Miller et al., 2015; Wang et al., 2016). In contrast, Class II strains comprise 21 genotypes (I–XXI) and five pathotypes, viscerotropic velogenic, neurotropic velogenic, mesogenic, lentogenic, and asymptomatic, which infect a broader range of avian hosts, including domestic poultry (Song et al., 2023). Of these, only the velogenic and mesogenic pathotypes are designated as notifiable diseases by the World Organisation for Animal Health (WOAH) and the European Union (EU), due to their capacity to induce severe disease and substantial economic loss (Fortin et al., 2023).

Virulence determination according to WOAH criteria relies on either an intracerebral pathogenicity index (ICPI) greater than 0.7 in day-old chicks (maximum score of 2.0), or the presence of a multi-basic cleavage site in the F protein, defined by at least three arginine or lysine residues at positions 112–116 followed by a phenylalanine at position 117 (112(R/K)-R-Q-(R/K)-R↓F117) (Liu et al., 2023; WOAH, 2022). While these molecular and in vivo standards reliably differentiate virulent from low‑pathogenic NDV strains, ICPI testing is labor-intensive, requiring live virus isolation and intracerebral inoculation of chicks under biosafety conditions. As a result, molecular diagnostics, particularly reverse transcription PCR (RT-PCR) and real-time RT-PCR, have become essential tools for rapid NDV detection (Alexander Morris et al., 2025; Farkas et al., 2009; Nguyen et al., 2013; Wise et al., 2004; Yi et al., 2011). Although several real-time RT-PCR assays have been developed for NDV detection and pathotyping (Kim et al., 2008; Wise et al., 2004), there remains a continuous need for optimized assays capable of accommodating the substantial genetic diversity of contemporary Class II isolates. Our method builds upon these established approaches by incorporating optimized primer sets targeting the HN and F genes, ensuring broad genetic coverage and reliable virulence-specific detection in the face of ongoing viral evolution. This represents a critical gap in diagnostics, especially in countries such as Korea, where ND is a legally regulated disease and rapid confirmation is required by the Animal and Plant Quarantine Agency (APQA). Therefore, this study aimed to develop and validate a multiplex real-time RT-PCR assay capable of concurrently detecting all Class II NDV genotypes and differentiating virulent from non-virulent pathotypes in a single assay platform.

Materials and methods

In silico design of primer sets

To design universal primers for the HN gene and virulence-specific primers for the F gene, full-length sequences of each target were retrieved from the National Center for Biotechnology Information (NCBI) GenBank database. These sequences, representing diverse Class II NDV genotypes (I–XXI), were subjected to multiple sequence alignment to identify highly conserved regions. To account for genetic diversity and minimize potential mismatches, primer binding sites were strategically selected based on the highest consensus scores, and degenerate bases were incorporated where necessary to accommodate known single-nucleotide polymorphisms (SNPs) among contemporary isolates. The alignments identified three highly conserved regions within the HN gene, enabling the universal detection of a broad range of Class II NDV strains, which encompass all major pathotypes (viscerotropic velogenic, neurotropic velogenic, mesogenic, lentogenic, and asymptomatic) (Supplementary Fig.1A).

Additionally, conserved motifs within the F gene were specifically targeted to allow for the selective amplification of virulent strains (velogenic and mesogenic) (Supplementary Fig.1B). The primer and probe sequences were finalized and optimized using OligoAnalyzer (Integrated DNA Technologies, USA). The resulting primer–probe sets amplified 198 bp fragments for the universal HN assay and 195 bp fragments for the virulence-specific F assay. All primer and probe sequences are listed in Table 1.

Table 1.

Primer sets optimized for detection and differentiation of Newcastle disease virus.

Primer set	Target gene	primer	Sequence (5′−3′)	Product Size (bp)	Accession number
Universal	HN	F	GTTGCACTCGGATACCCTCATT	198	OP921614, OP921615, OP921650, OP921697, OP921699, OP921616, OP921648, OP921706, HF969218, HF969194, OP921656, FJ410145, KY042135, MZ223437, JX012096, OP921653, AY741404, AY562991, OP921642, GQ918280, AF077761, EU289028
		R	GACTTCCGATTTTGGGTGTCATCTA
		P	HEX-AGGGTATTCTTTTCTACTCTGCGTTCC-TAMRA
Virulent	F	F	TCTATCCGCAGGATYCAAGAGTC	195	EU289028, AY562991, AF077761, MH996904, AY741404, KX857721, FJ436303, JN872151, JX915243, JX012096, OP818783, JX901351, KY042141, OP818787, HF969205, HF969194, JX518886, MZ306223, JN942043, OP818746, OP818811, OP818829, OP818789, FJ410145, OP818852, KT163262, HF969191, JN942024, AB853928, OP818834
		R	CATTGGTTGCRGCAATGCTCTC
		P	FAM-AGGAGACAAAAACGCTTYATAGGTGCC-TAMRA

Viral RNA preparation

For specificity testing, total RNA was extracted from a panel of NDV and non-NDV avian viruses, including APMV-1 (chicken, Korea, velogenic Kr005 and lentogenic LaSota/46), APMV-2 (chicken, Yucaipa, California, 56), APMV-3 (parakeet, Netherlands, 449/75), APMV-4 (duck, Hong Kong, D3/75), APMV-6 (duck, Hong Kong, 18/199/77), APMV-7 (dove, Tennessee, 4/75), APMV-8 (goose, Delaware, 1053/76), APMV-9 (duck, New York, 22/78), avian influenza virus H9N2 (chicken, Korea, 16AQ075), infectious bronchitis virus (ATCC VR21), avian metapneumovirus (chicken, Korea, 0915), avian encephalomyelitis virus (chicken, Korea, 14ABQ615), infectious bursal disease virus (ATCC VR478), and avian reovirus (chicken, Korea, 15ABD080). These viruses were obtained from the APQA. RNA extraction was performed using the Maelstrom 4000 DNA/RNA automated extraction system (TANBead, Taiwan) in accordance with the manufacturer’s protocol. Extracted RNA was eluted in 50 µL of elution buffer and stored at −70°C until further use.

Optimization of multiplex real-time RT-PCR assay

A one-step multiplex real-time RT-PCR assay was optimized in a 20 μL reaction containing 10 µL of BlackPCR RT-qPCR Master Mix (Ventech Science, South Korea), 5 µL of RNA template, 0.25 µM each of the universal HN primers and probe, and 0.50 µM each of the F gene primers with 0.25 µM of the F gene probe. Amplification was carried out on a QuantStudio 5 Real-Time PCR System (Applied Biosystems, USA) under the following cycling conditions: reverse transcription at 50°C for 10 min, initial denaturation at 95°C for 20 sec, and 40 cycles of 95°C for 5 sec and 56°C for 50 sec with fluorescence acquisition at 56°C. Data analysis was performed using the QuantStudio Design & Analysis Software (Applied Biosystems, USA). For each target, the baseline and fluorescence threshold settings were automatically determined by the software. A cycle threshold (Ct) cutoff value of 38 was applied, and samples showing a Ct value below 38 together with a typical sigmoidal amplification curve were considered positive. For result interpretation, a sample was classified as virulent NDV when both the HN and F genes were detected. In cases where only the HN gene was detected (HN+/F−), the sample was classified as non-virulent NDV.

Construction of standard plasmid for positive controls

Viral RNA from Kr005 and LaSota/46 strains was amplified using HN-specific primers (forward 5′-CAAGAACACCTGAATTTTATCCC-3′, reverse 5′-GATAGATAAGATGGCTTGCTGTA-3′) and F-specific primers (forward 5′-AATATGCCCAAGGATAAAGAGGC-3′, reverse 5′-TAATGAGCTG AGTTGGTTGTTCC-3′), which fully encompassed the sequences targeted by the multiplex assay, using a BlackPCR RT-PCR Premix (Ventech Science, South Korea). PCR cycling included 50°C for 10 min, 94°C for 5 min, 35 cycles of 94°C for 50 sec, 58°C for 50 sec, 72°C for 50 sec, and a final extension at 72°C for 5 min. Amplified products were purified and cloned into the pGEM-T Easy Vector (Promega, USA), yielding pGEM-HN and pGEM-F standard control plasmids. Recombinant clones were confirmed by Sanger sequencing (CosmoGenetech Co., Korea) and used to construct standard curves for assay quantification. Copy number calculation was performed as previously described for absolute quantitative PCR analysis (Whelan et al., 2003).

Specificity, sensitivity, and reproducibility testing

Specificity of the optimized multiplex real-time RT‑PCR assay was determined using 50 ng of RNA extracted from APMV-1 (including Kr005 and Lasota/46) and non-target avian viruses: APMV −2, −3, −4, −6, −7, −8, −9, AIV H9N2, IBV, AMPV, AEV, IBDV, and ARV.

Analytical sensitivity was evaluated using viral RNA extracted from allantoic fluid containing 10^7.0 EID₅₀/0.1 mL of Kr005 (velogenic) and LaSota/46 (lentogenic) strains, serially diluted tenfold. In parallel, recombinant pGEM-HN and pGEM-F plasmids were also serially diluted from 10^8.0 copies/tube to 10^1.0 copies/tube. From the definition of the limit of detection (LoD) follows that working at a 95% confidence level (Forootan et al., 2017), LoD is the measured concentration that produces at least 95% positive replicates. For consistency, sensitivity results were expressed in both EID₅₀/0.1 mL (for viral RNA) and copies/tube (for plasmid standards). The performance of the standard curves was evaluated using the coefficient of determination (R²) and amplification efficiency (E), calculated from the slope of the linear regression (%E = (10 ^(−1/slope) −1) × 100) (Bustin et al., 2025). Acceptable performance was defined as R² ≥ 0.99 and efficiencies between 90 and 110%.

Reproducibility was determined by testing three concentrations of each plasmid standard (10^6.0, 10^5.0, and 10^3.0 copies/tube) in triplicate within a single run (intra-assay variation). Inter-assay variation was evaluated through two independent runs performed on different days by two separate operators using different reagent lots. Ct values were used to calculate coefficients of variation (CV, CV = standard deviation/average Ct × 100), confirming the assay’s precision across different template concentrations and runs. Values of CV ≤ 10% were considered high repeatability and reproducibility, and CV ≤ 5% were indicative of very high R&R (Olveira et al., 2021).

Application of multiplex real-time RT-PCR assay

To evaluate the applicability of the multiplex real-time RT-PCR assay across genetically and geographically diverse NDV lineages, we assembled a panel of 104 isolates comprising 97 field strains and seven reference strains collected over a 75-year period (1946-2021) from multiple continents. This panel encompassed both historical and contemporary viruses to ensure broad representation of NDV diversity. Among the field strains, 77 velogenic isolates originated from Korea (n = 42, collected between 1982 and 2007), China (n = 2), Mongolia (n = 5), Vietnam (n = 5), Cambodia (n = 6), Pakistan (n = 3), and Malaysia (n = 13). In addition, 20 lentogenic field isolates were obtained in Korea between 2002 and 2007. The reference panel comprised three velogenic strains, Herts 33/56, Kr-KJW/49, and Kr005, and four lentogenic strains, Hitchner B1, Ulster/67, VG/GA, and LaSota/46. All isolates were propagated in 9-day-old specific-pathogen-free (SPF) embryonated chicken eggs via allantoic cavity inoculation, followed by confirmation of viral presence through hemagglutination activity (HA titres 2^7.0-2^10.0). Moreover, to assess the assay performance on field materials, 14 clinical samples were collected from naturally infected birds in Vietnam in 2023. These samples consisted of pooled trachea, liver, and kidney tissues, which were homogenized in 10% (w/v) phosphate-buffered saline (PBS). The homogenates were centrifuged at 10,000 × g for 5 min at 4°C. Viral RNA was extracted using the Maelstrom 4000 DNA/RNA automated system (TANBead, Taiwan) and stored at −70°C until analysis.

Validation of multiplex real-time RT-PCR assay

The genetically and geographically diverse NDV panel described above was evaluated in parallel using four established reference molecular assays: two RT-LAMP assays (one targeting the HN gene for universal NDV detection and one targeting the F gene for virulence determination) and two conventional RT-PCR assays (one each targeting the full-length HN and F genes), all performed as previously described (Song et al., 2023). Comparative analysis between the multiplex assay and these reference methods was used to determine diagnostic sensitivity, specificity, and overall concordance. Concordance was calculated as the percentage of samples with consistent results across all assays relative to the total number of samples tested.

Virulence classification was determined by detecting multi-basic cleavage site motifs in the F protein, with confirmation through sequencing. To verify strain identity and assess the assay’s genetic coverage across NDV diversity, full-length HN and F genes from all isolates were sequenced (CosmoGenetech Co., Korea). These sequences, along with representative reference strains retrieved from GenBank, were aligned using ClustalW, and subjected to phylogenetic analysis using the neighbor-joining method in MEGA 11 (Saitou and Nei, 1987). The evolutionary distances were calculated using the Kimura 2-parameter model, and the reliability of the phylogenetic tree was assessed with 1,000 bootstrap replicates (Kimura, 1980). This analysis confirmed that the primer/probe binding sites are highly conserved across various genotypes, supporting the broad diagnostic capability of the assay.

Results

Development of a multiplex real-time RT-PCR assay

We successfully developed a one-step multiplex real-time RT-PCR assay designed for the dual purpose of universal NDV detection and virulence differentiation. This dual-target system allows both the presence of NDV and its virulence status to be determined within a single reaction, streamlining diagnostic workflows. Under optimized reaction conditions, no primer-dimer formation or nonspecific amplification was observed. Universal NDV detection was achieved using a FAM-labeled probe targeting the HN gene, while virulence-specific amplification was facilitated by a HEX-labeled probe targeting the multi-basic cleavage site of the F gene. Both targets were efficiently amplified in representative velogenic (Kr005) and lentogenic (LaSota/46) APMV-1 strains (Fig. 1), demonstrating the assay’s capability to accurately detect and discriminate between genetically and pathotypically distinct NDV strains.

Fig. 1.

Specificity of the one-step multiplex real-time TaqMan RT-PCR assay. Amplification profiles of the assay evaluated using RNA from velogenic (Kr005) and lentogenic (LaSota) APMV-1 strains, 13 non-target avian viruses (APMV-2, −3, −4, −6, −7, −8, −9, AIV H₉N₂, IBV, AMPV, AEV, IBDV, and ARV), and a no-template control. Kr005 was detected by both FAM-labeled HN and HEX-labeled F probes, while LaSota yielded amplification only with the HN target. No amplification was observed for any non-NDV samples, demonstrating the high specificity of the multiplex assay.

Specificity of the multiplex real-time RT-PCR assay

The analytical specificity was rigorously evaluated using RNA from 13 non-target avian viruses, including APMV serotypes 2, 3, 4, 6, 7, 8, and 9, AIV H9N2, IBV, AMPV, AEV, IBDV, and ARV. This diverse panel represented a broad spectrum of avian pathogens that could potentially co-infect with NDV or cause similar clinical signs. No cross-reactivity was observed with any of these non-target RNA samples, demonstrating that the assay exclusively amplifies NDV RNA without false positives from other avian viruses (Fig. 1).

Sensitivity of the multiplex real-time RT-PCR assay

The assay reliably detected as few as 10^1.0EID₅₀/0.1 mL for the universal HN target and 10^2.0EID₅₀/0.1 mL for the virulence-specific F target when tested with RNA extracted from live virus preparations (Fig. 2A, Supplementary Table 1). Using recombinant plasmids (pGEM-HN and pGEM-F) as templates, the detection limits were approximately 10 copies/tube for the HN gene and 100 copies/tube for the F gene (Fig. 2B, Supplementary Table 2). These detection limits were comparable to those observed in corresponding singleplex assays, indicating that multiplexing did not compromise detection efficiency (Fig. 2C and 2D). Standard curves, generated by plotting log₁₀-transformed plasmid copy numbers against the corresponding cycle threshold (Ct) values, showed excellent linearity over eight orders of magnitude (10^8.0 to 10^1.0 copies), with a correlation coefficient (R²) of 0.998. PCR efficiencies were 100.2% for the HN target (slope = –3.318) and 104.7% for the F target (slope = –3.213) (Fig. 2E and 2F).

Fig. 2.

Analytical sensitivity and standard curves of the one-step multiplex real-time RT-PCR assay. (A) Sensitivity evaluation using 10-fold serial dilutions of Kr005 viral RNA, expressed as EID₅₀/0.1 mL. (B) Detection limits of the multiplex assay assessed using 10-fold serial dilutions of pGEM-HN and pGEM-F plasmid standards. (C) and (D) Detection limits of pGEM-HN and pGEM-F, respectively, evaluated using the singleplex real-time RT-PCR assays.

(E) and (F) Standard curves constructed by plotting Ct values against log₁₀-transformed copy numbers of pGEM-HN and pGEM-F, with the coefficient of determination (R²) and PCR efficiency indicated.

Reproducibility was confirmed through precision testing, with intra-assay CV ranging from 0.40% to 0.93%, and inter-assay CV ranging from 0.45% to 0.89% (Supplementary Table 3), demonstrating high precision across different template concentrations and runs.

Genetic characteristics of NDV panel

Sequence analysis of the fusion protein cleavage site (residues 112-117) revealed six distinct motifs: ¹¹²RRQKRF¹¹⁷ (54 strains: 40 from Korea, 5 from Mongolia, 3 from Pakistan, 2 from China, 3 from Malaysia, and 1 from Vietnam), ¹¹²RRRKRF¹¹⁷ (23 strains: 8 from Malaysia, 6 from Cambodia, 5 from Korea, and 4 from Vietnam), ¹¹²KRRKRF¹¹⁷ (2 strains from Malaysia), ¹¹²RRQRRF¹¹⁷ (1 strain from England), ¹¹²GKQGRL¹¹⁷ (20 strains from Korea), and ¹¹²GRQGRL¹¹⁷ (4 strains: 1 from the USA and 3 of unknown origin).

Phylogenetic analysis based on complete HN and F gene sequences was performed for all 97 field isolates. The HN-based phylogeny grouped isolates into classes A, B, D, E, and F (Fig. 3), whereas the F-based phylogeny classified them into Class II genotypes I, II, V, VII, IX, XI, and XX (Fig. 4A). Within genotype VII, strains further clustered into sub-genotypes VII.1.1 (formerly VIIb, VIId, VIIe), VII.1.2 (formerly VIIf), and VII.2 (formerly VIIh, VIIi) (Fig. 4B), with consistently high bootstrap values supporting the robustness of the branching patterns.

Fig. 3.

Phylogenetic tree of Class II NDV based on full-length HN gene sequences. The tree was constructed using the neighbor-joining method in MEGA v11.0, with bootstrap values calculated from 1,000 replicates; values ≥70% are shown at the corresponding nodes. Six genotypes (A–F) were identified. NDV isolates characterized in this study are indicated by green squares, and reference strains by red circles.

Fig. 4.

Phylogenetic analysis based on the complete F gene sequences of NDV strains. Phylogenetic trees were constructed using the neighbor-joining method in MEGA version 11.0 with bootstrap values ≥ 70% indicated at branch nodes. (A) Overall phylogeny of 97 field NDV isolates (green squares) and seven reference strains (red circles) retrieved from GenBank, grouped into Class II genotypes I, II, V, VII, IX, XI, and XX. (B) Sub-genotypic classification of genotype VII strains, illustrating clustering into subgenotypes VII.1.1 (formerly VIIb, VIId, VIIe), VII.1.2 (formerly VIIf), and VII.2 (formerly VIIh, VIIi).

Pairwise nucleotide distance analysis revealed that, among all 97 field isolates, the minimum identity in complete HN gene sequences (76.3%) occurred between 18D045 (class A, genotype VII) and GS20/05 (class F, genotype I) (Supplementary Fig. S1). Among the 77 virulent isolates, the lowest identity in complete F gene sequences (86.6%) was between UPM/559/2017 (sub-genotype VII.2, formerly VIIi) and Kr-1129/83 (genotype V) (Supplementary Fig. S2). These findings demonstrate the substantial genetic diversity among circulating NDV lineages.

Two Malaysian isolates, UPM/1035/2018 and UPM111, were newly sequenced in this study for both HN and F genes and deposited in GenBank under accession numbers OP921717 and OP921706 for HN, and OP818847 and OP818836 for F. Both isolates had HN genes of 1,716 bp and F genes of 1,662 bp (Supplementary Table 4).

Validation of a NDV panel with genetic and geographic diversity

To evaluate diagnostic performance, we tested it against a comprehensive panel of 104 NDV isolates that included 97 field strains and seven reference strains collected over a 75-year period (1946-2021) from multiple continents.

When tested using our newly developed multiplex assay, all 80 isolates carrying any of the four multi‑basic motifs (RRQKRF, RRRKRF, KRRKRF, or RRQRRF) were positive for both HN and F targets, whereas the 24 isolates harboring monobasic motifs (GKQGRL or GRQGRL) were positive only for the HN target, consistent with non-virulent NDV pathotypes. These results showed 100% concordance with four established molecular assays, RT-HN-LAMP, RT-F-LAMP, NDV-HN-PCR, and NDV-F-PCR, validating the high diagnostic accuracy and reliability of the multiplex assay (Table 2; see Supplementary Table 4 for the complete data set).

Table 2.

Diagnostic application of the multiplex real-time RT-PCR assay for 118 ClassⅡ samples in comparison with RT-LAMP and sequencing analysis.

No. of strains		F0 cleavage site	Multiplex Real-time RT-PCR		RT-LAMP		Full-length Sequencing
			Universal	Virulent	Universal	Virulent	HN	F
Isolates (104)	20	GKQGRL	P1	N2	P	N	P	P
	4	GRQGRL	P	N	P	N	P	P
	54	RRQKRF	P	P	P	P	P	P
	23	RRRKRF	P	P	P	P	P	P
	2	KRRKRF	P	P	P	P	P	P
	1	RRQRRF	P	P	P	P	P	P
Clinical Samples (14)	2	GKQGRL	P	N	P	N	NT3	P
	12	RRRKRF	P	P	P	P	NT	P

¹; positive result,.

²; negative result,.

³; not tested.

To further assess field applicability, 14 clinical samples collected in Vietnam in 2023 were analyzed. Twelve samples were positive for both HN and F targets, while two were positive only for the HN target. Sequencing identified these as subgenotype VIIh strains and genotype II vaccine strains, respectively (Table 2; see Supplementary Table 5 for the complete data set).

Discussion

NDV continues to circulate globally, with recurrent outbreaks underscoring the urgent need for rapid and accurate diagnostic tools. Since its first report in Indonesia in 1926, ND has become endemic worldwide. Genotypes II-IV were responsible for the first panzootic (1920s-1960s), followed by genotype V in Europe, genotype VIb in pigeons, and currently genotype VII as the major driver of the ongoing fourth panzootic. Contemporary genotype VII strains exhibit more than 20% nucleotide divergence from earlier lineages, contributing to antigenic drift and reduced vaccine efficacy (Akmyrzayev et al., 2025; Lee et al., 2004; Mao et al., 2022). Sporadic outbreaks caused by sub-genotype VII.2 strains (VIIh and VIIi) have been documented across Asia, Europe, Africa, and the Americas (Song et al., 2023). Despite extensive vaccination programs and biosecurity practices, NDV outbreaks continue to cause substantial economic losses. The 2002-2003 outbreak in the United States, for example, resulted in the depopulation of more than three million birds and losses estimated at $180-360 million (Susta et al., 2014). More recently, WOAH has documented re-emergence of NDV in countries such as Sweden, Israel, Poland, Chinese Taipei, and Botswana, some of which had remained ND-free for over a decade, highlighting the role of wild bird migration, global poultry trade, and viral evolution on the spread of virulent strains (Brown et al., 2017; Kalonda et al., 2024).

In Korea, the last reported ND outbreak occurred in 2010 and was caused by virulent sub-genotype VII.1.1 (VIId) strains. Since then, no further outbreaks have been confirmed. However, as a WOAH reference laboratory for ND since 2010, APQA has maintained continuous active and passive surveillance in wild birds, live bird markets (LBMs), and poultry farms, which represent key reservoirs of APMV-1, ensuring early detection capability in case of re-introduction.

Existing molecular diagnostic tools have notable limitations. The singleplex RT-PCR assay described by Wise et al. (2004) required different primer sets for pre- and post-1960s strains. Yi et al. (2011) improved sensitivity by incorporating a red blood cell adsorption step, but the method was incompatible with high-throughput workflows. In recent years, Alexander Morris et al. (2025) developed a virulence-specific assay that still required an additional universal detection test. Collectively, these limitations reduce the efficiency of rapid, large-scale testing during outbreak situations.

Our one-step multiplex real-time RT-PCR assay addresses these challenges by enabling universal NDV detection and virulence differentiation within a single reaction. The assay exhibited high analytical sensitivity, with detection limits of 10^1.0EID₅₀/0.1 mL for the HN target and 10^2.0EID₅₀/0.1 mL for the F target, and as few as 10 and 100 copies/tube for HN and F plasmid targets, respectively. No cross-reactivity was observed with RNA from non-target avian viruses. Importantly, the assay successfully detected virulent NDV lineages spanning more than 75 years of virus evolution, from historical strains such as Herts 33/56 and Kr-KJW/49 to recent isolates like UPM/1051/2018 and UPM111. Pairwise nucleotide distance analysis revealed substantial genetic diversity, with up to 23.7% divergence in complete HN sequences across all 104 isolates and up to 13.4% divergence in complete F sequences among virulent strains. These findings highlight the assay’s robustness and its suitability for both retrospective virus detection and real-time outbreak surveillance.

Finally, analysis of 14 clinical samples from Vietnam confirmed the assay’s field applicability. Twelve were identified as virulent subgenotype VIIh strains and two as genotype II vaccine strains. Notably, because these results were obtained directly from RNA of naturally infected tissues, they demonstrate that the assay enables rapid and reliable differentiation of virulent and vaccine-related NDV in real-world outbreak settings.

In addition, to ensure diagnostic accuracy and global applicability, the APQA has consistently participated in annual international proficiency testing (PT) programs for NDV conducted by the Istituto Zooprofilattico Sperimentale delle Venezie (IZSVe), an OIE/FAO Reference Laboratory. The robust performance of the developed assay was independently validated through these international and national PT programs conducted between 2024 and 2025. These evaluation panels included diverse NDV isolates originating from multiple geographical regions, including Nigeria, Italy, Poland, Spain, Belgium, and North Macedonia. In blind testing, the assay demonstrated 100% concordance, successfully detecting and pathotyping genetically diverse lineages far beyond East and Southeast Asian genotypes. These results provide strong evidence that the assay's optimized primer and probe sets are highly effective against a broad spectrum of globally circulating NDV strains.

As global poultry production and trade continue to expand, the demand for a rapid, sensitive, and comprehensive NDV diagnostic platform is more urgent than ever. Although next-generation sequencing has advanced the understanding of NDV diversity (Kapgate et al., 2015; Satam et al., 2023), it remains impractical for routine diagnostics. The multiplex assay developed in this study offers a practical, high-throughput solution with proven accuracy and specificity, making it valuable for national surveillance programs, such as those implemented in Korea, as well as for global outbreak response and vaccine strain monitoring.

Patent

A patent application related to this study has been filed with the Korean Intellectual Property Office (KIPO) (Application No. 1020230018740).

Funding

This study was supported by grants from the Animal and Plant Quarantine Agency (Grant B-1543084-2025-28-01 and Grant B-1543084-2024-26-02) and the Ministry of Agriculture, Food and Rural Affairs of the Republic of Korea.

CRediT authorship contribution statement

Hye-Soon Song: Writing – original draft, Funding acquisition, Data curation, Conceptualization. Hyeon-Su Kim: Validation, Methodology, Investigation, Formal analysis. Ji-Ye Kim: Funding acquisition. Yong-Kuk Kwon: Writing – review & editing. Moon Her: Writing – review & editing. Hye-Ryoung Kim: Writing – review & editing.

Disclosures

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.