Comparative analysis of 3 qPCR primer–probe sets for the detection of equid alphaherpesvirus 1

Abstract

With the revision of the World Organisation for Animal Health (WOAH) Terrestrial Manual on equine rhinopneumonitis in 2024, 3 recommended qPCR primer–probe sets were added for the detection of equid alphaherpesvirus 1 (EqAHV1; formerly equine herpesvirus 1 [EHV1]; family Orthoherpesviridae, taxon species Varicellovirus equidalpha1), also known as equine abortion virus. We compared the sensitivity and specificity of the 3 qPCR primer–probe sets to determine the most reliable set. Sets gB1H and gB1P, which target the glycoprotein B (gB) gene of EqAHV1, detected all 10 copies and even lower copy numbers. In contrast, set gC1 (ISO 17025–accredited method used at the WOAH reference laboratory), which targets the glycoprotein C (gC) gene, failed to detect ≤10 copies of EqAHV1. Our results showed the lower sensitivity of gC1, which was not improved by modification of primer and probe concentrations. gB1P detected not only EqAHV1 but also equid alphaherpesvirus 4 (EqAHV4; Orthoherpesviridae, Varicellovirus equidalpha4), likely owing to an erroneous amplification of the homologous EqAHV4 gB gene, indicating that gB1P is not suitable for the detection of EqAHV1 with high specificity. We then compared gB1H with gB1D, a set recommended in the previous version of the Manual, using 120 nasal swabs collected from febrile horses. gB1H had slightly higher sensitivity than gB1D. gB1H proved to be the most reliable primer–probe set for detecting EqAHV1, with high sensitivity and specificity. Nevertheless, individual laboratories are encouraged to validate these methods under their own conditions before implementation.

Equid alphaherpesvirus 1 (EqAHV1; formerly equine herpesvirus 1 [EHV1]; family Orthoherpesviridae, taxon species Varicellovirus equidalpha1), a double-stranded DNA virus, ¹⁸ is a major cause of equine rhinopneumonitis worldwide.^4,9,20,28 The clinical manifestations of equine rhinopneumonitis include 3 forms: respiratory illness, abortion in pregnant mares, and neurologic disorders. In the respiratory form, pyrexia can disrupt training and competition.^6,19 The abortive form poses a direct threat to equine reproductive health.^4,21,27 Outbreaks of the neurologic form have been reported at equestrian events, raising significant concerns among horse industry professionals.^8,20,23 The economic impact of EqAHV1 on the equine industry, including horse racing, equestrian sports, and breeding, is substantial. Equid alphaherpesvirus 4 (EqAHV4; formerly equine herpesvirus 4 [EHV4]; Orthoherpesviridae, Varicellovirus equidalpha4), which is genetically and serologically related to EqAHV1, is another causative agent of equine rhinopneumonitis.^20,30 However, EqAHV4 typically causes only mild respiratory illness and is rarely associated with abortion or neurologic disease. ¹⁵ Given these differences in pathogenicity, accurate differentiation between EqAHV1 and EqAHV4 is crucial for proper diagnosis and disease management. The development of rapid and reliable diagnostic techniques is therefore essential for the early detection of infected horses and the effective management of EqAHV1 outbreaks. Given its high sensitivity and throughput capacity, quantitative real-time PCR (qPCR) is often used as the gold standard test for the detection of viral disease agents.^7,31 qPCR is widely used for detecting EqAHV1 infection,^2,22 with various primer–probe sets.^3,12

Equine rhinopneumonitis is a listed disease in the World Organisation for Animal Health (WOAH) Terrestrial Manual, which outlines recommended diagnostic techniques and preventive measures for animal infectious diseases. In May 2024, the chapter for equine rhinopneumonitis was revised, and certain qPCR protocols were removed, including the primer–probe set that targets the glycoprotein B (gB) gene of EqAHV1 (called gB1D).^10,33 Equid alphaherpesvirus 8 (EqAHV8; formerly equine herpesvirus 8 (EHV8); Orthoherpesviridae, Varicellovirus equidalpha8), which is genetically close to EqAHV1, has been reported to be associated with fever and abortion in horses and donkeys.^14,17,32 The sequences of forward and reverse primers and probe of the gB1D set are identical to the corresponding gB gene regions of several EqAHV8 strains, suggesting that this set lacks specificity for EqAHV1. With the revision of the Terrestrial Manual, 3 recommended qPCR primer–probe sets designed to detect EqAHV1 were added. ³³ To our knowledge, their sensitivity and specificity have not been reported. To identify the most reliable assay with high sensitivity and specificity for the detection of EqAHV1, we evaluated and compared the performance of the 3 qPCR primer–probe sets using synthesized genes, cultured virus, and clinical specimens.

Materials and methods

Target sequences and qPCR protocols

The gB1H and gB1P sets amplify the EqAHV1 gB gene encoded by ORF33 (Table 1).^16,25 The gC1 set (an ISO 17025–accredited set used at the WOAH Reference Laboratory) targets the EqAHV1 glycoprotein C (gC) gene encoded by ORF16. All primers and probes were synthesized by a commercial provider of DNA synthesis services (Nippon Gene, Tokyo, Japan). Total nucleic acids for all qPCR assays were extracted (magLEAD 12gc automated nucleic acid extraction system with the MagDEA Dx SV; Precision System Science), according to the manufacturer’s instructions. All qPCR assays were conducted using TaqPath 1-step RT-qPCR master mix (ThermoFisher) on a QuantStudio 6 Pro system (ThermoFisher). Data were analyzed in Design and Analysis 2 software (ThermoFisher).

Table 1.

Primers and probe sequences of qPCR sets evaluated for the detection of equid alphaherpesvirus 1 (EqAHV1).

Target/Primer	Primer sequences (5′–3′)	Reference
EqAHV1 gB (gB1H)
Forward	CAT-ACG-TCC-CTG-TCC-GAC-AGA-T	16
Reverse	GGT-ACT-CGG-CCT-TTG-ACG-AA
Probe	FAM-TGA-GAC-CGA-AGA-TCT-CCT-CCA-CCG-A-BHQ
EqAHV1 gB (gB1P)
Forward	TAT-ACT-CGC-TGA-GGA-TGG-AGA-CTT-T	25
Reverse	TTG-GGG-CAA-GTT-CTA-GGT-GGT-T
Probe	FAM-ACA-CCT-GCC-CAC-CGC-CTA-CCG-BHQ
EqAHV1 gC (gC1)
Forward	GCG-GGC-TCT-GAC-AAC-ACA-A	ISO 17025 accredited at WOAH Reference Laboratory
Reverse	TTG-TGG-TTT-CAT-GGG-AGT-GTG-TA
Probe	FAM-TAA-CGC-AAA-CGG-TAC-AGA-A-BHQ
EqAHV1 gB (gB1D)
Forward	CAT-GTC-AAC-GCA-CTC-CCA	10
Reverse	GGG-TCG-GGC-GTT-TCT-GT
Probe	FAM-CCC-TAC-GCT-GCT-CC-MGB-NFQ

BHQ = black hole quencher; gB = glycoprotein B; gC = glycoprotein C; FAM = 6-carboxyfluorescein; MGB = minor groove binder; NFQ = non-fluorescent quencher.

We performed qPCR assays under identical conditions to ensure a fair and reproducible comparison across all assays. Although the primer–probe sets were derived from the WOAH Terrestrial Manual, ³³ the qPCR reaction conditions, including primer and probe concentrations as well as thermocycling steps, were standardized based on the instructions provided by the reagent manufacturer. Forward and reverse primer concentrations were set at 900 nM, and the probe concentration was set at 250 nM in a 20-µL reaction mixture. The thermocycling conditions consisted of an initial hold at 25°C for 2 min, 50°C for 15 min, and 95°C for 2 min, and then 40 cycles at 95°C for 3 s and 60°C for 30 s. Viral copy numbers were determined from standard curves automatically generated by the analysis software. RT-PCR–grade water (ThermoFisher) was used as a negative control in all tests.

Comparison of amplification efficiency and sensitivity

A 400-bp DNA fragment derived from the EqAHV1 Ab4 strain (GenBank AY665713.1), covering the amplification regions of the gB1H, gB1P, and gC1 sets, was synthesized by a commercial provider of DNA synthesis services (Fasmac, Kanagawa, Japan). A 200-bp DNA fragment covering the gB1D target region was synthesized by Takara Bio (Shiga, Japan). Ten-fold serial dilutions of the synthetic genes (10⁰–10⁷ copies/reaction) were used as positive controls to generate standard curves and to assess amplification efficiency for the gB1H, gB1P, and gC1 sets. Amplification efficiencies and R² values were automatically calculated by the qPCR analysis software. In contrast, to evaluate analytical sensitivity using viral genomic DNA, we used nucleic acids extracted from the EqAHV1 10-I-224 strain (GenBank LC109652.1). The number of viral copies in the elution was quantified using the gB1D set, which was not included in the sensitivity comparison. The elution containing 10⁴ copies/reaction was then serially diluted to obtain 10³, 10², 10¹, and 10⁰ copies/reaction. Each dilution was tested with 8 technical replicates, and the experiment was performed 3 times. Ct values <40 were regarded as positive. The number of positive results obtained from 24 replicates was compared among the gB1H, gB1P, and gC1 sets. To standardize the interpretation of qPCR results based on Ct values, we set the threshold to 0.2 for all tests.

To explore optimal conditions for the gC1 set, we tested different primer and probe concentrations, as no validated concentrations of gC1 were available in the literature. We tested the sensitivity using primers at 900, 600, and 300 nM with a probe concentration of 250 nM, and then tested probes at 250, 200, and 150 nM with primer concentrations of 900 nM. Each test was performed twice with 8 technical replicates.

Comparison of specificity

To confirm that the 3 primer–probe sets did not amplify genes from other pathogens, we evaluated their specificity by testing 9 viral and 7 bacterial species known to cause fever or respiratory disease in horses. The viruses were equine adenovirus 1, equine influenza virus, equine coronavirus, equid gammaherpesvirus 2 (EqGHV2; equine herpesvirus 2; Orthoherpesviridae, Percavirus equidgamma2), EqAHV4, equid gammaherpesvirus 5 (EqGHV5; equine herpesvirus 5; Orthoherpesviridae, Percavirus equidgamma5), equine rhinitis A virus, equine rhinitis B virus, and Getah virus (Table 2). The bacterial species tested were Escherichia coli, Pseudomonas aeruginosa, Rhodococcus equi, Staphylococcus aureus, Streptococcus equi subsp. equi, Streptococcus equi subsp. zooepidemicus, and Metamycoplasma (Mycoplasma) equirhinis (Table 2). Total nucleic acids were extracted from 200 µL of each sample as described above. All tests were performed in triplicate, and a sample was regarded as positive if gene amplification (Ct <40) was observed in at least 2 of the 3 replicates.

Table 2.

Strain names and titers of 9 viruses and 7 bacteria used to evaluate the specificity of the 3 qPCR sets for the detection of equid alphaherpesvirus 1.

Pathogen	Strain	Titer
Equine adenovirus 1	05C3	1.4 × 107 TCID50/mL
Equine influenza virus	A/equine/Ibaraki/1/2007(H3N8)	7.9 × 107 EID50/mL
Equine coronavirus	NC99	1.3 × 107 TCID50/mL
EqGHV2	08C3	3.7 × 103 TCID50/mL
EqAHV4	TH20p	2.5 × 107 TCID50/mL
EqGHV5	20-I-1208	2.6 × 104 TCID50/mL
Equine rhinitis A virus	NM11	1.3 × 107 TCID50/mL
Equine rhinitis B virus	1436/71	5.9 × 108 TCID50/mL
Getah virus	14-I-605	2.3 × 106 TCID50/mL
Escherichia coli	Entero-341	>1 × 108 CFU/mL
Pseudomonas aeruginosa	NE-316	>1 × 108 CFU/mL
Rhodococcus equi	R. equi-178	>1 × 108 CFU/mL
Staphylococcus aureus	stap-603	>1 × 108 CFU/mL
Streptococcus equi subsp. equi	S. equi-14	>1 × 108 CFU/mL
Streptococcus equi subsp. zooepidemicus	strep-540	>1 × 108 CFU/mL
Metamycoplasma (Mycoplasma) equirhinis	Myco-1	>1 × 108 CFU/mL

EID₅₀ = 50% egg infective dose; EqAHV4 = equid alphaherpesvirus 4; EqGHV2 and EqGHV5 = equid gammaherpesvirus 2 and 5, respectively.

During testing, we observed that the gB1P primer–probe set also detected EqAHV4. To investigate the specificity of gB1P in detail, we tested a 460-bp synthetic DNA fragment (Fasmac) using the gB1P set. The sequence of this fragment was based on that of the EqAHV4 TH20p strain (GenBank LC063142.1) and covered the genomic region corresponding to the EqAHV1 target amplified by the gB1P set (Suppl. Table 1). Ten-fold serial dilutions of the DNA fragment (10⁰–10⁷ copies/reaction) were subjected to qPCR with the gB1P set in triplicate. The synthetic EqAHV4 gene sequence had 6 nucleotide mismatches with the forward primer of gB1P, 4 with the probe, and 3 with the reverse primer (Suppl. Table 1).

Detection of EqAHV1 in clinical specimens

Based on the above experiments, the gB1H set had the highest sensitivity and specificity. We subsequently compared gB1H and gB1D using clinical specimens for EqAHV1 detection. A total of 120 nasal swab samples were collected from febrile Thoroughbred racehorses (defined as rectal temperature ≥38.5°C) between 2020 and 2023 at the Miho and Ritto training centers of the Japan Racing Association. Total nucleic acids were extracted from 200 µL of nasal swabs as described above. Each sample was tested in triplicate. A sample was regarded as positive if gene amplification (Ct <40) was observed in at least 2 of the 3 replicates. We compared the results between gB1H and gB1D. The threshold was set to 0.2 for both methods for the same reason as in the test comparing sensitivity. Paired sera were also collected, at presentation and 2–5 wk later, from all 120 horses. These sera were tested for antibody titers using complement fixation and an ELISA specific for EqAHV1 and EqAHV4. ⁵ A ≥4-fold increase in titer was considered seroconversion, indicative of recent infection.

Sequence comparison between 2 primer–probe sets and EqAHV8 strains

To assess the potential cross-reactivity of the tested primer–probe sets with EqAHV8, we compared the primer and probe sequences of the 2 sets (gB1D and gB1H) with the corresponding genomic regions of 5 EqAHV8 strains. These strains included Wh (GenBank JQ343919.1) from China, and IR/2003/19 (NC_075566.1), IR/2010/16 (MF431613.1), IR/2010/47 (MF431612.1), and IR/2015/40 (MF431614.1) from Ireland.^14,17 The alignments were manually inspected to identify sequence identities and mismatches.

Statistical analysis

The Cohen kappa coefficient was calculated for the comparison between gB1H and gB1D using clinical samples to assess their agreement. A 2 × 2 contingency table was constructed, and the observed agreement (P_o) and expected agreement (P_e) were manually calculated. The kappa value (κ) was computed as:

$$\kappa =\frac{P_o-P_e}{1-P_e}$$

Results

Amplification efficiency and sensitivity

Amplification efficiencies and R² values represent the mean and range (minimum–maximum) for 3 independent runs. gB1H had an amplification efficiency of 99.1% (98.4–99.5%) and R² = 0.998 (0.996–0.999); gB1P of 99.6% (97.2–101%) and R² = 0.999 (0.997–1.000); and gC1 of 99.0% (89.5–109%) and R² = 0.958 (0.938–0.988; Fig. 1).

Figure 1.

Comparison of amplification efficiencies of the 3 qPCR primer–probe sets for the detection of equid alphaherpesvirus 1. Amplification efficiencies were evaluated using 10-fold serial dilutions of synthesized DNA fragments comprising 400 bp of nucleotides (10⁰–10⁷ copies/reaction) with each set. Each graph presents 3 standard curves. The amplification efficiencies and R² values were automatically calculated by the qPCR analysis software and are shown as the average (minimal–maximum) of 3 independent runs.

gB1H and gB1P detected 10¹–10³ copies/reaction of the EqAHV1 DNA in all 24 replicates (Fig. 2). gB1H detected 10⁰ copy/reaction in 14 replicates, and gB1P detected 10⁰ copy/reaction in 8 replicates. gC1 detected 10² and 10³ copies/reaction in all replicates, but did not detect 10¹ and 10⁰ copies/reaction in any of them (Fig. 2). To investigate the lower sensitivity of the gC1 set, we compared sensitivities across different combinations of primers and probe concentrations. As a result, no improvement in sensitivity was observed (Suppl. Fig. 1).

Figure 2.

Comparison of detection sensitivity of the 3 qPCR primer–probe sets for the detection of equid alphaherpesvirus 1 (EqAHV1). Ten-fold serial dilutions of EqAHV1 10-I-224 strain (10⁰–10³ copies/reaction) were tested in 24 replicates (3 repeated tests of 8 technical replicates). Each result is shown as a dot plot. Solid lines within the dots are the median Ct value, and dotted lines are the upper Ct threshold of 40 for a positive result. ND = not detected.

Specificity test

gB1H and gC1 did not detect any pathogens other than EqAHV1, whereas gB1P also detected the EqAHV4 TH20p strain and 2 additional EqAHV4 strains isolated in Japan (data not shown).

In the additional test to evaluate the specificity of gB1P, the set detected the synthetic DNA fragments in 3 of 3 replicates at ≥10⁵ copies/reaction and in 2 of 3 replicates at 10⁴ copies, but failed to detect them at ≤10³ copies (Table 3).

Table 3.

Number of positive results, Ct values, and calculated copy number of equid alphaherpesvirus 4 synthetic DNA amplified with gB1P primer–probe set.

Copy number of synthetic DNA fragment	Positive results	Ct	Copy number calculated by qPCR, genome equivalent/reaction
1.0 × 107	3 of 3	28.2 ± 0.10	505 ± 35
1.0 × 106	3 of 3	31.2 ± 0.04	60.1 ± 1.8
1.0 × 105	3 of 3	34.1 ± 0.18	7.7 ± 1.0
1.0 × 104	2 of 3	37.7 ± 0.20	0.62 ± 0.09
1.0 × 103	0 of 3	ND	ND
1.0 × 102	0 of 3	ND	ND
1.0 × 101	0 of 3	ND	ND

The sequence of the synthetic DNA fragment is based on the equid alphaherpesvirus 4 TH20p strain. Each concentration was tested in triplicate. Ct values and calculated copy numbers are x– ± SD. ND = not detected.

Detection of EqAHV1 in clinical specimens

As gB1H proved the most ideal set for detecting EqAHV1, we then compared its performance with gB1D using clinical specimens. Among the 120 clinical samples tested, 39 were positive and 70 were negative by both sets (Table 4). Another 9 samples were positive only by gB1H, and 2 were positive only by gB1D. Calculation of the Cohen kappa coefficient to assess the agreement between the 2 primer–probe sets yielded κ = 0.804. All 50 samples that tested positive by either or both gB1H and gB1P sets were from horses that seroconverted specifically to EqAHV1. Conversely, several horses that tested negative by both qPCR sets also seroconverted.

Table 4.

Comparative results of the gB1H and gB1D primer–probe sets in testing 120 nasal swabs that were collected from febrile horses.

	gB1H positive	gB1H negative
gB1D positive	39 (32.5)	2 (1.7)
gB1D negative	9 (7.5)	70 (58.3)

Numbers in parentheses are the percentage of 120 samples. The Cohen kappa coefficient between the 2 methods was κ = 0.804.

Sequence identity analysis between the 2 primer–probe sets and EqAHV8 strains

The gB1D primer–probe set was identical to the gB gene of the 4 Irish EqAHV8 strains and had a single nucleotide mismatch in the forward primer compared with the Chinese strain. In contrast, the gB1H set had one mismatch in the forward primer, 3 in the reverse primer, and 3 in the probe region across all 5 EqAHV8 strains.

Discussion

Our comparison of the sensitivity and specificity of the 3 new primer–probe sets included in the 2025 WOAH Terrestrial Manual suggests that the set targeting the gB gene of EqAHV1 is the most reliable for the detection of EqAHV1. We found that the mean amplification efficiencies of the 3 sets were comparable, but gC1 had a much wider range of amplification efficiencies than the other 2 sets. Our finding indicates that gC1 has significant variability, which could compromise its reproducibility. In contrast, both gB1H and gB1P had amplification efficiencies of nearly 100% in all experiments, indicative of consistent performance with minimal variability. In testing of 10-fold serial dilutions of EqAHV1, gB1H and gB1P detected all samples with ≥10 copies of the virus and detected some with 1 copy. In contrast, gC1 was unable to detect any samples with ≤10 copies of EqAHV1. Hence, gB1H and gB1P have higher detection sensitivities than gC1. To ensure a more realistic assessment, these sensitivity tests were conducted using viral genomic DNA rather than synthetic constructs. To further investigate the potential for enhancing the sensitivity of gC1, we varied primer and probe concentrations but were unable to further improve sensitivity. Although gC1 is the ISO 17025–accredited method used at the WOAH Reference Laboratory, to our knowledge, its optimal conditions have not hitherto been reported. Although optimal conditions might yet be found, our results indicate that gC1 lacks sufficient sensitivity for the detection of EqAHV1.

gB1H and gC1 did not detect any viruses or bacteria other than EqAHV1, but gB1P also detected EqAHV4. It is well known that EqAHV1 and EqAHV4 are genetically closely related.^13,26,30 gB1P detected synthetic EqAHV4 gB gene constructs when the reaction mixture contained ≥10⁴ copies/reaction, yet the gB1P reference study reported that this set did not detect EqAHV4.^24,25 This discrepancy may be due to the lower amount of EqAHV4 DNA present in the reaction mixture in the reference study. Although we generated the synthetic sequences based on the TH20p strain, the sequence of this region is identical to that of the NS80567 strain (GenBank AF030027.1), which is commonly used as a reference gene for EqAHV4. ³⁰ Therefore, gB1P is likely to detect some EqAHV4 strains when samples contain a substantial amount of the virus. Despite the high sensitivity of gB1P, which is comparable to that of gB1H, our results indicate that gB1P is not suitable for detecting EqAHV1 with high specificity.

Although the gB1D set was removed from the WOAH Terrestrial Manual, it remains highly sensitive for EqAHV1 detection and is widely used in various studies.^1,8,11,29 Our experience also shows that it has high sensitivity for detecting EqAHV1 in clinical samples, including nasal swabs, peripheral blood, and aborted tissues. Comparison of the performance of gB1D and gB1H using nasal swabs collected from febrile horses gave κ = 0.804, indicating a strong agreement between the 2 methods in both sensitivity and specificity. More samples tested positive exclusively with gB1H than with gB1D, indicating the higher sensitivity of gB1H. Although gB1H may have lower specificity than gB1D, gB1H did not detect other pathogens commonly associated with fever in horses and gave no false-positive results in the negative control. Moreover, all samples that tested positive by either or both sets were from horses that had EqAHV1-specific seroconversion, supporting the validity of these qPCR results. Several samples that were negative by both qPCR sets had seroconversion. Our finding likely reflects the limited window of viral shedding detectable by qPCR. We conclude that the sensitivity of the gB1H set is comparable to, or even higher than, that of gB1D in detecting EqAHV1 in clinical specimens.

Our study had several limitations. Although the gB1D primer–probe set has the potential to detect EqAHV8, we did not perform functional testing using EqAHV8 isolates because they were not available to us. Instead, we aligned the primer and probe sequences of the gB1D and gB1H sets with 5 EqAHV8 strains. Although the number of nucleotide mismatches between gB1H and EqAHV8 is greater than that between gB1D and EqAHV8, sequence comparison alone is insufficient to definitively confirm assay specificity. EqAHV8 shares similar clinical manifestations with EqAHV1, including respiratory signs and abortion in equids.^14,17 Therefore, cross-reactivity of EqAHV1 primer–probe sets with EqAHV8 may result in misdiagnosis and overestimation of EqAHV1 prevalence. Accurate discrimination is crucial for outbreak management and epidemiologic surveillance. Further studies using EqAHV8 isolates from the field will be required to validate the specificity of the methods that we evaluated, particularly for gB1H.

Another limitation of our study is that we standardized the qPCR conditions, including primers and probe concentrations as well as thermocycling conditions based on the instructions provided by the qPCR reagent manufacturer, rather than following the original protocols described for the gB1H and gB1P sets.^16,25 We made this decision to allow a fair and reproducible comparison across all methods under uniform conditions. Importantly, gB1H and gB1P had high sensitivity and no background amplification under these conditions, suggesting that our standardized qPCR conditions did not compromise their performance. We performed all tests using a single qPCR reagent kit and qPCR apparatus to maintain consistency. However, qPCR performance can vary depending on factors such as the polymerase, buffer composition, and instruments used. Therefore, our results may not be fully generalizable to other reagent systems or platforms. We recommend that individual laboratories validate assay performance under their own test conditions before implementation.