The specific pair of primers and corresponding probe were designed targeting the highly conserved region of the B646L gene to simultaneously detect different genotypes of ASFV strains. The primers and probes were designed targeting the F1055L gene, and the E183L gene to specifically detect genotype I, and genotype II ASFV strains, respectively (Table 1). The purpose of the assay is to determine whether ASFV is present in the clinical sample, and if so, it will be further determined whether the strain belongs to genotype I and/or genotype II ASFV. The specificity of primers and probes were confirmed by GenBanK BLAST analysis.¹ The location of the primers and probes were indicated in Figure 1.

To verify the primers and probe ASFV-B646L-F/R/P could detect different genotypes of ASFV, the nucleotide sequences of the targeting region of 24 genotypes of ASFV were aligned, and confirmed that the primers and probe were located in the conserved region (Figure 1). Furthermore, the nucleotide sequences of the targeted fragments of the 24 genotypes were synthesized (Supplementary Table S1), and transformed into plasmids. The recombinant plasmids were used as templates to verify that the primers and probe ASFV-B646L-F/R/P could specifically detect these 24 genotypes of ASFV.

The vaccine strains of classical swine fever virus (CSFV, C strain), porcine reproductive and respiratory syndrome virus (PRRSV, JXA1-R strain), foot-and-mouth disease virus (FMDV, O/Mya98/XJ/2010 strain), and pseudorabies virus (PRV, Bartha-K61 strain) were bought from the China Animal Husbandry Industry Co. Ltd. (Beijing, China). The vaccine strains of swine influenza virus (SIV, TJ strain), porcine epidemic diarrhea virus (PEDV, CV777 strain), and porcine circovirus type 2 (PCV2, ZJ/C strain) were bought from the Huapai Biological Group Co. Ltd. (Chengdu, China). The positive samples of genotype I, and genotype II ASFV were provided by Guangxi Center for Animal Disease Control and Prevention (CADC), China. The vaccines and samples were stored at −80°C until used.

A total of 3,159 clinical samples, including whole blood, and the tissues of liver, spleen, lung, kidney, and lymph nodes from each pig (The pooled clinical tissue homogenate of liver, spleen, lung, kidney, and lymph nodes from each pig was considered as one sample when tested by the multiplex qPCR), were collected from 3,159 pigs in different pig farms, harmless disposal sites, slaughterhouses, and farmers markets in Guangxi province, southern China from March 2019 to February 2023. The pigs from pig farms, slaughterhouses, and farmers markets were clinically healthy pigs, and the pigs from harmless disposal sites were abnormal deaths of pigs with different manifestations. Before collecting the samples, the consent of the animal owners had been obtained to use the samples for this study. The samples were stored at −80°C until used.

The pooled clinical tissue homogenates (20%, W/V) were resuspended in phosphate-buffered saline (PBS, pH 7.2), freeze-thawed three times, vortexed, centrifuged at 12,000 × g at 4°C for 5 min, and 200 μL of supernatant was used to extract the total nucleic acids.

The anticoagulated whole blood samples containing EDTA-Na₂ (15 g/L) were spun for 10 s, centrifuged at 12,000 × g for 5 min at 4°C, and 200 μL of supernatant was used to extract the total nucleic acids.

The total nucleic acids were extracted by the GeneRotex 96 Automated Nucleic Acid Extractor (Tianlong, Xi’an, China) using the Viral DNA/RNA Isolation Kit 4.0 (Tianlong, Xi’an, China) following the manufacturer’s instructions. The total nucleic acids were stored at −80°C until used.

The total nucleic acids from the positive samples of genotype I, and genotype II ASFV were amplified by PCR using the specific primers (Table 1). The products of the B646L, F1055L, and E183L genes were purified, and used to generate the standard control plasmid constructs according to Liu et al. (16). The recombinant standard plasmid constructs were confirmed by sequence analysis, and named pASFV-B646L, pASFV-F1055L, and pASFV-E183L, respectively. The ultraviolet absorbance of the standard plasmid constructs at 260 nm and 280 nm was measured, and the following formula was used to determine their concentrations: plasmid copies / μL = 6.02 × 10 23 × X ng / μL × 10 − 9 plasmid length bp × 660 .

The QuantStudio 6 qPCR system (ABI, Carlsbad, CA, United States) was used to develop the triplex qPCR. The amplification conditions of 95°C for 30 s, 40 cycles of 95°C for 5 s, and 58°C for 30 s were used to determine the annealing temperature and the concentrations of the primers and probes. The reaction system in a total volume of 20 μL included Premix Ex Taq (Probe qPCR) (TaKaRa, Beijing, China), three pairs of the primers and probes, the mixtures of the three standard plasmid constructs, and distilled water. Different annealing temperatures (56°C, 57°C, 58°C, 59°C, 60°C) were used for amplification, different concentrations of primers and probes was adjusted by the matrix method, and the optimal reaction conditions were determined.

The three standard plasmid constructs pASFV-B646L, pASFV-F1055L, and pASFV-E183L were mixed at a ratio of 1:1:1, then 10-fold serially diluted from 1.55 × 10⁸ to 1.55 × 10² copies/μL (final concentration in the reaction system: 1.55 × 10⁷ to 1.55 × 10¹ copies/μL), and used as templates to generate the standard curves.

Analytical specificity refers to the ability of an assay to measure one particular organism or substance, rather than others, in a sample. This parameter in qPCR refers to the specificity of primers for target of interest. For evaluation of the analytical specificity of the developed triplex qPCR in this study, the total nucleic acids (RNA/DNA) of PRRSV, PEDV, FMDV, CSFV, SIV, ASFV, PRV, and PCV2 were extracted from vaccine solution or positive clinical samples using the Viral DNA/RNA Isolation Kit 4.0 (Tianlong, Xi’an, China) following the manufacturer’s instructions. The RNA of PRRSV, PEDV, FMDV, CSFV, and SIV was reverse transcribed to cDNA using PrimeScript II 1st Strand cDNA Synthesis Kit (TaKaRa, Beijing, China) according to the manufacturer’s instructions, respectively. The DNA of ASFV, PRV, and PCV2 was used to analyze the specificity of the assay directly.

To analyze the specificity of the developed triplex qPCR, the DNA of ASFV, PRV, and PCV2, and the cDNA of PRRSV, PEDV, FMDV, CSFV, and SIV were used as templates, and the distilled water was used as negative control.

Analytical sensitivity (Limit of Detection, LOD) represents the smallest amount of substance in a sample that can accurately be measured by an assay. The LOD in qPCR is defined as the spike amount of target organism in dilution that can be detected in 95% of replicates. For evaluation of the analytical sensitivity of the developed triplex qPCR in this study, the three standard plasmid constructs pASFV-B646L, pASFV-F1055L, and pASFV-E183L were mixed at a ratio of 1:1:1, then 10-fold serially diluted from 1.55 × 10⁸ to 1.55 × 10⁰ copies/μL (final concentration in the reaction system: 1.55 × 10⁷ to 1.55 × 10⁻¹ copies/μL), and used as templates to determine the LOD of each plasmid. The Probit regression was used to analyze the LOD of the developed triplex qPCR.

The three standard plasmid constructs pASFV-B646L, pASFV-F1055L, and pASFV-E183L were mixed at a ratio of 1:1:1, then 10-fold serially diluted. Three concentrations of 1.55 × 10⁷, 1.55 × 10⁵, and 1.55 × 10³ copies/μL (the final reaction concentration) were used as templates to evaluate the intra-assay variation and the inter-assay variation. The intra-assay was repeated in three times, and the inter-assay was done on three different days, and the coefficients of variation (CVs) were determined.

The established triplex qPCR was used to test 3,519 clinical samples from 3,519 pigs in Guangxi province, southern China. At the same time, these samples were also tested by the duplex qPCR developed by Cao et al. (28), and the agreement of both methods were analyzed using Chi-square-test analysis.

In addition, the clinical and diagnostic sensitivity refers to a measure of a test’s ability to recognize positives in a sample, and the clinical and diagnostic specificity refers to a measure of a test’s ability to recognize negatives in a sample. The clinical and diagnostic sensitivity, and specificity are used to compare novel qPCR methods with reference methods. In this study, the clinical sensitivity and clinical specificity of the developed triplex qPCR was determined by comparing the detection results of the clinical samples using this assay with those detection results of the clinical samples using the reference method (28).

To further validate the detection results of the developed assay, all the samples were also tested by the singleplex qPCR for detection of ASFV described in the Chinese national standard: Diagnostic techniques for African swine fever (GB/T 18648-2020).²

The B646L, F1055L, and E183L genes of ASFV were amplified by PCR, and the products were used to generate the standard control plasmid constructs. The concentrations of the three standard plasmid constructs named pASFV-B646L, pASFV-F1055L, and pASFV-E183L were determined to be 2.35 × 10¹⁰, 1.55 × 10¹⁰, and 2.08 × 10¹⁰ copies/μL (original concentration), respectively. The three standard plasmid constructs were diluted to 1.55 × 10¹⁰ copies/μL, and stored at −80°C until used.

The primers and probe ASFV-B646L-F/R/P (Table 1) were used to detect the recombinant plasmid constructs with the target fragments of 24 genotypes of ASFV. The results showed that all 24 genotypes were detected by ASFV-B646L-F/R/P (Figure 2), indicating that ASFV-B646L-F/R/P could be used as universal primers and probe to specifically detect different genotypes of ASFV.

After optimizing different annealing temperatures, primer and probe concentrations, the developed triplex qPCR in a total volume of 20 μL included Premix Ex Taq (TaKaRa, Beijing, China), ASFV-B646L-F/R/P, ASFV-F1055L-F/R/P, ASFV-E183L-F/R/P, total nucleic acids, and RNase-free distilled water (Table 2). The amplification procedure was: 95°C for 30 s, followed by 40 cycles of 95°C for 5 s and 58°C for 30 s. At the end of each cycle, the fluorescence signals were recorded. The sample with a Ct value ≤36 was considered as positive sample.

The three standard plasmid constructs pASFV-B646L, pASFV-F1055L, and pASFV-E183L were 10-fold serially diluted from 1.55 × 10⁸ to 1.55 × 10² copies/μL (final concentration in the reaction system: 1.55 × 10⁷ to 1.55 × 10¹ copies/μL), and used to generate the standard curves of the developed triplex qPCR. The amplification efficiencies, slopes, and correlation coefficients (R²) were 104.9%, −3.21, and 0.999 for the B646L gene; 99.63%, −3.333, and 0.999 for the F1055L gene; 101.5%, −3.287, and 0.999 for the E183L gene (Figure 3).

The DNA/cDNA of genotypes I and II ASFV, PRRSV, PEDV, FMDV, CSFV, SIV, PRV, PCV2, and RNase-free distilled water were used as templates for amplification. The results showed that there were amplification curves of genotypes I and II ASFV, while no amplification curve was obtained from other swine viruses, indicating good specificity of the developed triplex qPCR without cross-reactivity with other viruses (Figure 4).

The three standard plasmid constructs pASFV-B646L, pASFV-F1055L, and pASFV-E183L were 10-fold serially diluted from 1.55 × 10⁸ to 1.55 × 10⁰ copies/μL (final concentration in the reaction system: 1.55 × 10⁷ to 1.55 × 10⁻¹ copies/μL), and used to evaluate the LOD of each plasmid construct. The results showed that the LOD of pASFV-B646L, pASFV-F1055L, and pASFV-E183L were 1.55 × 10¹ copies/μL (the final reaction concentration), respectively (Figure 5).

The LOD of the developed triplex qPCR was determined by testing serial dilutions of the standard plasmid constructs. Four serial dilutions of pASFV-B646L, pASFV-F1055L, and pASFV-E183L were used as templates to measure the Ct values and hit rates, and the results were shown in Table 3. The LODs of pASFV-B646L, pASFV-F1055L, and pASFV-E183L were analyzed using the PROBIT regression, and were determined to be 399.647 copies/reaction (95% confidence interval 339.947–554.984 copies/reaction), 374.409 copies/reaction (95% confidence interval 316.950–492.742 copies/reaction), and 355.083 copies/reaction (95% confidence interval 294.493–468.449 copies/reaction), respectively (Figure 6).

Three standard plasmid constructs with 1.55 × 10⁷, 1.55 × 10⁵, and 1.55 × 10³ copies/μL (the final reaction concentration) were used as templates to evaluate the repeatability. The results showed that the coefficients of variation (CVs) of the intra-assay variation and the inter-assay variation were 0.22–1.88% and 0.16–1.68%, respectively (Table 4).

A total of 3,519 clinical samples from Guangxi province in China were tested by the developed triplex qPCR. The results showed that 287 (8.16%) samples were positive for ASFV, of which 6 (0.17%), 253 (7.19%), 28 (0.80%) samples were positive for genotype I, genotype II, and both genotypes I and II, respectively (Table 5). The representative amplification curves of ASFV genotype I positive sample, genotype II positive sample, and co-infection of genotypes I and II positive sample were shown in Figure 7.

These 3,519 samples were also tested by the qPCR established by Cao et al. (28), and showed that the positivity rates of ASFV, genotype I, genotype II, and both genotypes I and II were 8.13% (286/3,519), 0.20% (7/3,519), 7.13% (251/3,519), and 0.80% (28/3,519), respectively. Compared to the detection results of the reported reference method (28), the developed triplex qPCR showed a clinical sensitivity of 100% (98.67–100% at 95% confidence interval), and a clinical specificity of 99.97% (99.83–100% at 95% confidence interval) (Table 6).

Agreement with the reported reference method (28) was also tested. The results of the developed triplex qPCR showed more than 99.94% agreement for both positive and negative samples with those of the reference method (Table 7), and there was no significant difference between the results of these two methods using Chi-square-test analysis (p > 0.05).

The 3,519 clinical samples were also tested by the qPCR described in the Chinese national standard of Diagnostic techniques for African swine fever (GB/T 18648-2020). The positivity rate of this method for ASFV was 8.16% (287/3,519), which was consistent with the results of the developed triplex qPCR in this study.