African swine fever (ASF), caused by the African swine fever virus (ASFV), is an acute, hemorrhagic, and contagious disease of domestic pigs and wild boars.  The disease is notifiable and listed by the World Organization for Animal Health (WOAH) (Wang N et al. 2019).  The outcomes of ASF infection can be peracute, acute, subacute, and chronic, depending on the virulence of ASFVs.  According to the report of WOAH (https://www.woah.org/app/uploads/2022/12/asf-report24.pdf), from January 2020 to December 2022, ASF led to more than  2 million pig losses.  Currently, ASFV persists continuously in more than 23 countries and poses a serious threat to the global swine industry.  ASF invaded China on 3 August, 2018, caused by genotype II virulent Georgia-07-like ASFVs (Wen et al. 2019; Zhao et al. 2019; Wang et al. 2020; Wang L et al. 2022).  An experimental study showed that Georgia-07-like ASFV HLJ/18 isolated in China is highly lethal and efficiently transmissible in domestic pigs (Zhao et al. 2019; Jiang et al. 2021).  During the past four years, genotype II Georgia-07-like ASFVs dominantly spread in China.  However, the low virulent genotype II and I ASFVs have been successively reported in China in 2020 and 2021, respectively (Sun et al. 2021a, b; Shi et al. 2022).  Compared with the high virulent genotype II HLJ/18 strain, the low virulent genotype I and II ASFVs had lower virulence and high transmissibility in pigs and induced persistent and chronic infection showing irregular virus shedding at low levels (Sun et al. 2021a, b; Tsegay et al. 2022; Wang P et al. 2022).  Notably, when different genotype I and genotype II viruses infect the same pig in the field, a novel virus may be generated through viral genome recombination, which brings new problems and challenges for the prevention and control of ASF in China.  Thus, a diagnostic method that differentiates genotype I and II ASFVs with high sensitivity and stability is urgently needed and will be helpful for the prevention and control of ASF in China.  

ASFVs have been divided into at least 24 genotypes based on the C-terminus of the B646L gene with 478 nt (Bastos et al. 2003).  B646L gene is one of the most used target genes for ASF diagnosis, which is also the target gene for the WOAH recommended PCR and fluorescent quantitative PCR assays (Agüero et al. 2003; King et al. 2003).  Sanger sequencing of targeted amplification of the B646L genes is the main genotyping approach for ASFVs.  Recently, Li et al. (2022) developed the duplex real-time PCR assay based on the ASFV E296R gene, and Cao et al. (2022) established the TaqMAN-MGB probe assay based on the N-terminal sequences of the B646L gene (Cao et al. 2022; Li et al. 2022), which could distinguish genotype I and II ASFVs with detection limits of 10 copies.  However, the target genes or regions in their methods were out of ASFV genotyping regions.  

Single nucleotide polymorphism (SNP) is a single base change at a specific position in the genome of different individuals and can be used as a genotyping marker for the detection of different individual genotypes (Gut 2001).  The amplification refractory mutation system (ARMS), also named Allele-specific PCR (AS-PCR), relies on the extension of primer only when its 3´ end has a perfect complement to the template (Wang M et al. 2019).  ARMS-qPCR technology has been developed and widely used in SNP detection and genotyping (Ochsenreither et al. 2010; Shi et al. 2013; Wang M et al. 2019).  Compared with other assays for SNP detection and genotyping, ARMS-qPCR has the advantage of low-cost, simple operation, high sensitivity, and rapid and real-time detection.

Here, 126 complete or partial B646L genes of ASFVs, including 78 genotype I and 48 genotype II viruses, were obtained from the GenBank database, and their information is shown in Appendix A.  After analyzing these genes by the MegAlign Software (DNAStar), there were 4 SNPs in the C-terminus of the B646L gene, differentiating genotype I viruses from genotype II viruses (Fig. 1-A).  Two SNPs at sites 1 656 and 1 710 were used to design primers and probes for differential detection of genotype I and II ASFVs (Fig. 1-A).  As previously described (Huang et al. 1992; Liu et al. 2012), primers (I F, II F and R) and probes (probe 1 and probe 2) were designed with the targeted gene sequences using Primer 5 Software (Fig. 1-B; Appendix B).  The duplex ARMS-qPCR reaction system volume was 25 μL: 12.5 μL of 2× HyperProbe Mixture (GENFINE), 0.5 μL of I F, II F and R primers (10 μmol L^–1), 0.5 μL of probe 1 and probe 2 (10 μmol L^–1), 5 μL of template DNA, and 5 μL of ddH₂O.  The duplex ARMS-qPCR was performed by using the Bio-Rad CFX96 Touch Real-Time PCR Detection System with the following reaction conditions: 95°C for 30 s, followed by 40 cycles of 95°C for 10 s, and 60°C for 30 s.  Fluorescence signal was detected at the end of each cycle of extension step.  For the positive sample of genotype I ASFV, FAM and Cy5 fluorophores could be detected; however, for the positive sample of genotype II ASFV, only FAM fluorophore could be detected (Fig. 1-B).  

The standard curve test revealed that for the standard plasmids of genotype I ASFV, the slopes were –3.3825 for Cy5 and –3.1906 for FAM; the correlation coefficient R² was 0.999 for Cy5 and 0.998 for FAM; the amplification efficiency was 97.53% for Cy5 and 100.06% for FAM, respectively (Fig. 1-C); for the standard plasmids of genotype II ASFV, the slope was –3.2983 for FAM, the correlation coefficient R² was 0.992 for FAM, the amplification efficiency was 100.01% for FAM, whereas Cy5 fluorophore could not be detected (Fig. 1-C).  In addition, the sensitivity of the duplex ARMS-qPCR was 10 copies per reaction for both genotype I and II ASFVs (Fig. 1-D).  Thus, these results indicated that the duplex ARMS-qPCR assay has high efficiency and sensitivity.  

We then evaluated the specificity of the duplex ARMS-qPCR.  The nucleic acids of 7 other swine viruses, including PRRSV, CSFV, PRV, PCV2, PEDV, TGEV, and PoRV, were used as templates.  There were 3 amplification curves obtained for genotype I ASFV (FAM and Cy5 signals) and II ASFV (FAM signal), whereas no amplification curve was recorded for the nucleic acids of PRRSV, CSFV, PRV, PCV2, PEDV, TGEV, and PoRV, as well as genotype II ASFV (Cy5 signal) and ddH₂O (Fig. 1-E).  The results demonstrated that the duplex ARMS-qPCR has a good specificity without cross-reactivity with other swine viruses.

The results of the stable detection limit test showed that for the standard plasmids of genotype I ASFV, all 12 replicates were tested positive at the dilution of 10 copies, while 7/12 replicates were tested positive at the dilution of 5 copies (Fig. 1-F); for the standard plasmids of genotype II ASFV, all 12 replicates were tested positive at the dilution of 10 copies, while 6/12 replicates were tested positive at the dilution of 1 copy (Fig. 1-F).  Thus, the stable detection limit of the duplex ARMS-qPCR was 10 copies per reaction for both genotype I and II ASFVs (Fig. 1-F).

We further assessed the repeatability and reproducibility of the duplex ARMS-qPCR.  The assay tested the standard plasmids of 3 concentrations (10⁶, 10⁴, and 10² copies).  For the standard plasmids of genotype I ASFV, the intra- and inter-assay variation of Ct value for the duplex ARMS-qPCR ranged from 0.07 to 0.93% and 1.2 to 2.17% in FAM fluorescence channel and from 0.38 to 1.02% and 0.85 to 1.27% in Cy5 fluorescence channel, respectively (Table 1).  For the standard plasmids of genotype II ASFV, the intra- and inter-assay variation of Ct value for the duplex ARMS-qPCR ranged from 0.27 to 0.61% and 0.77 to 1.07% in FAM fluorescence channel (Table 1).  These findings suggested that the duplex ARMS-qPCR assay has satisfactory repeatability and reproducibility.

Finally, we evaluated the duplex ARMS-qPCR compared with WOAH-qPCR.  A total of 40 samples were detected using both assays, including blood, oral and rectal swabs, tissues, and cell cultures from pigs or PAMs infected by genotype I and II ASFVs.  Animal studies have evaluated the virulence and transmissibility of genotype I ASFV SD/DY-I/21 and genotype II virus HLJ/18 (Zhao et al. 2019; Sun et al. 2021a), respectively.  The results showed that 36 samples, including 18 of genotype I ASFV and 18 of genotype II ASFV were detected to be positive and differentiated by the duplex ARMS-qPCR, which were consistent with the results of the WOAH-qPCR (Appendix C).  

In summary, we developed a duplex ARMS-qPCR assay based on ASFV genotyping region of B646L gene, which can effectively differentiate genotype I and II ASFVs.  The assay had high sensitivity and specificity and exhibited good results in detecting samples, including blood, oral and rectal swabs, tissues, and cell culture.  Whether our method could be used for differentiating other genotypes of ASFVs is needed for further evalution.  However, just genotype I and II ASFVs are spreading outside Africa.  Thus, our method will provide an additional epidemiological investigation tool to implement effective ASFV control and prevention.