Oral infectivity and 50% porcine infectious dose of classical swine fever virus JPN/1/2018 strain isolated from the first case in Japan

Abstract

We determined the 50% porcine infectious dose (PID₅₀) of the classical swine fever virus (CSFV) JPN/1/2018 strain currently circulating in Japan. Twelve piglets were orally inoculated with 10^1.0, 10^3.0, or 10^5.0 50% tissue culture infectious dose (TCID₅₀) of the virus. All piglets receiving 10^5.0 TCID₅₀ developed typical clinical signs with viral RNA detection in their blood from 2 days post-inoculation (dpi). One piglet receiving 10^3.0 TCID₅₀ was confirmed to be infected at 4–5 dpi, while the remaining piglets in this group became infected later, with evidence of infection detected at 11–14 dpi. These results suggest that only the first infected piglet was directly infected by the viral inoculum, while the remaining 3 piglets were indirectly infected through horizontal transmission from the first infected piglet. No piglets inoculated with 10^1.0 TCID₅₀ became infected. Based on these infection patterns, the PID₅₀ was calculated as 10^3.5 TCID₅₀, providing essential data for infection risk assessment and establishing target reduction titers for disinfection against the JPN/1/2018 strain.

Classical swine fever virus (CSFV), a member of the genus Pestivirus in the family Flaviviridae, possesses a positive single-stranded RNA genome of approximately 12,300 nucleotides and comprises three major genotypes [4]. The virus causes severe hemorrhagic disease in domestic pigs and wild boar, resulting in substantial mortality and economic losses through both direct damage and trade restrictions [4]. Following its reemergence in Japan in September 2018 [11], CSFV has affected 98 domestic pig farms across 24 prefectures and infected over 8,400 wild boars in 40 prefectures as of April 2025 [15]. Control measures implemented have included bait vaccination of wild boar starting in March 2019 and vaccination of domestic pig with the GPE⁻ live attenuated vaccine strain developed in Japan in 1969 [13], which commenced in October 2019.

The determination of minimum infectious doses for CSFV strains is crucial for developing effective countermeasures, as it enables assessment of infection risk from infected animal excreta and establishes target reduction titers for disinfectants. Despite its importance, minimum infectious doses remain poorly characterized for most CSFV strains [1, 8]. This study determined the 50% porcine infectious dose (PID₅₀) of the CSFV JPN/1/2018 strain currently circulating in Japan to approximate its minimum infectious dose.

Infectious virus experiments were conducted in 14 m² cubicles within a high-containment facility at the National Institute of Animal Health. The facility, maintained at 25°C with 10–15 air changes per hour, meets WOAH group 4 pathogen containment requirements [14]. The Animal Care and Use Committee of the National Institute of Animal Health approved all animal procedures before the initiation of this study (authorization number: R4-I004-NIAH). We established humane endpoints for the study, and animals were euthanized if they exhibited any of the following conditions: (i) watery diarrhea or fever exceeding 41°C persisting for more than two days; (ii) emaciation to the extent that vertebral or pelvic bone prominences were visible through the skin due to subcutaneous fat reduction; or (iii) depression accompanied by complete loss of appetite or inability to stand.

Cloned porcine kidney (CPK) cells [7] were cultured in Dulbecco’s Modified Eagle Medium: Nutrient Mixture F12 (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% pestivirus-free fetal bovine serum. The CSFV JPN/1/2018 strain [9], isolated from Japan’s first case and passaged twice in CPK cells, has demonstrated moderate virulence in a previous study [5].

This study utilized twelve 60-day-old White Yorkshire × White Yorkshire crossbred piglets from a specific pathogen-free pig farm. Neither the sow nor the piglets had received GPE⁻ vaccination, and the piglets showed no antibodies against CSFV as confirmed using a CSF ELISA kit II (Nippon Gene, Tokyo, Japan) [12]. The piglets were randomly assigned to three groups (Groups 1 to 3) of 4 animals each. Groups 1, 2, and 3 were orally inoculated with 2 mL of the JPN/1/2018 strain at 10^1.0, 10^3.0, and 10^5.0 50% tissue culture infectious dose (TCID₅₀), respectively, following previously described methods [2]. Back-titration of the inoculum confirmed the administration of viruses at the intended titers. Serum, whole blood, and oral swabs were collected daily throughout the experimental period.

Viral RNA was extracted using the MagMAX CORE Nucleic Acid Purification Kit and KingFisher Flex Purification System (Thermo Fisher Scientific), then detected using the CSFV/ASFV Direct RT-qPCR Mix & Primer/Probe Set (Takara Bio, Kusatsu, Japan) [10]. Samples were considered positive for CSFV when virus gene-specific fluorescence amplification was detected, regardless of Ct values. For virus titration, samples were serially diluted and inoculated onto CPK cells, then incubated for 7 days at 37°C in 5% CO₂. After acetone fixation, viruses were detected by immunofluorescence using WH303 anti-CSFV E2 monoclonal antibody (Animal and Plant Health Agency, Surrey, UK) and Alexa Fluor 488-conjugated secondary antibody (Thermo Fisher Scientific), then visualized using an LSM700 microscope (Zeiss, Baden-Württemberg, Germany). Titers were calculated using the Spearman-Kärber method [6].

Antibody titers against the JPN/1/2018 strain were determined by virus neutralization test. Sera were treated with chloroform and heat-inactivated, then serially diluted and mixed with 100 TCID₅₀ of virus. After incubation with CPK cells for 7 days, virus neutralization was assessed by immunostaining using the same method as for virus titration, with titers expressed as the highest serum dilution inhibiting virus growth in 50% of wells [3].

PID₅₀ of the JPN/1/2018 strain was calculated using the Spearman-Kärber method by the following formula [6]:

$$\log {\mathrm{PID}}_{50}=\mathrm{A}-\mathrm{B}\times \left(\sum \frac{\mathrm{C}}{\mathrm{D}}-0.5\right)$$

A=logarithm of the highest dilution showing complete infection (100% infection)

B=logarithmic interval between dilutions

C=number of infected animals at each dilution

D=number of animals used at each dilution

∑=sum (summation over all dilutions)

Piglets in Group 1 remained clinically normal, showing neither fever (>40°C) nor leucopenia (<10,000 cells/µL) throughout the experimental period (Fig. 1A and 1B). In contrast, one piglet (#5) of 4 piglets in Group 2 developed fever (>40°C) from 7 days post-inoculation (dpi) and leucopenia (<10,000 cells/µL) from 5 dpi (Fig. 1C and 1D). Subsequently, the remaining 3 piglets in Group 2 developed fever (>40°C) from 13 to 16 dpi, and 2 of these piglets (#7 and #8) showed leucopenia (<10,000 cells/µL) from 12 to 13 dpi, while the third piglet (#6) exhibited a tendency toward leucopenia. However, these four piglets in Group 2 exhibited no clinical signs other than fever and leucopenia. Additionally, all piglets in Group 3 developed fever (>40°C) from 3 to 7 dpi and leucopenia (<10,000 cells/µL) from 4 to 7 dpi (Fig. 1E and 1F). These animals exhibited multiple clinical signs including anorexia, decreased vitality, diarrhea, tremors, ataxia, and pile up.

Fig. 1.

Body temperature and leucocyte counts in piglets inoculated with different doses of classical swine fever virus (CSFV) JPN/1/2018 strain. (A) Body temperature in Group 1 (10^1.0 50% tissue culture infectious dose (TCID₅₀)) remained with normal range (<40°C) throughout the experimental period. (B) Leucocyte counts in Group 1 similarly showed no decline, maintaining levels above 10,000 cells/µL. (C) Body temperature in Group 2 (10^3.0 TCID₅₀) showed biphasic fever patterns, with one piglet developing fever (>40°C) from 7 days post-inoculation (dpi), followed by the remaining three piglets showing elevated temperatures from 13–16 dpi. (D) Leucocyte counts in Group 2 exhibited a corresponding pattern, with progressive leucopenia in the first infected piglet from 5 dpi, and subsequent decreases in the remaining piglets from 12–13 dpi. (E) Body temperature in Group 3 (10^5.0 TCID₅₀) demonstrated rapid and consistent fever development in all piglets from 3 dpi. (F) Leucocyte counts in Group 3 showed marked and synchronous leucopenia across all piglets beginning at 4 dpi. These dose-dependent clinical responses provide critical data for determining the 50% porcine infectious dose of the JPN/1/2018 strain.

In Group 1, no viral genes were detected in serum, whole blood, and oral swab samples throughout the experimental period, and no infectious viruses were isolated from any clinical samples during this time (Fig. 2A to 2B , Supplementary Tables 1 and 2).

Fig. 2.

RT-qPCR detection of viral genes in clinical samples from Groups 2 and 3 inoculated with classical swine fever virus (CSFV) JPN/1/2018 strain. Ct values for whole blood (A, C, E) and oral swab samples (B, D, F) from Group 1 (10^1.0 50% tissue culture infectious dose (TCID₅₀), piglets #1–4, panels A and B), Group 2 (10^3.0 TCID₅₀, piglets #5–8, panels C and D), and Group 3 (10^5.0 TCID₅₀, piglets #9–12, panels E and F) are shown. In Group 2, viral RNA was initially detected in piglet #5 beginning at 4–5 days post-inoculation (dpi) in all samples, while the remaining piglets showed delayed detection from 7–10 dpi in oral swabs, and from 11–14 dpi in serum and whole blood samples. This biphasic pattern aligns with the clinical observations in Fig. 1. In contrast, Group 3 exhibited synchronous viral replication with consistent RNA detection across all piglets starting from 1 to 3 dpi in all sample types, demonstrating more rapid and uniform infection at higher viral doses. Lower Ct values indicate higher viral loads.

In Group 2, viral genes were initially detected in serum, whole blood, and oral swab samples collected from one piglet (#5) of 4 piglets from 4 to 5 dpi (Fig. 2C to 2D, Supplementary Table 1). Subsequently, viral genes were detected in oral swab samples collected from the remaining 3 piglets from 7 to 10 dpi (Fig. 2D, Supplementary Table 1); however, since viral shedding had already occurred in the oral swabs of piglet #5, and piglets in Group 2 shared food and water containers, the viral genes detected in oral swab samples from the remaining 3 piglets may have originated from piglet #5. Viral genes were later detected in serum and whole blood samples from the remaining 3 piglets from 11 to 14 dpi onward (Fig. 2C, Supplementary Table 1), confirming their infection during this period. Consistent with viral gene detection, infectious viruses were isolated from clinical samples collected from piglet #5 from 6 to 7 dpi (Supplementary Table 2). Infectious viruses were also isolated from all clinical samples, except for the oral swab sample from piglet #8, collected from the remaining 3 piglets from 11 to 16 dpi onward (Supplementary Table 2). These results suggest that only piglet #5 was directly infected by the viral inoculum, while the remaining 3 piglets were indirectly infected through horizontal transmission from piglet #5.

In Group 3, viral genes were detected in serum and whole blood from 2 dpi, and in oral swab samples from 1 dpi (Fig. 2E to 2F, Supplementary Table 1). Infectious viruses were isolated from clinical samples collected from piglets in Group 3 at 5, 4, and 6 dpi, respectively (Supplementary Table 2).

Neutralization antibodies against the JPN/1/2018 strain were detected in all piglets in Group 3 at the end of the experimental period, but not in any piglets in Groups 1 and 2 throughout the study period (data not shown). In experimental infections using CSFVs currently prevalent in Japan, neutralization antibodies have been reported to first appear in some pigs relatively late, at 14–17 dpi [3, 16]. Therefore, neutralization antibodies might have been detected in piglet #5 of Group 2, which was considered to be directly infected via oral inoculation, had the experiment continued longer. In particular, slower antibody production might be expected in piglets in Group 2 since they were inoculated with a lower viral titer of 10^3.0 TCID₅₀, which is close to the PID₅₀. This further supports the possibility that neutralization antibodies might have been developed in this piglet if the experiment had continued for a longer period.

Taken together, our results suggest that one out of four piglets in Group 2 (piglet #5) inoculated with 10^3.0 TCID₅₀ of the JPN/1/2018 strain and all four piglets in Group 3 inoculated with 10^5.0 TCID₅₀ of the same virus were directly infected by the viral inoculum. In contrast, none of the piglets in Group 1 inoculated with 10^1.0 TCID₅₀ showed clinical signs of infection. Based on these infection rates (0%, 25%, and 100% respective doses), the minimum infectious dose, namely PID₅₀ was calculated as 10^3.5 TCID₅₀ using the Spearman-Kärber method.

This study determined the PID₅₀ of the CSFV JPN/1/2018 strain currently circulating in Japan, providing crucial insights into its minimum infectious dose.

A previous study reported that the minimum infectious dose of the highly virulent Alfort strain was less than 10 TCID₅₀ per pig, although the route of inoculation was not specified [8]. Our findings indicate that oral infection with the JPN/1/2018 strain requires a dose higher than 10^1.0 TCID₅₀, suggesting different infectious characteristics compared to the Alfort strain, although direct comparison is challenging due to the unspecified inoculation route in the previous study. Additionally, the PID₅₀ of the Alfort strain could not be calculated because infection rates at other dilutions were not reported.

In contrast, another previous study reported substantially different minimum infectious doses for other CSFV strains compared to the aforementioned study [1]. The latter study reported that three out of six pigs inoculated with 10^4.2 TCID₅₀ of the moderately virulent UK2000/7.1 strain became infected, as did five out of the six pigs inoculated with 10^5.2 TCID₅₀ of the same strain. In contrast, none of the pigs inoculated with 10^1.2, 10^2.2, or 10^3.2 TCID₅₀ of the UK2000/7.1 strain developed infection. Based on these reported data, we calculated the PID₅₀ of the UK2000/7.1 strain as 10^4.4 TCID₅₀ using the Spearman-Kärber method. Additionally, the study reported that two out of six pigs inoculated with 10^5.3 TCID₅₀ of the highly virulent Brescia strain became infected, as did all six pigs inoculated with 10^6.3 TCID₅₀ of the same strain. In contrast, none of the pigs inoculated with 10^1.3, 10^2.3, 10^3.3, or 10^4.3 TCID₅₀ of the Brescia strain developed infection. Based on these reported data, we calculated the PID₅₀ of the Brescia strain as 10^5.5 TCID₅₀ using the Spearman-Kärber method.

The PID₅₀ of the JPN/1/2018 strain was lower than those of both previously mentioned strains. In this study, the JPN/1/2018 strain was inoculated directly onto the piglets’ tonsils as described previously [2]. In contrast, both the UK2000/7.1 and Brescia strains were packed into corn-covered blister pack baits and fed to the pigs. The difference in PID₅₀ values may be attributable to these distinct oral inoculation methods.

Despite these findings, it is important to acknowledge that for a more accurate determination of PID₅₀, an experiment should ideally include five groups with 10-fold dilution intervals ranging from 10^1.0 TCID₅₀ to 10^5.0 TCID₅₀. Due to facility constraints, we were limited to conducting experiments with only three groups simultaneously. Therefore, our PID₅₀ calculation based on these three groups represents a limitation of our study. Future experiments with more dilution groups would allow for more precise determination of the minimum infectious dose for this virus strain.

In conclusion, this study determined the PID₅₀ of the CSFV JPN/1/2018 strain, currently prevalent in Japan, as 10^3.5 TCID₅₀. This information provides essential data for assessing infection risks from infected animal excreta and establishing target reduction titers for disinfectants, contributing to the development of more effective control strategies against CSFV in Japan.

CONFLICTS OF INTEREST

There are no conflicts of interest to declare by any of the authors. None of the authors have any financial or personal relationships that could inappropriately influence or bias the content of the paper.