Foot-and-mouth disease (FMD) is a contagious viral disease of cloven-hoofed animals that can cause severe economic losses. The causative agent, foot-and-mouth disease virus (FMDV) is a non-enveloped single-stranded RNA virus of the species Aphthovirus vesiculae in the Picornaviridae family. The characteristic clinical manifestations of FMD, including erosions and vesiculation of the cornified epithelium within the oral cavity, on the udder and on the feet, are observed across a broad spectrum of susceptible host species, encompassing both domestic and wild ruminants as well as domestic and wild pigs. [1, 2].

There are seven antigenically distinct serotypes, which are further classified into different topotypes based on a sequence analysis of VP1, the major viral antigen [3]. Serotype O is the most prevalent and economically significant and is responsible for the majority of global outbreaks [4, 5] including the recent events in Europe. In 2025 alone, outbreaks have been reported in three European countries, including a small outbreak in Brandenburg, Germany, affecting 14 water buffalo. In contrast, several farms with thousands of cattle have been affected in Slovakia and Hungary. Notably, these countries had been officially free of FMD for 37 and over 50 years, respectively.

In FMD-free countries, the consequences of an FMD outbreak can be devastating with substantial economic losses due to trade restrictions and killing and disposal of affected and at-risk animals [6]. Inactivated vaccines are available, but there is no cross-protection between the serotypes and sometimes antigenic variability within serotypes limits protection within a serotype, especially when using vaccines with a low potency [7].

FMDV can be transmitted by various routes: infected animals can spread the virus through direct contact and indirect contact within the same holding. Indirect transmission can also occur through products of infected animals such as meat, milk or semen or by contaminated inanimate objects and people [6]. Infection by products of infected animals fed to pigs was most likely the cause of the huge FMD outbreak in the UK in 2001 but might also have been the route of introduction for the 2025 outbreak in water buffalo in Germany.

Since products of infected animals can spread FMDV, oral exposure to FMDV has been studied in the past [8–14]. In those studies, a wide variation in the dose required to achieve infection in 50 % of xposed animals was observed, primarily depending on the virus strain, but also influenced by whether the virus was applied as a liquid suspension (resulting in a lower dose needed for infection) or mixed into solid feed. The extent to which lymphoid tissue in the oropharynx is exposed to the virus has been suggested as a key factor contributing to these differences [15]. This assumption was also the basis for Stenfeldt and colleagues to introduce intra-oropharyngeal (IOP) inoculation as an experimental method that might reflect a natural route of FMDV infection in pigs, while retaining a degree of standardization and reproducibility [12, 16].

Validation experiments with the IOP method were conducted independently at the Friedrich-Loeffler-Institut (FLI) in Germany, The Pirbright Institute in the United Kingdom, and Wageningen Bioveterinary Research (WBVR) in the Netherlands. The objective of this work is to combine the data from these studies retrospectively to evaluate the dose-response relationship of different FMDV serotype O strains following exposure targeting the oropharyngeal lymphoid tissue in pigs.

All animal studies at Pirbright were conducted in accordance with the project licence approved by the Home Office. The pigs used in this study were 15 clinically healthy, female pigs (commercial Large White/Landrace × Hampshire breed), 12 weeks of age at the time of challenge, ranging from 30–40 kg. Pigs were sourced from a standard indoor finisher pig farm in the UK, thus FMDV-free and not vaccinated against FMD. Pigs were randomly allocated into three treatment groups of five pigs each.

The pigs were housed in a large-animal isolation facility at The Pirbright Institute, a SAPO4 (BSL3-ag) high-containment facility which meets the ethical standards laid out in both the UK legislation (Animals Scientific Procedures Act 1986) and the EU Directive 2010/63/EU. The rooms in which the pigs were kept were microbiologically separate from one another, with a separate air supply, such that cross contamination between rooms was not possible. The environmental conditions were maintained at 20 ± 1°C and 55 ± 5% RH. Full decontamination of clothing was applied with chemical disinfectant (1:240 dilution of FAM 30; Evans Vanodine International) upon exit from a room. Additionally, post challenge, personnel disinfected footwear, gloves, snare and waterproof suits when moving between pens to prevent cross contamination within rooms and washed with water afterwards. Each group of five pigs was housed together in one room until the day of challenge, at which point the pigs were separated from each other immediately post challenge within the same room by installation of a set of gating and transparent barriers with rubber seals along the bottom (Fig. 2). This arrangement resulted in prevention of nose-to-nose contact as well as direct contact with excretions and secretions from other pigs in the room. Air space in the room was still shared, and pigs were able to see and hear each other which reduced stress upon separation. The pigs were housed on straw bedding at all times, had free access to water and were fed a ration appropriate to their weight of complete pig grower feed split into two feeds per day.

Figure 2: Housing of experimental pigs in high-containment isolation facility at The Pirbright Institute.

Immediately post-challenge, pigs were separated from each other within the same room using transparent barriers with rubber seals along the bottom to prevent direct contact and contact with excretions or secretions from other pigs. Visual and auditory contact was maintained to reduce stress.

Experimental Design

The aim of this study was to validate the IOP method of FMDV infection using the O/UKG/2001 strain and to include the standard IDHB inoculation route as a control. FMDV O/UKG/34/2001 was administered to three groups with five pigs each. The doses were calculated by diluting the original virus stock based on the viral concentration determined by titration on BVDV-free LFBK αvβ6 cells [30]. For the IDHB route, 1 × 10⁵ TCID₅₀/0.2 ml per pig was used (n = 5). IOP inoculation was assessed at two different doses: 1 × 10⁵ TCID₅₀/2 ml per pig (n = 5) and 1 × 10⁶ TCID₅₀/2 ml per pig (n = 5).

Virus preparation

The O/UKG/34/2001 virus preparation used for inoculation of all pigs at The Pirbright Institute was generated by homogenising the epithelial component of FMD foot lesions collected at necropsy from two pigs (VO56 and 57) inoculated with an original O/UKG/34/2001 field isolate at The Pirbright Institute in 2007. An aliquot of this virus batch was titrated on LFBK-αvβ6 cells and found to have a titre of 1.6 × 10⁸ TCID₅₀/ml. This aliquot was also sequenced using Sanger sequencing in order to give assurance that there was no evidence of cell culture adaptations. The remaining aliquots of this virus were maintained at -80°C until the day of challenge. Virus stock was diluted in pH-verified (7.4) M25 buffer (a solution [in water] of 34.3 mmol/l Na₂HPO₄ and 5.7 mmol/l KH₂PO₄) to achieve the relevant concentrations of inoculum. All dilutions were made and stored on ice.

Inoculation procedure

For IOP inoculation, pigs were anesthetized by intramuscular injection of 2 mg/kg azaperone, 0.625 mg/kg zolazepam, 0.625 mg/kg tiletamine, and 1 mg/kg xylazine. Pigs were placed in dorsal recumbency on a metal V-trough and 2 ml of the inoculum was dispensed on the tonsil of the soft palate using a stainless steel 20 cm cannula and kept in situ for one minute, monitored using a digital timer. After the minute had passed, the head was tilted forward such that excess inoculum was removed and not swallowed or aspirated. For IDHB inoculation, pigs were anesthetized by intramuscular injection of 4 mg/kg azaperone, 2.5 mg/kg zolazepam and 2.5 mg/kg tiletamine, and 2 mg/kg xylazine. Four injections of 0.1 ml each were administered into the dermis of the heel bulb of the left hind foot using a 0.8 × 16 mm needle.

Monitoring and Sampling

All pigs were weighed 2 days after arrival and on the day of challenge. Following challenge, all pigs were monitored daily including rectal temperature, examination of the snout, mouth and feet as well as assessment of lameness, nasal discharge, salivation, and anorexia. Nasal and oral swabs as well as blood samples were collected on -1, 1, 2, 3, 4, 5, 6 and 7 dpi. Quantification of lesion scores was performed according to the method used at FLI.

Euthanasia Criteria and Necropsy

Pigs were euthanised at 7 dpi (end of study) or immediately if humane or scientific endpoints were reached. The scientific endpoint was defined as virus generalization evidenced by development of FMD lesions on at least one foot (other than the inoculation site, for IDHB pigs).

Humane endpoints were based on the following criteria:

The humane endpoint was also applied if the duration of individual endpoints was not reached but the pig showed three or more of the above clinical signs combined on a single day.

Pigs were euthanized under anaesthesia with an overdose of pentobarbital. Once confirmed dead by auscultation of the heart, pigs were exsanguinated and examined for evidence of FMD lesions.

RNA Extraction and Quantification of Viral RNA

Blood samples were centrifuged into serum aliquots. Nasal swabs and mouth swabs were vortexed in 1 ml of M25 buffer. RNA was extracted from 105 µl of serum, nasal swab eluate, and mouth swab eluate via robotic extraction method with the KingFisher extraction robot (Thermo Fisher Scientific) using the MagMAX CORE Nucleic Acid Purification Kit (Life Technologies). In total, 5 µl of the eluted nucleic acid was added to a RT-qPCR assay targeting the 3D region of the FMDV genome [29]. A standard curve generated by serially diluted RNA transcripts was used for virus load quantification. The exponential part of the amplification plots was read to give a Ct value which is proportional to the quantity of FMDV RNA present in the sample.

Serology

Detection of antibodies against FMDV serotype O was carried out using the FMDV PrioCHECK Type O ELISA kit (Thermo Fisher Scientific) following the manufacturer’s instructions.

All animal studies were conducted according to the licence approved the Dutch authorities (AVD401002015265). The pigs used in this study were 15 clinically healthy male pigs from a multiplier herd, 9 weeks of age at the time of challenge, born and raised in the Netherlands, which are FMD-free. The pigs were not vaccinated against FMDV. Pigs were randomized into three groups of five pigs each.

The pigs were housed in a large animal containment isolation facility at Wageningen Bioveterinary Research in Lelystad, under conditions described in the “Minimum Biorisk Management Standards for FMD Laboratories” as recently revised in 2025 by the European Commission for the Control of FMD. The stable had its own individual air ventilation unit, with HEPA-filtered inlet and outlet. Within the stable, the pigs were housed individually separated by walls (1.16 m high), which allowed vocal and visual contact (through a window), but no direct physical contact.

Experimental Design

The IOP method of FMDV infection was compared to the standard IDHB inoculation. FMDV strain O/Manisa/TUR/69 was administered to three groups with five pigs each. IOP inoculation was assessed at two different doses, 4.5 × 10² TCID₅₀/2 ml per pig (n = 5) and 4.5 × 10⁴ TCID₅₀/2 ml per pig (n = 5). For the IDHB route, 0.9 × 10⁴ TCID₅₀/0.4 ml per pig (n = 5) was used by intradermal injection of 0.1 ml at four sites in the bulb of the heel.

Virus preparation

The virus used for inoculation was a suspension of lesion material harvested from pigs infected by intradermal injection in the bulb of the heel with O/Manisa/TUR/69. Titrations of the virus were done on primary porcine kidney cells [31]. For conversion from PFU/ml to TCID₅₀/ml, the titre in PFU/ml was divided by log_e(2) = 0.69 (based on the Poisson distribution) [32] which is equal to adding 0.16 to the log₁₀ titre (PFU/ml). The virus was titrated before the start of the experiment, but also during the study. The actual dose was based on the titre of the virus used in the study.

Inoculation procedure

For IOP inoculation, pigs were anesthetized by intramuscular injection of a mixture containing 2 mg/kg azaperone, 10 mg/kg ketamine and 1 mg/kg xylazine. Pigs were placed in dorsal recumbency, and 2 ml of the inoculum was dispensed on the tonsil of the soft palate using a stainless-steel cannula. The pigs were kept in this position for at least 1 minute. For IDHB inoculation, pigs were anesthetized as described for IOP inoculation. The inoculum was injected into 4 places, 0.1 ml per site into the dermis of the heel bulb of the left hind foot using a 0.7 × 38 mm needle.

Monitoring and Sampling

All pigs were monitored daily after inoculation, including rectal temperature. The clinical infection was evaluated 3 days after inoculation under anaesthesia (the same used as during inoculation). All feet as well as nose, tongue and mouth were inspected for FMD-specific lesions, and the quantifications of the lesion scores were done identical to the method used at FLI. All lesions were recorded. Blood and oral swabs were collected on 0, 1, 2, 3, 4, 5, 6 and 7 dpi.

Euthanasia Criteria and Necropsy

Pigs were euthanised by an overdose of pentobarbital if they had FMD-specific lesions at least 3 feet other than the one used for inoculation at 3 dpi. Pigs that did not show lesions at 3 dpi were euthanized at 7 dpi (end of study).

Sample preparation and virus quantification by plaque count

Blood samples were centrifuged and serum was aliquoted. Mouth swabs were vortexed in 1.5 ml of Dulbecco’s Minimal Essential Medium supplemented with 5% foetal bovine serum and antibiotics. The swabs were kept for 15 minutes at room temperature and then centrifuged (10 minutes at 1400 × g). Virus was titrated on primary ovine kidney cells, and titres were expressed as log₁₀ PFU/ml [31].

The intra-oropharyngeal (IOP) method was used to expose the lymphoid tissue in the pharynx of pigs to FMDV to evaluate the dose-response characteristics of different virus strains. The established method of intradermal heel bulb (IDHB) inoculation was used as a control. The studies were performed with different strains and at different locations, therefore the results of the trials are presented per research institute.

Tables 1 and 2 present infection outcomes following IOP (Table 1) and IDHB (Table 2) inoculations with various FMDV strains and doses. The applied dose was either determined by titration after infection or calculated based on the titres of the virus stock. Data include results for O/FRA/2001-P1(O/BUL/2011) (derived directly from an infectious clone or from pig passages), O/TAW/97 (field isolate), O/UKG/34/2001 (from pig passage) and O/TUR/Manisa/69 (from pig passages). The number of pigs inoculated, and the number of pigs tested positive by FMDV RT-PCR (FLI and the Pirbright Institute) or virus isolation (WBVR), as well as the time of onset of infection (immediate or delayed), is reported for each experimental condition.

With the recombinant virus O/FRA/2001-P1(O/BUL/2011), the IOP method resulted in infection only at a dose of 1.3 × 10⁶ TCID₅₀/2 ml. At this dose, all six pigs ultimately became infected (Fig. 3A), probably caused by transmission from the first infected pig. Clinical signs appeared in one animal at 3 dpi, whereas the remaining five pigs developed disease at later time points: three at 5 dpi, one at 7 dpi, and one at 9 dpi (Table 1). Serum from the pig with immediate onset was positive in the FMDV RT-qPCR on 2 dpi and the animal developed the first detectable lesion at 3 dpi. The lowest Ct value in serum in the group was recorded at 4 dpi (Ct = 23.6) in this pig. Of the co-housed pigs, one tested RT-qPCR positive in serum at 4 dpi, another at 8 dpi (in a swab sample), and one more at 10 dpi. The remaining pigs were only RT-qPCR positive at necropsy (see Supplementary Material). Due to limited sampling points (serum was only collected at 0, 2, 4 and 10 dpi), an earlier onset of viremia in some pigs cannot be excluded. No FMDV infections (no clinical signs nor positive RT-qPCR results, see Supplementary Material) were seen in the groups inoculated with doses of 1.0 × 10⁴ and 1.1 × 10⁵ TCID₅₀/2 ml (Table 1). The 1.3 × 10⁶ TCID₅₀/2 ml group of infected pigs was initially treated with nonsteroidal anti-inflammatory drugs (meloxicam, 0.4 mg/kg) to alleviate clinical signs. Nevertheless, pigs were euthanized in accordance with welfare regulations at different time points: one pig at 8 dpi, two at 9 dpi, and the remaining pigs at 10 dpi at the planned end of the experiment.

IDHB inoculation of 2.9 × 10⁵ TCID₅₀/0.2 ml resulted in immediate infection of all pigs, with clinical signs observed in all pigs at 2 dpi (Table 2). Affected pigs exhibited lameness, increased body temperature, and depression. All pigs in this group were positive in the FMDV PCR at 2 dpi, with the lowest Ct value of 18.2 detected in serum (see Supplementary Material). These pigs were euthanized on 2 dpi.

As the first study did not find efficient infection after IOP inoculation, we used lesion material obtained from IOP-inoculated pigs from the first study for a subsequent experiment. Here, IOP inoculation of 1.3 × 10⁶ TCID₅₀/2 ml resulted in infection in five out of six pigs. One pig developed clinical signs as early as 2 dpi, while the remaining four showed delayed onset at 4 dpi (Table 1). The pig with immediate onset showed increased body temperature and lameness starting at 2 dpi and was PCR-positive in serum on the same day. Three pigs were RT-qPCR positive in serum at 4 dpi and one at 6 dpi. The pig that remained clinically unaffected was consistently PCR-negative in all serum samples. The lowest Ct value (Ct = 21.0) was recorded at 4 dpi in one of the pigs with delayed onset (see Supplementary Material). Sequence analysis of the viral inoculum and viruses recovered from the first passage in pigs revealed two silent mutations in the VP2- and VP3-coding regions and one nonsynonymous mutation in VP1. Specifically, in VP1, an asparagine (N) at residue 143 was replaced by lysine (K). This substitution is located immediately upstream of the conserved RGD motif within the VP1 GH loop, with only a single amino acid separating the mutated position from the RGD sequence (original: NVRGD, mutated: KVRGD). No additional mutations were detected after the second passage in pigs.

From the study using the 1st pig passage of O/FRA/2001-P1(O/BUL/2011), vesicular material was recovered and called the 2nd pig passage of O/FRA/2001-P1(O/BUL/2011); IOP inoculation with this material using 4.6 × 10², 2.6 × 10³, 1.7 × 10⁴, and 8.0 × 10⁶ TCID₅₀/2 ml, respectively, did not result in infection (Table 1). In contrast, IDHB inoculation of 7.2 × 10² TCID₅₀/2 ml led to clinical infection (Table 2) as well as positive RT-qPCR results in serum of all 6 pigs. Among the IDHB-inoculated pigs, clinical signs developed early in three pigs (two at 2 dpi, one at 3 dpi), whereas the remaining three showed delayed onset at 4 dpi (n = 1) and 5 dpi (n = 2). The first two pigs developed lesions at 2 dpi (Fig. 3B) and were PCR-positive in serum on the same day. The lowest Ct value was recorded in a pig with early onset of infection at 2 dpi (Ct = 22.9) (see Supplementary Material).

Figure 3 Lesion scores per pig over time for different inoculation groups.

Lesion scores were recorded daily for each animal following the inoculation. Each data point represents the lesion score of an individual pig at a given day post-inoculation (dpi). The solid lines indicate the mean lesion score per group per day, illustrating lesion progression within each group over time. Each IOP inoculation group is shown alongside a corresponding IDHB control group for comparison.

(A) O/FRA/2001-P1(O/BUL/2011) (infectious clone) – Lesion scores following IOP or IDHB inoculation with O/FRA/2001-P1(O/BUL/2011) derived from an infectious clone.

(B) O/FRA/2001-P1(O/BUL/2011) (pig passages) – Lesion scores following IOP or IDHB inoculation with O/FRA/2001-P1(O/BUL/2011) after one or two passages in pigs.

(C) O/TAW/97 (field isolate) – Lesion scores following IOP or IDHB inoculation with the field isolate O/TAW/97.

Symbols represent individual pigs, with the symbol size indicating the number of pigs within a group that had the same lesion score on a given day. Larger symbols denote multiple pigs with identical scores on that day. Pigs that were euthanized due to FMD-related welfare concerns are no longer included in the graph after the day of euthanasia.

The objective of this study was to combine data from several experiments retrospectively to evaluate the dose-response relationship of different FMDV serotype O strains following exposure targeting the oropharyngeal lymphoid tissue in pigs. One would expect a dose-response relationship if infection of lymphoid tissue was the point of entry in these experiments. The tonsil of the soft palate has been suggested as a possible primary site of infection in studies reporting successful infection of pigs by direct exposure of this tissue to FMDV [15]. We used the same inoculation method as described previously [16] to assess the reproducibility of this route of infection at different research institutes with different FMDV isolates. No significant dose-response relation was found in these experiments.

The FMDV isolates used all belong to serotype O and included a recombinant O Bulgaria virus (O/FRA/2001-P1(O/BUL/2011)), O/TAW/97, which is a porcinophilic strain [18], O/UKG/34/2001, and O/Manisa/TUR/69, which was previously used in successful IOP experiments at Plum Island Animal Disease Center [16]. To confirm the infectivity of our virus isolates for pigs, they were also administered by intradermal heel bulb (IDHB) injection [26, 27], the gold standard for inoculation of pigs with FMDV. All virus isolates used in this study consistently produced clinical disease following IDHB inoculation, even at low virus doses (Table 2). Using the same virus isolates, IOP inoculation did not reliably cause infection at any of the participating research institutes (Table 1). No dose could be determined that consistently infected pigs by the oral route and statistical analysis did not find a significant dose-response relationship following IOP inoculation.

To explore whether the outcome of oral application could be improved by host adaptation of the virus, we also tested preparations of O/FRA/2001-P1(O/BUL/2011) obtained from sequential pig passages. A dose of 1.3 × 10⁶ TCID₅₀/2 ml of virus, either directly from cell culture or after a single passage in pigs, resulted in infection of at least one pig. However, a higher dose of 8 × 10⁶ TCID₅₀/2 ml, derived from a second pig passage, failed to cause infection. Conversely, material from the same second passage led to early infection of 50% of the pigs after IDHB inoculation, using a dose of 7.2 × 10² TCID₅₀/2 ml. These findings suggest that pig passages do not enhance infectivity via the IOP route, despite the assumption that prior replication in the porcine host might lead to adaptation and increase the likelihood of overcoming mucosal barriers.

Each research institute used slightly different protocols for anaesthesia, animal handling, housing, and inoculation, which may have influenced the results, but none of the institutes achieved a more consistent outcome. Moreover, no significant differences in infection outcomes were observed between them.

Pigs at FLI were anesthetized using only azaperone and ketamine, which suppressed the swallowing reflex, but was occasionally associated with increased salivation. The Pirbright Institute used a combination of azaperone, tiletamine, zolazepam, and xylazine and no swallowing occurred during the one-minute contact time. At WBVR, where the pigs were anesthetized by intramuscular injection of a mixture containing azaperone, ketamine, and xylazine, some swallowing was observed before the contact time was over. In previous studies that observed infection after IOP inoculation of pigs, the protocols were slightly different. At the Plum Island Animal Disease Center in the United States, tiletamine and zolazepam had been used in combination with ketamine and xylazine [16], while the National Institute of Animal Health in Japan [10] used xylazine in combination with pentobarbital. It is tempting to assume that the anaesthesia protocol is relevant to the outcome, but we cannot make a final conclusion since the previously used anaesthesia protocols were not followed in any of the studies reported in this paper.

The virus suspension was applied directly onto the tonsil of the soft palate and left in contact for at least one minute. At FLI, the virus suspension was diluted in tissue culture medium containing a phenol red pH indicator, enabling visual confirmation of contact with the mucosal surface of the soft palate. At Pirbright, residual inoculum was removed after the contact time, whereas at FLI and WBVR, it remained in the oral cavity and was then swallowed. Feed was offered at least one to two hours post-inoculation to avoid potential interference with virus uptake.

Another aspect that may have influenced infection outcomes is the sampling done in previous studies. For example, Stenfeldt et al. [16] sampled the oropharyngeal tonsils immediately after virus exposure using large cotton swabs. This procedure might have impacted infection success in multiple ways, e.g. by mechanically distributing the virus deeper into the crypts or by causing minor mucosal abrasions. In our experiments, low virus doses were consistently able to cause infection in pigs when inoculated into the skin by the IDHB method. This suggests that mucosal or skin damage occurring before or during the inoculation can make a huge difference.

Taken together, the duration and nature of contact between the virus and the tonsillar tissue differed slightly between the research institutes in this study and the original description of the method [16]. But since there were no significant differences between the research institutes in the current analysis, it is assumed that variations in the procedures alone cannot explain the discrepancy with previous studies.

Housing animals independently is essential in inoculation studies (as it is in potency tests). Pigs at FLI were housed in group pens, permitting horizontal transmission and complicating the differentiation between primary infection by IOP and secondary pig-to-pig infection, while pigs at the Pirbright Institute and WBVR were housed individually with visual and auditory contact only. Notably, in previous studies reporting successful infection by IOP [10, 16, 29], pigs were also kept in groups. Although group housing reflects more natural conditions, it prevents the reliable exclusion of horizontal transmission and introduces methodological and statistical limitations due to the lack of independent observations.

Taken together, our findings suggest that minor procedural differences cannot account for the failure to consistently induce infection via the IOP route. Ultimately, we could not determine a dose-response relationship when using IOP inoculation, nor did we observe virus strain differences. While the method has shown promising results, our findings indicate that its robustness is more context-dependent than initially assumed.