Intranasal inoculation with African swine fever virus genotype II reveals dose-dependent infectivity, pathogenesis and diagnostic challenges in pigs

Abstract

The African swine fever virus (ASFV) surveillance strategies, including the early diagnosis, are key for prevention and control of African swine fever (ASF). This work studies the effects of different doses of ASFV Georgia strain: high dose (10⁴ HAU); moderate dose (10^2.5 HAU) and low dose (10 HAU), inoculated intranasally in three groups with 20 pigs each. The qPCR was used to follow the infection dynamics in a wide sample panel (clinical, tissues, non-invasive and environmental), together with LAMP, serological tests and cytokine detection. After inoculation, the Georgia strain was lethal for all animals that resulted infected, including those in the low-dose group, where such low ASFV doses, as used in this study, have rarely been tested before. However, in the low dose group only one pig developed the ASF, while the rest remained negative despite being in contact with the infected one. Here, blood was confirmed as the gold standard sample but other matrices such as serum, spleen, tonsils, bone marrow, ear tip and tongue could be useful for the early ASFV diagnosis. We also showed the air and non-invasive samples as helpful monitoring tools for ASFV. No ASFV-specific antibodies were detected in most of the animals. Our data support the relevance of both molecular and serological diagnostic strategies for ASF control. The rapid detection together with strict biosecurity measures can efficiently minimize the ASFV dissemination.

1. Introduction

African swine fever (ASF) is a viral disease affecting both domestic and wild swine. The causative agent of ASF is the ASF virus (ASFV), a complex double-stranded DNA virus, of the Asfarviridae family (Wang et al. 2023). ASF causes an extremely negative impact worldwide from a sanitary and socioeconomic point of view, being classified as a notifiable disease by the World Organization for Animal Health (WOAH) (Wang et al. 2023). ASF was detected in the Caucasus region of the Republic of Georgia in 2007, rapidly spreading to western Russia and to other countries into Eastern Europe. The characterization of the circulating virus revealed the high-virulence level of this strain that continued spreading worldwide (Rowlands et al. 2008). In 2018, China declared an ASF outbreak, that was rapidly expanding to other countries in Southeast Asia (Dixon et al. 2020; Food and Agriculture Organization of the United Nations 2024). Years later, the disease was notified in Timor-Leste and Papua New Guinea (Mighell and Ward 2021). In 2021, the disease was detected in the Caribbean Island of Hispaniola (Dominican Republic and Haiti) after a nearly 40-year absence. Nowadays, the number of territories affected by ASF has increased in recent years, being endemic in sub-Saharan Africa, parts of West Africa and Europe, becoming the major problem for the global pork industry (Wang et al. 2023).

Research into the mechanisms of ASFV infection and transmission has expanded in recent decades, with focus on understanding its complex pathogenesis. ASFV can infect pigs through multiple routes, including oral, intramuscular, and intranasal administration, as well as via direct contact (Heuschele 1967; Olesen et al. 2017; Oh et al. 2023; Cho et al. 2024; Sánchez-Cordón et al. 2024). Previous studies have shown that lymph nodes in the oronasal tract have been key portals for ASFV infection and for the systemic spread to the tonsils and blood after intranasal administration (Heuschele 1967; Howey et al. 2013; Sánchez-Cordón et al. 2017; Walczak et al. 2021; Sánchez-Cordón et al. 2024). Therefore, intranasal route is gaining attention as represent one of the major natural routes of infection in pigs and then a potential transmission source within herds (Olesen et al. 2017; Oh et al. 2023; Cho et al. 2024; Sánchez-Cordón et al. 2024). Experimental studies have demonstrated that intranasal inoculation results in acute disease with efficient virus replication and shedding, enabling sequential transmission to within-pen, between-pen, and aerosol-contact pigs. Viral DNA has been consistently detected in nasal, oral, and rectal swabs, as well as in air samples from pens housing infected animals. Direct contact and nose-to-nose interactions have also been identified as major contributors to virus spread within herds. These findings underscore the relevance of direct contact and intranasal exposure as major natural pathways for ASFV transmission within pig herds, highlighting their importance for understanding disease dynamics and developing effective surveillance strategies (Olesen et al. 2017; Oh et al. 2023; Cho et al. 2024). The route of ASFV inoculation may influence not only the pathogenesis and the host immune response, but also viral replication kinetics and shedding dynamics. In parallel, the virus’s remarkable stability under diverse environmental conditions and its capacity for transmission via multiple routes, including direct contact, fomites, and aerosols, have been extensively documented (Karger et al. 2019; Mazur-Panasiuk et al. 2019; Li et al. 2023; Cho et al. 2024). Then route of ASFV inoculation may influence not only the pathogenesis and immune response to the virus, but also its replication and shedding. In addition, the ASFV resistance to a wide range of environmental conditions and its transmission capacity by a variety of routes have been also extensively studied (Karger et al. 2019; Mazur-Panasiuk et al. 2019; Li et al. 2023). These aspects highlight the need for infection models that simulate natural exposure routes and realistic viral doses to better reflect field conditions and improve early detection strategies.

Previous studies have shown experimental infections at different doses, showing that high doses produce an immediate immune response but also rapid disease progression, whereas low doses may allow a more sustained immune response (Sánchez-Cordón et al. 2017; Sehl et al. 2020; Sánchez-Cordón et al. 2024). There is other several factors such as the nature and virulence of the infecting strain, the route of infection and the genetic factors of the animal that underlying the ASF progression and clinical manifestation (Oh et al. 2021; Bergmann et al. 2022). These make it difficult to interpret the previous results on these aspects. Although it is known that the infection generated in pigs infected with highly virulent ASFV strains genotype II is rapid and lethal (Pietschmann et al. 2015), the early stages of the disease have not yet been extensively addressed from a diagnostic perspective.

For ASF diagnosis, serological tests are valuable for herd-level surveillance, particularly from 7 to 14 days post-infection. Virus isolation remains the most reliable confirmatory test; however, it requires high expertise, stringent biosafety conditions, and considerable time. In contrast, molecular techniques such as qPCR are highly sensitive, specific, and rapid. Blood is considered the most reliable sample for early and accurate molecular diagnosis of ASFV by qPCR, compared to other clinical and non-invasive samples (Guinat et al. 2014; Kosowska et al. 2020; Ramirez-Medina et al. 2022; Walczak et al. 2022; World Organisation for Animal Health 2024). However, it is not always possible to access blood samples, for example, in animals found dead whose blood cannot be extracted.

Therefore, in the absence of an effective global vaccine against ASF virus, prevention in disease-free countries depends on the strict implementation of epidemiological surveillance, including the fast and accurate diagnosis, biosecurity measures, stringent import policies, among others World Organization for Animal Health (WOAH) recommendations. Despite extensive research on ASFV pathogenesis, most experimental infections have employed relatively high inoculation doses and have focused on limited sample types, primarily blood or organ tissues. However, under natural conditions, pigs may be exposed to low viral doses through environmental contamination or indirect contact, scenarios in which the infection dynamics and early diagnostic performance remain poorly characterized. Moreover, studies reproducing ASFV infection using low dose inoculum while systematically evaluating viral detection across an extensive and diverse range of sample matrices, including clinical, tissue, non-invasive, and environmental specimens, are notably scarce. Such comprehensive approaches are essential to better understand early ASFV transmission events and to improve diagnostic strategies, particularly for surveillance in field settings where access to conventional samples may be limited.

In this context, the present study aims to evaluate the impact of intranasal inoculation with high, moderate, and low doses of the highly virulent ASFV Georgia 2007/1 strain on infection dynamics, clinical outcomes, and diagnostic detectability in domestic pigs, through an extensive panel of sample types and molecular and serological techniques. The qPCR and loop-mediated isothermal amplification (LAMP) molecular tools were used to follow the infection in clinical, tissues, non-invasive and environmental samples dynamics for the early and sensitive detection of ASFV. Along with this objective, the modulation of immune responses and disease progression by monitoring immunological parameters such as antibody response, anti-inflammatory and pro-inflammatory cytokine were also evaluated in the infected pigs.

2. Material and methods

2.1. Cells and virus

The ASFV Georgia 2007/1 strain (genotype II) was kindly provided by Dr. Linda Dixon (The Pirbright Institute, Ash Road, Pirbright, Surrey GU24 0NF, UK). The Georgia 2007/1 strain was produced in porcine alveolar macrophages (PAMs) obtained from bronchoalveolar lung lavage from healthy pigs. The virus titer was obtained as 50% hemagglutination activity units (HAU50)/mL, according to the Reed and Müench method (Reed and Muench 1938).

2.2. Experimental design

Three groups of healthy, 6-week-old pigs were used, with 20 animals per group (group A: 10 male and 10 females, group B: 20 males and group C: 10 males and 10 females). The animals were housed in the Biosafety Level 3 (BSL3) animal facilities at IRTA-CReSA (Barcelona, Spain) and numbered from 1 to 20 per group. Pigs generally develop full immunocompetence between 5 to 8 weeks of age (Harayama et al. 2022; Martínez-Boixaderas et al. 2022). After 7 days of acclimatization period, at 7-week-old, all animals were inoculated intranasally, using the intranasal mucosal atomization device (MAD, Castle Bolognese, Italy), with different doses of ASFV Georgia strain: group A,10⁴ HAU; group B 10^2.5 HAU and group C, 10 HAU. Animals were monitored for 23 days post-infection (dpi) by a trained veterinarian who recorded clinical signs daily in a blind manner, according to previously defined methodology (Galindo-Cardiel et al. 2013). The clinical status of the animals was scored from 0 to 5: 0, absence of clinical signs; 1, mild fever (<40.5 °C) or mild diarrhea; 2, mild diarrhea, mild apathy and fever (40.5–40.9 °C); 3, mild to moderate clinical signs, including fever; 4, moderate clinical signs and high fever (≥41 °C); 5, moderate to severe clinical signs, including high fever. For ethical reasons, the animals were euthanized when the clinical score reached 5, when exhibiting a fall of the hindquarters, when there was inability to drink or feed, when prostration occurred, or when exhibiting moderate nervous disorders. Moreover, three animals were euthanized at 3 and 4 dpi, respectively, for groups A and B and 2 animals at 3 and 8 dpi, respectively, for group C. Euthanasia was performed with 60 to 100 mg pentobarbital per kg body weight administered via the jugular vein in accordance with European Directive 2010/63/EU. After euthanasia, necropsies were carried out to confirm or discard the presence of ASFV-compatible lesions. The experiment was approved by the Ethics Committee from the Government of Catalonia under animal experimentation project number 12122, in accordance with Spanish and European regulations.

Environmental samples (box feces, wall swabs and box air) were collected daily from 3 dpi until the end of the study. Oral and rectal swabs were collected non-invasively on a daily basis from 3 dpi until the end of the study across all three experimental groups. To facilitate voluntary sampling, pigs underwent gradual positive-trained familiarization with the swab device shaft-mounted absorbent-tip applicator. For oral swabs, animals were conditioned to approach, sniff, explore, and lick the device before sample collection. Rectal swabs were collected contemporaneously with routine rectal temperature measurements, for which the pigs had been daily habituated. This approach aligns with previously described positive-reinforcement training protocols that enable voluntary oral and rectal sampling with minimal stress in pig (Layton et al. 2025). Clinical samples (blood, sera and nasal swabs) and tissues samples (bone marrow, tonsils, spleen, lungs, gastrohepatic lymph node, ear, tongue and muscle) were collected after euthanasia. In group A, the muscle was not collected and in group C the spleen of the animals was not collected. Oral, rectal, nasal and wall swabs were collected in 1 mL of phosphate-buffered saline (PBS) supplemented with penicillin 10,000 U/mL and streptomycin 10,000 μg/mL. Blood was collected in tubes with EDTA and blood for serum collection was centrifugated at 2500 rpm, 10 min. Tissue samples were homogenized in sterile distilled water (1 g of tissue in 9 mL) supplemented with 2% penicillin (10,000 U/mL) and streptomycin (10,000 U/mL) (Cultek, Madrid, Spain), and then centrifuged at 13,000 rpm for 10 min. The air samples were obtained using the Airport MD8 portable air sampler in conjunction with a disposable gelatin filter with a nominal pore size of 3 μm (Sartorius Göttingen, Germany). The sample was collected at a flow rate of 100 cubic meters of air over a period of 20 min and filter was dissolved in 1 mL PBS. The feces from the box were collected in feces containers and 1 mg was dissolved in 1 mL PBS.

2.3. ASFV detection by qPCR

An initial volume of 200 µL of environmental, non-invasive, clinical and tissues samples was used for DNA extraction, using the IndiMag Pathogen Kit (Indical, Leipzig, Germany) according to the manufacturer’s instructions. Finally, the DNA was eluted in a volume of 100 µL and was stored at −80 °C until used. The presence of ASFV DNA was analyzed by real-time PCR (qPCR) (Fernández-Pinero et al. 2013), using the modified protocol that uses the ASF-VP72P1 probe, according to the WOAH guidelines 2024 (World Organisation for Animal Health 2024). Cycle threshold (Ct) values equal to or less than 40 were considered positive and samples in which fluorescence was undetectable were considered negative. As established previously, the standard curve for the ASFV Georgia 2007/1 strain was established to correlate Ct values with corresponding viral loads and to categorize Ct results into three defined positivity ranges, high (Ct < 23), moderate (Ct = 23–29) and low (Ct > 29) viral DNA loads (Wang et al. 2022; Bohorquez et al. 2023; Coronado et al. 2025).

2.4. Thermal stability of bone marrow samples

Bone marrow samples were collected from three pigs (one representative animal from each experimental group) and subjected to a long-term incubation study to assess the stability of ASFV DNA under different temperature conditions. Initially, each sample was placed in sterile tubes and incubated at room temperature (25 °C) for a period of 49 days. Following this, the same samples were transferred to an incubator set at 37 °C for an additional 21 days, resulting in a total incubation period of 70 days.

To monitor ASFV DNA persistence over time, aliquots were taken from each sample at 7-day intervals throughout the incubation period. ASFV DNA was extracted from each aliquot using the IndiMag Pathogen Kit (Indical, Leipzig, Germany) according to the manufacturer’s instructions. DNA amplification was performed by real-time PCR (qPCR) following the protocol described previously in Section 2.3. The cycle threshold (Ct) values obtained were used to assess viral DNA load and potential degradation over time (Fernández-Pinero et al. 2013).

2.5. Detection of ASFV DNA by colorimetric LAMP in samples from group A

The blood, serum, spleen and bone marrow together with nasal, rectal and oral swabs samples collected from animals in group A at 3 and 4 dpi were used for DNA extraction (section 2.3), as well as the air and wall swabs samples. The DNA obtained was tested using the ASFV qPCR (Fernández-Pinero et al. 2013) and colorimetric loop-mediated isothermal amplification (LAMP) (Bohorquez et al. 2023). Likewise, blood, serum, spleen and bone marrow samples collected at 4 dpi in group A also underwent heating extraction for DNA release, as an alternative method for DNA extraction. After tissue maceration (section 2.3), the samples were diluted at 1:100 ratio in sterile water and then subjected to heating at 94 °C for 10 min before either qPCR or LAMP amplification (Bohorquez et al. 2023). LAMP reactions were considered positive (bright yellow), doubtful (pale pink) or negative (bright pink), according to the color of the mastermix at the end of the amplification.

2.6. Serum anti-ASFV antibodies detection

ASFV-specific antibodies in pig sera were detected using a competitive enzyme-linked immunosorbent assay (ELISA) (INgezim^® PPA Compac, Madrid, Spain). The optical density was measured at 450 nm and the absorbance values were then converted to blocking percentage values following manufacturer’s instruction. Values below 40% of blocking percentage were considered negative, between 40% and 50% were considered doubtful, and above 50% were considered positive. Samples that were positive or doubtful by ELISA were confirmed by the indirect immunoperoxidase test (IPT) (Gallardo et al. 2015).

2.7. Serum cytokines detection by Luminex assay

The cytokines IFN-α, IL-1β, IL-4, IL-6, IL-8, IL-10, IFN-γ, and TNF-α were quantified in the sera of animals with cytokine and chemokine 9-plex porcine ProcartaPlex panel 1 (Invitrogen, Bender MedSystems GmbH, Vienna, Austria) according to the manufacturer’s instructions. Results were quantified on a Luminex^® 200^™ and cytokine concentrations were expressed as pg/mL, based on their mean fluorescence intensity, calculated according to the standard curve.

3. Results

3.1. Clinical manifestations after inoculation with different ASFV doses

Mild clinical signs, including fever, diarrhea and apathy, were registered progressively from 1 to 3 dpi in animals from group A (high dose,10⁴ HAU). At 4 dpi, 16 out of 17 animals developed mild or moderate clinical signs and from 5 to 7 dpi, all the animals met the endpoint ethical criteria showing high fever (≥41 °C), moderate to severe apathy, dyspnea, weakness of hindquarters or prostration, and were euthanized for ethical reasons (Figure 1(A)). In group B (moderate dose, 10^2.5 HAU), clinical signs started at 3 dpi, and until 6 dpi animals developed mild and moderate fever, apathy and diarrhea. From 7 to 10 dpi, all animals showed clinical signs compatible with the endpoint criteria and were euthanized for ethical reasons (Figure 1(B)). In group C (low dose, 10 HAU), fever peaks (<40.5 °C), but no other clinical signs, were registered from 1 to 12 dpi in some animals (Figure 1(C)). At 7 dpi, Animal number 12 had to be euthanized due to severe dyspnea and prostration. Notably, at 5 dpi onwards, animal 9 developed fever, mild-moderate apathy, progressing with diarrhea, dyspnea, weakness of hindquarters and prostration. This animal was euthanized at 9 dpi for ethical reasons, according to the endpoint criteria. No clinical signs were registered in the rest of animals form group C during the trial. After the necropsy, similar severe lesions in most tissues, characteristics of acute ASF disease were found in all animals from group A and B, as well as, in the pig number 9 from group C. The pig 12 in group C revealed bacterial suppurative bronchopneumonia at post-mortem examination. The remaining pigs in group C necropsied at either 3, 8 or 23 dpi were unremarkable.

Figure 1.

Clinical signs monitored after ASFV infection. The individual clinical signs were recorded daily in pigs from: A) group A (inoculated with 10⁴ HAU), B) group B (inoculated with 10^2.5 HAU) and C) group C (inoculated with 10 HAU) during 23 dpi or until euthanasia. Clinical signs were scored from 0 to 5 and represented by a colour barcode according to the figure legend. *indicates that animal was euthanized due to bacterial infection.

3.2. ASFV DNA detection in clinical samples collected from pigs infected with different viral doses

ASFV DNA was detected in the blood collected from the three animals in group A that were euthanized at 3 dpi. The Ct values correlated with high (blood from pigs 10 and 20) and low DNA load (blood from pig 13) (Figure 2(A)). In serum samples, ASFV DNA was detected only in 2 out of 3 samples analyzed with moderate DNA load. While no detection was found by qPCR in the nasal swabs at 3 dpi. At 4 dpi, ASFV was detected in all clinical samples collected from animals 12, 15 and 17 (in blood and serum with high DNA load and in nasal swab with low DNA load) and continued to be detectable until the end of the study with high viral load for all samples tested (7 dpi) (Figure 2(A)). In group B, ASFV DNA was detected in the blood and serum of one animal at 3 dpi. At 4 dpi, ASFV DNA was detected with moderate load in all blood samples and in 2 out of 3 sera collected from pigs number 9, 15 and 19. At this time point, all the nasal swab samples were negative by qPCR. From 7 to 9 dpi, all animals were positive for ASFV detection in all clinical samples tested with high viral DNA load, while at the end of the study (10 dpi) one animal tested negative in blood and serum (Figure 2(B)). In group C, ASFV DNA was only detected in the clinical samples collected from pig number 9 at 9 dpi, with high DNA load (Figure 2(C)).

Figure 2.

The ASFV DNA detection in clinical samples after ASFV infection. The ASFV DNA load was determined by qPCR in the blood, sera and nasal swabs after the pig euthanasia: A) group A (inoculated with 10⁴ HAU), B) group B (inoculated with 10^2.5 HAU) and C) group C (inoculated with 10 HAU). Ct values were represented by a colour barcode according to the figure legend, where Ct values > 40 were considered as negative, > 29 as low DNA load, 23–29 as moderate DNA load and < 23 as high DNA load. *indicates that nasal swab samples were not collected at 3 dpi.

3.3. ASFV DNA detection in tissues

In samples from group A, at 3 dpi, only positive animals with high viral DNA load in blood were also positive in spleen with high DNA load (animals 10 and 20). Likewise, the Lung, bone marrow and tonsil were positives for these animals albeit with moderate ASFV DNA load. In gastrohepatic lymph node and tongue, low viral DNA load was detected. In the case of ear tip samples, only one of the blood-positive animals was positive with low ASFV DNA load (Figure 3(A)). Notably, animal with low ASFV DNA load in blood was negative for all the tissues tested at 3 dpi. From 4 dpi onwards, the detection in tissues turned out to be highly positive in all the spleen and bone marrow tissues, as well as in most lung and gastrohepatic lymph node samples. All ear tip and tongue samples were positive from 4 dpi, reaching values of moderate and high viral DNA load at 7 dpi (Figure 3(A)).

Figure 3.

The ASFV DNA detection in tissue samples after ASFV infection. The ASFV DNA load was determined by qPCR in the bone marrow, tonsil, spleen, lungs, gastrohepatic lymph node, ear tip and tongue after the euthanasia of pigs from: A) group A (inoculated with 10⁴ HAU), B) group B (inoculated with 10^2.5 HAU) and C) group C (inoculated with 10 HAU). Ct values were represented by a colour barcode according to the figure legend, where Ct values 40 were considered as negative, > 29 as low DNA load, 23–29 as moderate DNA load and < 23 as high DNA load.

For group B samples, ASFV DNA was only detected in the spleen of 2 out of 3 animals at 3 dpi with Ct values of 27 (moderate DNA load) and 34 (low DNA load), respectively (Figure 3(B)). At 4 dpi, more positive samples were detected, although not in the samples of all animals tested. ASFV DNA was detected in 2 out of 3 animals in the spleen, with high and moderate DNA load, and in the lung and gastrohepatic lymph node with low DNA load. Also, low DNA load was detected in the bone marrow of one animal. By 7 dpi, ASFV DNA detection progressed, with high viral DNA levels in bone marrow, spleen and lung samples, while moderate viral DNA loads were found in some tonsil and gastrohepatic lymph node samples. At this time post infection, ear tip and tongue were also positive with Ct values that correlated mainly with moderate ASFV DNA load (Figure 3(B)). At 8 and 9 dpi, values that correlated with a high viral DNA load were detected in most tissues analyzed, including the ear tip and the tongue. The highest viral DNA loads were again detected in the spleen, bone marrow, and tonsil samples (Figure 3(B)). Notably, at 8 dpi, time in which most animals in this group were euthanized by endpoint criteria, the highest values of viral DNA load were recorded. At 10 dpi, while one animal presented high DNA load in all the tissues, the animal number 16 that previously tested negative in blood, was also negative in spleen and showed moderate or low DNA load in the rest of tissues (Figure 3(B)). Finally, for group C, ASFV DNA was only detected in animal 9 at 9 dpi, with high DNA load for all tissues samples. The rest of the animals remaining negative during the trial (Figure 3(C)).

In the case of muscle (collected only in groups B and C), was detected at 7 dpi with low DNA load in 2 out of 3 samples in group B. Likewise, from 8 dpi onwards, this tissue was positive with high DNA load for all animals, except animal number 16 at 10 dpi. This animal also was negative for blood, serum and spleen. In group C, the muscle was only positive for animal 9 at 9 dpi with high DNA load.

3.4. ASFV DNA detection capacity in non-invasive samples

Oral and rectal swabs were collected noninvasively daily from 3 dpi until the end of the study in the three experimental groups. At 3 dpi, the ASFV DNA was first detected only in the oral swab from one animal in group A at 3 dpi (Ct = 30.8). Subsequently, at 4 dpi, 6 oral swabs were positive in 8 out of 17 animals with moderate load in two of them. From day 5 post infection onwards, the detection levels were increasingly higher and in a greater number of animals. All oral swabs collected at 7 dpi were positive and had a high viral DNA load in most of them. Rectal swabs began to be positive at 4 dpi (6 out of 17 samples), most of the positive samples had moderate ASFV DNA load. From this time point on, the number of positive samples and their respective viral DNA load increased progressively, detecting 10 out of 11 positive samples. At 7 dpi, all rectal swabs were positive whit high viral DNA load in 4 out of 7 tested samples (Figure 4(A)).

Figure 4.

The ASFV DNA detection in non-invasive samples after ASFV intranasal inoculation. The ASFV DNA load was determined by qPCR in rectal and oral swabs daily after 3 dpi and during 23 dpi or until the euthanasia in pigs from: A) group A (10⁴ HAU), B) group B (10^2.5 HAU) and C) group C (10 HAU). Ct values were represented by a colour barcode according to the figure legend, where Ct values > 40 were considered as negative, > 29 as low DNA load, 23–29 as moderate DNA load and < 23 as high DNA load.

In group B, ASFV DNA was first detected in oral and rectal swabs from four animals at 5 dpi with low DNA load. As of this time, the detection levels were increasingly higher and in a greater number of animals in both types of samples. From 7 dpi to the end of the study (10 dpi), the oral swabs were ASFV positive in all the samples collected from all the animals in group B, with moderate to high viral DNA loads. In contrast, from 7 to 10 dpi several animals were negative, or several samples were positive with low viral DNA load when rectal swab samples from group B animals were analyzed by qPCR (Figure 4(B)).

In group C, ASFV DNA was first detected at 6 dpi in the rectal swab from animal 9 and in the oral swab sample from animal 20. At 7 and 8 dpi, samples collected from animal 9 were positive with moderate ASFV DNA load. In addition, three animals with low DNA load were positives (two oral and one rectal swabs) at 7 dpi Figure 4(C)). Other four pigs were also positive with moderate to low DNA load (one rectal and three oral swabs) at 8 dpi. Notably, 13 out of 15 oral swabs collected at 9 dpi were test positive with low DNA load, except for the sample from pig 9, which turned out to have a moderate DNA load. Rectal swab detection began to disappear from 9 dpi and for oral swab disappear at 10 dpi. From day 11 dpi until the end of the study (23 dpi), all oral and rectal swab samples from group C were negative (Figure 4(C)).

3.5. ASFV detection in samples collected in the pen environment

To assess the utility of environmental matrices as early detection tools for ASFV, air samples, wall swabs, and pen feces were systematically analyzed. The ASFV DNA was detected in the air of the boxes of all three experimental groups. In group A, viral DNA was first detected at 5 dpi with high DNA load, and in group B at 6 dpi and continued detectable until the end of the study with a moderate DNA load for both groups. Whereas, in group C, ASFV DNA was only detected in the air of the box from 9 to 11 dpi with a low DNA load (Figure 5). Likewise, ASFV DNA was first detected in the walls swabs at 4 and at 6 dpi in groups A and B, respectively, and remained detectable until the end of the study with moderate DNA load in both groups. In the case of group C, ASFV DNA was not detected in the wall swabs from group C during the trial. Finally, the ASFV DNA was detected in the feces from group A at 5 dpi and from group B at 7 and 10 dpi with a low DNA load, while no viral DNA was detected in feces from group C (Figure 5).

Figure 5.

The ASFV DNA detection in environmental samples after ASFV intranasal inoculation. The ASFV DNA load was determined by qPCR in the air, feces and wall swabs collected in the pen daily after 3 dpi onwards. A) group A (10⁴ HAU), B) group B (10^2.5 HAU) and C) group C (10 HAU). Ct values were represented by a colour barcode according to the figure legend, where Ct values > 40 were considered as negative, > 29 as low DNA load, 23–29 as moderate DNA load and < 23 as high DNA load.

3.6. Thermal stability of ASFV in bone marrow of infected pigs

The ASFV DNA detection in the bone marrow from three pigs (pig 15 from group A, pig 11 from group B and pig 9 from group C) remained detectable after 70 days of incubation, first at 25 °C and then at 37 °C. During the 49 days of incubation at 25 °C, the Ct values of all three samples remained stable, detecting high ASFV DNA load. Subsequently, from day 49 to 70, the incubation at 37 °C resulted in increased Ct values for group A and C, while group B remained stable with the detection of viral DNA during the 70 days (Figure 6).

Figure 6.

Thermal stability of ASFV DNA in bone marrow samples. The bone marrow of three pigs, one of each group, was submitted to incubation at 25 °C for 42 days (white zone) and after at 37 °C for 20 additional days (blue zone). Every 7 days, ASFV DNA was detected by qPCR in the bone marrow of pigs from group A (black circle), group B (empty square) and group C (grey triangle). Ct values above 40 were considered negative.

3.7. ASFV DNA detection by colorimetric LAMP and qPCR in group a

Consistent results were found at 3 dpi for ASFV DNA detection by both assays, LAMP and qPCR. The sample panel analyzed included: blood, sera, spleen, bone marrow, nasal, oral and rectal swabs. From these samples, only in one blood (pig 13) that showed Ct value higher than 30 (low DNA load) by qPCR, the LAMP reaction was negative (Figure 7(A)). In addition, both assays were coincident at 4 dpi for all samples tested except for two oral swabs that showed doubtful results by LAMP. These LAMP-doubtful samples from pigs 15 and 17, were positive and negative by qPCR, respectively (Figure 7(B)).

Figure 7.

ASFV DNA detection in samples from group a by qPCR and LAMP, using magnetic or heating extraction. A) The ASFV DNA detection was analyzed in the blood, serum, spleen bone marrow, nasal swab and rectal swabs of three pigs (Pietschmann et al. 2015; Walczak et al. 2021; Oh et al. 2023) infected with 10⁴ HAU at 3 dpi by qPCR and LAMP, using magnetic extraction. B) The ASFV DNA detection was analyzed in the blood, serum, spleen, bone marrow, nasal swab and rectal swabs of three pigs (Sánchez-Cordón et al. 2017; Karger et al. 2019; Sehl et al. 2020) infected with 10⁴ HAU at 4 dpi by qPCR and LAMP, using magnetic extraction. C) The ASFV DNA detection was analyzed in the air and wall swabs of group A (10⁴ HAU), from 3 to 7 dpi by qPCR and LAMP, using magnetic extraction. D) The ASFV DNA detection was analyzed in the blood, serum, spleen and bone marrow of three pigs (Sánchez-Cordón et al. 2017; Karger et al. 2019; Sehl et al. 2020) infected with 10⁴ HAU at 4 dpi by qPCR and LAMP using heating extraction. The qPCR result is expressed as Ct value, with a downward facing bar. Dotted lines represent the limit of detection =40. Ct values above 40 were considered negative for qPCR. Each downward facing bar corresponds to the colour bar of LAMP results. Colour bars represent the colorimetric LAMP result for each sample. Results are shown in a qualitative scale where bright yellow was positive, pale pink was doubtful and bright pink was negative result, according to the colour legend. Empty bars correspond with samples that had been magnetic-extracted, whereas filled bars represent boiling-extracted matrices.

Similarly, both assays were coincident for ASFV DNA detection in air samples, showing positive result from 5 dpi onwards. In the case of wall swabs, coincident results were found by both assays at 3, 5 and 7 dpi. However, doubtful and negative results were obtained by LAMP at 4 and 6 dpi, respectively, despite that these samples tested positive by qPCR. Notably, the discordant results were found always in the case that lower DNA load was detected by qPCR (Figure 7(C)).

After heating extraction of blood, serum, spleen and bone marrow samples at 4 dpi, 100% of coincidence for ASFV DNA detection was found by both test, LAMP and qPCR. All sample tested were positive regardless the test used (Figure 7(D)).

3.8. ASFV-antibody response and cytokine profile in the three experimental groups

In general, absence of ASFV-specific antibody response was found during the trial in the three experimental groups. A doubtful response was only detected by ELISA in animal 9 in group C, showing 44.1% of blocking percentage. This result was further confirmed as positive in 1/40 dilution by the IPT assay (data no shown). In addition, the pro-inflammatory (IL-1β, IL-6, IL-8 and TNF-α) and anti-inflammatory (IL-4 and IL-10) cytokines profile, together with IFN-α and IFN-γ in sera, showed a distinctive profile in the three groups. In group A, an increase of IFN-α, IL-8, TNF-α and IL-10 was found as early as 3 and 4 dpi, increasing from 5 dpi onwards for all the animals. Subsequently, the IL-6 started increasing from 6 dpi, while IL-1β increased at 7 dpi for some animals (Figure 8(A)). In the case of group B, the levels of IFN-α, IL-1β, IL-6, IL-8, TNF-α and IL-10 cytokines were observed in animals from 7 days dpi until the end of the study (Figure 8(B)). Contrary, low or undetectable cytokines levels were detected in group C during the whole study, except the animal 9, at 9 dpi, who exhibited increased levels of IFN-α, IL-6, IL-10, and TNF-α (Figure 8(C)). Notably, low IFN-γ and IL-4 levels were detected during the study in the three experimental groups.

Figure 8.

Cytokine profile in the sera of ASFV inoculated pigs. The cytokine levels of IFN-α, IL-1β, IL-4, IL-6, IL-8, IL-10, IFN-γ, and TNF-α were quantified in the sera after the euthanasia of pigs from: A) group A (10⁴ HAU), B) group B (10^2.5 HAU) and C) group C (10 HAU). Cytokine levels were expressed in pg/mL (from 0 to 8000) and represented by a colour barcode according to the figure legend.

4. Discussion

The rapid spread of ASFV in the recent years has caused a major crisis for the pork industry with massive pig populations losses and dramatic economic consequences (World Organisation for Animal Health 2024). In the absence of a globally available commercial vaccine, good surveillance strategies, including the early diagnosis are key for prevention and control of the disease (Urbano and Ferreira 2022). In this study we investigate the dynamics of infection of the highly virulent ASFV Georgia strain in domestic pigs inoculated intranasally with different doses: high, moderate and low. A wide range of matrices, clinical, non-invasive, tissue and environmental samples, collected from each experimental group, were analyzed to assess the detection dynamic of ASFV DNA at different stages of the disease.

Our results showed that pigs infected by intranasal route with different doses of a highly virulent ASFV strain developed ASF acute disease. Likewise, as previously demonstrated, no sex-related differences were observed in clinical signs, pathogenesis, or ASFV detection dynamics throughout the study, corroborating findings that ASFV causes systemic disease in pigs irrespective of sex (Oh et al. 2021; Ekakoro et al. 2025). Notably, high and moderate doses of ASFV developed disease severity in all animals from both groups, as previously reported (Zhou et al. 2023). However, the low dose was not sufficient to guarantee the infection in most of the inoculated animals, since most animals lacked clinical signs, ASFV gross lesions and ASFV DNA load. This could be explained by the fact that such low doses of ASFV as those in this study have never been used. However, the infection was lethal for the unique animal (1 out of 20) that showed clinical signs and ASFV DNA load. Accordingly, this was the only animal in this group which revealed ASFV gross lesions upon necropsy. This points out that highly virulent ASFV Georgia strain led to lethality in pigs even at low doses (Pietschmann et al. 2015; Walczak et al. 2020).

The high dose of ASFV resulted in early clinical signs that worsened during time. For this group, the ASFV DNA was rapidly detected in blood and in most of the serum and tissue samples as early as 3 dpi, while detection in non-invasive samples was delayed. In pigs infected with moderate dose, the clinical signs appeared later, and so did the detection of ASFV DNA in clinical, tissues and non-invasive samples. Overall, the infection was lethal for both groups, and the blood was the best sample for early ASFV detection. This is in accordance with previous studies in which detection in blood occurred earlier than in other samples such as non-invasive samples and tissues (Guinat et al. 2014; Olesen et al. 2017; Walczak et al. 2020; 2022). However, to understand the ASF pathogenesis and diagnosis, it is important to investigate the ASFV DNA load in a wide range of matrices. In the case of diagnosis, it is not always possible to have access to blood (i.e. animals, such as wild boars, found dead in nature) and then other tissues need to be investigated as an alternative source for good sampling. According to previous studies, the spleen represents the tissue with the highest viral DNA load from the early stages of the infection to the end of the study (Cafariello et al. 2024). Likewise, the bone marrow, gastrohepatic lymph nodes and tonsils also showed high DNA load, which is consistent with ASFV targeting myeloid cells (Gómez-Villamandos et al. 2013; Sánchez-Cordón et al. 2018; Ruedas-Torres et al. 2024). On the other hand, the ear tip and tongue ranged from low-moderate DNA loads at early stages to moderate-high loads at the end of the study. Considering the relatively easy access of these tissues, which were highly positive in infected animals when moderate to severe clinical signs developed, they could constitute a good and rapid alternative as a test samples.

Moreover, one of the challenges facing molecular diagnostic in-field is the detection of the etiological agent in animals founds dead in advanced stages of autolysis. The sample used for the detection can be determinant for the diagnosis in these circumstances. We took the bone marrow of three animals to determine the ASFV DNA stability at 25 °C and 37 °C, as an example of what can be found in nature. While the tissues of animals can undergo rapid autolysis or decomposition, DNA often persists in bones much longer than in the soft tissues because of the rigid structure of bones that provide some protection against DNA degradation (Latham and Miller 2019). Our results showed that ASFV DNA was detected in the bone marrow samples after incubation for 70 days, noting that this tissue represent a good source for ASFV diagnosis.

In the case of low dose group, the only infected animal was unable to effectively spread the ASFV to the rest, however low viral DNA load was detected in non-invasive samples of some animals. Even when we cannot totally exclude ASFV shedding and transmission via oral and rectal (Kittawornrat and Zimmerman 2011; Ramirez-Medina et al. 2022), the DNA viral load detected in the pen was not sufficient for effective transmission of ASFV among pigs. The single animal that developed disease with a low inoculation dose, was the first one to yield detectable DNA viral load in rectal and oral swabs. The rest of animals showed ASFV DNA load at punctual time points during the time that this infected animal was in the pen. When this animal was euthanized, all non-invasive samples were negative in the remaining animals, suggestion viral absence from the pen. The ASFV genome has been previously detected in oral samples from the contact pigs before they developed viremia (Guinat et al. 2014) but in our case pigs didn’t developed viremia probably due to the low dose administered and circulating in the pen. This data also showed that ASFV infection is more difficult to establish via oral than intranasal (Olesen et al. 2025). In the present study, ASFV DNA was intermittently detected in non-invasive samples (oral and rectal swabs) from pigs that did not develop viremia or clinical disease, particularly in the low-dose group. These transient detections were not associated with systemic infection, as confirmed by the absence of viral DNA in blood and tissues. This pattern suggests that ASFV DNA presence in these animals may be attributable to environmental contamination, likely resulting from exposure to viral traces shed by the single infected pen-mate, rather than true infection. Our findings underscore the potential for detecting non-infectious viral DNA residues in the environment or on mucosal surfaces of uninfected animals, highlighting the importance of cautious interpretation of positive results in low-prevalence contexts and the value of complementary diagnostic approaches to confirm infection status (Olesen et al. 2017). The ASFV can be efficiently transmitted via contact and through the air (Olesen et al. 2017; Li et al. 2023). In the present study, the ASFV DNA was detected in the air of the boxes from the three experimental groups. This constitutes evidence that viral excretions could be easily spread in the environment as previously reported (Olesen et al. 2017) and could constitute an interesting sample to detect ASFV DNA circulation. Even in the case of animals infected with very low ASFV doses, where only 1 out of 20 animals resulted infected, air from the box was ASFV DNA positive, indicating the possibility of detecting viral DNA in the environment during the time the infected animals were kept. Shortly after the euthanasia of the infected animal at 9 dpi, air sample tested negative showing virus clearance in the pen. It is highly interesting the positive detection of ASFV DNA in the air from box and in non-invasive samples from pigs which resulted negative for clinical and tissue samples, despite being in contact with an infected animal for at least 9 days. In addition, although ASFV DNA detection in feces and pen walls was positive (especially in wall swabs) in the groups infected with high and moderate doses, the ASFV DNA was not detected in the group infected with low dose. Therefore, non-invasive and environmental air samples may result crucial as surveillance tools for rapid detection, when large population of pigs needs to be screened or even for tracking of infection course and clearance.

Similarly to the selection of the type of sample, the molecular test used for the detection of ASFV is essential for the successful diagnosis of the disease. Currently, PCR is the most sensitive and recommended technique to detect ASFV DNA with high specificity and sensitivity from a very early stage of infection (World Organisation for Animal Health 2024), while gold standard sample for an early and reliable detection of ASFV is the blood (Walczak et al. 2022; World Organisation for Animal Health 2024) as was corroborated by our results. The selection of the test also depends on the disease situation and laboratory diagnostic capacity (World Organisation for Animal Health 2024). Recently, the LAMP test has been noted as a promising point-of-care technology for the diagnosis of ASFV in field application by the WOAH (Inui and Williams 2022). One of the main inconveniences of LAMP, as well as qPCR technique, is the relatively high cost of equipment. Recently, our lab developed a colorimetric ASFV-LAMP assay that can be performed with minimal equipment, using a water bath for DNA amplification and heating the sample for DNA extraction (Bohorquez et al. 2023). In this study, we compared the detection by qPCR and LAMP using conventional magnetic-based extraction, proving that both assays have high sensitivity and specificity for ASFV DNA detection in most of the samples analyzed, especially with high and moderate DNA load. Notably, only one clinical sample at 3 dpi, that showed a low DNA load by qPCR, showed a negative result by LAMP. However, at 4 dpi all samples showed consistent results across both LAMP and qPCR assays. We also compare ASFV detection by qPCR and LAMP using heating extraction, in an attempt to reduce the equipment needed for diagnosis in field. Both assays, qPCR and LAMP, efficiently detected ASFV DNA in blood, serum, spleen and bone marrow at 4 dpi after heating DNA extraction, proving that this procedure could constitute a good and rapid step in the method for ASFV detection in field applications.

As was previously explained, the highly virulent ASFV strain used was lethal for the 100% of animals, using high or moderate doses, and before the appearance of ASFV-specific antibodies in group A and B. Then, the association between inoculation doses (high, moderate or low) and the severity level, using highly virulent ASFV strains, such as Georgia, may be inaccurate (Pietschmann et al. 2015; Sánchez-Vizcaíno et al. 2015). No significant differences in pathogenicity were observed between pigs inoculated intranasally with high or moderate doses of ASFV, in line with previous findings (Zhou et al. 2023). Notably, within the low-dose group (10 HAU), only a single animal (pig 9) developed infection, yet its clinical course and disease severity were indistinguishable from those observed in animals exposed to higher doses. These findings underscore that while the inoculation dose critically determine the likelihood of infection establishment, particularly at low thresholds, it does not influence the progression or severity of disease once infection is initiated. This supports the notion that highly virulent ASFV strains, such as Georgia 2007/1, consistently induce acute and lethal disease regardless of the infectious dose, whenever infection occurs. In this regard, this animal (pig 9) was the only able to activate antibody immunity to ASFV during the trial, despite developing severe clinical signs, indicating the absence of the protecting role of antibodies. Interestingly, the rest of animals in this group (low dose group) reached the end of the study (23 dpi) without antibodies against ASFV, confirming absence of infection.

The ASF severity could also be reflected in the cytokine pattern registered in the three experimental groups. In agreement with previous studies (Cabezón et al. 2017; Franzoni et al. 2023; Zuo et al. 2024), the present data showed a high increased of pro- (TNF-α, IFN-α, IL-1β, IL-6 and IL-8) and anti- (IL-10) inflammatory cytokine levels in the ASFV infected animals, regardless the dose. Whereas animals that tested negative by qPCR showed low or undetectable levels of this cytokine panel, confirming the absence of infection (Zakaryan et al. 2015; Wang et al. 2020). Similarly to our data, the IFN-α levels have been extensively related to the cytokine storm following ASFV infection, especially in animals about to died (Fishbourne et al. 2013; Lacasta et al. 2015; Cabezón et al. 2017). The TNF-α, IL-8 and IL-10 levels increased after 7 days, but were found earlier coinciding with the appearance of clinical signs. The increase of these cytokines at the terminal phase of infection has been previously reported (Wang et al. 2020). Interestingly, the anti-inflammatory cytokine IL-10 was detected in all infected animals, whereas uninfected animals showed low or undetected levels, as previously reported (Reis et al. 2016; Cabezón et al. 2017; Sánchez-Cordón et al. 2020; Franzoni et al. 2023). The increased IL-10 has been associated as part of cytokine storm during acute ASF, as an attempt to fight against the inflammatory state, and has been linked to a lethal outcome (Wang et al. 2020). Although IL-10 can modulate IFN-γ production, low or undetectable levels of IFN-γ, in both infected and non-infected pigs, suggested that highly virulent ASFV may block IFN-γ responses (Wang et al. 2020). Likewise, low levels of the anti-inflammatory cytokine IL-4 are consistent with previous studies that found little or no change in serum levels after immunization or infection with different ASFV strains (Sánchez-Cordón et al. 2017; Franzoni et al. 2023).

Taken together, these data showed that high and moderate intranasal doses of highly virulent ASFV pandemic Georgia strain resulted in 100% mortality. On the contrary, a very low dose of intranasal infection does not guarantee the infection in most of the inoculated animals, since no clinical sign or DNA load was detected. Notably, the infection was lethal for the unique animal infected with low dose, that showed clinical signs and ASFV DNA load during the trial. This points out that Georgia strain led to lethality in pigs even at low doses. The high number of animals and the large panel of samples analyzed in the present study has allowed us to conclude that intranasal doses of highly virulent ASFV may influence viral infectivity and early detection but not disease severity in domestic pigs. In this regard, the correlation of infectious doses using highly virulent strains with the disease severity may be subjective. Our data corroborates that blood is the gold standard sample for ASFV diagnosis. Likewise, our results also show the usefulness for the ASFV early diagnosis of samples such as serum, spleen, tonsils, the highly stable bone marrow and the easy-access ear tip or tongue. In addition, ASFV detection in air and non-invasive samples can help monitoring virus detection in an installation, even when there is very little virus circulating. The serology would not be a sufficient method for ASF surveillance when highly virulent strains are circulating, confirming once again the need to implement both molecular and serological diagnostic strategies in ASFV surveillance. Finally, the LAMP test offers an alternative method for fast, reliable and economic ASFV DNA detection with potential for field applications. Our data strongly supports that early and fast diagnostic together with the implementation of strict biosecurity measures can efficiently minimize the ASFV spread.

Funding Statement

This research was funded by the National Animal Health Laboratory Network from the United States Department of Agriculture under grant AP21VSD&B000C011 and by the Spanish Ministry of Science and Innovation grant PID2021-125599OB-100. AMA was supported by the ICA grant 001–2022. LC was supported by the Juan de la Cierva Program (2022), the Spanish Ministry of Science and Innovation. PM was supported by FPI PRE2022-101808, the Spanish Ministry of Science and Innovation. SH was supported by PTA2022-022356-I, the Spanish Ministry of Science and Innovation.

Author contributions

LG designed experiments; AMA and SPM drafted the original paper; AMA performed the most experiments; AMA, SPM and LG analyzed the data; CR, SH, PM, LC extracted the DNA samples and performed the immune response assays; AC carried out the necropsies; JIN, SPM, LG, LC and AC helped revise the manuscript; LG supervised the study and provided the resources. All the authors have approved the final manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.