Selected ASFV Georgia 2007/1 open-reading-frames (Gene bank Accession FR682468) were used to design and generate multicistronic expression cassettes (Table 1). The pp220 polypeptide was split into two due to its large size. The polypeptide sequences were used to generate codon-optimized synthetic genes (GenScript, NJ, USA) that were cloned into pcDNA3.1+ vector (Invitrogen, K8300001, CA, USA) with N-terminal HA and C-terminal FLAG tags, respectively. Protein expression by the recombinant plasmid constructs was evaluated by immunocytometric analyses as previously described (21–23, 37). The outcome was used to select the best expressers that were used as templates to PCR amplify the genes, which were then used to assemble cognate recombinant replication-incompetent adenoviruses using the Invitrogen ViraPower Adenoviral Expression System (K493000, CA, USA) as previously described (21–23, 37). A recombinant replication-incompetent adenovirus expressing luciferase (Ad5-Luciferase) was similarly generated. The recombinant adenoviruses were scaled up, and viral titers (IFU/mL) were determined by immunocytometric analyses as previously described (21, 22).

Flow cytometry was used to evaluate protein expression by each recombinant adenovirus construct. Briefly, human embryonic kidney (HEK) 293A cells were infected with Ad5-ASFV constructs and harvested after 24 h using 2 mM EDTA to lift cells from the bottom of the culture vessel. Cells were then distributed at 1 × 10⁶ cells per 5-mL polystyrene snap cap tube for each construct, followed by washing 2X using the Cyto-Fast™ Fix/Perm Buffer Set (BioLegend, 426803, CA, USA). Briefly, following the manufacturer's instructions, 1 mL wash buffer was added to each tube, gently vortexed, and centrifuged at 900 rpm for 5 min, after which the supernatant was discarded. Duplicate tubes were probed with a 1:200 dilution of ASFV-specific convalescent swine serum (26–28, 38) for 20 min at 4°C (kept in the dark). After washing as above, the cells were then probed with goat anti-porcine IgG FITC secondary antibody (Jackson ImmunoResearch, 114-095-003, PA, USA) diluted 1:250 in wash buffer for 20 min in the dark at 4°C. The cells were washed and fixed as per the manufacturer's instructions and resuspended in 2% goat sera (prepared in 1X PBS) before data acquisition using the BD LSRFortessa™ flow cytometer and BD FACSDiva™ followed by analyses using FlowJo software (BD Biosciences, OR, USA) (Figure 1). Negative mock-infected cells (media alone) were included and probed with ASFV convalescent serum and secondary antibody as controls for infection and gating.

Immunogenicity of the Ad5-ASFV immunogens in pigs was determined by tracking pp62-specific IgG responses by ELISA as this antigen is highly immunogenic and sufficient amounts of mammalian cell-expressed recombinant protein can readily be generated (23). The ELISA was conducted using a pp62 concentration of 1 μg/mL (100 μL/well) to coat microplates and a 1:100 dilution of serum samples as previously described (21–23, 37). In brief, plates were first blocked with 5% non-fat dry milk (diluted in PBST: PBS + 0.1% Tween 20) followed by the addition of diluted serum (in triplicate) for 1 h at 37°C. After six repeated washes (using PBST), a 1:5000 dilution of anti-porcine IgG-POD (Jackson ImmunoResearch, 114-035-003, PA, USA) secondary antibody was added to each well and the plates incubated at 37°C for 1 h. The plates were washed 6× with PBST followed by 3× with PBS, and the POD colorimetric reaction was developed by adding Sure Blue tetramethylbenzidine (TMB) substrate (53-00-02, KPL, MA, USA). A 1N HCl solution was added to stop color development after 10 min, and the optical density (OD) was measured at 450 nm using a BioTek Epoch spectrophotometer (VT, USA). An irrelevant antigen, TMSP7, was used as a background control to establish the baseline for each sample which was subtracted from the pp62 OD value. Additionally, a previously validated ASFV convalescent serum was included in each plate as a positive control alongside a validated naïve swine serum.

Recognition of wild-type ASFV by the antibodies induced by the Ad5-ASFV immunogens was determined by indirect fluorescence antibody (IFA) using ASFV (Georgia 2007/1)-infected and mock-infected naïve swine peripheral blood mononuclear cells (PBMCs) as previously described with the following modifications (21, 22). Briefly, PBMCs were thawed from liquid nitrogen, washed using cold complete RPMI (cRPMI) plated at a density of 1 × 10⁶ cells/well in a 96-well plate and incubated overnight at 37°C. The cells were infected using a MOI of 0.1 of ASFV (Georgia 2007/1) prepared in cRPMI and incubated at 37°C for 1 h after which infection media were discarded and replaced with fresh cRPMI. Non-infected cells were included to serve as negative controls. After 48 h, the cells were washed twice using PBS, fixed using ice-cold methanol (100%), and air-dried before blocking.

To perform the IFA, the plates were washed 2× using PBS and incubated with blocking buffer (2% bovine serum albumin (BSA) in PBS) for 1 h at 37°C. After blocking, the cells were probed with a 1:20 dilution of serum (from 1 week after the final boost) prepared in the blocking buffer for 1 h at 37°C. The ASFV-specific convalescent serum diluted at 1:500 was used as a positive control, and a 1:20 dilution of normal swine serum (Vector Laboratories, S-4000-20, CA, USA) was used as a negative control. Following three rinses with PBS, the wells were then incubated with 1:200 goat anti-swine IgG FITC (Jackson ImmunoResearch, 114-095-003, PA, USA) alongside additional wells probed with FITC only as controls for the secondary antibody for 45 min at 37°C. Two more washes were performed using PBS followed by the addition of 100 μL PBS to each well before microscopic examination. The cells were visualized, and images were acquired using an EVOS fluorescent imaging system (Thermo Fisher, MA, USA).

Two weeks after the final boost, the pigs were moved to the ABSL-3Ag biocontainment facility within the BRI on the KSU campus, acclimatized, and then challenged by exposure to contact spreaders infected with ASFV (Georgia 2007/1) (Figure 2). The two contact spreaders in each group were inoculated intramuscularly (IM) with a dose of 10² TCID₅₀/mL as determined by our previous study (23). The challenge of the comingled vaccinees in each group (n = 5) occurred via the natural transmission where infection occurs through direct contact with the infected spreaders or virus shedding in the pens which mimic the natural course of infection in field settings (23, 36).

Prior to initiation of challenge in the ABSL-3Ag biocontainment, baseline nasal swabs, whole blood, and blood for serum were collected. Following challenge, blood, nasal swabs, weights, and rectal temperatures were collected on days 0, 5, 7, 11, and 14 with clinical signs monitored twice daily using a scoring rubric (Supplementary Table 2). The surviving pig had additional collections on days 18, 21, 25, 28, 32, 35, and 37. Pigs that developed severe ASF were humanely euthanized followed by necropsy. The following tissues were collected: cranial mediastinal, gastrohepatic, mandibular, mesenteric, and renal lymph nodes; kidney, liver, lung, spleen, and tonsil.

Formalin-fixed tissues were processed for histological analysis using standard procedures (39). Histological lesions were evaluated and scored for all animals utilizing the standardized parameters established previously by Galindo-Cardiel et al. with adaptations as per Sunwoo et al. (20, 39). Systematic histological assessment was performed blindly on the spleen, tonsil, mandibular, cranial mediastinal, mesenteric, gastrohepatic, renal lymph nodes, lung, liver, and kidney. The scoring matrix was as follows: absent (0) or minimal (1), mild (2), moderate (3), or severe (4). Lesion categories scored included the following: tissue necrosis (such as infarction) and cellular necrosis (such as lymphocytolysis); fibrin thrombi, fibrin deposition in tissue or fibrinoid degeneration of vessels, congestion and/or hemorrhage, and inflammation such as infiltrates of macrophages, eosinophils, or neutrophils. Overall total organ and lymphoid organ score ranges are represented as minimal (0–30), mild (31–60), moderate (61–90), or severe (above 90). The total score ranges for viscera (lung, liver, and kidney tissues) are indicated as minimal (0–10), mild (11–20), moderate (21–30), or severe (above 30).

The detection and quantification of the ASFV genome in blood, swabs, and tissue samples from all pigs were performed using DNA purification followed by viral load assessment using a validated ASFV p72-specific quantitative real-time PCR (qPCR) assay. Briefly, ASFV DNA was extracted and purified from samples using the automated magnetic bead KingFisher Flex equipment (ThermoFisher, 5400610, MA, USA) that utilizes the Total DNA Extraction Kit (GeneReach, Taiwan) as previously described (20). Quantitative real-time PCR for the detection of the ASFV gene encoding the p72 antigen (primers: Integrated DNA Technology, IA, USA, and probe: ThermoFisher, MA, USA) was performed in duplicate wells using PerfeCTa^® FastMix^®II (Quanta Biosciences, MA, USA) and following reagent and cycle parameters as previously described on the Stratagene Mx3005p Real-Time PCR Detection System (Agilent, CA, USA) (20, 40).

For the quantification of ASFV copy number (CN), serial dilutions (10×) of the positive control (ASFV p72 plasmid) were used to generate an eight-point standard curve (10⁹ to 10¹ copies) using 16 qPCR well replicates performed as two scientific replicates. The ASFV p72 CN/reaction was mathematically determined using the mean cycle threshold value (Ct) and the slope and intercept of the DNA standard curve. A cycle threshold (Ct) cutoff of 38 for both PCR replicates was used as the positive cutoff. Each extraction and qPCR run included a standardized ASFV sample for the extraction and PCR-positive control as well as a negative control sample (molecular grade water) as an extraction negative and PCR no template control.

GraphPad Prism, version 9.5.0, was used to analyze the data as follows: The percent survival significance was determined by the Mantel–Cox test; antibody responses were assessed using the Ordinary one-way ANOVA followed by Tukey's multiple comparisons test, and virus titers and histological scores were analyzed using the two-way ANOVA followed by Tukey's multiple comparisons test. Each comparison evaluated the treatment groups against the mock-immunized negative control (Ad5-Luciferase) group as well as each immunization group against each other (ASFV-No adjuvant vs. ASFV-ISA-201 and ASFV-BioMize; ASFV-ISA-201 vs. ASFV-BioMize). For viremia and histological evaluation, all groups including the contact spreaders were compared for statistical significance. FlowJo, version 10.8.1, was used to gate and calculate the protein expression in the cells infected with the recombinant Ad5 virus constructs.

Kansas State University Institutional Animal Care and Use Committee (IACUC) (Protocol # 3871 and #4411) and Institutional Biosafety Committee (IBC) (registration #1481) follows the regulations, policies, and guidelines put forth by the Animal Welfare Act, United States Department of Agriculture (USDA) Animal Care Resource Guide, and the Public Health Service (PHS) Policy on Humane Care and Use of Laboratory Animals. All protocols outlined in this document were followed including the use of clinical scoring for daily monitoring and assessment of animal health with weight and temperature included in this evaluation.

Recombinant pcDNA3 plasmid constructs encoding 42 ASFV expression cassettes, with HA and FLAG tags at the 5′ and 3′ ends, respectively, were generated (Table 1). The pp220 polyprotein was split into two constructs due to its large size, whereas one construct was generated for each one of the next two largest antigens, NP1450L and G1340L (Table 1). The pcDNA3 constructs expressed the encoded antigens as judged by immunocytometric analyses of transfected HEK293A cells probed with tag-specific mAbs or ASFV convalescent serum [expression data obtained as before (21, 22, 37)] with additional pp220 polyprotein expression previously demonstrated by Western Blot (37). The ASFV antigen expression cassettes from the pcDNA3 constructs were used to generate recombinant replication-incompetent adenoviruses (Table 1), and protein expression was confirmed by flow cytometric analyses of infected HEK293A cells probed with the ASFV convalescent serum (Figure 1). Protein expression by the recombinant adenoviruses was heterogeneous with low, medium, and high expression levels noted (Figure 1, Supplementary Table 1). Notable low expressers include Ad5-05 and Ad5-13, medium expressers include Ad5-04 and Ad5-29, whereas Ad5-09 and Ad5-19 were high expressers (Supplementary Table 1).

Homologous prime-boost immunization of piglets with a cocktail of the recombinant adenoviruses expressing multiple ASFV antigens, designated Ad5-ASFV, alone or formulated in adjuvant (Figure 2, Table 2), was well tolerated as there was no adverse effect observed. Notably, the vaccinees, the negative controls, and the comingled non-immunized piglets that were included to serve as contact spreaders at the challenge stage were healthy, had normal weight gain, and maintained normal body temperatures throughout the immunization phase (Figure 3).

Immunization of piglets with the Ad5-ASFV cocktail, but not the negative control Ad5-Luciferase, primed ASFV-specific humoral immune responses in all the vaccinees as judged by pp62-specific IgG responses, which were used to evaluate immunogenicity (Figures 4A, B). Compared with the negative control Ad5-Luciferase Montanide ISA-201™ immunogen, the priming dose of the Ad5-ASFV cocktail alone or formulated with the Montanide ISA-201™ adjuvant elicited significantly (p < 0.0001) higher pp62-specific IgG responses, but the IgG responses primed by the Ad5-ASFV cocktail formulated in BioMize^® adjuvant was not significant (Figure 4B). The pp62-specific IgG responses primed by the Ad5-ASFV cocktail alone or the Ad5-ASFV cocktail formulated with the Montanide ISA-201™ adjuvant were significantly (p = 0.002 and p = 0.005, respectively) higher than the pp62-specific IgG responses induced by the Ad5-ASFV cocktail formulated in BioMize^® adjuvant (Figure 4B). The IgG responses primed by all the three formulations were significantly (p < 0.0001) amplified in all the vaccinees following the second booster dose (Figure 4C). Compared with the pp62-specific IgG responses induced by the priming dose, the second booster dose significantly (p < 0.0001) amplified the primary response (Figures 4B, C). However, following boosting, there was no significant difference in pp62-specific IgG responses between the pigs immunized with the Ad5-ASFV cocktail alone and the pigs immunized with the Ad5-ASFV cocktail formulated with either the Montanide ISA-201™ or the BioMize^® adjuvant, suggesting that these adjuvants did not have a booster effect on the pp62-specific IgG responses (Figure 4C). Immunization of piglets with the Ad5-ASFV cocktail, but not the Ad5-Luciferase immunogen, elicited antibodies that strongly recognized primary swine cells infected with wild-type ASFV (Georgia 2007/1) as judged by IFA using post-boost sera (Figure 5).

Following challenge by comingling with naïve contact spreaders, which had received IM injection of the ASFV (Georgia 2007/1), the mean pig weights in all the groups were stable for 1 week (Figure 6A). As expected, the contact spreaders developed a high fever, rapid onset of high viremia in blood as well as nasal fluids, and severe clinical disease by the time they were terminated on day 8 post-challenge (Figures 6B, 7). Histopathology of two representative contact spreaders showed tissue lesions that were consistent with severe acute ASF (Figure 8). Multiple organ and tissue samples had high virus loads, which was consistent with acute ASF observed in naïve pigs, and these values were statistically significant (p < 0.0001) compared with all other groups (Figure 9). Additionally, viral loads for multiple individual tissues were also found to be statistically significant when compared to the contact spreaders (see Supplementary Figures 1, 2). One week after the initiation of contact challenge, the pigs immunized with the Ad5-ASFV cocktail formulated in BioMize^® adjuvant had rapid weight loss that was accompanied by elevated rectal temperature, viremia in blood, and nasal fluid, and they were terminated after developing severe clinical disease (Figures 6, 7). The ASFV-BioMize immunized pigs had significantly lesser (p = 0.0439) viremia in their blood as compared to the contact spreaders (Figure 7A). Histopathology of one representative pig from this group revealed lesions that were consistent with moderate subacute ASF (Figure 10), but all the tissue samples collected at termination had high levels of ASFV (Figure 9). The mean weight of the negative control pigs immunized with the Ad5-Luciferase Montanide ISA-201™ formulation and the pigs immunized with the Ad5-ASFV cocktail formulated in Montanide ISA-201™ remained unchanged post-challenge, but the pigs developed a high fever in the 2^nd week and severe clinical disease as well as high viremia in blood and nasal fluids that necessitated termination (Figures 6, 7). Viremia in blood was found to be significantly (p = 0.0417) lower for the pigs that received the Ad5-ASFV cocktail formulated in Montanide ISA-201™ compared with contact spreaders, whereas the negative control pigs did not have significant levels of blood viremia compared with any other treatment groups (Figure 7A). High levels of ASFV were also detected in multiple organs and tissues (Figure 9).

Although the mean weight of the pigs immunized with the Ad5-ASFV cocktail without adjuvant was stable for 13 days, it declined thereafter as the pigs developed a high fever and rapid onset of viremia in blood and nasal fluids that peaked on day 7 after the initiation of contact challenge (Figures 7A, B). Four of the five vaccinees developed a clinical disease that necessitated termination by day 14 and histopathology of one representative pig revealed lesions that were consistent with moderate subacute ASF (Figures 7C, 10). However, tissue samples collected from the four pigs at termination had high levels of ASFV that mirrored outcomes from the other treatment groups (Figure 9). It was noted that, compared with the other groups, pigs from this group had the lowest mean viremia in the spleen (Figure 9). It was also noted that the pigs immunized with the Ad5-ASFV formulated with or without adjuvant had the lowest mean viremia in mandibular lymph nodes (Figure 9). Overall, there was no significant difference in total tissue viral load in the pigs immunized with the Ad5-ASFV formulated with or without adjuvant, when compared to the Ad5-Luciferase-negative controls (Figure 9). As expected, contact spreaders total tissue viremia was significantly higher (p < 0.0001) when compared with the pigs immunized with the Ad5-ASFV or the Ad5-Luciferase-negative controls (Figure 9).

One of the five pigs (No. 6892) immunized with the Ad5-ASFV cocktail without adjuvant seroconverted and had a rapid increase in antibody response within the first 2 weeks, but unlike the other four vaccinees in this group, boosting did not significantly amplify pp62-specific IgG responses (Figure 11A). Interestingly, this pig had the lowest post-boost pp62-specific IgG responses compared with the other vaccinees in this group (Figure 11A). Following exposure to contact spreaders, viremia in blood and nasal fluid peaked on day 7 in this pig, whereas fever peaked on the 11 day (Figure 11B). The weight of the pig declined subtly from day 7, but on day 14, when the other four vaccinees in this group were terminated due to severe clinical disease, the pig started to gain weight and sustained weight gain until the study was terminated at 39 days after initiation of contact challenge (Figures 7C, 11B). The weight gain was accompanied by the resolution of fever and gradual decrease in viremia in blood until study termination (Figure 11B). Although the pig was clinically healthy at termination, it had recurrent episodes of high viremia in nasal fluid suggesting that shedding of virus and reinfection was likely still occurring (Figure 11B). Necropsy revealed that the pig had evident inflammation and marked thickening of the epicardium with an accumulation of fibrino-hemorrhagic fluid in the pericardial sac due to chronic-active pericarditis (Figure 12). In addition, histopathology revealed lesions consistent with moderate subacute ASF, but viremia in tissue samples collected at termination had much lower levels of ASFV compared with the average tissue viremia detected in the contact spreaders as well as the other treatment and control pigs (Figures 9, 11C, Supplementary Table 3). The viremia in tissue samples for the survivor was significantly lower than the contact spreaders in the mandibular LN (p = 0.0012), cranial mediastinal LN (p = 0.0029), gastrohepatic LN (p = 0.0344), and liver (p = 0.0088) (Figures 9, 11C). Significantly lower viremia in the mandibular LN of the survivor was also found compared with the pigs that received the Ad5-ASFV cocktail formulated in Montanide ISA-201™ (p = 0.005) and when compared to the four pigs that were also immunized without adjuvant (p = 0.0421) who succumbed earlier in the study (Figures 9, 11C).

Following initiation of challenge by contact with naïve ASFV-infected pigs, which were terminated on day 8, the mean survival of the negative control pigs (Ad5-Luciferase Montanide ISA-201™ formulation) was 11.4 days compared with 12.2 to 17 days for Ad5-ASFV vaccinees (Ad5-ASFV cocktail without adjuvant, Ad5-ASFV cocktail Montanide ISA-201™ formulation, and Ad5-ASFV cocktail BioMize^® formulation) with the exemption of one vaccinee that survived (Ad5-ASFV cocktail without adjuvant) until study termination on day 39 (Figure 7C). Typical symptoms for all animals that succumbed to the disease consisted of high fever, lethargy, reddening of the skin, anorexia, and weight loss. Analysis of clinical presentation of ASF at termination showed that the vaccinees, which were immunized with the Ad5-ASFV cocktail formulated with Montanide ISA-201™ or BioMize^® had an average clinical score of 3.2, while the vaccinees that were immunized with the Ad5-ASFV cocktail without adjuvant had an average clinical score of 3.4 (Supplementary Table 4). Although these scores were less than the average clinical score of 3.6 for the naïve contact spreaders, the difference was not significant (Supplementary Table 4). The negative controls (Ad5-Luciferase Montanide ISA-201™ formulation) were a notable exemption with an average clinical score of 1.4 and the lone survivor (Ad5-ASFV cocktail without adjuvant) had no clinical symptoms at termination (Supplementary Table 4).

Compared with the contact spreaders, the mean histopathological scores of lymphoid tissues from the Ad5-Luciferase controls and the vaccinees immunized with the Ad5-ASFV formulated with or without adjuvant were significantly (p < 0.0001) lower (Figure 13A). It was also noted that the mean histopathological scores of lymphoid tissues from the pigs immunized with Ad5-ASFV without adjuvant or the Ad5-ASFV Montanide ISA-201™ vaccinees were significantly (p < 0.0001) lower than the mean scores for lymphoid tissues from the Ad5-Luciferase controls and the Ad5-ASFV BioMize^® vaccinees (Figure 13A). There was no difference between the mean scores for the lymphoid tissues from the pigs immunized with the Ad5-ASFV without adjuvant and the Ad5-ASFV Montanide ISA-201™ vaccinees (Figure 13A). The mean histopathological scores for visceral organs (liver, lungs, and kidney) from the contact spreaders were significantly (p = 0.0014) higher than the mean score for Ad5-Luciferase controls and the Ad5-ASFV Montanide ISA-201™ vaccinees as well as the mean score for the vaccinees that received the Ad5-ASFV without adjuvant (p = 0.0057) (Figure 13B). There was no difference between the contact spreaders and the Ad5-ASFV BioMize^® vaccinees (Figure 13B). However, analyses of individual lymphoid tissues as well as visceral organs showed that there was extreme variation in histopathological scores between the groups and among the controls as well as the vaccinees (Supplementary Figures 3, 4).