All of the methods were approved by the Seoul National University Institutional Animal Care and Use and Ethics Committee (SNU-210518-3).

The clinical field trial was performed on an 800-sow, farrow-to-finish swine farm that carries out an all-in-all-out production system. A stable status of the porcine reproductive and respiratory syndrome (PRRS) was maintained with the absence of active porcine reproductive and respiratory syndrome virus (PRRSV) circulation (only multiple parity sows were seropositive animals). Sows were not previously immunized against PCV2 and M. hyopneumoniae. The farm was selected on the basis of subclinical PCV2 infection and enzootic pneumonia. The farm consistently had respiratory problems due to poor growth in the late post-weaning and growing stages. Clinical signs first appeared at approximately 8–11 weeks of age and reached peak mortality (~3–5%) between 10 and 15 weeks of age. PCV2d was detected in serum from 3 pigs with poor growth, where log₁₀ DNA copies/ml ranged from 2.46 to 3.34. These values were consistent with the definition of subclinical PCV2 infection (21, 22). Through pre-trial investigations, a PCV2 serological profile was identified that presented an increase in antibody titers starting around 8 weeks of age; 7–16-week-old pigs with poor growth were also PCV2 PCR-positive in tested blood samples. M. hyopneumoniae serology was positive in 8–16-week-old pigs with respiratory signs. Moreover, the laryngeal swabs of the same pigs were PCR-positive for M. hyopneumoniae on the farm. Histologically, mycoplasmal lung lesions characterized by peribronchiolar and peribronchial lymphoid tissue hyperplasia were observed in 4 out of 5 submitted pigs. Lymphoid depletion without granulomatous inflammation was also observed in 3 out of 5 submitted pigs. Pre-trial diagnostic results indicated subclinical PCV2 infection and enzootic pneumonia.

To reduce sow variation, six pigs (21-day-old) per sow were randomly selected by the random number generator function (Excel, Microsoft Corporation, Redmond, WA, USA) and evenly allocated to each of the three groups. A total of 180 pigs were randomly divided into 3 groups (60 pigs per group; male pigs = 30 and female pigs = 30) within the same software and function (Table 1).

At 0 day post-vaccination (dpv, 21 days of age), pigs in the VacA group were intramuscularly vaccinated with a 2.0 ml dose of the bivalent vaccine containing PCV2a and M. hyopneumoniae (Porcilis^® PCV M Hyo, lot no. A114A01, expiration date: 23 June 2022, MSD Animal Health, Boxmeer, Netherlands) at the right side of the neck in accordance with the manufacturer's directions. According to the manufacturer's instruction, pigs in the VacB group were intramuscularly vaccinated with a 1.0 ml dose of the bivalent vaccine containing PCV2b and M. hyopneumoniae (Circo/MycoGard^®, serial no: CMG-21006, expiration date: 20 January 2023, Pharmgate Animal Health, Wilmington, NC, USA) in the same anatomical location. Pigs in the unvaccinated (UnVac) group were injected with 2.0 ml of phosphate-buffered saline (PBS, 0.01 M, pH 7.4) in the same anatomical location. Sample collection of blood and laryngeal swabs were performed at 0 (21 days old), 28 (49 days old), 49 (70 days old), and 91 (112 days old) dpv.

The pigs were daily monitored for abnormal clinical signs and weekly scored using scores ranging from 0 (normal) to 6 (severe dyspnea and abdominal breathing) (23). The mortality rate was calculated as the number of dead pigs by the number of pigs initially assigned to the group within the batch. Necropsy was performed on dead or culled pigs throughout the study. Including palpation, injection site reaction was evaluated 24 h post-vaccination. Observers were blinded to the type of vaccine status and vaccination.

The live weight of each pig was measured at 21 (0 dpv), 70 (49 dpv), and 175 (154 dpv) days of age. Over two time periods, (i) between 21 and 70 days old and (ii) between 70 and 175 days old, the ADWG (grams/pig/day) was analyzed. During the different production stages, ADWG was calculated as the difference between the starting and final weight divided by the duration of the stage. Data for culled or dead pigs were also included in the calculation.

For DNA extraction from collected serum samples, the commercial kit was used (QIAamp DNA Mini Kit, QIAGEN). An extracted DNA sample was used for the quantification of PCV2d genomic DNA copy numbers by real-time PCR (24). The forward and reverse primers (5′-GTA TTC AAA GGG CAC AGT GAG G-3′ and 5′-GCA CCA TCG GTT ATA CTG TCA AGA AA-3′) and probe specific for PCV2d (5′-FAM^TM-CAT CAT GTC CAC ATT CCA G-3′ Black Hole Quencher) were designed for detecting specific Capsid-coding region of PCV2d only (24).

For DNA extraction from collected laryngeal swabs, the commercial kit was used (QIAamp DNA Mini Kit, QIAGEN). An extracted DNA sample was used for the quantification of M. hyopneumoniae genomic DNA copy numbers by real-time PCR (25). The forward and reverse primers (5′-TTG ACT GCT ATC TTT GCA CGA TAA G-3′ and 5′- ACA ATA ATT GCT GAC CGT GGC-3′) and probe (5′-FAM-TGT CCA CTG CTG CAA ATA TTC GAT TTC TTG AA-TAMRA-3′) were used to detect M. hyopneumoniae (25).

Enzyme-linked immunosorbent assay (ELISA) was performed to measure antibodies against M. hyopneumoniae (M. hyo. Ab test, IDEXX Laboratories Inc.) and PCV2 (Ingezim CIRCO IgG, Ingenasa, Madrid, Spain). According to the instruction of the manufacturer for each kit, serum samples were recorded as positive for M. hyopneumoniae antibody if the S/P ratio (sample-to-positive ratio) was ≥0.4. For anti-PCV2 antibodies, samples were recorded as positive if the reciprocal ELISA titer was >350. A serum virus neutralization test against PCV2d was used to test the serum sample (26–28).

The number of PCV2d-specific and M. hyopneumoniae-specific interferon-γ secreting cells (IFN-γ-SC) in peripheral blood mononuclear cells (PBMCs) was measured by ELISPOT assay (24, 29). Briefly, 100 ml containing 2 × 10⁶ PBMCs in Gibco Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% fetal bovine serum (HyClone Laboratories, Inc., SelectScience, Bath, UK) were seeded onto plates and precoated overnight with anti-porcine IFN-γ monoclonal antibody (5 μg/ml, Mabtech, Mariemont, OH, USA) and incubated with PCV2d (20 mg/ml), M. hyopneumopniaee (4 mg/ml), phytohemagglutinin (10 mg/ml, Roche Diagnostics GmbH, Mannheim, Germany) as a positive control, or PBS as a negative control for 20 h at 37°C in a 5% humidified CO₂ atmosphere. The wells were washed five times with PBS (200 ml per well) and thereafter, the procedure followed instructions of the manufacturer using a commercial ELISpot assay kit (Mabtech). The spots on the membranes were read by an automated ELISpot reader (AID ELISpot Reader, AID GmbH, Strassberg, Germany). The results were expressed as the number of responding cells/million PBMC.

To estimate the percentage of the lung affected by pneumonia, two pathologists (Chae and one graduate student) at the Seoul National University (Seoul, Republic of Korea) scored the severity of macroscopic lung lesions as previously described (23). The collected lung and lymphoid tissue sections were examined and scored by two blinded veterinary pathologists; the severity of peribronchiolar and perivascular lymphoid tissue hyperplasia by mycoplasmal pneumonia lesions, on a scale of 0–6 (30). Based on lymphoid depletion and granulomatous inflammation, lymphoid lesion severity was also scored on a scale of 0–5 (31).

Real-time PCR and neutralizing antibody data were recalculated to log₁₀ and log₂ values, prior to statistical analysis. The collected data were assessed by the Shapiro-Wilk test for a normal distribution. Then, one-way ANOVA was performed to determine if there were statistically significant differences between different groups at each time point. For further evaluation, a post-hoc test for a pairwise comparison with Tukey's adjustment was conducted with a statistical significance result from a one-way ANOVA test. In case of the normality, assumption was not met and the Kruskal-Wallis test was performed. Results, which showed a statistical significance from the Kruskal-Wallis test, were further evaluated with the Mann-Whitney test to include Tukey's adjustment to compare the differences among the groups. The value of p < 0.05 was considered significant and reported in p-values.

Two vaccinated animals from the VacA and VacB groups had significantly (p < 0.05) lower respiratory signs than those of unvaccinated animals from the UnVacA group at 14–112 dpv. There were no differences in respiratory signs between the two vaccinated groups (VacA and VacB).

No difference in mean body weight was observed between the vaccinated and unvaccinated animals at the time of vaccination. During the growing period (70–175 days of age), the ADWG of the vaccinated groups (VacA and VacB) was significantly (p < 0.05) higher than that of the unvaccinated group (UnVac). Overall (3–175 days of age), the ADWG of the vaccinated groups (VacA and VacB) was significantly (p < 0.05) higher than that of the unvaccinated group (UnVac) (Table 1).

A total of 2 pigs died in the VacA group of severe pneumonia as determined by a combination of M. hyopneumoniae and PCV2d that was detected with PCR testing, and Pasteurella multocida was isolated from the lungs at 77 and 85 days of age. One pig from the VacB group died of pleuropneumonia caused by coinfection with Actinobacillus pleuropneumoniae and Glaesserella parasuis that was determined through isolation from the lungs at 81 days of age. A total of 4 pigs were died in the UnVac group; two pigs were died of bronchopneumonia as determined by a combination of M. hyopneumoniae that was detected with PCR, and P. multocida and Trueperella pyogenes that were isolated from the lungs at 69 and 85 days of age. The other two pigs were died from bronchopneumonia as determined by a combination of M. hyopneumoniae and PCV2d that were detected with PCR, and G. parasuis that was isolated from the lungs at 81 and 86 days of age.

The PCV2 DNA blood loads from vaccinated groups (VacA and VacB) were significantly (p < 0.05) lower than that of unvaccinated groups (UnVac) at 28, 49, and 91 dpv (Figure 1). Two of the vaccinated groups (VacA and VacB) had comparable PCV2 DNA loads in their blood throughout the entire field trial.

The amount of M. hyopneumoniae DNA loads in the larynx was significantly (p < 0.05) lower in the vaccinated groups (VacA and VacB) than in those of the unvaccinated group (UnVac) between 28 and 91 dpv (Figure 2). Throughout the entire field trial, two vaccinated groups (VacA and VacB) had comparable M. hyopneumoniae DNA loads in their larynx.

The vaccinated groups (VacA and VacB) were measured significantly (p < 0.05) higher in their PCV2 ELISA titers (Figure 3A), neutralizing antibody titers (Figure 3B), and IFN-γ-SC (Figure 3C) than that of the unvaccinated (UnVac) group at 28, 49, and 91 dpv. No significant differences in PCV2 ELISA titers, neutralizing antibody titers, or IFN-γ-SC were observed in the two vaccinated (VacA and VacB) groups throughout the entire field trial.

Both vaccinated groups (VacA and VacB) were measured significantly (p < 0.05) higher in their M. hyopneumoniae ELISA S/P ratios (Figure 4A) and IFN-γ-SC levels (Figure 4B) than animals from the UnVac group at 28, 49, and 91 dpv. No significant differences in M. hyopneumoniae ELISA S/P ratios or IFN-γ-SC were observed in the two vaccinated (VacA and VacB) groups throughout the entire field trial.

Vaccination of animals from both vaccinated groups (VacA and VacB) effectively reduced macroscopic lung lesion scores, microscopic lung, lymphoid lesion scores, and the numbers of lymphoid PCV2-positive cells when compared to the unvaccinated group (UnVac) at 154 dpv (Table 2). Throughout the entire field trial, there were no significant differences in overall scores for macroscopic and microscopic lung lesions, microscopic lymphoid lesions, and the numbers of lymphoid PCV2-positive cells between the two vaccinated groups (VacA and VacB).