From March 30th to May 23rd, 2024, 404 lung specimens with confirmed pathological lesions were collected from slaughtered pigs at a slaughterhouse in Chongqing. These pigs were sourced from multiple distinct pig farms (Table 1). Prior to sample processing, the lung surface was disinfected with 3% hydrogen peroxide, after which a rice-grain-sized internal tissue fragment was homogenized in 1 ml of complete KM2 medium (KM2 broth medium containing 20% porcine serum and 0.01% NAD, pH 7.5) (23) supplemented with 100 μg/ml ampicillin. Homogenates were incubated at 37 °C for 7–14 days, with a red-to-yellow color change indicating medium acidification. After three passages, cultures were plated on KM2 agar and incubated at 37 °C for 5–10 days. Fried egg-like colonies were selected, expanded in complete KM2 broth, and re-plated on KM2 agar; this purification was repeated a minimum of three times to ensure monoclonality.

DNA was extracted from the purified final culture using rapid boiling DNA extraction. Each broth sample was centrifuged at 13,500 × g for 10 min and discarded the supernatant. The bacterial pellet was resuspended in 50 μl of sterile double-distilled water. The suspended cells were treated with dry heat block at 105 °C for 5 min, placed on ice for 10 min, and centrifuged at 13,500 × g for 5 min. The supernatant containing the DNA template was collected. A real-time TaqMan PCR assay was conducted by amplification of p37 gene of M. hyorhinis (24) using ApexHF HS DNA Polymerase FS Master Mix (Accurate Biology, Changsha, Hunan, China). Moreover, a partial sequence of the 16S rRNA gene of purified final culture was amplified by PCR using a primer pair (forward:5′-ACCCGAGAACGTATTCAC3′; reverse:5′-GCAAGCGTTATCCGGA3′). The PCR was performed in a 25 μl reaction volume consisting of 0.5 μl of DNA template, 0.5 μl of each primer (10 μm), 12.5 μl of ApexHF HS DNA Polymerase FS Master Mix, and 11 μl of ddH₂O. The amplification protocol consisted of an initial denaturation step at 94 °C for 5 min, followed by 30 cycles of amplification at 98 °C for 10 s, annealing at 55 °C for 10 s, and extension at 72 °C for 45 s, with a final extension at 72 °C for 5 min. The PCR products were subsequently sequenced in both directions by Sangon Biotech Co., Ltd.

MLST primers for six housekeeping genes were synthesized according to published methods (Supplementary Table S1) (25). PCR was performed in 25 μl reactions with ApexHF HS DNA Polymerase FS Master Mix, following the method of Trüeb et al. (25), and products were sequenced by Sangon Biotech (Shanghai, China). The DNA sequences of the six housekeeping genes were submitted to the PubMLST database (https://pubmlst.org/organisms/mycoplasma-hyorhinis) for allele IDs assignment, and an allelic profile (six loci) was generated for each strain to determine STs. STs were compared with 176 reference strains (from 9 countries across Asia, Europe, North America) in the database. BURST analysis (26), in which allelic profiles matched at 4 loci, was applied to divide the strains into various lineage groups. A phylogenetic tree was constructed by MEGA 12.0 and visualized and annotated using iTOL v7.2 (https://itol.embl.de/).

M. hyorhinis strains were inoculated into complete KM2 medium at a ratio of 1:9, followed by incubation at 37 °C. When the medium turned yellow, the culture was inoculated into two portions of fresh complete KM2 medium portion (1:9 ratio) and incubated at 37 °C. At 0 h, 12 h, 24 h, 36 h, 48 h, 60 h, 72 h, 84 h, and 96 h, 200 μl and 100 μl aliquots were collected from each culture. The optical density at 600 nm (OD₆₀₀) of each 200 μl sample was measured using an ELISA plate reader (Thermo Fisher Scientific, Ratastie 2, FI-01620 Vantaa, Finland). Serial tenfold dilutions were performed on 100 μl samples: 100 μl sample was added to 900 μl complete KM2 medium, then 100 μl of this suspension was transferred to a second tube with 900 μl medium. Dilutions were continued from the 2nd to the 22nd tube, with the 23rd tube as a negative control. Tubes were incubated at 37 °C for 14 days; no visible growth (no red to yellow color change) was confirmed. The color changing unit (CCU) the bacterial fluid at each time point was the average of three repeats. Three growth-related curves were plotted for each isolate: (1) h-OD₆₀₀ curve (X-axis: time points; Y-axis: average OD₆₀₀ values), (2) h-lgCCU curve (X-axis: time points; Y-axis: lgCCU values), and (3) growth curve (X-axis: average OD₆₀₀ values; Y-axis: lgCCU values).

The minimum inhibitory concentration (MIC) of M. hyorhinis isolates for 15 antimicrobials and 12 Chinese herbal monomers (Supplementary Table S2) was determined via broth microdilution (27). Kanamycin, gentamicin, amikacin, apramycin, chlortetracycline, doxycycline, tylosin, tylvalosin, lincomycin, tiamulin, enrofloxacin and ciprofloxacin were diluted with ddH₂O at room temperature. Berberine hydrochloride was diluted with ddH₂O at 40 °C. Florfenicol, erythromycin, tilmicosin, allicin, gingerenone A, baicalin, trans-cinnamic acid, cinnamaldehyde, gallic acid, glycyrrhizic acid, quercetin, curcumin, sodium houttuyfonate and ferulic acid were diluted with DMSO. The broth microdilution test was performed in 96-well microliter plates (31121, Labselect, Beijing, China) with two-fold serial dilutions (antimicrobials: 0.25–128 μg/ml; herbal monomers: 0.25–256 μg/ml) in 100 μl complete KM2 medium with sterility, pH and growth control. M. hyorhinis cultures (105 CCU/ml, 100 μl) were inoculated into wells and incubated at 35 ± 2 °C; MICs were read after 1 week (no further growth). All tests were run in triplicate. Notably, official breakpoint criteria for in vitro susceptibility testing of animal mycoplasmas are currently lacking. The M. hyorhinis reference strain BTS7 (ATCC 17981) was used as a quality control for MIC determination. MIC50 and MIC90 values were defined as the lowest concentrations inhibiting the growth of 50% and 90% of the tested strains, respectively (27).

PCR was used to detect the mutations within four QRDRs of gyrA, gyrB, parC and parE, and domains II and V of the 23S rRNA genes to analyze the resistance mechanisms. The primer sequences and corresponding PCR amplification conditions are presented in Supplementary Table S1, as described in previous studies (15, 22). The PCR products were subjected to electrophoresis on a 1.5% agarose gel, and positive amplicons were bidirectionally sequenced. The sequencing results were aligned and analyzed using BLAST (http://www.ncbi.nlm.nih.gov/BLAST/). The resulting DNA sequences and the corresponding amino acid sequences of all the PCR products were compared with the gyrA, parC and parE sequences (accession number: NZ_KB911491) and the 23S rRNA sequence (accession number: AB182581) of the M. hyorhinis BTS7 (ATCC 17981) strain and the gyrB sequence of the M. hyorhinis HUB-1 strain (accession number: CP002170).

A total of 28 M. hyorhinis isolates were obtained from 404 lung samples (6.9%). M. hyorhinis isolates were confirmed by single colony selection, amplification of the M. hyorhinis-specific p37 gene, and sequence alignment between the 16S rRNA of the isolates and that of M. hyorhinis strains. After the nucleotide sequences were submitted to the PubMLST database for homology alignment, which revealed that none of the STs of the 28 isolates matched any entries in the database. Nucleotide sequences of the isolates were submitted to the PubMLST database for homology alignment, which revealed that none of the STs of the 28 isolates matched any entries in the database. These results indicated that the isolates identified in this study represent novel STs that have not been previously recorded in the database. Following reannotation by the database curator, the 28 isolates were classified into 11 distinct ST patterns (Tables 2, 3). Among these, ST226 and ST229 were the most prevalent, each accounting for 25.0% (7/28) of the total isolates. Three isolates (10.7%, 3/28) were assigned to ST227, while ST230, ST260, and ST262 each comprised two strains (7.1%, 2/28). Additionally, five isolates were distributed across five unique STs, namely ST228, ST261, ST263, ST264, and ST265.

Three novel alleles were identified among the isolates, designated as dnaA-38 (5), rpoB-31 (4), and gyrB-18 (n). Specifically, three strains carried two novel gene sequences, whereas four strains harbored a single novel allelic variant. As summarized in Table 4, the dnaA gene exhibited five allelic variants: dnaA-1 was the dominant allele, detected in 39.3% (11/28) of isolates, followed by dnaA-10 (35.7%, 10/28), dnaA-38 (17.9%, 5/28), dnaA-2 (3.6%, 1/28), and dnaA-26 (3.6%, 1/28). The rpoB gene was represented by three alleles: rpoB-1 (46.4%, 13/28), rpoB-3 (39.3%, 11/28), and rpoB-31 (14.3%, 4/28). For the gyrB gene, five alleles were identified across the 28 strains: gyrB-1 (28.6%, 8/28), gyrB-3 (28.6%, 8/28), gyrB-4 (25.0%, 7/28), gyrB-11 (14.3%, 4/28), and gyrB-18 (3.6%, 1/28). Notably, gltX-4 was the sole allelic variant of the gltX gene detected in all isolates. Regarding the adk gene, adk-2 was the most abundant allele (78.6%, 22/28), followed by adk-8 (14.3%, 4/28), adk-1 (3.6%, 1/28), and adk-15 (3.6%, 1/28). For the gmk gene, gmk-3 was the predominant allele (42.9%, 12/28), with gmk-1 (25.0%, 7/28) and gmk-5 (32.1%, 9/28) as the other two variants.

To investigate the genetic clustering of the isolates, BURST analysis was performed with a threshold of 4 shared allelic loci among group members. A total of 204 STs were included in the analysis, consisting of the 28 STs identified in this study and 176 reference STs from domestic and international sources. Notably, all STs were clustered into one group (data not shown). ST40 was identified as the central genotype, with ST13, ST17, ST24, ST30, ST32, ST41, ST88, and ST171 forming eight distinct subgroups (Figure 1). Sequence divergence analysis showed that ST40 differed from ST265 at 3 loci and from ST227 and ST228 at 4 loci. ST13 differed from ST226 at 3 loci. Seven additional STs were not visualized in Figure 1, as these did not cluster within any of the eight established subgroups. The inability of these isolates to form independent subgroups separate from other isolates is likely due to the insufficient data present in the current database.

Furthermore, a phylogenetic tree was constructed based on the concatenated nucleotide sequences of six housekeeping genes using the neighbor-joining method (Figure 2). The 28 isolates from this study, marked with green pentagrams, formed a distinct clade that was separated from all reference isolates in the database.

The h-OD₆₀₀ curves, h-lgCCU curves, and growth curves of all isolates are presented in Table 5, Figure 3, Supplementary Figures S1, S2. Overall, the OD600 values of isolates CQ001–CQ024, CQ026, and BTS07 exhibited a gradual decline with prolonged cultivation time. In contrast, isolates CQ025, CQ027, and CQ028 showed an increase in OD600 values up to 24 h of cultivation, followed by a gradual decrease thereafter. With respect to lgCCU, all isolates displayed a phase of increase until a specific time point, after which lgCCU values declined. The peak lgCCU was observed at 24 h for isolates CQ001, CQ003, CQ010, CQ012, CQ016, CQ021, and CQ026–CQ028; at 36 h for isolates CQ004–CQ006, CQ008, CQ009, CQ011, CQ013–CQ015, CQ017, CQ020, CQ023–CQ025, and BTS07; and at 48 h for isolates CQ002, CQ007, CQ018, CQ019, and CQ022. Growth curves were plotted with the average OD600 values at 0 h, 12 h, 24 h, 36 h, and 48 h on the X-axis and the corresponding lgCCU values on the Y-axis. As shown in Figure 3, Supplementary Figures S1, S2, a negative correlation between OD600 and lgCCU values was observed for isolates CQ001–CQ024, CQ026, and BTS07. In contrast, isolates CQ025, CQ027, and CQ028 exhibited a positive correlation between these two parameters during the 0–24 h cultivation period. Significant inter-isolate variability was observed in the maximum CCU achieved in complete KM2 medium, with values ranging from 10²² to 10²0 CCU/ml (Table 5).

The MIC values, as well as the MIC50 and MIC90 values, for all tested antimicrobials and Chinese herbal monomers are summarized in Table 6 and Supplementary Table S3. The isolates demonstrated high susceptibility to tiamulin (MIC ≤ 0.25 μg/ml), doxycycline (MIC ≤ 0.25–1 μg/ml), ciprofloxacin (MIC ≤ 0.25–2 μg/ml), florfenicol (MIC 0.5–2 μg/ml), kanamycin (MIC ≤ 0.25–4 μg/ml), and enrofloxacin (MIC 0.5–4 μg/ml). The MICs of gentamicin (96.4%, 27/28), chlortetracycline (92.9%, 26/28), tylvalosin (85.7%, 24/28) and apramycin (82.1%, 23/28) for most of the isolates were lower than 8 μg/ml. In contrast, all isolates exhibited high MIC values (32–≥128 μg/ml) to erythromycin. A bimodal distribution of MIC values was observed for tylosin and tilmicosin: 50.0% (14/28) of isolates had a tylosin MIC ≤ 1 μg/ml, and 39.3% (11/28) of isolates had a tilmicosin MIC ≤ 2 μg/ml.

Overall, compared with antimicrobials, most Chinese herbal monomers exhibited higher MIC values against M. hyorhinis isolates (Table 6, Supplementary Table S3). The MIC ranges of the tested herbal monomers were as follows: ferulic acid (128 to ≥256 μg/ml); cinnamaldehyde, gallic acid, glycyrrhizic acid, quercetin, curcumin, and sodium houttuyfonate (64 to ≥256 μg/ml); trans-cinnamic acid (32 to ≥256 μg/ml); baicalin (64–128 μg/ml); gingerenone A (16–128 μg/ml); and allicin (16– 64 μg/ml). Notably, the MICs of 85.7% (24/28) of the isolates for berberine hydrochloride were < 8 μg/ml.

All isolates were phenotypically susceptible to the two tested fluoroquinolones, although the MIC values of enrofloxacin for four isolates were equal to 4 μg/ml. To explore the potential association between QRDR mutations and MIC values, we amplified four genes encoding QRDRs by using the primers designed by Li et al. (15). PCR amplification results showed that gyrA and parE were successfully amplified in all 28 isolates, whereas gyrB and parC were detected in only 8 and 23 isolates, respectively.

Nucleotide sequence alignment was performed using the gyrA, parC, and parE sequences of the BTS7 strain and the gyrB sequence of the HUB-1 strain as references. Several nonsynonymous mutations were identified in one or two QRDR genes across the isolates (Table 7, Supplementary Table S3). A conserved nucleotide substitution at position 185 (A185G) in the gyrA gene resulted in an I62S amino acid substitution in all the strains. A nucleotide mutation at position 274 (A274G) in the parC gene led to an I92V amino acid change in 5 strains. For the parE gene, a nonsynonymous mutation at position 1098 (C1098A) caused an E366D amino acid substitution in 6 strains; additionally, isolate CQ005 carried two mutations in parE (T873A and C1098A), which resulted in amino acid substitutions N291K and E366D, respectively. No mutations were detected in the amplified gyrB gene fragments.

Notably, none of the mutations identified in this study matched those previously reported by Li et al. (15). To further explore genetic variations in gyrB, we aligned the gyrB sequences used for MLST analysis (25) against the HUB-1 reference sequence. A nonsynonymous mutation at position 1708 (G1708A) was detected in isolates CQ021, CQ022, and CQ023, leading to a V570I substitution. Collectively, these results indicate that no key QRDR mutations in gyrA, gyrB, parC, or parE were associated with elevated enrofloxacin or ciprofloxacin MIC values in the tested isolates.

Nucleotide sequence alignment of the 23S rRNA gene against the BTS7 reference strain revealed no mutations in domain II among the 28 isolates. In contrast, two distinct mutations were identified in domain V: an A1553G substitution was detected in 15 isolates, and a C2014T substitution was uniquely present in isolate CQ013 (Supplementary Table S3). The A1553G mutation appeared to be strongly correlated with high MIC values for tylosin, tylvalosin, tilmicosin, and lincomycin, despite the fact that all isolates exhibited high MICs of erythromycin (MIC: 32–≥128 μg/ml). Specifically, all isolates carrying the A1553G mutation displayed elevated MIC values for tylosin (≥16 μg/ml), tylvalosin (≥2 μg/ml), tilmicosin (≥32 μg/ml), and lincomycin (≥32 μg/ml), with three exceptions: isolate CQ017 showed low tylosin susceptibility (MIC ≤ 0.25 μg/ml), while isolates CQ003 and CQ004 exhibited low tylvalosin MIC values (0.5 and 1 μg/ml, respectively).

The mutation at A1553G of the 23S rRNA in M. hyorhinis was equivalent to the mutation at A2058G on the basis of 23S rRNA of E. coli (22). The C2014T mutation in domain V of the 23S rRNA in the CQ013 strain was related at least to the high MICs to tilmicosin, as the MIC for CQ013 (MIC: 32 μg/ml) was greater than that for strains without mutation in domain V of the 23S rRNA (MIC: ≤ 0.25 μg/ml) to tilmicosin.