Targeted sequencing analysis of Mycoplasma gallisepticum isolates in chicken layer and breeder flocks in Thailand

Mycoplasma gallisepticum (MG) is one of the most economically important pathogens worldwide. MG affects the respiratory system and impairs growth performance in poultry. In developing countries, the most widely used technique to identify MG is the conventional PCR assay. In this study, 24 MG isolates collected from Thailand farms with unvaccinated chickens during 2002–2020 were characterized by gene-targeted sequencing (GTS), followed by phylogenetic analysis using unweighted pair group method with arithmetic mean. These 24 Thai MG isolates differed from vaccine strains, including the F, ts-11 and 6/85 strains. One isolate showed 99.5–100% genetic similarity to the F strain with 4 partial gene analyses. This result may have been due to contamination from vaccinated flocks because the F strain is the most commonly used vaccine strain in Thailand. However, the GTS analysis using the partial MG genes in this study showed that the isolates could be grouped into different patterns based on individual gene sequences. The phylogenetic analysis of partial mgc2, gapA, pvpA and lp gene sequences classified the Thai MG isolates into 7, 11, 7 and 2 groups, respectively. In conclusion, at least 2 partial MG genes, especially partial gapA and mgc2 genes, are needed to differentiate MG isolates.

www.nature.com/scientificreports/ However, RAPD has low reproducibility, and results from different laboratories cannot be compared 14,15 . Sequencing is a potential technique for MG classification. MG strains can be differentiated with partial DNA sequences and compared among laboratories in different areas or countries 4,16,17 . In addition, gene targeted sequencing (GTS) is a cost-efficient and affordable method for use in developing countries, including Thailand, where advanced techniques are not generally feasible.
The important genes of MG, including gapA, mgc2, pvpA and MGA_0319 (lp), have been investigated in several epidemiological studies 4,18,19 . In Thailand, Limsatanun et al. 20 classified MG strains with partial mgc2 gene sequences; thus, the partial mgc2 gene can be used to classify Thai MG strains from vaccine strains and various strains from different countries. However, partial mgc2 gene classification is not a reliable method for MG characterization 4 .
The aim of this study was to determine a GTS technique for differentiating field and vaccine MG strains in commercial chicken flocks from different regions in Thailand. This is the first study to use 4 partial MG gene sequences for commercial MG classification in Thailand.

Results
PCR amplification. All twenty-four Thai MG isolates were detected by MG-specific PCR amplification following the Lauerman method 21 . To amplify partial mgc2 genes, which were 615 bp in size, 22 Thai MG isolates were successfully amplified and sequenced. According to the specific partial gapA PCR with 306 bp, 21 Thai MG isolates were positive and included in the phylogenetic analysis, while 20 samples of Thai MG isolates were successfully amplified using the pvpA and MGA_0319 (lp) primers with lengths of 456 and 495 bp, respectively. All nucleotide sequences from Thai MG isolates in this study were submitted to GenBank and given accession numbers (Table 1).

Phylogenetic analysis.
The phylogenetic tree based on the partial mgc2 gene demonstrated that 3 Thai MG isolates were closely related to the F strain. AHRU/2014/CU4508.1 was grouped together with the F strain, while AHRU/2020/CU0143.1 and AHRU/2020/0147.1 showed 97.6% genetic similarity to the F strain ( Fig. 1). According to the phylogenetic tree based on the partial gapA gene, AHRU/2014/CU4508.1 showed 99.5% genetic similarity to the F strain. AHRU/2020/CU3704.1 was also grouped with the F strain (Fig. 2). The phylogenetic analysis of the partial pvpA gene placed all Thai MG isolates in the same cluster except the reference strain S6 (Fig. 3). Four Thai MG isolates showed 94.3% genetic similarity to the 6/85 strain. AHRU/2014/CU450 8.1 was grouped with the F strain with 100% similarity. The partial lp gene sequences of Thai MG were compared with reference strains. The 6/85 strain was grouped into different Clusters. AHRU/2014/CU4508.1 and AHRU/2020/ CU0143.1 had 100% genetic similarity to the F strain and 99.2% genetic similarity to the ts-11 strain (Fig. 4). The genetic similarity of Thai MG strains and F strain is shown in Table 2. The lp gene showed the highest similarity of genetic sequences (98.2-100%) between the F strain and Thai MG strains. The phylogenetic trees with DNA sequence data are available in the Supplementary Information.

Discussion
Avian mycoplasmosis is an important disease-causing pathogen in the poultry industry with substantial economic impacts. Live, inactivated, and recombinant MG vaccines have been used in Thailand for a long time. Due to the increased use of MG vaccines, differentiation between field and vaccine strains is needed. Molecular characterization of MG has been investigated in many countries 4,5,7,22,23 . This study is the first to use the GTS technique on Thai MG strains with 4 partial MG gene sequences. The partial mgc gene has been used for MG characterization in many epidemiological studies 6,20,24 . It encodes the MGC2 protein, which coordinates with the gapA gene-encoded protein for cell attachment 25 and is involved in MG immunogenicity 12,26 .
In a previous study, Armour et al. 22 investigated MG isolates from South Africa using intergenic spacer regions (IGSRs), mgc2 and gapA genes. Thirty-six MG isolates were classified into 8 types by the mgc2 gene and 2 types by the gapA gene. Thus, the mgc2 gene had a higher discriminatory power than the gapA gene. Another study in Russia conducted an epidemiological investigation of MG 7 . The results showed that mgc2 gene had good discriminatory power, while gapA did not provide a good discriminatory index for MG classification. However, the use of only a single gene for classification could not determine the similarity between MG isolates. Additionally, www.nature.com/scientificreports/ some MG isolates were negative for the mgc2 gene according to the PCR assay, resulting in a failure to obtain mgc2 sequences 5,6,19,22 ; thus, using only one partial gene sequence is insufficient to characterize MG.
In the present study, the lp gene of Thai MG isolates was more conserved than the gapA, mgc2 and pvpA genes, as 18 out of 20 Thai MG isolates showed 100% genetic similarity on this gene. The use of partial gapA showed the highest genetic variation among Thai isolates. These results contradicted those of previous studies 4,7,22 , which indicated that MG isolates from the same area would have lower genetic diversity than MG isolates from different regions 22 . In the present study, Thai MG isolates were identified with 4 genes using the phylogenetic tree (UPMGA) method. AHRU/2014/CU4508.1 had the closest genetic relationship to the F strain. The UPMGA results showed that AHRU/2014/CU4508.1 was grouped with the F strain on all 4 partial gene analyses. Interestingly, all Thai MG isolates in this study were collected from farms with unvaccinated flocks. In Thailand, poultry breeders and layers are widely vaccinated with the F strain. Interestingly, the AHRU/2014/CU4508.1 isolate from these farms might have been contaminated from other farms with vaccinated flocks. The F strain from the live MG vaccine can be transmitted both horizontally and vertically [27][28][29] . Furthermore, several epidemiological studies have shown that the F strain can cause MG outbreaks if it spreads from vaccinated to nonvaccinated flocks 5,24,30 . Other Thai MG isolates in this study varied in genetic classification depending on the gene analysed. The results of gapA and mgc2 gene analysis showed that AHRU/2003/CU5113.2 and AHRU/2003/CU5808.2 were grouped with the S6 strain with 97% and 99.4% genetic similarity, respectively. In contrast, using the partial pvpA gene  In conclusion, the Thai MG isolates in this study could be differentiated with partial MG genes, including the gapA, mgc2, pvpA and MGA_0319 (lp) genes. All Thai MG isolates could be classified with at least 2 out of 4 partial gene sequences, especially the partial gapA and mgc2 genes, which had satisfactory discriminatory power for Thai MG characterization. Using partial DNA sequencing for MG characterization is an effective and reproducible method for establishing the genetic relationship between MG strains and differentiating between vaccine and field strains. In addition, this study was the first epidemiological study of Thai MG strains to use 4 partial MG gene sequences, demonstrating the genetic diversity of circulating MG strains in Thailand. In future studies, the GTS technique should be implemented along with other molecular techniques, including a multilocus www.nature.com/scientificreports/ sequence typing scheme, to provide more epidemiological and evolutionary data and improve the system for monitoring MG outbreaks in poultry farms in Thailand. Reference sequences. Four reference strains were used in this study. The F strain was the vaccine strain provided by a local distributor (MSD, Thailand). The S6 strain was obtained from ATCC (15302). The ts-11 and DNA sequence analysis. Amplified PCR products of MG-targeted gene-positive extracts were submitted to determine the sequence. Partial mgc2 gene sequences (Accession Numbers KX268616-KX268632) from 16 Thai MG isolates had been submitted to GenBank in a previous study 20 (Table 3.) All sequences were analysed with the Editseq program (Lasergene, DNASTAR Inc., USA), and a consensus was constructed with the Seqman program (Lasergene, DNASTAR Inc., USA). Thai MG isolates and reference gene sequence data were aligned to construct a phylogenetic tree in Bionumeric version 7.6 software (Applied Maths, Sint-Martens-Latem, Belgium). Cluster analysis was performed with the UPGMA method. The similarity coefficients of Thai MG isolates and reference strains were determined from multiple sequence alignments. www.nature.com/scientificreports/  www.nature.com/scientificreports/