All chemicals used in this study were purchased from Sigma-Aldrich unless stated otherwise. Antibodies were obtained from Abcam (rabbit anti-lipoic acid antibody, goat anti-mouse secondary antibody, HRP conjugated) and LI-COR (IRDye 680RD goat anti-mouse secondary antibody, IRDye 800CW goat anti-rabbit secondary antibody). PCR amplification was performed using Taq (New England Biolabs) PrimeSTAR HS (Premix) (TaKaRa) polymerase. Restriction enzymes and T4 DNA ligase were supplied by TaKaRa. A ClonExpress II One Step Cloning Kit was purchased from Vazyme. GE Healthcare provided a Ni-NTA agarose column.

The bacterial strains and plasmids used in this study are listed in Table 1. The E. coli strains MG1655, DH5α, and BL21(DE3) were used for genome mutation, DNA manipulation and protein expression, respectively. The vectors pET32a and pET22b were used to express recombinant proteins in BL21(DE3), while pET-HITES was gifted by Qiang Li of Tsinghua University and used to express apo lipoate-requiring substrates in the MG1655-derived strain DE0626. The plasmid pBAD322G was gifted by John Cronan of University of Illinois and used in complementation assays. The plasmids pKD3, pKD46, and pCP20 were used for disrupting chromosomal genes in MG1655.

All plasmids used and constructed in this study are shown in Table 1. The primers used are shown in Table 2. The M. hyopneumoniae pdhD gene and Mhp-lplJ were synthesized with the TGA stop codons in the open reading frames (ORFs) replaced with TGG. The synthesized pdhD was amplified with the primer pair pA1 and inserted into pET32a between the NdeI and NotI sites to obtain the plasmid pEA001. The synthesized Mhp-lplJ was amplified with the primer pair pB1 and inserted into pET22b between the SalI and NotI sites to obtain the recombinant plasmid pEB001. The Mhp-LplJ_G77A/S and Mhp-LplJ_K131A/S mutations were generated by site-directed mutagenesis using the primer pairs MJ77A, MJ77S, MJ131A, and MJ131S, respectively.

The gene fragments of M. hyopneumoniae gcvH, M. hyopneumoniae pdhD, E. coli gcvH, E. coli sucB, and E. coli aceF were inserted into pET-HITES between the BseRI and EagI sites after amplification with primer pairs pH1–pH5 from the corresponding genomic DNAs. These plasmids, named pEH001–pEH005, were used to express apo lipoate-requiring substrates. E. coli lplA, E. coli lipB, Mhp-lplJ, and Mhp-lpl were amplified with the primer pairs pG1–pG4 and ligated into pBAD322G between the EcoRI and HindIII sites to obtain plasmids pBG001–pBG004, respectively.

All E. coli mutants were derivatives of E. coli MG1655 and were constructed by the method reported by Datsenko and Wanner (2000). The lplA gene was replaced with a chloramphenicol resistance cassette by transformation of PCR products generated with the primer pairs DElplA. The resistance gene was removed using FLP recombinase encoded by the temperature-sensitive plasmid pCP20. The genes lipB, lipA, and gcvH were deleted by transformation of PCR products generated with the primer pairs DElipB, DElipA, and DEgcvH, respectively. To construct a strain in which the aceF gene of E. coli is replaced with the wild-type pdhD gene of M. hyopneumoniae, a fragment containing the intact pdhD fused with an antibiotic cassette and the homologous region of aceF was amplified by overlap PCR with the primer pairs REaceF1 and REaceF2. All the mutant strains are described in Table 1.

Mycoplasma hyopneumoniae was cultured in mycoplasma medium (BasalMedia, China) containing 20% (v/v) swine serum for 6–7 days at 37°C. E. coli strains were grown in Luria-Bertani (LB) rich or M9 minimal medium (Miller, 1972). Antibiotics were used at the following concentrations: chloramphenicol, 40 μg/mL; ampicillin sodium, 100 μg/mL; kanamycin sulfate, 50 μg/mL; and gentamicin sulfate, 50 μg/mL. L-arabinose was used at a final concentration of 0.2%. Lipoic acid was added at a final concentration of 1 mM.

To express and purify the hexa-histidine-tagged E. coli LplA, Mhp-Lpl, Mhp-LplJ, Mhp-LplJ mutants and holo-PdhD, the recombinant plasmids pX3, pX1, pEB001-005, and pEA001 were transformed into E. coli BL21 (DE3) cells. The cells were cultured at 37°C in Luria broth. When the OD₆₀₀ reached 0.5, the cells were induced to express the target proteins with 1 mM isopropyl 1-thio-β-D-galactopyranoside (IPTG) at 25°C for 15 h. The cells were collected by centrifugation and resuspended in lysis buffer (50 mM Tris-HCl, 300 mM NaCl, 1 mM dithiothreitol and 20 mM imidazole, pH 8.0). Following sonication, crude lysates were centrifuged at 12,000 g for 30 min to remove the cell debris. The supernatants were applied to purify the recombinant proteins with a Ni-NTA-agarose affinity chromatography column according to the instructions. The purified proteins were concentrated by ultrafiltration (3- or 10-kDa cutoff), and concentrations were measured by extinction coefficients and spectrophotometric determination. Protein purity was monitored by SDS-PAGE. To obtain apo lipoate-proteins, the plasmids pEH001–pEH005 were transformed into E. coli lipoic acid auxotroph strain DE0626. Protein expression and purification were performed as described above.

Specific monoclonal antibodies against M. hyopneumoniae PdhD were produced by hybridoma technology. Briefly, 6- to 8-week-old female mice were immunized with purified protein emulsified with complete or incomplete Freund’s adjuvant. After three immunizations with an interval of 2 weeks, the mice were sacrificed to isolate spleen cells. Splenocytes were then fused with mouse SP2/0 myeloma cells using PEG 1450. The resulting hybridoma cells were plated onto ten 96-well plates and selected with hypoxanthine-aminopterin-thymidine (HAT)-conditioned medium for 2 weeks. The antibodies specific for M. hyopneumoniae PdhD in the supernatant from each well were detected by ELISA. The positive hybridoma cells were harvested and injected into the abdominal cavities of the mice to produce monoclonal antibodies (mAbs). The animal experiment was approved by the Institutional Animal Care and Use Committee of Harbin Veterinary Research Institute (IACUC#181122-02, approval date: 22 November 2018).

Conservation analysis of Mhp-LplJ amino acid residues was based on Clustal Omega (Fábio et al., 2019). Homology modeling predicting the three-dimensional (3D) structure of Mhp-LplJ was conducted using Swiss-Model (Jürgen and Torsten, 2006; Waterhouse et al., 2018). Sequence alignment was performed using ClustalX, and amino acid residues were colored by the ESPript server (Xavier and Patrice, 2014). Molecular docking analyses of Mhp-LplJ were performed using lipoic acid or lipoyl-AMP as ligands to study the mode of interaction within the binding site. The 3D structures of lipoic acid and lipoyl-AMP were obtained from the Protein Data Bank (PDB) database. PyMOL, AutoGrid, and Autodock Tools (version 1.5.6) were used during the molecular docking analysis. LigPlot⁺ was used to generate 2D ligand-protein interaction diagrams.

Lipoate ligation reactions (100 μL) contained 50 mM sodium phosphate (pH 7.0), 5 mM disodium ATP, 5 mM dithiothreitol, 1 mM MgCl₂, 1 mM lipoic acid, 20 μM apo-substrate, and 4 μM lipoate-protein ligase. After incubation at 37°C for 3 h, the lipoylation of the indicated proteins was analyzed by SDS-PAGE followed by western blot using rabbit anti-lipoic acid antibodies (Anti-LA). To analyze the active sites of Mhp-LplJ, Mhp-LplJ_G77A/S, or Mhp-LplJ_K131A/S mutants were added to the reaction as the lipoate-protein ligase. The effect of 8-BrO and 6,8-diClO on Mhp-LplJ lipoylation activity was assessed by adding analogs dissolved in dimethylsulfoxide (DMSO) at a final concentration of 1 mM. The reaction was carried out under the conditions described above using apo-PdhD as a substrate. Octanoate ligation assay was performed as lipoate ligation assay except free octanoic acid was added to reactions instead of lipoic acid. The results were analyzed by gel shift assay.

The reaction (100 μL) contained 50 mM sodium phosphate (pH 7.0), 5 mM disodium ATP, 1 mM MgCl₂, 1 mM lipoic acid, 20 μM apo-PdhD and 4 μM Mhp-LplJ. After incubation at 37°C for 3 h, the proteins in the reaction system were loaded on an 8% native polyacrylamide gel and separated by electrophoresis.

Nano-LC-LTQ-Orbitrap XL MS/MS was performed to detect lipoate modification of PdhD. Chymotrypsin and trypsin-digested peptides were separated by a C₁₈ reversed-phase column (filled with 3 μm ReproSil-Pur C₁₈-AQ from Dr. Maisch GmbH) and loaded by a C₁₈ reversed-phase column (filled with 5 μm ReproSil-Pur C₁₈-AQ from Dr. Maisch GmbH) onto the nanoLC-LTQ-Orbitrap XL system (Thermo). Data were analyzed by Proteome Discoverer (version 1.4.0.288, Thermo Fischer Scientific). The MS2 spectra were searched in the PdhD sequence plus Contaminants (cRAP) database using the SEQUEST search engine. Lipoylation of lysine and oxidation of methionine were set as variable modifications. The matching of searched peptide and MS spectra was filtered by Percolator calculation.

Mycoplasma hyopneumoniae genes were inserted into the arabinose-inducible plasmid pBAD322G and transformed into E. coli lipoic acid auxotroph strain DE0626 (Christensen and Cronan, 2010). To prevent carryover lipoic acid, all plasmid-carrying strains were grown for 36 h in M9 minimal medium containing 5 mM acetate, 5 mM succinate, 0.4% glycerol and appropriate antibiotics to bypass the lipoic acid-requiring aerobic pathways. Strains were then grown for 2 days on M9 minimal plates with or without supplementation of lipoic acid.

To analyze the lipoate modification of the indicated protein in the cells complementary with different genes, the E. coli-derived strain DE0626/RE0403 was transformed with a plasmid expressing a candidate lipoate-protein ligase (pBG001–pBG004). Transformants were grown in 5 mL Luria broth at 37°C for 16 h with or without lipoic acid. Cells were harvested by centrifugation and resuspended in 0.5 mL PBS. The total proteins of the cells were analyzed by western blot as described for the lipoate ligation assay.

Western blot assays were carried out using the following procedure. Briefly, the samples were loaded and separated on a 12% SDS-polyacrylamide gel and transferred by electrophoresis to nitrocellulose (NC) membranes (Millipore) for 30 min at 25 V. The membranes were first blocked with TBS buffer (100 mM Tris base and 0.9% NaCl, pH 7.4) containing 0.1% Tween 20 and 5% non-fat milk powder. Then, they were washed three times and probed for 1 h with Anti-LA (1:8000) or mouse anti-PdhD monoclonal antibodies (Anti-PdhD) (1:10000) for 1 h. Next, the membranes were analyzed by using Odyssey CLx Image Studio software after incubation for 1 h with an IRDye 680RD goat anti-mouse secondary antibody or IRDye 800CW goat anti-rabbit secondary antibody (1:10000).

The binding kinetics between Mhp-LplJ and lipoic acid or analogs were analyzed in real time by surface plasmon resonance (SPR) on a Biacore 8K machine with CM5 chips (GE Healthcare) at room temperature. Mhp-LplJ was diluted to 50 μg/mL with 10 mM acetate (pH = 5.0) and immobilized through a standard amine-coupling protocol with an amine coupling kit (GE Healthcare). The molecules (lipoic acid, 8-BrO and 6,8-diClO) were serially diluted using PBS-P buffer (GE Healthcare) containing 5% DMSO. The analytes passed through chip surface flow cell 1 (fc1, activate/deactivate during immobilization) and flow cell 2 (fc2, immobilized by Mhp-LplJ) at a rate of 30 μL/min. The response units (RU) were measured in real-time and are shown in the sensorgram. Association phases were monitored for 120 s, dissociation phases were monitored for 300 s, and the experiments were performed in 5% DMSO PBS-P buffer consisting of 20 mM phosphate, 2.7 mM KCl, 137 mM NaCl, 0.005% (v/v) P20 and 5% (v/v) DMSO. The results were analyzed using Biacore Insight Evaluation Software.

Mycoplasma hyopneumoniae strain J was cultured in the medium described above at 37°C and 100 rpm for 6 days. The seed liquid was diluted 100-fold into fresh medium containing an indicated final concentration of analogs dissolved in DMSO or the same volume of DMSO. After incubation at 37°C for 10 days, the copy titers of all M. hyopneumoniae samples were determined by a color changing units (CCU) assay (Stemke and Robertson, 1990; Calus et al., 2010).

PDH, OGDH, BCOADH, AoDH, and Gcs are the common lipoate-dependent enzymes in bacteria, but we failed to identify the genes encoding these enzymes except PDH and Gcs in the genome sequence of M. hyopneumoniae. To determine the lipoylated protein in M. hyopneumoniae, the total proteins of the cells from M. hyopneumoniae strain J and strain 168-L were analyzed using a western blot assay with Anti-LA. As shown in Figure 2A, a distinct band appeared in the scanned NC membrane. To identify this lipoylated protein in M. hyopneumoniae, the band in the gel was excised and analyzed by LC-MS, and the results indicate that the lipoylated protein might be the PdhD encoded by pdhD (Supplementary File S1). To further confirm this conclusion, Anti-PdhD was produced, and the PdhD protein expressed in M. hyopneumoniae J strain was analyzed using a western blot assay with both Anti-LA and Anti-PdhD simultaneously. As shown in Figure 2B, PdhD was finally verified to be the lipoylated protein in M. hyopneumoniae.

Previously, we reported that Mhp-Lpl catalyzes the attachment of lipoic acid to the lipoyl domains of GcvH in vitro (Zhu et al., 2020). We wondered whether the lipoate modification of PdhD is catalyzed by Mhp-Lpl. However, analysis showed that Mhp-Lpl failed to catalyze the lipoylation of PdhD in vitro (Figure 3A). The functions of lipoate-protein ligases among organisms are highly conserved. As E. coli LplA could catalyze lipoic acid attachment to the dihydrolipoyllysine-residue acetyltransferase component of PDH (AceF), a similar lipoate-dependent substrate as PdhD in E. coli, it is theoretically confirmed for E. coli LplA to catalyze the lipoylation of PdhD if purified PdhD retains its functional structure. Therefore, we used E. coli LplA to detect whether the purified PdhD could be lipoylated in vitro. As shown in Figure 3B, PdhD was successfully modified by E. coli LplA in vitro in the presence of lipoic acid. These results suggest that the failure of Mhp-Lpl to modify PdhD was not because of structure collapse but because Mhp-Lpl cannot recognize this substrate, and a new enzyme must exist in M. hyopneumoniae to complete the lipoylation of PdhD.

To determine the enzyme modifying PdhD, BLASTp searches against the M. hyopneumoniae strain 232 genome (accession number: NC_006360) were carried out with the amino acid sequences of E. coli LplA (UniProt code: P32099) and B. subtilis LplJ (UniProt code: O07608), which are required for lipoic acid salvage. Apart from MHP_RS01680, previously reported as Mhp-Lpl by us (Zhu et al., 2020), we found that MHP_RS00640 (renamed it as Mhp-LplJ) shared 33.9% identity with E. coli LplA and 39% identity with B. subtilis LplJ. Mhp-LplJ may be a new lipoate-protein ligase in M. hyopneumoniae. Lipoate-protein ligases mostly consist of a large N-terminal domain and a small C-terminal domain. We first model the structure of the Mhp-LplJ protein using the Swiss-Model server with LplA of Streptococcus pneumoniae (S. pneumoniae) (PDBid: 1vqz) as the modeling template, which shows the highest reliability with a global model quality estimation (GMQE) value of 0.72. As expected, the Mhp-LplJ prediction model has a typical structure of lipoate-protein ligase, consisting of a large N-terminal domain (residues 1–245) and a small C-terminal domain (residues 252–336), and the two domains are linked by a short polypeptide chain (residues 246–252) (Figure 3C). The similarity between Mhp-LplJ and other lipoate-protein ligases, Mhp-Lpl, B. subtilis LplJ and E. coli LplA protein sequences, were further analyzed. The multiple sequence alignments showed that Mhp-LplJ contained the same conserved sequence motifs as E. coli LplA and B. subtilis LplJ (Supplementary Figure S1).

To characterize the enzymatic properties and compare the lipoate ligase activities of Mhp-LplJ with Mhp-Lpl in vitro, M. hyopneumoniae Mhp-LplJ, Mhp-Lpl and E. coli LplA proteins were expressed in the E. coli BL21 (DE3) strain and purified (Supplementary Figure S2A), apo lipoate-dependent proteins of M. hyopneumoniae (GcvH and PdhD) and E. coli (GcvH, dihydrolipoyllysine-residue succinyltransferase component of OGDH SucB encoded by sucB, and AceF encoded by aceF) were expressed in strain DE0626 in which the lplA and lipB genes were deleted to shut down lipoic acid metabolism and purified (Supplementary Figure S2B). We tested the lipoate ligase activities of Mhp-LplJ and Mhp-Lpl in vitro with E. coli LplA as a positive control. Mhp-Lpl lipoylates M. hyopneumoniae GcvH as previously reported; however, it does not lipoylate PdhD and three substrates of E. coli under the same conditions (Figure 4A). By contrast, Mhp-LplJ can lipoylate M. hyopneumoniae GcvH, PdhD, E. coli SucB and AceF, but cannot lipoylate E. coli GcvH. To further verify the lipoylation of PdhD catalyzed by Mhp-LplJ in vitro, a gel shift assay and liquid chromatography tandem mass spectrometry (LC-MS/MS) were carried out. Loss of the positive lysine charge upon lipoylation resulted in the more rapid migration of PdhD on native polyacrylamide gel electrophoresis (Figure 4B). The LC-MS/MS results further confirmed that the K₄₂ of the PdhD protein is modified by lipoate attachment (Figure 4C and Supplementary Figure S3). These results indicated that Mhp-LplJ is a new lipoate-protein ligase responsible for the lipoate modification of PdhD in M. hyopneumoniae.

As there is no technique for knocking out genes in M. hyopneumoniae so far, complementation analysis could not be carried out in M. hyopneumoniae. To analyze the functions of Mhp-LplJ in vivo, we tested its ability to restore growth of the E. coli ΔlplA ΔlipB strain DE0626. In this strain, the lipoic acid salvage pathway and synthesis pathway were blocked, it was able to grow on M9 minimal agar plates only when the synthesis pathway is completed, or the salvage pathway is rebuilt with a plasmid encoding a lipoate-protein ligase at the presence of lipoic acid. As shown in Figure 5A, upon compensation with Mhp-LplJ, the model strain grew well in M9 minimal plate when lipoic acid was added. Mhp-Lpl, as a control group, could not restore the growth of strain DE0626 under the same conditions. These results indicated that Mhp-LplJ can catalyze the salvage pathway of lipoic acid and has a different function from Mhp-Lpl in vivo. Western blot assays further indicated that SucB and AceF of the strains that could grow on M9 plates are lipoylated (Figure 5B).

To further confirm the lipoate ligase activity of Mhp-LplJ on PdhD in vivo, the homologous aceF gene in the DE0626 genome was replaced with pdhD of M. hyopneumoniae. The strain was transformed with pBAD322G-derived plasmids, and the lipoylated proteins were analyzed using western blot assay. As shown in Figure 6A, Mhp-LplJ can catalyze the lipoylation of PdhD and SucB in the presence of lipoic acid. However, to our surprise, in the absence of lipoic acid, PdhD can also be lipoylated by Mhp-LplJ, whereas SucB cannot. To determine the proteins involved in the lipoylation of PdhD without free lipoic acid, we disrupted lipA and gcvH, which function in the lipoate synthesis pathway of other species (Miller et al., 2000; Christensen et al., 2011b), in the genome of RE0403 strain to generate strains RE1011 and RE1009, respectively. As shown in Figure 6B, lipA gene deletion prevented the lipoylation of PdhD in strain RE1011, while gcvH removal did not differ in strain RE1009. These results suggest that lipoyl synthase LipA cooperates with Mhp-LplJ to complete the lipoylation of PdhD in the absence of lipoic acid, while GcvH does not play any role in this pathway. Referring to the octanoate ligase activity of E. coli LplA (Hermes and Cronan, 2009), we hypothesized that Mhp-LplJ first scavenged free cytosolic octanoic acid from E. coli, and then LipA catalyzed the insertion of two sulfur atoms into the C-6 and C-8 positions of the octanoyl moiety bound to PdhD, converting the octanoyl moiety into lipoylated derivatives. We further verified the octanoate ligase activity of Mhp-LplJ by octanoate ligation assay in vitro. As shown in Supplementary Figure S4, attachment of octanoic acid to PdhD results in loss of a positive charge, which causes the modified PdhD to migrate more rapidly in gel shift assay. This result provided convincing evidence that Mhp-LplJ could able to catalyze the attachment of free octanoic acid to PdhD.

To identify the key residues of Mhp-LplJ, computational molecular docking was used to characterize the favored binding models of Mhp-LplJ with lipoic acid or lipoyl-AMP (Supplementary Figures S5A,B). The alignment of amino acid sequences showed that the residues of Mhp-LplJ forming the lipoic acid and lipoyl-AMP pockets are conserved with E. coli LplA and B. subtilis LplJ. We further identified the key residues of Mhp-LplJ using site-directed mutagenesis according to the favored binding models of Mhp-LplJ with lipoic acid or lipoyl-AMP, and the results implied that G⁷⁷ and K¹³¹ are the key amino residues for Mhp-LplJ activity (Supplementary Figure S5C).

Biotin (hexahydro-2-oxo-1H-thieno[3,4-d]imidazole-4-penta- noic acid, vitamin B₇, vitamin H) and lipoic acid share many similarities in properties, metabolism and function. Biotin protein ligase (BPL) was proven to be a new drug target. Some biotin analogs have potential as antibacterial agents by inhibiting BPLs from Staphylococcus aureus (S. aureus), E. coli, Mycobacterium tuberculosis (M. tuberculosis), and Homo sapiens. To investigate whether lipoic acid analogs affect the function of Mhp-LplJ and the growth of M. hyopneumoniae, analogs 8-BrO and 6,8-diClO were selected for the tests. First, the binding characteristics of lipoic acid, 8-BrO and 6,8-diClO to Mhp-LplJ were analyzed by SPR. As shown in Figure 7, the sensorgrams of lipoic acid and 6,8-diClO are fitted to a single-site kinetics model, while the sensorgram of 8-BrO is suitable for the single-site affinity model. 8-BrO and lipoic acid exhibit similar equilibrium dissociation constants (K_Ds) of 29.80 ± 3.62 μM and 24.14 ± 1.61 μM, respectively, whereas 8-BrO displays a quick association and disassociation pattern for Mhp-LplJ. 6,8-diClO shows a lower binding affinity to Mhp-LplJ, and the K_D value is one order of magnitude higher than that of lipoic acid (Figure 7B). Then, we determined the effects of these two lipoic acid analogs on Mhp-LplJ activity in vitro. Consistent with the measured binding affinity, 8-BrO with higher affinity has a more obvious interference effect on the lipoate ligase activity of Mhp-LplJ than 6,8-diClO (Figure 7D).

The attachment of lipoic acid is necessary for activation of PDH, and the activity of PDH affects the energy metabolism of microorganisms. We further investigated whether lipoic acid analogs that could disrupt lipoate ligase activity of Mhp-LplJ affect the growth of M. hyopneumoniae. M. hyopneumoniae was grown in mycoplasma medium supplemented with analogs at different concentrations. After culturing for an indicated time, the color of cultures presented a gradient change along the analog concentration (Figure 8A). The numbers of viable M. hyopneumoniae showed a negative correlation with the concentrations of lipoate analogs when determined with the CCU assay. As shown in Figure 8B, the lipoate analogs inhibit M. hyopneumoniae growth in a dose-dependent manner. Compared with 6,8-diClO, 8-BrO has a more obvious inhibitory effect. When the concentration of 8-BrO increased from 10 to 50 μM, the number of viable bacteria decreased sharply, and the inhibition rate increased from 28.57 to 70%. To verify that the growth inhibition of M. hyopneumoniae by the two analogs is correlated with the function disruption of Mhp-LplJ in vivo, we tested the lipoylation level of PdhD in these cultures. As shown in Figures 8C,D, the degree of lipoylation of PdhD was decreased with increasing analog concentration, but what puzzled us was that a certain number of different size bands appeared when 8-BrO was added. We also performed an inhibitory analysis of the two lipoate analogs in the E. coli strain RE1011. Similarly to the results found in M. hyopneumoniae, both analogs disrupted the enzymatic activity of Mhp-LplJ. However, multiple bands of PdhD also appeared when 8-BrO was added to the strain RE1011 (Supplementary Figure S6). To explain the multiple bands of PdhD, we further explored it through an in vitro lipoate ligation assay. As shown in Supplementary Figure S7, the PdhD bands with high-molecular weight cannot be probed when 8-BrO or Mhp-LplJ exists alone, but can only be detected when they are added to the reaction at the same time. Based on the above results, we speculate that the PdhD bands with high-molecular weight are due to the nucleophilic reactions of bromoalkanes, which leads to crosslinking of two proteins in the process of 8-BrO interfering with PdhD lipoylation catalyzed by Mhp-LplJ. The PdhD bands with high-molecular weight were analyzed by LC-MS and the results showed that both PdhD and Mhp-LplJ were present in these bands (data not shown).

The above results strongly suggest that the activity of Mhp-LplJ could be disrupted by lipoic acid analogs, and disruption of Mhp-LplJ function can inhibit the in vitro growth of M. hyopneumoniae.