Bacterial strains used in this study are listed in Supplementary Table S1. M. bovis HB0801 (GenBank sequence CP002058.1; NCBI Reference Sequence NC_018077.1) was grown in pleuropneumonia-like organism (PPLO) medium (BD Company, Sparks, MD, United States) with or without 100 μg/mL gentamicin, as previously described (Han et al., 2015). Mycoplasma titers were determined based on colony counts after 2 to 5 days of incubation at 37°C (Zhu et al., 2020). Escherichia coli DH5α and BL21 were grown in Luria Bertani (LB) broth medium with appropriate antibiotics.

Since mycoplasmas use UGA as a tryptophan codon, the MbovGdpP nucleotide sequence was modified by converting UGA codons into UGG to avoid premature translation stops in E. coli. The modified MbovGdpP nucleotide sequence was synthesized by the Beijing Tianyi Huiyuan Bioscience & Technology Inc. The synthetic gene was digested by NcoI and XhoI (Takara, Dalian, China), and then was cloned into the plasmid vector pET28b (+) using T4 DNA Ligase (Takara, Dalian, China) to generate plasmid pMbovGdpP (Supplementary Table S1). DNA constructions were validated by DNA sequencing before transformation into E. coli BL21 for protein expression. Recombinant proteins were purified with the nickel affinity chromatography method using 1 L of E. coli cultures (Supplementary Table S1) induced by 0.8 mM b-D-1-thiogalactopyranoside (IPTG) for 20 h at 16°C. Purified recombinant proteins were analyzed by SDS-PAGE, and their concentration was determined by using the BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, United States). Purified proteins were stored at −80°C.

The ssDNA was synthesized by the Wuhan TsingKe Biological Technology Inc. with the sequence listed in Supplementary Table S1. For ssDNase assay, 5 μM recombinant proteins were incubated with ssDNA at 37°C overnight in 20 mM Tris-HCl (pH 7.0) containing 2.5 mM MnCl₂. The reaction was stopped by boiling for 10 min. After centrifuging at 20,800 × g, samples were analyzed by electrophoresis on 12% polyacrylamide gels and stained with ethidium bromide before visualization with an imaging system (ProteinSimple, Santa Clara, CA, United States). For enzymatic degradation efficiency assays, 50 μM c-di-AMP, 50 μM pApA, and 10 μM ssDNA were used as substrates to measure PDE, nanoRNase, and ssDNase activities, respectively. The reaction mixtures were incubated from 0 to 240 min at 37°C and analyzed by HPLC, as previously described. Briefly, the soluble components were separated on an RP C18 column (250 × 4.6 mm id, 5 μm; Thermo Fisher Scientific, Waltham, MA, United States). The mobile phase consisted of 90% phosphate buffer (30 mM K₂HPO₄, 20 mM KH₂PO₄; pH 6.0) and 10% methanol at a flow rate of 1 mL/min. The wavelength of the UV detector was set up at 254 nm and the injection volume of the autosampler was 10 μL. For ssDNase activity, the incubation time was extended to 480 min, and reaction mixtures were analyzed by polyacrylamide gels. The conversion rate c-di-AMP was calculated by converting the corresponding HPLC peak regions to concentrations according to the standard curve.

To test the ability of ppGpp in inhibiting the MbovGdpP PDE activity, the reaction mixture containing 50 μM c-di-AMP was incubated with increasing concentrations of ppGpp (0 to 200 μM; Biolog, Bremen, Germany). The reaction products were analyzed by HPLC.

Tolerance to osmotic stress was carried out in axenic conditions. Briefly, M. bovis (10⁴ CFUs) was grown in PPLO medium containing increasing concentrations of KCl or NaCl (0 to 250 mM; Sinopharm, Shanghai, China). Mycoplasma titers were determined each 12 h incubation at 37°C.

Mycoplasma cells were harvested from 3 mL of PPLO cultures by centrifugation at 12,000 × g for 5 min. Total RNA was isolated from M. bovis pellets using the TRIzol reagent (Invitrogen Corporation, Carlsbad, CA, United States) (Flores-Valdez et al., 2018). RNA was stored at −80°C. Quantitative RT-PCR was performed as described (Zhu et al., 2021).

The expression profiles of differentially expressed genes of M. bovis were determined by RNA-sequencing (RNA-seq). The cDNA libraries were constructed by using TruSeq Stranded Total RNA with Ribo-Zero Gold (Illumina, San Diego, CA, United States) according to the manufacturer’s instructions. The quality and purity of RNA samples were determined with an Agilent Technologies 2100 Bioanalyzer (Agilent, Santa Clara, CA, United States). The RNA-seq library was sequenced on an Illumina HiSeq X10 sequencer (Illumina, San Diego, CA, United States) to generate paired-end (2 × 150 bp) reads. Raw reads generated during high-throughput sequencing were fastq format sequences. High-quality clean reads were generated by using Trimmomatic software to remove the linker and filtered out low-quality bases, N-bases or low-quality reads. Rockhooper2 was used to align clean reads to the M. bovis HB0801 genome (NCBI Reference Sequence NC_018077.1). Both M. bovis wild-type strain and mutant were set up for 3 biological replicates. Differentially expressed genes (DEGs) were defined by 1.5-fold change with a p-value <0.05, a commonly used cut-off value in transcriptomoic studies (Pletzer et al., 2020; Brochado et al., 2021; Tremblay et al., 2021). RNA-seq data (raw fastq files and read counts) are deposited in the Gene Expression Omnibus (GEO) repository under accession number GSE233141.

Protein sequences alignments between RecJ and RecJ-like proteins were performed using National Center for Biotechnology Information (NCBI) Cobalt multiple alignment tool¹ combined with ESPript 3.0 software. ²

Statistical analyses were performed with SPSS software (SPSS, Inc., Chicago, IL, United States). The unpaired student’s t-test was used for the comparisons of two groups, while one-way ANOVA was used for multiple comparison. The differences were considered to be significant when p-value was lower than 0.05.

Unlike typical members of the GdpP family, such as B. subtilis GdpP, MbovGdpP is only characterized by a GGDEF-like motif and a DHH-DHHA1 domain, but no PAS motif (Figure 1). Yet, sequence comparison with other members of the DHH superfamily revealed similarity with prokaryotic subfamily I members including cyclic nucleotide phosphodiesterase, nanoRNases, and exonuclease RecJ (Srivastav et al., 2019; Wang et al., 2020). In particular, MbovGdpP sequence alignment with RecJ and RecJ-like proteins revealed an important level of similarity (Supplementary Figure S1). This led us to test the ssDNase activity of MbovGdpP with a synthetic oligonucleotide (Supplementary Table S1). The MbovGdpP_120–666 and MbovGdpP_158–666 were successfully expressed in E. coli, and the molecular weight of these proteins were 61.8 kDa and 57.0 kDa, respectively (Supplementary Figure S2). Polyacrylamide gel electrophoresis (PAGE) revealed a progressive degradation of the ssDNA concentration upon incubation with a recombinant MbovGdpP having the N-terminal transmembrane domain deleted amino acid residues 0 to 119 (MbovGdpP_120–666) (Zhu et al., 2020), but not with the truncated MbovGdpP_158–666 having the N-terminal region deleted up to amino acid 157 (Zhu et al., 2020) (Figures 1A–C). The catalytic efficiency of MbovGdpP on ssDNA was estimated to be less than 0.02 μM L⁻¹ min⁻¹, with a nearly complete degradation of ssDNA within 480 min. These data suggest that MbovGdpP has ssDNase activity and that amino-acid residues 120–158 are essential for MbovGdpP ssDNase activity.

To further characterize MbovGdpP enzymatic activity, we have analyzed the degradation efficiency of c-di-AMP and pApA upon incubation with the recombinant protein MbovGdpP. Quantitative analyses revealed a complete degradation of c-di-AMP and pApA within 5 min and 120 min, respectively (Figure 1D). The catalytic rate of MbovGdpP on c-di-AMP and pApA was nearly 10 μM L⁻¹ min⁻¹ and 0.42 μM L⁻¹ min⁻¹, respectively. These results suggest that c-di-AMP and pApA may be the preferential substrates of MbovGdpP when compared to ssDNA.

To test the capacity of ppGpp to inhibit MbovGdpP enzymatic activities, c-di-AMP degradation was analyzed following complete and incomplete reaction conditions. Data revealed that c-di-AMP degradation into AMP includes the formation of a pApA intermediate (Figures 2A,B). Remarkably, HPLC analysis only identified the accumulation of the pApA intermediate product upon c-di-AMP degradation in the presence of ppGpp, even after overnight incubation (Figures 2C–F). This result indicates that ppGpp is only able to inhibit the nanoRNase activity of MbovGdpP, but not its PDE activity. Altogether, these data highlight the unique enzymatic properties of MbovGdpP.

Deletion of GdpP can alter bacterial resistance to osmotic stress (Pham et al., 2018; Teh et al., 2019). In the present study, M. bovis resistance to salt stress was tested by using the MbovGdpP knock-out mutant T6.290 (Zhu et al., 2020). When compare to wild-type strain HB0801, T6.290 growths in PPLO medium containing increasing concentrations of KCl/NaCl was found to be affected by KCl concentrations higher or equal to 50 mM, but not by NaCl (Figure 3). This result indicates that MbovGdpP may play important role in M. bovis resistance to potassium (K⁺) stress.

To better understand the role of MbovGdpP in cellular processes, RNA-seq was used to determine the transcriptional profile of T6.290 and HB0801 grown to the stationary phase. Differential gene expression analysis identified up to 161 genes with significant changes in T6.290, with 74 mRNA up-regulated and 87 down-regulated (Figure 4A and Supplementary Table S2). The accuracy of RNA-seq was further validated by qRT-PCR. Upregulated (n = 8) and down-regulated (n = 10) genes were selected among the most significant differentially expressed genes (DEGs) in T6.290. The qRT-PCR analysis confirmed changes in mRNA levels for 17 out of the 18 genes selected, including 9 up-regulated genes Mbov_0022 (deoxyguanosine kinase), Mbov_0023 (deoxyguanosine kinase), Mbov_0147 (hypothetical protein), Mbov_0215 (hypothetical protein), Mbov_0426 (glycine cleavage system protein H), Mbov_0476 (PTS sugar transporter subunit IIA), Mbov_0523 (DNA-binding protein WhiA), Mbov_0709 (DNA adenine methylase), Mbov_0725 (Cof-type HAD-IIB family hydrolase), and 8 down-regulated genes Mbov_0049 (hypothetical protein), Mbov_0242 (S8 family peptidase), Mbov_0277 (50S ribosomal protein L9), Mbov_0278 (replicative DNA helicase), Mbov_0279 (DUF21 domain-containing protein), Mbov_0292 (variable surface lipoprotein), Mbov_0395 (hypothetical protein), Mbov_0639 (hypothetical protein) (Figure 4B).

Interestingly, tRNA accounted for over 26% (23/87) of the down-regulated genes (Table 1), representing 67% of the total number of tRNA genes (23/34). The tRNA Arg, Gly, Ile, Leu, Lys, Met, Ser, Thr, and Trp are encoded by more than one gene. Among them, 7 tRNA genes (tRNA Arg, Leu, Lys, Met, Ser, Thr, and Trp) were down-regulated, while tRNA Gly and Ile remained unchanged. The remaining tRNAs (n = 9) were encoded by one gene, among them, six were differentially expressed in T6.290. These results indicate that MbovGdpP plays an important role in tRNA biosynthesis which may further influence gene expression in M. bovis.

Up to 57 DEGs in T6.290 were of unknown functions. GO enrichment was carried out to analyze the functions of the 81 remaining DEGs in T6.290. According to GO annotation, DEGs significantly enriched were involved in biological process (cell septum assembly), and molecular function (nucleoside kinase activity and nucleotidyltransferase activity) (Figure 5). KEGG pathway analysis revealed many DEGs participating in genetic information processing including translation, replication and repair (Figure 6A) and a vast group of DEGs participating in metabolism such as nucleotide, amino acid, cofactors and vitamins, as well as carbohydrate metabolism (Figure 6A). Up to 7 DEGs were involved in central metabolism including upregulated genes Mbov_0150 (Phosphate acetyltransferase, Pta), Mbov_0151 (Acetate kinase, AckA) and Mbov_0312 (Alcohol dehydrogenase, Adh), as well as down-regulated genes Mbov_0155 (Pyruvate kinase, Pk), Mbov_0160 (Lactate dehydrogenase, LdhA), Mbov_0286 (NADH oxidase (NOXASE), HcaD) and Mbov_0338 (Alcohol dehydrogenase, Adh) (Figure 6B). Among them, 4 genes (Mbov_0150, Mbov_0151, Mbov_0155 and Mbov_0160) are involved in pyruvate metabolism (Figure 6C). Indeed, genes involved in acetate production were up-regulated in the mutant, while genes that participate in the production of lactate were down-regulated. The Mbov_0286 product is the NADH oxidase (NOXASE) which is involved in the H₂O₂ production. These data illustrate the broad transcriptomic response of mycoplasma cells to the loss of GpdP activity and indirectly the central regulatory role played by GpdP and its secondary messenger substrates in the biology of M. bovis.