Mycoplasma pneumoniae was handled under L2 laboratory safety conditions. The M. pneumoniae strain used in this study was M. pneumoniae M129 (ATCC 29342). M. pneumoniae was grown at 37°C in 175 cm² tissue culture flasks containing 100 ml of modified Hayflick medium as described previously (Halbedel et al., 2004). Carbon sources were added to a final concentration of 1% (w/v). Escherichia coli XL1blue and BL21(DE3)/pLysS (Sambrook et al., 1989) were used as host for cloning and recombinant protein expression, respectively. E. coli strains were cultivated in Luria Bertani broth (Sambrook et al., 1989) containing ampicillin 100 μg ml^-1. Bacillus subtilis 168 was used to extract chromosomal DNA.

Chromosomal DNA from B. subtilis and M. pneumoniae was isolated using the Blood and Tissue Kit according to the manufacturer’s instructions (Qiagen, Hilden, Germany). Transformation of E. coli and plasmid DNA extraction was performed using standard procedures (Sambrook et al., 1989).

Plasmids for the overexpression and purification of the enzymes potentially involved in c-di-AMP metabolism from M. pneumoniae were constructed as follows. The coding sequence of each gene was amplified by PCR with gene specific primers (listed in Supplementary Table S1) using chromosomal DNA of M. pneumoniae M129 as the template. As the M. pneumoniae genes contain TGA codons that code for tryptophan in M. pneumoniae but for an opal stop codon in E. coli, the PCR fragments were used as templates for mutagenesis by the multiple mutation reaction (Hames et al., 2005) using the amplification primers and 5′-phosphorylated mutagenic primers (listed in Supplementary Table S1) to introduce TGA to TGG transitions. The cytoplasmic portion of the B. subtilis gdpP gene (GdpP_84-659) was amplified from B. subtilis chromosomal DNA. The PCR products of genes potentially encoding phosphodiesterases were digested with BamHI and NdeI and cloned into the expression vector pGP574 (Schilling et al., 2006). This plasmid allows the expression of enzymes carrying a C-terminal Strep-tag. The resulting plasmids are pGP2717 (mpn140), pGP2718 (mpn549), and pGP2720 (gdpP). The PCR product for the potential diadenylate cyclase-encoding gene mpn244 was cleaved with NdeI and BamHI and cloned into the vector pET3c (Novagene, Darmstadt, Germany). The resulting plasmid was pGP2036. All plasmid inserts were verified by DNA sequencing.

The Strep-tagged proteins were overexpressed in E. coli BL21(DE3). Expression (1 l culture, 37°C, 200 rpm, baffled flasks) was induced by the addition of 1 mM IPTG to exponentially growing cultures (OD₆₀₀ of 0.6 to 0.8). After expression for 3 hours, the cells were pelleted at 4°C for 20 min at 4,000 rpm and washed once with 20 ml PDE buffer (100 mM Tris-HCl pH8.3, 150 mM NaCl, 1 mM EDTA, 5 % glycerol). Cells were lysed using a French press (18,000 p.s.i., 138,000 kPa, three passes, SLM Aminco, United States). After lysis, the crude extracts were centrifuged at 35 000 r.p.m for 30 min. The crude extract was then passed over a StrepTactin column (IBA, Göttingen, Germany). The recombinant proteins were eluted with desthiobiotin (IBA, final concentration 2.5 mM). After elution, the fractions were tested for the desired protein using 12% SDS-PAGE. Only fractions which contained the desired protein in apparent homogeneity (content of the specific protein >95%) were used for further purification by running the proteins over a 10 ml HiTrap Heparin HP column (GE Healthcare) using the ÄKTA prime plus system (GE Healthcare, flow rate 2 ml/min). With 1 M NaCl in the PDE buffer, proteins were eluted from the column. Fractions containing pure protein were concentrated and dialyzed to 1.5 ml without NaCl in PDE buffer using Vivaspin^®Turbo15 ultrafiltration spin columns (MW 5,000 Da; Sartorius, Göttingen, Germany). Protein concentrations were determined using the Bio-Rad dye-binding assay where bovine serum albumin served as the standard. Aliquots of the proteins were frozen in liquid nitrogen and stored at –80°C prior to further experiments.

The assays for phosphatase and phosphodiesterase activities of Strep-tagged proteins were performed as described previously (Diethmaier et al., 2014). Phosphatase activity against para-nitrophenol phosphate (pNPP) was assayed in a buffer containing 100 mM Tris-HCl pH8.3, 10 mM NaCl, 0.1 mM MnCl₂, 25 mM pNPP and different amounts of purified protein in a total volume of 100 μl. Phosphodiesterase activity against bis-pNPP was assayed in a buffer containing 100 mM Tris-HCl pH8.3, 10 mM NaCl, 0.1 mM MnCl₂, or MgCl₂, 1–7.5 mM bis-pNPP and purified protein in a total volume of 100 μl. The reactions were initiated by the addition of the protein, and the reaction mixture was incubated at 37°C for 4 h. Relative substrate cleavage was quantified by measuring the OD₄₁₀ using a microplate reader (EPOCH| 2, BioTek, Winooski, United States).

Phosphodiesterase activity with cyclic dinucleotides was measured using a quantitative assay based on the interaction of c-di-AMP with the fluorescent dye coralyne (Zhou et al., 2014). Briefly, reaction mixtures (150 μl) consisted of 100 mM Tris-HCl pH8.3, 10 mM NaCl, 0.1 mM MnCl₂, and 100 μM c-di-AMP. The reaction was initiated by addition 100 nM enzyme (for MPN549 additionally 50 and 10 nM), and the reaction mixture was incubated at 37°C for 4 h. At different time points the reactions were stopped by addition of 1 μl EDTA (0.5 M), shock-freezing and subsequent boiling for 10 min. To quantify the remaining c-di-AMP, the samples were spun to remove protein precipitate. KBr and coralyne were added to the sample to final concentrations of 250 mM and 10 μM, respectively. After 20 min of incubation at 37°C in the dark, relative c-di-AMP cleavage was determined as described (Zhou et al., 2014) by measuring the fluorescence emission at 475 nm using a microplate reader (SYNERGY Mx, BioTek, Winooski, United States).

In all kinetic experiments, K_m and V_max were determined by nonlinear curve fitting from Lineweaver-Burk plots.

The concentrations of c-di-AMP in E. coli and in M. pneumoniae cells were determined by a liquid chromatography tandem MS (LC-MS/MS) method, essentially as described previously (Mehne et al., 2013; Gundlach et al., 2015b). Briefly, E. coli or M. pneumoniae cells were grown in LB (20 ml) or MP (100 ml) medium, respectively. For E. coli two additional aliquots (1 ml each) of each sample were harvested for total protein determination. For M. pneumoniae, the wet weight of each sample was determined. The pellets were resuspended in 300 μl extraction mixture (acetonitrile/methanol/water 40/40/20 v/v/v) and shock-frozen in liquid nitrogen, followed by boiling for 10 min. The boiled samples were centrifuged for 10 min at 4°C and 20,800 g. The supernatants were stored on ice, and the remaining pellets were used for two more extraction steps with 200 μl extraction mixture. For this purpose, the samples were mixed, incubated on ice for 15 min, and centrifuged again. The obtained supernatants were pooled. The samples were incubated at –20°C over night, and centrifuged again (20 min, 4°C, 20,800 g). Then, the supernatant was transferred to a fresh reaction tube, and dried in a speed vac at 50°C for 2 h and resuspended in 200 μl of water. After repeated centrifugation and addition of the internal standard ([¹³C,¹⁵N]c-di-AMP), part of the extract was analyzed by LC-MS/MS.

Chromatographic separation was performed on a series 200 high-pressure liquid chromatography (HPLC) system (PerkinElmer Life Sciences) as described previously (Mehne et al., 2013). The analyte detection was performed on an PI4000 triple-quadrupole mass spectrometer equipped with an electrospray ionization source (AB Sciex), using selected reaction monitoring (SRM) analysis in positive ionization mode. The SRM transitions labeled “quantifier” were used to quantify the compound of interest, whereas “identifier” SRM transitions were monitored as confirmatory signals. The quantifier SRM transitions were most intense and were therefore used for quantification.

To isolate c-di-AMP-binding proteins from M. pneumoniae, we made use of immobilized c-di-AMP (2′-O-(6-aminohexylcarbamoyl)-cyclic diadenosine monophosphate on agarose (2′-AHC-c-di-AMP-agarose; BIOLOG, Bremen, Germany). We prepared M. pneumoniae crude cell extracts from 150 mg cells per assay. The cells were lysed in pulldown buffer (100 mM Tris-HCl pH7.5, 100 mM KCl, 150 mM NaCl, 5 mM MgCl₂, 0.5 mM DTT, 0.1% (v/v) Tween-20, and 1x cOmplete^TM protease inhibitor (Roche Diagnostics, Mannheim, Germany)). Cells were lysed in a tissue lyser with 0.1 mm glass beads (2 × 2.5 min, 30 Hz, cooled block) followed by ultracentrifugation for 30 min at 35,000 rpm and 4°C. To prepare the matrix, 500 μl 2′-AHC-c-di-AMP agarose or ethanolamine-coupled agarose (negative control) were equilibrated three times with 1 ml of pulldown buffer. After equilibration, the cell lysate and the agarose matrix were incubated rotating over night at 4°C. The supernatant was discarded and the matrix was washed three times (1 ml 1x PBS; 137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.76 mM KH₂PO₄, pH7.4). Elution of binding proteins was performed with 200 μl of 1x PBS, pH7.4 containing 1 mM c-di-AMP and 15 min incubation at RT and 1,400 rpm. The eluted proteins were concentrated using acetone precipitation (1:1 sample:acetone) over night at –20°C. The precipitate was centrifuged 1 h at 14,800 rpm and 4°C. The remaining supernatant was discarded and the sample dried for 10 min at RT. The remaining pellet was solved in 40 μl 1x PBS and analyzed by SDS-PAGE. Proteins of interest were identified by mass spectrometry as described previously (Meyer et al., 2011). Briefly, after destaining of gel slices, proteins underwent in-gel digestion with trypsin (10 ng/ μl trypsin in 20 mM ammonium bicarbonate). Peptides were separated by C18 reverse phase liquid chromatography (nanoAcquity UPLC system, 10 cm, Waters, Manchester, United Kingdom) in a linear gradient of 0.1 % acetic acid in acetonitrile from 5 % up to 25 % within 65 min (flow rate: 400 nl/min). MS analysis was performed on a LTQ-Orbitrap Velos hybrid mass spectrometer (Thermo Electron, Bremen, Germany) operated in data-dependent MS/MS mode. Proteins were identified by searching all MS/MS spectra against a M. pneumoniae M129 protein database (687 entries, extracted from Uniprot rel. 05-2014) using Sequest algorithm on a SorcererTM software platform. Initial mass tolerance for peptide identification on MS and MS/MS peaks were 10 ppm and 1 Da, respectively. Up to two missed tryptic cleavages were allowed. Methionine oxidation (+15.99492 Da) and propionamide modification on cysteine (+71.037109 Da) were set as variable modifications. Protein identification results were evaluated by determination of probability for peptide and protein assignments provided by PeptideProphet and ProteinProphet (ISI Seattle, WA, United States) incorporated in the Scaffold software package rel. 4.3.2 (Proteome Software, Portland, OR, United States). Proteins were identified by at least two peptides with peptide probability >90% reflecting protein probability of >99%.

For the isolation of M. pneumoniae mutants, we used an ordered collection of strains carrying insertions of transposon Tn4001 (Halbedel et al., 2006). The presence of the desired mutant was assayed by a PCR screen using one primer that hybridizes to the transposon (directed outward), and a second primer specific for the gene of interest (see Supplementary Table S1).

All Firmicutes studied so far produce c-di-AMP, and in most of them, this second messenger is essential for the growth of these bacteria (Commichau et al., 2015). However, the presence of this signaling nucleotide has never been analyzed in representatives of the Mollicutes. To address this question for M. pneumoniae, we cultivated cells, extracted nucleotides, and analyzed these extracts for the presence of c-di-AMP. The total amount of c-di-AMP (nM) extracted from the pellet was divided by the total cell volume (mg) to yield the c-di-AMP concentration per mg cells. In three independent biological experiments (with three technical replicates each), we determined concentrations of 489 ± 107 nM, 958 ± 153 nM, and 859 ± 81 nM. Such differences among biological replicates are not unusual when assaying c-di-AMP concentrations (Gundlach et al., 2015b). Importantly, our results demonstrate that M. pneumoniae does possess c-di-AMP, and imply the existence of enzymes that synthesize and degrade this second messenger.

To identify candidate genes involved in M. pneumoniae c-di-AMP metabolism, we analyzed the genome for genes encoding proteins possessing a c-di-AMP synthesizing DAC domain as well as for proteins containing c-di-AMP degrading DHH-DHHA1 and HD domains. A single protein containing a DAC domain was found. The DAC domain of this protein, MPN244, is most similar to the domain of the B. subtilis sporulation-specific cyclase CdaS. However, the N-terminal domain of this protein is not similar to any domains found in other characterized diadenylate cyclases. According to the UniProt database, the amino acids 6 to 26 form a transmembrane helix. Thus, if MPN244 is endowed with diadenylate cyclase activity, this protein would represent a novel class of diadenylate cyclases. A comparison of the domain arrangement of the different classes of diadenylate cyclases is shown in Figure 1A.

An analysis of potential c-di-AMP degrading phosphodiesterases in M. pneumoniae failed to identify a protein containing a HD domain similar to that of the B. subtilis phosphodiesterase PgpH (Huynh et al., 2015). In contrast, the search for proteins containing a DHH–DHHA1 domain resulted in the identification of two proteins, MPN140 and MPN549. These proteins have been annotated as nano-RNase (NrnA) and probable single-stranded-DNA-specific exonuclease (RecJ) (Postic et al., 2012; Wodke et al., 2015), respectively. Two classes of c-di-AMP specific DHH-DHHA1 phosphodiesterases have been identified: a class of large membrane-bound proteins, as exemplified by B. subtilis GdpP. This protein contains an N-terminal transmembrane domain, a PAS domain, and a degenerate GGDEF domain possibly involved in c-di-GMP sensing in addition to the DHH-DHHA1 domain. In contrast, the second class, represented by Pde2 of Streptococcus pneumoniae, consists only of the DHH-DHHA1 domain (Bai et al., 2013). Both potential phosphodiesterases of M. pneumoniae are similar to Pde2 rather than to GdpP (see Figure 1B for domain organization).

In order to test whether MPN244 does indeed exhibit diadenylate cyclase activity, we took advantage of the inability of E. coli to produce c-di-AMP (Corrigan et al., 2011). The mpn244 gene was cloned into the expression vector pET3c, and the amounts of c-di-AMP produced in E. coli BL21(DE3) carrying the corresponding plasmid pGP2036, were compared to those of the same strain carrying the empty vector pET3c. As observed previously (Mehne et al., 2013), no c-di-AMP was present in the strain with the empty vector. In contrast, the expression of MPN244 resulted in the detection of 47.3 ± 11.8 pmol of c-di-AMP/mg of protein. This result provides unequivocal evidence for the diadenylate cyclase activity of MPN244, and we renamed the protein accordingly CdaM (c-di-AMP synthase of Mycoplasma), and the gene cdaM.

As c-di-AMP is essential in many Gram-positive bacteria, we wondered whether this is also the case in M. pneumoniae. To address this question, we attempted the isolation of a cdaM mutant from our transposon mutant library. This library contains 2,976 independent individual transposon mutants, and the probability to find mutants for any non-essential gene is as high as 99.999% (Halbedel et al., 2006). In agreement with previous results (Lluch-Senar et al., 2015), no mutant could be isolated suggesting that the cdaM gene and thus c-di-AMP is essential in M. pneumoniae.

As mentioned above, two potential c-di-AMP degrading phosphodiesterases, MPN140 and MPN549, are encoded in the genome of M. pneumoniae. To assess their activity, we expressed and purified these enzymes carrying a C-terminal Strep-tag. Initially, we determined the phosphatase and phosphodiesterase activities of both enzymes. To determine phosphatase activities, we used pNPP, a synthetic phosphatase substrate, and the M. pneumoniae protein phosphatase PrpC (Halbedel et al., 2006; Schmidl et al., 2010) as a positive control. While PrpC exhibited a strong phosphatase activity, only negligible activity was detected for MPN140 and MPN549 (data not shown). Phosphodiesterase activity was assayed against bis-pNPP as the substrate, and the B. subtilis phosphodiesterase GdpP served as the control. As shown in Table 1, all three enzymes exhibited phosphodiesterase activity against this non-specific substrate. Thus, MPN140 and MPN549 are specifically active against phosphodiesters. In order to explore the activity of these enzymes against c-di-AMP, we determined c-di-AMP degradation using binding of c-di-AMP to the fluorescence dye coralyne as an assay. As shown in Figure 2, B. subtilis GdpP was capable of degrading the nucleotide. For MPN140, no activity against c-di-AMP was observed. In contrast, MPN549 exhibited high activity against c-di-AMP. As little as 10% of the enzyme amount used in the assay with GdpP degraded the nucleotide more rapidly than GdpP. To conclude, our results clearly demonstrate that MPN140 is a phosphodiesterase, but with a substrate distinct from c-di-AMP whereas MPN549 is the c-di-AMP-degrading phosphodiesterase in M. pneumoniae. Thus, MPN549 was renamed PdeM (phosphodiesterase of M. pneumoniae), and the gene accordingly pdeM.

In B. subtilis, c-di-AMP is not only essential but also toxic in the absence of the phosphodiesterases that degrade the nucleotide (Gundlach et al., 2015b). Thus, we tested whether a pdeM mutant could be isolated from our transposon mutant collection. As observed in a global transposon mutant study (Lluch-Senar et al., 2015), no mutant could be isolated indicating that the pdeM gene is essential in M. pneumoniae.

Our results indicate that c-di-AMP is synthesized in M. pneumoniae, and that this second messenger is essential for the growth of the bacteria. This suggests an important physiological function of c-di-AMP. To identify this function, we decided to isolate c-di-AMP-binding proteins, which are excellent candidates for c-di-AMP-mediated regulation. For this purpose, we used c-di-AMP immobilized via a linker to agarose and incubated it with a protein extract of M. pneumoniae. As a control, we used agarose containing ethanolamine immobilized via the same linker. After extensive washing (see Materials and Methods), c-di-AMP-binding proteins were eluted with free c-di-AMP. The proteins were analyzed by SDS-PAGE (see Figure 3), and a band exhibiting increased intensity in the binding assay identified by mass spectrometry as MPN461. This protein is similar to the B. subtilis low affinity potassium transporter regulatory subunit KtrC (Holtmann et al., 2003). The mpn461 gene forms a gene cluster with the upstream gene mpn460, which encodes a protein most similar to the integral membrane subunit KtrD of the KtrC-KtrD potassium channel. An alignment of the c-di-AMP-binding region of S. aureus (Kim et al., 2015) with the relevant residues of M. pneumoniae and related proteins from other bacteria indicates that the amino acids interacting with c-di-AMP are conserved in the M. pneumoniae protein (see Figure 4). Taken together, these results suggest that c-di-AMP controls potassium transport in M. pneumoniae.