Fig 1.
Ion-exchange chromatography of the cyclic diadenylate phosphodiesterase present in soluble lysates of E. coli BL21: coincidence with phosphohydrolytic activities over other substrates.
This corresponds to the second step of the purification, where a sample obtained by gel-filtration chromatography was fractionated by chromatography on a Q-Sepharose column. This experiment was performed when the protein associated with the activities have been already identified as CpdB by peptide mass fingerprinting. In the graph, the activities on cyclic dinucleotides are multiplied by 3000.
Fig 2.
Identification of the endogenous cyclic diadenylate phosphodiesterase.
Correlation of cyclic diadenylate phosphodiesterase activity with a ≈ 66-kDa protein band. Two identical lanes of the same denaturing electrophoresis gel were loaded each with 24 μg of partially purified cyclic diadenylate phosphodiesterase (Fig 1). One of them was stained with Coomassie blue (top), and the other was cut into pieces and extracted under conditions that favored renaturation. In the extracts of these pieces, the phosphohydrolytic activities on the indicated substrates were assayed. They are plotted in the main graph. The activities on cyclic dinucleotides are multiplied by 5000. The protein band coinciding with the activities was processed to obtain its tryptic fingerprint. Out of 65 peptide masses recorded, 21 coincided with tryptic peptides of CDPB_ECOLI with a 50 ppm tolerance.
Fig 3.
Expression and purification of mature CpdB.
The recombinant protein was expressed in BL21 cells transformed with pGEX-6P-3-cpdB and induced by IPTG. The GST-CpdB fusion protein was purified by affinity adsorption to GSH-Sepharose. Mature CpdB separated from the GST tag was recovered from the gel after in-column digestion with PreScission. (A) Denaturing gel electrophoresis analysis. L–and L+, bacterial lysates before and after IPTG induction. S and P, supernatant and precipitate of the L+ lysate. M, molecular weight markers (phosphorylase B, 97400; bovine serum albumin, 66200; ovoalbumin, 45000). E, fraction excluded from the GSH-Sepharose column during application of the PreScission protease. 1–4, fractions collected after intracolumn proteolysis. X and Y, bands corresponding to the GST-CpdB fusion and to GPLGS-CpdB obtained by proteolysis. (B) Phosphohydrolytic activities on the indicated substrates, measured in the fractions collected from the GSH-Sepharose column.
Fig 4.
Response of recombinant CpdB activities to pH.
The activities were assayed on the indicated substrates under standard conditions except that 100 mM Tris-HCl or 100 mM Tris-acetate were used as buffers. The values of pH were taken directly from reaction mixtures at 37°C. Each data point is a mean value ± standard deviation (n = 3).
Fig 5.
Response of recombinant CpdB activities to divalent cations.
(A) Response of the phosphohydrolytic activity on 2´,3´-cAMP to varying concentration of Mn2+. Data points are mean values ± standard deviations of 3–6 measurements in six different experiments. (B) Effects of cations alternative to Mn2+ as activators. In every case the cation was used at a 2 mM concentration of the chloride salt. Cyclic dinucleotides were used at 20 μM, 2´,3´-cAMP and 3´-AMP at 750 μM. The results are expressed as percentages of the activities measured with 2 mM Mn2+: c-di-AMP, 0.15 ± 0.015 U mg-1 (100 ± 10%); c-di-GMP, 0.05 ± 0.005 U mg-1 (100 ± 10%); 2´,3´-cAMP, 130 ± 17 U mg-1 (100 ± 13%); 3´-AMP, 120 ± 16 U mg-1 (100 ± 13%). The bars are mean values ± standard deviations of 3–5 measurements in five different experiments. All the differences with respect to the Mn2+-dependent controls were significant according to the Dunnett test (P < 0.01 or, in the case of the Co2+-dependent activity on 3´-AMP, P < 0.05).
Table 1.
The substrates are listed in order of decreasing catalytic efficiency (kcat/Km). The kcat and Km data are mean values ± standard deviations of 3–6 experiments.
Fig 6.
Reaction products of recombinant CpdB by HPLC.
All the incubations were performed at 37°C. (A) Hydrolysis of 500 μM 2´,3´-cAMP by 0.23 μg ml-1 of CpdB in a 12-min incubation. (B) Hydrolysis of (left panels) 10 μM pApA by 1.5 μg ml-1 of CpdB in a 4-min incubation, or (right panels) 5 μM pGpG by 1 μg ml-1 of CpdB in a 20-min incubation. (C) Hydrolysis of (left panels) 20 μM c-di-AMP by 0.3 μg ml-1 of CpdB in a 15-min incubation, or (right panels) 20 μM c-di-GMP by 15 μg ml-1 of CpdB in an 8-min incubation.
Fig 7.
Architectures of cyclic dinucleotide phosphodiesterases of different classes.
For comparison with CpdB, six proteins representative of previously known cyclic dinucleotide phosphodiesterases are shown with their catalytic domains highlighted (grey): two EAL domain proteins of different architecture, a membrane-bound and a soluble DHH–DHHA1 proteins, a HD-GYP protein, and a 7TMR-HD protein. CpdB does not contain any of those domains. TM, transmembrane helix. SP, cleavable signal peptide addressing the precursor protein to the periplasm; not present in the mature protein. The numbers indicate the length of the proteins.